Commodore 128 Programmers Reference Guide

T .M .
COMMODORE cfl
Programmer’s
Reference Guide
COMMODORE
BUSINESS MACHINES, INC.
BANTAM
COMPUTER
B O O K S A Bantam Computer Book
COMMODORE 128
PROGRAMMER’S
REFERENCE GUIDE
Bantam Computer Books
Ask your bookseller for the books you have missed
THE AMIGADOS MANUAL THE MACINTOSH

by Commodore-Amiga, Inc. by Bill O ’Brien
THE APPLE //c BOOK THE NEW jr. A GUIDE TO IB M ’S PCjr

by Bill O ’Brien by Winn L. Rosch
THE COM MODORE 64 SURVIVAL ORCHESTRATING SYMPHONY

MANUAL by The Waite Group/Dan Shafer
by Winn L. Rosch
PC-DOS/MS-DOS
COM M ODORE 128 PROGRAM M ER’S User’s Guide to the M ost Popular Operating
REFERENCE GUIDE System fo r Personal Computers
by Commodore Business Machines, Inc. by Alan M. Boyd
EXPLORING ARTIFICIAL INTELLIGENCE POWER PAINTING: COMPUTER GRAPHICS

ON YOUR APPLE II ON THE MACINTOSH
by Tim Hartnell by Verne Bauman and Ronald Kidd/
illustrated by Gasper Vaccaro
EXPLORING ARTIFICIAL INTELLIGENCE
ON YOUR COM MODORE 64 SMARTER TELECOMMUNICATIONS
by Tim Hartnell Hands-On Guide to On-Line Computer Services
by Charles Bowen and Stewart Schneider
EXPLORING THE UNIX ENVIRONMENT
by The Waite Group/Irene Pasternack SWING WITH JAZZ: LOTUS JAZZ ON THE
MACINTOSH
FRAMEWORK FROM THE GROUND UP
by Datatech Publications Corp./Michael McCarty
by The Waite Group/Cynthia Spoor and
Robert Warren USER’S GUIDE TO THE AT&T PC 6300
PERSONAL COMPUTER
HOW TO GET THE MOST OUT OF
by David B. Peatroy, Ricardo A. Anzaldua,
COM PUSERVE, 2d ed.
H. A. Wohlwend, and Datatech Publications
by Charles Bowen and David Peyton
Corp.
HOW TO GET THE MOST OUT OF THE
SOURCE
by Charles Bowen and David Peyton
CO M M ODORE 128
PROGRAMMER’S
REFERENCE GUIDE
COMM ODORE BUSINESS MACHINES, INC.
BANTAM BOOKS
TORONTO • NEW YORK • LONDON • SYDNEY • AUCKLAND
COM M O D O RE 128 PROGRAMMER S REFERENCE GUIDE
A Bantam Book / February 1986
Commodore 64 and Commodore 128 are registered trademarks o f Commodore

Electronics. Ltd.
CPIM and CPIM Plus Version 3.0 are registered trademarks o f Digital Research
Inc.
Perfect is a registered trademark o f Perfect Software.
TouchTone is a registered trademark o f AT&T.
WordStar is a registered trademark o f MicroPro International Corporation.
Grateful acknowledgment is made fo r permission to reprint two bars o f Invention

13 (Inventio 13) by Johann Sebastian Bach. Sheet music copyright © C. F. Peters,
Corp., New York.
Book design by Ann Gold.
Cover design by Jo Ellen Temple.
All rights reserved.

Copyright © 1986 by Commodore Capital, Inc.
This book may not be reproduced in whole or in part, by
mimeograph or any other means, without permission.
For information address: Bantam Books, Inc.
ISBN 0-553-34292-4
Published simultaneously in the United States and Canada
Bantam Books are published by Bantam Books, Inc. Its trademark, consisting o f
the words “Bantam Books" and the portrayal o f a rooster, is Registered in U.S.
Patent and Trademark Office and in other countries. Marca Registrada. Bantam
Books, Inc., 666 Fifth Avenue, New York, New York 10103.
P R IN T E D IN T H E U N IT E D S T A T E S O F A M E R IC A
HL 0 9 8 7 6 5 4 3 2 1
CONTENTS
C hapter 1
Introduction 1
C hapter 2
B A SIC Building Blocks and BASIC 7.0 Encyclopedia 11
C hapter 3
O ne Step Beyond Simple BASIC 91
C hapter 4
C om m odore 128 G raphics Program m ing 109
C hapter 5
M achine Language 123
C hapter 6
H ow to Enter M achine Language Program s Into the
C om m odore 128 181
C hapter 7
M ixing M achine Language and BASIC 197
C hapter 8
The Pow er Behind Com m odore 128 Graphics 207
C hapter 9
Sprites 265
C hapter 10
Program m ing the 80-Colum n (8563) Chip 291
C hapter 11
Sound and M usic on the Com m odore 128 335
C hapter 12
Input/O utput Guide 371
Chapter 13
The Com m odore 128 O perating System 401
Chapter 14
CP/M 3.0 on the Commodore 128 A ll
Chapter 15
The Com m odore 128 and Com m odore 64 M em ory M aps 501
Chapter 16
C l 28 H ardw are Specifications 555
Appendixes 643
G lossary 731
Index 739
ACKNOWLEDGMENTS
W ritten by Larry Greenley

and
Fred Bowen
Bil Herd
Dave Haynie
Terry Ryan
Von Ertwine
Kim Eckert
Mario Eisenbacher
Norman McVey
The authors are deeply indebted to the many people who have contributed to the
preparation of this book. Special thanks go to Jim Gracely of Commodore Publications,
who reviewed the entire manuscript for technical accuracy and provided important
corrections, clarifications, and user-oriented suggestions, and to Steve Beats and Dave
Middleton of Commodore Software Engineering for their programming assistance and
expertise.
We also want to recognize the contributions of Frank Palaia of Commodore Hardware

Design, who provided expertise in the operation of the Z80 hardware, and of Dave
DiOrio of Commodore Integrated Circuit Design, who provided insight into the design
of the Memory M anagement Unit and the C l28 VIC chip enhancements.
For their extensive technical reviews of the manuscript, we wish to thank Bob Albright,
Pete Bowman, Steve Lam and Tony Porrazza of Commodore Engineering. We also
thank Dan Baker, Dave Street and Carolyn Scheppner of Commodore Software Techni
cal Support for providing an always available source of technical assistance. In addition,
we want to acknowledge the valuable contributions of members of Commodore Soft
ware Quality Assurance, especially Mike Colligon, Karen Mackenzie, Pat McAllister,
Greg Rapp, Dave Resavy, and Stacy English.
We also thank Carol Sullivan and Donald Bein for carefully proofreading various
sections of the text, Michelle Dreisbach for typing the manuscript, Marion Dooley for
preparing the art, Jo-Ellen Temple for the cover design, and Nancy Zwack for overall
coordination assistance.
Finally, we would like to acknowledge the unflagging support and guidance provided by
senior Commodore executives Paul Goheen, Harry McCabe and Bob Kenney.
p
; J___ _
INTRODUCTION
I
2 COMMODORE 128
■i
The Commodore 128 Personal Computer is a versatile, multimode computer. The

Commodore 128 is the successor to the commercially successful Commodore 64 com
puter. The principal features of the Commodore 128 are:
■ 128K bytes of RAM, optionally expandable to 256K or 640K

■ 80-column horizontal screen display
■ Hardware and software compatibility with Commodore 64
» CP/M 3.0 operation
■ Enhanced BASIC language
As this Guide shows, the Commodore 128 has many other new or expanded
capabilities and features. Those listed above, however, are the most significant when
assessing the Commodore 128’s capabilities against those of the Commodore 64 and
other microcomputers.
The Commodore 128 is actually three computers in one, with the following three
primary operating modes:
■ C l 28 Mode
■ C64 Mode
a CP/M Mode
Two of these primary modes (C l28 and CP/M) can operate using either a 40- or
80-column screen display. Following is a summary of the major features of each of the
three primary operating modes.
C l28 MODE
In C128 M ode, the Commodore 128 Personal Computer provides the capabilities and
memory needed to run sophisticated applications, such as word processing, spreadsheets,
and database programs.
C l 28 Mode features include:
■ 8502 processor running at 1.02 or 2.04 MHz

■ New, enhanced C l 28 Kernal
■ Built-in machine language monitor
■ Commodore BASIC 7.0 language, with over 140 commands and functions
■ Special new BASIC 7.0 commands that provide better, quicker and easier ways
to create complex graphics, animation, sound and music programs
■ 40-column text and bit map screen output using VIC II chip
■ 80-column text screen output using 8563 chip
IN TR O D U C TIO N 3
NOTE: The 40- and 80-column screen displays can be used either singly
or simultaneously with two monitors.
■ Sound (three voices) using SID chip

■ A 92-key keyboard that includes a full numeric keypad and ESCAPE, TAB,
ALT, CAPS LOCK, HELP, LINE FEED, 40/80 DISPLAY, and NO SCROLL
keys
■ Access to the full capabilities of the new peripheral devices from Commodore
(1571 fast disk drive, 1902 dual 40/80-column RGBI monitor, etc.)
■ Access to all standard Commodore serial peripherals
■ RAM expansion to 256 or 640K with optional RAM expansion modules
C64 MODE
In C64 Mode, the Commodore 128 retains all the capabilities of the Commodore 64,
thus allowing you to use the wide range of available Commodore 64 software.
C64 Mode features include:
■ 8502 processor running at 1.02 MHz

■ Standard C64 Kernal
■ BASIC 2.0 language
■ 64K of RAM
■ 40-column output using VIC II chip
■ Sound (three voices) using SID chip
■ Standard Commodore 64 keyboard layout except for function keys
■ All standard Commodore 64 keyboard functions
■ Access to all Commodore 64 graphics, color and sound capabilities, used
as on a Commodore 64
■ Compatibility with standard Commodore 64 peripherals,including user port and
serial devices, Datasette™, joysticks, com posite video m onitors, and
RF (TV) output devices
NOTE: The 1571 disk drive will function in C64 Mode, but only
at standard 1541 speed. C64 compatibility requirements make it impossi
ble for the 1571 to operate in C64 Mode at fast speed.
4 COMMODORE 128
CP/M MODE
In CP/M M ode, an onboard Z80 microprocessor gives you access to the capabilities of
Digital Research’s CP/M Version 3.0, plus a number of new capabilities added by Commodore.
CP/M Mode features iclude:
■ Integral Z80 processor running at 2.04 MHz

■ Disk-based CP/M 3.0 System
■ 128K bytes of RAM (in 64K banks)
■ 40-column screen output using VIC II chip
■ 80-column screen output using 8563 chip
■ Access to the full keyboard, including the numeric keypad and special keys
■ Access to the new fast serial disk drive (1571) and the standard serial peripherals
■ Ability to redefine almost any key
■ Ability to emulate several terminals (Lear-Siegler ADM31, ADM3A)
■ Support for various MFM disk formats (IBM, Kaypro, Epson, Osborne)
■ RAM expansion to 256 or 640K RAM with optional RAM expansion modules
The incorporation of CP/M 3.0 (also called CP/M Plus) into the Commodore 128
makes thousands of popular commercial and public domain software programs available
to the user.
HARDWARE COMPONENTS
The Commodore 128 Personal Computer incorporates the following major hardware
components:
PROCESSORS
8502: Main processor in C128, C64 Modes; I/O support for CP/M; 6502 software-
compatible; runs at 1.02 or 2.04 MHz
Z80: CP/M Mode only; runs at 2.04 MHz
MEMORY
R O M : 64K standard (C64 Kernal plus BASIC; C l 28 Kernal plus BASIC, character
ROMs and CP/M BIOS); one 32K slot available for software
RAM: 128K in two 64K banks; 16K display RAM for 8563 video chip; 2K X 4 Color RAM
VIDEO
8564: 40-column video (separate versions for NTSC and PAL TV standards)
8563: 80-column video
IN T R O D U C TIO N 5
SO UND
6 581: SID Chip
IN P U T /O U T P U T
6526: Joystick ports/keyboard scan/cassette
6526: User and serial ports
MEMORY M ANAGEMENT
8921: PLA (C64 plus C l28 mapping modes)
8922: MMU (Custom gate array)
For details on these and other hardware components see Chapter 16, Commodore
128 Hardware Specifications.
COM PATIBILITY W ITH

COMMODORE 64
The Commodore 128 system is designed as an upgrade to the Commodore 64. Accord
ingly, one of the major features of the Commodore 128 design is hardware and software
compatibility with the Commodore 64 when operating in C64 Mode. This means that in
C64 Mode the Commodore 128 is capable of running Commodore 64 application
software. Also, the Commodore 128 in C64 Mode supports Commodore 64 peripherals
except the CP/M 2.2 cartridge. (NOTE: The Commodore 128’s built-in CP/M 3.0
capability supersedes that provided by the external cartridge. This cartridge should not
be used with the Commodore 128 in any mode.)
The C128 Mode is designed as a compatible superset to the C64. Specifically, all
Kernal functions provided by the Commodore 64 are provided in the C l28 Kernal.
These functions are also provided at the same locations in the jump table of the C l 28
Kernal to provide compatibility with existing programs. Zero page and other system
variables are maintained at the same addresses they occupy in C64 Mode. This simpli
fies interfacing for many programs.
Providing Commodore 64 compatibility means that the new features of the Com
modore 128 cannot be accessed in C64 Mode. For example, compatibility and memory
constraints preclude modifying the C64 Mode Kernal to support the 1571 fast serial disk
drive. As noted previously, C64 Mode sees this drive as a standard serial disk drive. For
the same reason, C64 Mode does not have an 80-column screen editor, and C64 Mode
BASIC 2.0 cannot use the second 64K bank of memory.
6 COMMODORE 128
SW ITCHING FROM MODE TO MODE

As mentioned before, in the C128 and CP/M Modes the Commodore 128 can provide
both 40-column and 80-column screen displays. This means that the Commodore 128
actually has five operating modes, as follows:
■ C l 28 Mode with 80-column display

■ C l 28 Mode with 40-column display
■ C64 Mode (40-column display only)
■ CP/M Mode with 80-column display
■ CP/M Mode with 40-column display
Figure 1-1 summarizes the methods used to switch from mode to mode.
FRO M
TO
O FF C128 C128 C64 CP/M CP/M

40 COL 80 COL 40 COL 80 COL
C l 28 Check that 1. Press ESC I . Check that 1. Check that

40 COL 40/80 key key; 40/80 key 40/80 key 40/80 key
is UP. release. is UP. is UP. is UP.
Make sure 2. Press X 2. Turn com 2. Turn com 2. Turn com
that: key. puter OFF, puter OFF, puter OFF,
a)No CP/M OR then ON. then ON. then ON.
system 1. Check that 3. Remove
disk is 40/80 key cartridge
in drive is UP. if present
b)No C64 2. Press
cartridge RESET
is in ex button.
pansion
port
Turn com
puter ON.
C l 28 1. Press 1. Press 1. Press 1. Press 1. Check that
80 COL 40/80 key ESC key; 40/80 key 40/80 key 40/80 key
DOW N. release. DOW N. DOW N. is DOWN.
2. Turn com 2. Press X 2. Turn com 2. Remove 2. Remove
puter ON. key. puter OFF, CP/M sys CP/M sys
OR then ON. tem disk tem disk
1. Press 3. Remove from from
40/80 key cartridge drive, if drive, if
DOWN. if present. necessary. necessary.
2. Press 3. Turn com 3. Turn com
RESET puter OFF, puter OFF,
button. then ON. then ON.
Figure I —I . C o m m o d o re 128 M ode Sw itching C h a rt

FRO M
TO
OFF C128 C128 C64 CP/M CP/M

40 COL 80 COL 40 COL 80 COL
C64 1. Hold 1. Type GO 1. Type GO 1. Turn com 1. Turn com

“ C1 ’’ key 64; press 64; press puter OFF. puter OFF.
DOW N. RETURN. RETURN. 2. Check that 2. Check that
2. Turn com 2. The com 2. The com 40/80 key 40/80 key
puter ON. puter re puter re is UP. is UP.
OR sponds: sponds: 3. Hold 3. Hold
1. Insert C64 ARE YOU ARE YOU DOWN DOW N
cartridge. SURE? SURE? C‘ key C1 key
2. Turn com Type Y; Type Y; while while turn
puter ON. press press turning ing com
RETURN. RETURN. com puter puter ON.
ON. OR
OR 1. Turn com
1. Turn com puter OFF.
puter OFF. 2. Insert C64
2. Insert C64 cartridge.
cartridge. 3. Turn
3. Turn power
power ON. ON.
CP/M 1. Turn disk 1. Turn disk 1. Turn disk 1. Check that 1. Insert
40 COL drive ON. drive ON. drive ON. 40/80 key CP/M util
2. Insert 2. Insert 2. Insert is UP. ities disk
CP/M sys CP/M sys CP/M sys 2. Turn disk in drive.
tem disk tem disk tem disk drive ON. 2. At screen
in drive. in drive. in drive. 3. Insert prom pt,
3. Check that 3. Check that 3. Check that CP/M sys A > type:
40/80 key 40/80 key 40/80 key tem disk DEVICE
is UP. is UP. is UP. in drive. CONOUT: =
4. Turn com 4. Type: 4. Type: 4. Turn com 40 COL
puter ON. BOOT BOOT puter OFF. 3. Press
5. Press 5. Press RETURN.
RETURN. RETURN.
CP/M 1. Turn disk 1. Turn disk 1. Turn disk 1. Press 1. Insert

80 COL drive ON. drive ON. drive ON. 40/80 key CP/M util
2. Insert 2. Insert 2. Insert DOWN. ities disk
CP/M sys CP/M sys CP/M sys 2. Turn disk in drive.
tem disk tem disk tem disk drive ON. 2. At screen
in drive. in drive. in drive. 3. Insert prom pt,
3. Press 3. Press 3. Check that CP/M sys A > type:
40/80 key 40/80 key 40/80 key tem disk DEVICE
DOW N. DOW N. is DOW N. in drive. CONOUT =
4. Turn com 4. Type: 4. Type: 4. Turn com 80 COL
puter ON. BOOT BOOT. puter OFF. 3. Press
5. Press 5. Press RETURN.
RETURN. RETURN.
Figure I —I . C o m m o d o re 128 Mode Sw itching C h a rt (continued)

8 COMMODORE 128
NOTE: If you are using a Commodore 1902 dual monitor, remember to

move the video switch on the monitor from COMPOSITE or SEPA
RATED to RGBI when switching from 40-column to 80-column display;
reverse this step when switching from 80 to 40 columns. Also, when
switching between modes remove any cartridges from the expansion port.
You may also have to remove any disk (e.g., CP/M) from the disk drive.
CP/M 3.0 SYSTEM RELEASES
When you send in your C l28 warranty card, your name will be added to a list
which makes you eligible for CP/M system release dates.
H O W TO USE THIS GUIDE

This guide is designed to be a reference tool that you can consult whenever you need
detailed technical information on the structure and operation of the Commodore 128
Personal Computer. Since many of the design features of the Commodore 128 can be
viewed from various aspects, it may be necessary to consult several different chapters to
find the information you want. Note that certain groups of chapters form logical sequences
that cover in detail an extended topic like BASIC, graphics, or machine language.
The following chapter summaries should help you pinpoint what chapter or
chapters are most likely to provide the answer to a specific question or problem.
C H A P T E R 2. BASIC BUILDING BLOCKS AND BASIC 7.0 ENCYCLOPEDIA—

Defines and describes the structural and operational components of the BASIC
language, including constants, variables and arrays, and numeric and string ex
pressions and operations.
C H A P T E R 3. ONE STEP BEYOND SIMPLE BASIC— Provides routines (menu,
keyboard buffer, loading, programming function keys) and techniques (“ crunch
ing” or saving memory; debugging and merging programs; relocating BASIC)
that can be incorporated in your own programs. Provides modem-related informa
tion (how to generate TouchTone® frequencies, how to detect telephone ringing,
etc.) plus technical specifications for Commodore Modem/1200 and Modem/300.
C H A P T E R 4. COMMODORE 128 GRAPHICS PROGRAMMING— Describes the
general BASIC 7.0 graphics commands (COLOR, GRAPHIC, DRAW, LO
CATE, BOX, CIRCLE, PAINT) and gives annotated examples of use, including
programs. Describes the structure and general function of the color modes and
character and bit map graphics modes that are fundamental to Commodore 128
graphics.
C H A P T E R 5. MACHINE LANGUAGE ON THE COMMODORE 128— Defines,

with examples, machine language (ML) and associated topics, including the
Kernal; the 8502 registers, binary and hexadecimal numbers, and addressing
modes. Defines, with examples, types of ML instructions (op codes, etc.).
Includes 8502 instruction and addressing table.
C H A P T E R 6. HOW TO ENTER MACHINE LANGUAGE PROGRAMS INTO THE
COMMODORE 128— Describes, with examples, how to enter ML programs by
using the built-in Machine Language Monitor or by POKEing decimal op-code
values with a BASIC program. Defines, with examples, the ML Monitor commands.
C H A P T E R 7. MIXING MACHINE LANGUAGE AND BASIC— Describes, with
examples, how to combine BASIC and ML instructions in the same program by
using BASIC READ, DATA, POKE and SYS commands. Shows where to place
ML programs in memory.
C H A P T E R 8. THE POWER BEHIND COMMODORE 128 GRAPHICS— Describes
the C l 28 Mode memory banking concept and tells how to manage banked
memory. Defines the use of shadow registers. Describes how screen, color and
character memory are handled in BASIC and machine language, for both character
and bit map modes. Shows how to redefine the character set. Describes use of
split-screen modes. Includes a tabular graphics programming summary.
C H A P T E R 9. SPRITES— Describes programming of sprites or MOBs (movable object
blocks). Defines and shows how to use the BASIC 7.0 sprite-related commands
(SPRITE, SPRDEF, MOVSPR, SSHAPE, GSHAPE, SPRSAV). Provides anno
tated examples of use, including programs.
C H A P T E R 10. PROGRAMMING THE 8O-COLUMN (8563) CHIP— Defines the
8563 registers and describes, with machine language examples, how to program
the 80-column screen in character and bit map modes.
C H A P T E R I I . SOUND AND MUSIC ON THE COMMODORE 128— Defines the
BASIC 7.0 sound and music commands (SOUND, ENVELOPE, VOL, TEMPO,
PLAY, FILTER). Describes how to code a song in C128 Mode. Defines in detail
the Sound Interface Device (SID) and how to program it in machine language.
C H A P T E R 12. INPUT/OUTPUT GUIDE— Describes software control of peripheral
devices connected through I/O ports, including disk drives, printers, other User
Port and Serial Port devices, the Datassette, and Controller Port devices. Provides
pin-out diagrams and pin descriptions for all ports.
C H A P T E R 13. THE COMMODORE 128 OPERATING SYSTEM— Describes, with
examples, the operating system (Kernal), which controls the functioning of the
Commodore 128; includes the Kernal Jump Table, which lists the ROM entry
points used to call the Kernal routines; defines each Kernal routine; defines the
C128 Screen Editor. Describes the Memory Management Unit (MMU), defines
the MMU registers, tells how to select and switch banks in BASIC and ML, and
tells how to predefine memory configurations.
C H A P T E R 14. CP/M 3.0 ON THE COMMODORE 128— Summarizes the Commo
dore version of CP/M 3.0. Defines the general system layout and the operating
system components (CCP, BIOS and BDOS). Describes the Commodore enhance
ments to CP/M 3.0. (Additional details on CP/M 3.0 are given in Appendix K.)
10 COMMODORE 128
C H A P T E R 15. COMMODORE 128 AND COMMODORE 64 MEMORY MAPS—

Provides detailed memory maps for C l 28 and C64 modes. (The Z80 memory
map is shown in Appendix K.)
C H A P T E R 16. HARDWARE SPECIFICATIONS— Includes technical specifications for
Commodore 128 hardware components (8563, 8564, etc.).
A P P E N D IX E S A through L— Provide additional technical information and/or a more
convenient grouping of information supplied elsewhere in the Guide (e.g., pinout
diagrams).
G L O S S A R Y — Provides standard definitions of technical terms.
2
BASIC
BUILDING
BLOCKS AND
BASIC 7.0
ENCYCLOPEDIA
12 COMMODORE 128
The BASIC language is composed of commands, operators, constants, variables, arrays

and strings. Commands are instructions that the computer follows to perform an
operation. The other elements of BASIC perform a variety of functions, such as
assigning values to a quantity, passing values to the computer, or directing the computer
to perform a mathematical operation. This section describes the structure and functions
of the elements of the BASIC language.
COMMANDS A N D STATEMENTS
By definition, commands and statements have the following distinctions. A command is
a BASIC verb which is used in immediate mode. It is not preceded by a program line
number and it executes immediately after the RETURN key is pressed. A statement is
a BASIC verb which is contained within a program and is preceded by a line number.
Program statements are executed with the RUN command followed by the RETURN key.
Most commands can be used within a program. In this case the command is
preceded by a line number and is said to be used in program mode. Many commands
also can be used outside a program in what is called direct mode. For example, LOAD
is an often-used direct mode command, but you can also include LOAD in a program.
GET and INPUT are commands that only can be used in a program; otherwise, an
ILLEGAL DIRECT ERROR occurs. While PRINT is usually included within a
program, you can also use PRINT in direct mode to output a message or numeric value
to the screen, as in the following example:
PRINT “ The Commodore 128” RETURN
Notice that the message is displayed on the screen as soon as you press the return
key. The following two lines display the same message on the screen. The first line is a
program mode statement; the second line is a direct mode command.
10 PRINT “ The Commodore 128” RETURN
RUN RETURN
It is important to know about the concepts behind memory storage before examin
ing the Commodore BASIC language in detail. Specifically, you need to understand
constants, variables and arrays.
NUMERIC MEMORY STORAGE:

CONSTANTS, VARIABLES A N D ARRAYS
There are three ways to store numeric information in Commodore BASIC. The first way
is to use a constant. A constant is a form of memory storage in which the contents
remain the same throughout the course of a program. The second type of memory
storage unit is a variable. As the name indicates, a variable is a memory storage cell in
BASIC BUILDING BLOCKS A N D BASIC 7.0 ENCYCLOPEDIA 13
which the contents vary or change throughout the course of a program. The last way to store
information is to use an array, a series of related memory locations consisting of variables.
Each of these three units of memory storage can have three different types of
information or data assigned. The three data types are INTEGER, FLOATING-POINT
or STRING. Integer data is numeric, whole number data— that is, numbers without
decimal points. Floating-point is numeric data including fractional parts indicated with a
decimal point. String data is a sequential series of alphanumeric letters, numbers and
symbols referred to as character strings. The following paragraphs describe these three
data types and the way each memory storage unit is assigned different data type values.
C O N STAN TS: INTEGER,

F LO A T IN G -P O IN T A N D STRING
INTEGER C O N STA N TS
The value assigned to a constant remains unchanged or constant throughout a program.
Integer constants can contain a positive or negative value ranging from -32768 through
+ 32767. If the plus sign is omitted, the C128 assumes that the integer is positive.
Integer constants do not contain commas or decimal points between digits. Leading
zeros are ignored. Integers are stored in memory as two-byte binary numbers, which
means a constant requires 16 bits or two bytes of memory to store the integer as a base
two number. The following are examples of integer constants:
1
1000
-3 2
0
-32767
F L O A TIN G -P O IN T CO N STA N TS
Floating-point constants contain fractional parts that are indicated by a decimal
point. They do not contain commas to separate digits. Floating-point constants may be
positive or negative. If the plus sign is omitted, it is assumed that the number is
positive. Again, leading zeros are unnecessary and ignored. Floating-point constants are
represented in two ways depending on their value:
1. Simple Number Notation

2. Scientific Notation
In simple number notation, the floating-point number is calculated to ten digits of

precision and stored using five bytes, but only nine digits are displayed on the screen or
printer. If the floating-point number is greater than nine digits, it is rounded according to
the tenth digit. If the tenth digit is greater than five, the ninth digit is rounded to the next
higher digit. If the tenth digit is less than five, the ninth digit is rounded to the next
lower digit. The rounding of floating-point numbers may become a factor when calculat
14 COMMODORE 128
ing values based upon floating-point numbers greater than nine digits. Your program
should test floating-point results and take them into consideration when basing these
values on future calculations.
As mentioned, floating-point numbers are displayed as nine digits. If the value of a
floating-point constant is less than .01 or greater than 999999999, the number is
displayed on the screen or printer in scientific notation. For example, the number
12345678901 is displayed as 1.23456789E+ 10. Otherwise, the simple number notation
is displayed. A floating-point constant in scientific notation appears in three parts:
1. The mantissa is the leftmost number separated by a decimal point.

2. The letter E indicates that the number is displayed in exponential (scientific)
notation.
3. The exponent specifies the power of ten to which the number is raised and the
number of places the decimal point is moved in order to represent the number
in simple number notation.
The mantissa and exponent can be positive or negative. The exponent can be
within the range -3 9 to + 3 8 . If the exponent is negative, the decimal point moves to
the left representing it as a simple number. If the exponent is positive, the decimal
point moves to the right representing it in simple number notation.
The Commodore 128 limits the size of floating-point numbers. The highest
number you can represent in scientific notation is 1.70141183E + 38. If you try to
represent a number larger than that, an OVERFLOW ERROR occurs. The smallest
number you can represent in scientific notation is 2.93873588E-39. If you try to
represent a number smaller than that, no error occurs but a zero is returned as the value.
You should therefore test floating-point values in your programs if your calculations are
based on very small numbers and the results depend on future calculations. Here are
examples of floating-point constants in simple number notation and others in scientific
notation:
SIM PLE NUMBER SCIENTIFIC
9.99 22.33E + 20
.0234 99999.234E -23
+ 10.01 ^ t5 .8 9 E -ll
-9 0 .2 3 -3 .1 4 E + 17
NOTE: The values in either column are not equivalent.
STRING C O N STA N TS
A string constant, as mentioned, is a sequential series of alphanumeric characters
(numbers, letters and symbols). A string constant can be as long as a 160-character input
line, minus the line number and any other information appearing on that program line.
By concatenating strings together, you may form a string as long as 255 characters. The
string may contain numbers, letters, and even decimal points and commas. However,
the string cannot contain the double quote (“ ) character, since this character delimits or
marks the beginning or ending of the string. You can represent a double quote character
within a string using CHR$(34). You can omit the closing double quote character of a
string if it is the last statement in a line of a program.
A string can even be assigned a null value, meaning no characters are actually
assigned to it. Assign a string a null value by omitting characters between the double
quotes and follow the opening double quote directly with a closing double quote. Here
are some examples of string constants:
“ Commodore 128”
“ q w erl2 3 4 !# $ % ()* .:,”
“ ” (null string)
“ John and Joan”
VARIABLES: INTEGER,
FLO A TIN G -P O IN T A N D STRING
Variables are units of memory storage that represent varying data values within a
program. Unlike constants, variables may change in value throughout the course of a
program. The value assigned to a variable can be an integer, a floating-point number, or
a string. You can assign a value to a variable as the result of a mathematical calculation.
Variables are assigned values using an equals sign. The variable name appears to the left
of the equals sign and the constant or calculation appears to the right. When you refer to
a variable in a program before you assign it a value, the variable value becomes zero if
it is an integer or floating-point number. It becomes a null string if the variable is a
string.
Variable names can be any length, but for efficiency you should limit the size
of the variable to a few characters. Only the first two characters of a variable name
are significant. Therefore, do not begin the names of two different variables with
the same two characters. If you do, the C l 28 will interpret them as the same variable
name.
The first character o f a variable name must be a letter. The rest of the
variable name can be any letter or number from zero to nine. A variable name
m ust not contain any BASIC keyw ord. If you include a BASIC keyw ord in
a variable nam e, a SYNTAX ERROR occurs. BASIC keyw ords include all
BASIC statements, commands, function names, logical operator names and reserved
variables.
You can specify the data type of a variable by following the variable name with
a percent sign (%) if the variable is an integer value, or a dollar sign if the
variable is a string. If no character is specified, the C l28 assumes that the variable
value is a floating-point number. Here are some examples of variables and how they are
assigned:
16 COMMODORE 128
A = 3.679 (floating-point)
Z% = 714 (integer)
F$ = “ CELEBRATE THE COMMODORE 128” (string)
T = A + Z% (floating-point)
Count % = Count % + 1 (integer)
G$ = “ SEEK A HIGHER LEVEL OF CONSCIOUSNESS” (string)
H$ = F$ + G$ (string)
ARRAYS: INTEGER,
FLO A TIN G -P O IN T A N D STRING
Although arrays were defined earlier in this chapter as series of related variables or
constants, you refer to them with a single integer, floating point or string variable name.
All elements have the same data type as the array name. To access successive elements
within the array, BASIC uses subscripts (indexed variables) to refer to each unique storage
compartment in the array. For example, the alphabet has twenty-six letters. Assume an
array called “ ALPHA” is constructed and includes all the letters of the alphabet. To
access the first element of the array, which is also the first letter of the alphabet (A),
label Alpha with a subscript of zero:
ALPHA$(0) A
To access the letter B, label Alpha with a subscript of one:
ALPHA$( 1) B
Continue in the same manner to access all of the elements of the array ALPHA, as in
the following:
ALPHA$(2) C
ALPHA$(3) D
ALPHA$(4) E
ALPHA$(5) Z
Subscripts are a convenient way to access elements within an array. If subscripts

did not exist, you would have to assign separate variables for all the data that would
normally be accessed with a subscript. The first subscript within an array is zero.
Although arrays are actually stored sequentially in memory, they can be multi
dimensional. Tables and matrices are easily manipulated with two-dimensional arrays.
For example, assume you have a matrix with ten rows and ten columns. You need 100
storage locations or array elements in order to store the whole matrix. Even though
your matrix is ten by ten, the elements in the array are stored in memory one
after the other for 100 hundred locations.
You specify the number of dimensions in the arrays with the DIM statement. For
example:
10 DIM A(99)
dimensions a one-dimensional floating-point array with 100 elements. The following are
examples of two-, three- and four-dimensional integer arrays:
20 DIM B(9, 9) (100 elements)

30 DIM C(20,20,20) (9261 elements)
40 DIM D( 10,15,15,10) (30976 elements)
In theory the maximum number of dimensions in an array is 255, but you cannot
fit a DIMension statement that long on a 160-character line. The maximum number of
DIMension statements you can fit on a 160-character line is approximately fifty. The
maximum number of elements allowed in each dimension is 32767. In practice, the size
of an array is limited to the amount of available memory. Most arrays are one-, two- or
three-dimensional. If an array contains fewer than ten elements, there is no need for a
DIM statement since the C l28 automatically dimensions variable names to ten elements.
The first time you refer to the name of the undimensioned array (variable) name, the
C128 assigns zero to the value if it is a numeric array, or a null string if it is a string
array.
You must separate the subscript for each dimension in your DIMension statement
with a comma. Subscripts can be integer constants, variables, or the integer result of an
arithmetic operation. Legal subscript values can be between zero and the highest
dimension assigned in the DIMension statement. If the subscript is referred to outside of
this range, a BAD SUBSCRIPT ERROR results.
The type of array determines how much memory is used to store the integer,
floating-point or string data.
Floating-point string arrays take up the most memory; integer arrays require the
least amount of memory. H ere’s how much memory each type of array requires:
5 bytes for the array name

+ 2 bytes for each dimension
+ 2 bytes for each integer array element
OR + 5 bytes for each floating-point element
OR + 3 bytes for each string element
AND + 1 byte per character in each string element
Keep in mind the amount of storage required for each type of array. If you only
need an integer array, specify that the array be the integer type, since floating-point
arrays require much more memory than does the integer type.
Here are some example arrays:

A$(0) = “ GROSS SALES” (string array)
MTH$(K%) = “ JA N ” (string array)
G2%(X) = 5 (integer array)
CNT%(G2%(X)) = CNT%( 1)-2 (integer array)
FP(12*K%) = 24.8 (floating-point array)
18 COMMODORE 128
SUM(CNT%( 1)) = FP*K% (floating-point array)

A(5) = 0 Sets the 5th element in the 1 dimensional array
called “ A ” equal to 0
B(5,6) = 26 Sets the element in row position 5 and column
position 6 in the 2 dimensional array called “ B ”
equal to 26
C (l,2 ,3 )= 1 0 0 Sets the element in row position 1, column
position 2, and depth position 3 in the 3 dimen
sional array called “ C ” equal to 100
EXPRESSIONS AND OPERATORS
Expressions are formed using constants, variables and/or arrays. An expression can be a
single constant, simple variable, or an array variable of any type. It also can be a
combination of constants and variables with arithmetic, relational or logical operators
designed to produce a single value. How operators work is explained below. Expres
sions can be separated into two classes:
1. ARITHMETIC
2. STRING
Expressions have two or more data items called operands. Each operand is
separated by a single operator to produce the desired result. This is usually done by
assigning the value of the expression to a variable name.
An operator is a special symbol the BASIC Interpreter in your Commodore 128
recognizes as representing an operation to be performed on the variables or constant
data. One or more operators, combined with one or more variables and/or constants
form an expression. Arithmetic, relational and logical operators are recognized by
Commodore 128 BASIC.
ARITHM ETIC EXPRESSIONS

Arithmetic expressions yield an integer or floating-point value. The arithmetic operators
( + ,-,* ,/, t ) are used to perform addition, subtraction, multiplication, division and
exponentiation operations, respectively.
ARITHM ETIC OPERATIONS

An arithmetic operator defines an arithmetic operation which is performed on the two
operands on either side of the operator. Arithmetic operations are performed using
floating-point numbers. Integers are converted to floating-point numbers before an
arithmetic operation is performed. The result is converted back to an integer if it is
assigned to an integer variable name.
A D D IT IO N ( + )
The plus sign ( + ) specifies that the operand on the right is added to the operand on the
left.
EXA M PLES:
2+2
A + B+C
X%+ 1
BR + 10E-2
SU BTR A C TIO N (-)

The minus sign (-) specifies that the operand on the right is subtracted from the operand
on the left.
EX A M P LE S:
4 -1
100-64
A-B
55-142
The minus also can be used as a unary minus which is the minus sign in front of a
negative number. This is equal to subtracting the number from zero (0).
EXA M PLES:
-5
-9E 4
-B
4 - (-2 ) (same as 4 + 2)
M U LTIP LIC A TIO N (*)

An asterisk (*) specifies that the operand on the left is multiplied by the operand on the
right.
EXA M PLES:
100*2
50*0
A*X1
R%* 14
D IV IS IO N (/)
The slash (/) specifies that the operand on the left is divided by the operand on the
right.
20 COMMODORE 128
EX A M P LE S:
10/2
6400/4
A/B
4E2/XR
E X P O N E N T IA T IO N ( | )
The up arrow ( f ) specifies that the operand on the left is raised to the power specified
by the operand on the right (the exponent). If the operand on the right is a 2, the number
on the left is squared; if the exponent is a 3, the number on the left is cubed, etc. The
exponent can be any number as long as the result of the operation gives a valid
floating-point number.
EXA M PLES:
2| 2 Equivalent to 2*2
3| 3 Equivalent to 3*3*3
4| 4 Equivalent to 4*4*4*4
AB f CD
3| - 2 Equivalent to V3*lA
RELATIONAL OPERATORS
The relational operators ( < , = , > , < = ,> = ,< > ) are primarily used to compare the
values of two operands, but they also produce an arithmetic result. The relational
operators and the logical operators (AND, OR, and NOT), when used in comparisons,
produce an arithmetic true/false evaluation of an expression.If the relationship stated in
the expression is true, the result is assigned aninteger value o f - 1 . If it’s false a value of
0 is assigned. Following are the relational operators:
< LESS THAN

EQUAL TO
> GREATER THAN
< = LESS THAN OR EQUAL TO
> = GREATER THAN OR EQUAL TO
<> NOT EQUAL TO
EX A M P LE S:
5 -4 = 1 result true (-1)

14>66 result false (0)
15 > = 15 result true (-1)
Relational operators may be used to compare strings. For comparison purposes,

the letters of the alphabet have the order A < B < C < D . etc. Strings are compared by
evaluating the relationship between corresponding characters from left to right (see
string operations).
EXA M PLES:
“ A” < “ B” result true (-1)

“ X ” = “ YY” result false (0)
BB$ < > CC$ result false (0) if they arethe same
Numeric data items can only be compared (or assigned) with other numeric items.
The same is true when comparing strings; otherwise, the BASIC error message ?TYPE
MISMATCH occurs. Numeric operands are compared by first converting the values of
either or both operands from integer to floating-point form, as necessary. Then
the relationship between the floating-point values is evaluated to give a true/false
result.
At the end of all comparisons, you get an integer regardless of the data type
of the operand (even if both are strings). Because of this, a comparison of two
operands can be used as an operand in performing calculations. The result will
be -1 or 0 and can be used as anything but a divisor, since division by zero is
illegal.
LOGICAL OPERATORS
The logical operators (AND, OR, NOT) can be used to modify the meaning of the
relational operators or to produce an arithmetic result. Logical operators can produce
results other than -1 and 0, although any nonzero result is considered true when testing
for a true/false condition.
The logical operators (sometimes called Boolean operators) also can be used to
perform logical operations on individual binary digits (bits) in two operands. But when
you’re using the NOT operator, the operation is performed only on the single operand to
the right. The operands must be in the integer range of values (-32768 to + 32767)
(floating-point numbers are converted to integers) and logical operations give an integer
result.
Logical operations are performed bit-by-corresponding-bit on the two operands.
The logical AND produces a bit result of 1 only if both operand bits are 1. The logical
OR produces a bit result of 1 if either operand bit is 1. The logical NOT is the opposite
value of each bit as a single operand. In other words, “ If it’s NOT 1 then it is 0. If it’s
NOT 0 then it is 1.”
The exclusive OR IF (XOR) doesn’t have a logical operator but it is performed as
part of the WAIT statement or as the XOR function. Exclusive-OR means that if the
bits of two operands are set and equal, then the result is 0; otherwise the result is 1.
Logical operations are defined by groups of statements which, when taken to
gether, constitute a Boolean “ truth table” as shown in Table 2-1.
22 COMMODORE 128
The AND operation results in a 1 only if both bits are 1:

1 AND 1 = 1
0 AND 1 = 0
1 AND 0 = 0
0 AND 0 = 0
The OR operation results in a 1 if either bit is 1:

1 OR 1 = 1
0 OR 1 = 1
1 OR 0 = 1
0 OR 0 = 0
The NOT operation logically complements each bit:

NOT 1 = 0
NOT 0 = 1
The exclusive OR (XOR) is a function (not a logical operator):

1 XOR1= 0
1 XOR 0 = 1
0 XOR 1 = 1
0 XOR 0 = 0
T a b le 2-1 Boolean T ru th T ab le
The logical operators AND, OR and NOT specify a Boolean arithmetic operation
to be performed on the two operand expressions on either side of the operator. In the
case of NOT, only the operand on the right is considered. Logical operations (or
Boolean arithmetic) aren’t performed until all arithmetic and relational operations in an
expression have been evaluated.
EXA M P LE S:
IF A = 100 AND B = 100 THEN 10 (if both A and B have a value of 100 then
the result is true)
A = 96 AND 32: PRINT A (A = 32)
IF A = 100 OR B = 100 THEN 20 (if A or B is 100 then the result is true)
A = 64 OR 32: PRINT A (A = 96)
X = NOT 96(result is -9 7 (two’s complement))
HIERARCHY OF OPERATIONS
All expressions perform the different types of operations according to a fixed hierarchy.
Certain operations have a higher priority and are performed before other operations. The
normal order of operations can be modified by enclosing two or more operands within
parentheses ( ), creating a “ subexpression.” The parts of an expression enclosed in pa

rentheses will be reduced to a single value before evaluating parts outside the parentheses.
When you use parentheses in expressions, pair them so that you always have an
equal number of left and right parentheses. If you don’t, the BASIC error message
7SYNTAX ERROR will occur.
Expressions that have operands inside parentheses may themselves be enclosed in
parentheses, forming complex expressions of multiple levels. This is called nesting.
Parentheses can be nested in expressions to a maximum depth of ten levels— ten
matching sets of parentheses. The innermost expression has its operations performed
first. Some examples of expressions are:
A+ B
C f (D + E)/2
( ( X - C t (D + E)/2)*10) + 1
G G $>H H $
JJ$ + “ M ORE”
K % = 1 AND M O X
K% = 2 OR (A = B AND M <X )
NOT (D = E)
The BASIC Interpreter performs operations on expressions by performing arithme

tic operations first, then relational operations, and logical operations last. Both arithme
tic and logical operators have an order of precedence (or hierarchy of operations) within
themselves. Relational operators do not have an order of precedence and will be
performed as the expression is evaluated from left to right.
If all remaining operators in an expression have the same level of precedence, then
operations are performed from left to right. When performing operations on expressions
within parentheses, the normal order of precedence is maintained. The hierarchy of
arithmetic and logical operations is shown in Table 2-2 from first to last in order of
precedence. Note that scientific notation is resolved first.
OPERATOR DESCRIPTION EXAMPLE
T Exponentiation BASE t EXP

- Negation (Unary Minus) -A
*/ Multiplication AB * CD
Division E F / GH
+ Addition CNT + 2
- Subtraction JK-PQ
> = < Relational Operations A <= B
NOT Logical NOT NOT K%
(Integer T w o’s Complement)
AND Logical AND JK AND 128
OR Logical OR PQ OR 15
T a b le 2 -2 H ierarch y o f O peratio ns P erfo rm ed on Expressions

24 COMMODORE 128
STRING OPERATIONS
Strings are compared using the same relational operators ( = , < > , < = , > = , < , > )
that are used for comparing numbers. String comparisons are made by taking one
character at a time (left-to-right) from each string and evaluating each character
code position from the character set. If the character codes are the same, the char
acters are equal. If the character codes differ, the character with the lower CBM ASCII
code number is lower in the character set. The comparison stops when the end of either
string is reached. All other factors being equal, the shorter string is considered less than
the longer string. Leading or trailing blanks are significant in string evaluations.
Regardless of the data types, all comparisons yield an integer result. This is
true even if both operands are strings. Because of this, a comparison of two string
operands can be used as an operand in performing calculations. The result will
be -1 or 0 (true or false) and can be used in any mathematical operation but division
since division by zero is illegal.
STRING EXPRESSIONS
Expressions are treated as if an implied “ < > 0 ” follows them. This means that if an
expression is true, the next BASIC statement on the same program line is executed. If
the expression is false, the rest of the line is ignored and the next line in the program is
executed.
Just as with numbers, you can perform operations on string variables. The only
arithmetic string operator recognized by BASIC 7.0 is the plus sign ( + ) which is used
to perform concatenation of strings. When strings are concatenated, the string on the
right of the plus sign is appended to the string on the left, forming a third string. The
result can be printed immediately, used in a comparison, or assigned to a variable name.
If a string data item is compared with (or set equal to) a numeric item, or vice-versa, the
BASIC error message ?TY PE M ISM A TC H occurs. Some examples of string expres
sions and concatenation are:
10 A$ = “ FILE” : B $ = “ NAM E”
20 NAM$ = A$ + B$ (yields the string “ FILENAME” )
30 RES$ = “ N EW ” + A$ + B$ (yields the string “ NEW FILENAME” )
O R G A N IZA TIO N OF THE

BASIC 7.0 ENCYCLOPEDIA
This section of Chapter 2 lists BASIC 7.0 language elements in an encyclopedia
format. It provides an abbreviated list of the rules (syntax) o f Commodore 128
BASIC 7.0, along with a concise description of each. Consult the Commodore 128
System Guide BASIC 7.0 Encyclopedia (Chapter 5) included with your computer for a
BASIC BUILDING BLOCKS A N D BASIC 7.0 ENCYCLOPEDIA 2S
more detailed description of each command. BASIC 7.0 includes all the elements of
BASIC 2.0.
The different types of BASIC operations are listed in individual sections, as
follows:
1. C om m ands and Statem ents: the commands used to edit, store and erase
programs, and the BASIC program statements used in the numbered lines of a
program.
2. Functions: the string, numeric and print functions.
3. Reserved W ords and Symbols: the words and symbols reserved for
use by the BASIC 7.0 language, which cannot be used for any other
purpose.
COMMAND AND
STATEMENT FORMAT
The command and statement definitions in this encyclopedia are arranged in the follow
ing format:
Command name— A U TO
Brief definition— Enable/disable automatic line numbering
Command format— AUTO [line#]
Discussion of This command turns on the automatic line-numbering feature.
format and use— This eases the job of entering programs, by automatically typing
the line numbers for the user. As each program line is entered by
pressing RETURN, the next line number is printed on the screen,
and the cursor is positioned two spaces to the right of the line
number. The line number argument refers to the desired incre
ment between line numbers. AUTO without an argument turns off
the auto line numbering, as does RUN. This statement can be
used only in direct mode (outside of a program).
E XA M PLES:
AUTO 10 Automatically numbers program lines in incre

Example(s)— ments of 10.
AUTO 50 Automatically numbers lines in increments of 50.
AUTO Turns off automatic line numbering.
26 COMMODORE 128
The boldface line that defines the format consists of the following elements:
DLOAD “ program nam e” [,D0,U8]
t t
keyword argument additional arguments
(possibly optional)
The parts of the command or statement that must be typed exactly as shown are in
capital letters. Words the user supplies, such as the name of a program, are not
capitalized.
When quote marks (“ ” ) appear (usually around a program name or filename), the
user should include them in the appropriate place, according to the format example.
Keywords are words that are part of the BASIC language. They are the central part of a
command or statement, and they tell the computer what kind of action to take.
These words cannot be used as variable names. A complete list of reserved words
and symbols is given at the end of this chapter.
Keywords, also called reserved words, appear in upper-case letters. Key
words may be typed using the full word or the approved abbreviation. (A full list
of abbreviations is given in Appendix I). The keyword or abbreviation must be
entered correctly or an error will result. The BASIC and DOS error messages are
defined in Appendices A and B, respectively.
A rgum ents, also called parameters, appear in lower-case letters. Arguments comple
ment keywords by providing specific information to the command or statement.
For example, the keyword LOAD tells the computer to load a program while the
argument “ program name” tells the computer which specific program to load. A
second argument specifies from which drive to load the program. Arguments
include filenames, variables, line numbers, etc.
Square Brackets [ ] show optional arguments. The user selects any or none of the
arguments listed, depending on requirements.
A ngle Brackets < > indicate the user MUST choose one of the arguments listed.
A V ertical Bar | separates items in a list of arguments when the choices are limited to
those arguments listed. When the vertical bar appears in a list enclosed in
SQUARE BRACKETS, the choices are limited to the items in the list, but the
user still has the option not to use any arguments. If a vertical bar appears within
angle brackets, the user MUST choose one of the listed arguments.
Ellipsis . . . (a sequence of three dots) means an option or argument can be repeated more
than once.
Q u o tatio n M arks “ ” enclose character strings, filenames and other expressions.
When arguments are enclosed in quotation marks, the quotation marks must be
included in the command or statement. In this encyclopedia, quotation marks are
not conventions used to describe formats; they are required parts of a command or
statement.
Parentheses ( ) When arguments are enclosed in parentheses, they must be included in
the command or statement. Parentheses are not conventions used to describe
formats; they are required parts of a command or statement.
V ariab le refers to any valid BASIC variable names, such as X, A$, T%, etc.
Expression refers to any valid BASIC expressions, such as A + B + 2,.5*(X + 3),
etc.
NOTE: For all DOS commands, variables and expressions used as

arguments must be endorsed in parentheses.
BASIC COMMANDS A N D
STATEMENTS
APPEND
Append data to the end of a sequential file
APPEND #logical file num ber,“ filenam e” [,Ddrive num ber][<O N |,>U device]
EXA M PLES:
Append # 8, “ M YFILE” OPEN logical file 8 called “ M YFILE” , and prepare
to append with subsequent PRINT # statements.
Append #7,(A $),D 0,U 9 OPEN logical file named by the variable in A$ on
drive 0, device number 9, and prepare to APPEND.
A U TO
Enable/disable automatic line numbering
AUTO [line#]
EXA M P LE S:
AUTO 10 Automatically numbers program lines in increments of 10.

AUTO 50 Automatically numbers lines in increments of 50.
AUTO Turns off automatic line numbering.
BACKUP
Copy the entire contents from one disk to another on a dual disk drive
BACKUP source Ddrive number TO destination Ddrive number [< O N |,>
Udevice]
NOTE: This command can be used only with a dual-disk drive.

28 COMMODORE 128
EX A M P LE S:
BACKUP DO TO D1 Copies all files from the disk in drive 0 to the disk
in drive 1, in dual disk drive unit 8.
BACKUP DO TO D1 ON U9 Copies all files from drive 0 to drive 1, in disk
drive unit 9.
BANK
Select one of the 16 BASIC banks (default memory configurations), numbered 0-15 to
be used during PEEK, POKE, SYS, and WAIT commands.
BANK bank number
Here is a table of available BANK configurations in the Commodore 128 memory:
BANK CONFIGURATION
0 RAM(O) only
L R A M (l) only
2 RAM(2) only (same as 0)
4 Internal ROM , RAM(O), I/O
5 Internal ROM , R A M (l), I/O
6 Internal ROM , RAM (2), I/O (same as 4)
8 External ROM , RAM(O), I/O
9 External ROM , R A M (l), I/O
10 External ROM , RAM (2), I/O (same as 8)
12 Kernal and Internal ROM (LOW ), RAM(O), I/O
13 Kernal and External ROM (LOW ), RAM(O), I/O
14 Kernal and BASIC ROM, RAM(O), Character ROM
15 Kernal and BASIC ROM, RAM(O), I/O
Banks are described in detail in Chapter 8, The Power Behind Commodore 128
Graphics and Chapter 13, The Commodore 128 Operating System.
BEGIN / BEND
A conditional statement like IF . . . THEN: ELSE, structured so that you can include
several program lines between the start (BEGIN) and end (BEND) of the structure.
H ere’s the format:
IF condition THEN BEGIN : statement

statement
statement BEND : ELSE BEGIN
statement
statement BEND
EXAMPLE
10 IF X = 1 THEN BEGIN: PRINT “ X = 1 is True”
20 PRINT “ So this part of the statement is performed”
30 PRINT “ When X equals 1”
40 BEND: PRINT “ End of BEGIN/BEND structure” :GO to 60
50 PRINT “ X does not equal 1” :PRINT “ The statements between BEGIN/
BEND are skipped”
60 PRINT “ Rest of Program”
BLOAD
Load a binary file starting at the specified memory location
BLOAD “ filenam e” !,Ddrive num ber][< O N I,U >device number] [,Bbank

number] [,Pstart address]
where:
■ filename is the name of your file

a bank number selects one of the 16 BASIC banks (default memory con
figurations)
■ start address is the memory location where loading begins
EXA M PLES:
BLOAD “ SPRITES” , B0, P3584 LOADS the binary file "SPR IT E S”

starting in location 3584 (in BANK 0).
BLOAD “ DATA1” , DO, U8, B l, P4096 LOADS the binary file “ DATA 1”
into location 4096 (BANK 1) from
Drive 0, unit 8.
BO O T
Load and execute a program which was saved as a binary file
BOOT “ filenam e” [,Ddrive num ber][<O N |,>U device][,Palt LOAD ADD]
E X A M P LE :
BOOT BOOT a bootable disk (CP/M Plus for ex

ample).
30 COMMODORE 128
BOOT “ GRAPHICS 1” , DO, U9 LOADS the binary program “ GRAPHICS 1”

from unit 9, drive 0, and executes it.
BOX
Draw box at specified position on screen
BOX [color source], X I, Y l[,X 2,Y 2][,angle][,paint]
where:
color source 0 = Background color

1 = Foreground color (DEFAULT)
2 = Multi-color 1
3 = Multi-color 2
X I,Y 1 Corner coordinate (scaled)
X2,Y2 Corner diagonally opposite X I, Y l, (scaled); default is the PC

location.
angle Rotation in clockwise degrees;default is 0 degrees
paint Paint shape with color

0 = Do not paint
1 = Paint
(default = 0)
EX A M P LE S:
BOX 1, -I- 10, -I- 10 Draw a box 10 pixels to the right and 10 down from
the current pixel cursor location.
BOX 1, 10, 10, 60, 60 Draws the outline of a rectangle.
BOX , 10, 10, 60, 60, 45, 1 Draws a painted, rotated box (a diamond).
BOX , 30, 90, , 45, 1 Draws a filled, rotated polygon.
Any parameter can be omitted but you must include a comma in its place, as in the last
two examples.
NOTE: Wrapping occurs if the degree is greater than 360.

BSAVE
Save a binary file from the specified memory locations
BSAVE “ filenam e” [,D drive num ber][< O N I,U >device number] [,Bbank
number],Pstart address TO Pend address
where:
■ start address is the starting address where the program is SAVEd from
a end address is the last address + 1 in memory which is SAVEd
This is similar to the SAVE command in the Machine Language MONITOR.
EXA M P LE S:
BSAVE “ SPRITE DATA” ,BO, Saves the binary file named “ SPRITE DATA” ,
P3584 TO P4096 starting at location 3584 through 4095 (BANK
0).
BSAVE “ PROGRAM .SCR” ,DO, Saves the binary file named “ PROGRAM.
U9,B0,P3182 TO P7999 SCR” in the memory address range 3182
through 7999 (BANK 0) on drive 0, unit 9.
CATALOG
Display the disk directory
CATALOG [Ddrive num ber][<O N |,>U device number][,wildcard string]
E X A M P LE :
CATALOG Displays the disk directory on drive 0.
CHAR
Display characters at the specified position on the screen
CHAR [color source],X,Y[,string][,RVS]
This is primarily designed to display characters on a bit mapped screen, but it can also
be used on a text screen. H ere’s what the parameters mean:
color source 0 = Background

1 = Foreground
X Character column (0-39) (VIC screen)

(0-79) (8563) screen
32 COMMODORE 128
Y Character row (0-24)

string String to print
reverse Reverse field flag (0 = off, 1 = on)
E X A M P LE :
10 COLOR 2,3: REM MULTI-COLOR 1 = RED

20 COLOR 3,7: REM MULTI-COLOR 2 = BLUE
30 GRAPHIC 3,1
30 CHAR 0,10,10, “ TEX T” ,0
CIRCLE
Draw circles, ellipses, arcs, etc., at specified positions on the screen
CIRCLE [color source],X,Y[,Xr][,Yr] [,sa][,ea][,angle][,inc]

where:
color source 0 = background color

1 = foreground color
2 = multi-color I
3 = multi-color 2
X,Y Center coordinate of the CIRCLE
Xr X radius (scaled); (default = 0)
Yr Y radius (sealed default is Xr)
sa Starting arc angle (default 0 degrees)
ea Ending arc angle (default 360 degrees)
angle Rotation is clockwise degrees (default is 0 degrees)
inc Degrees between segments (default is 2 degrees)
EX A M P LE S:
CIRCLE1, 160,100,65,10 Draws an ellipse.

CIRCLE1, 160,100,65,50 Draws a circle.
CIRCLE1, 60,40,20,18,,,,45 Draws an octagon.

CIRCLE 1, 2 60,40,20,,,,.90 Draws a diamond.
CIRCLE 1, 60,140,20.18,,,,120 Draws a triangle.

CIRCLE 1, + 2, + 2,50,50 Draws a circle (two pixels down and two to the
right) relative to the original coordinates of the
pixel cursor.
CIRCLE 1, 30;90 Draws a circle 30 pixels and 90 degrees to the
right of the current pixel cursor coordinate
position.
You may omit a parameter, but you must still place a comma in the appropriate
position. Omitted parameters take on the default values.
CLOSE
Close logical file
CLOSE file number
EX A M P LE :
CLOSE 2 Logical file 2 is closed.
CLR
Clear program variables
CLR
CMD
Redirect screen output to a logical disk or print file.
CMD logical file number [,write list]
EX A M P LE :
OPEN 1,4 Opens device 4 (printer).
CMD 1 All normal output now goes to the printer.
LIST The LISTing goes to the printer, not the screen— even the word
READY.
PR IN T # 1 Sends output back to the screen.
CLOSE 1 Closes the file.

34 COMMODORE 128
COLLECT
Free inaccessible disk space
COLLECT [Ddrive num ber][<O N |,>U device]
E X A M P LE :
COLLECT DO Free all available space which has been incorrectly allocated to
improperly closed files. Such files are indicated with an asterisk
on the disk directory.
C O LLISIO N
Define handling for sprite collision interrupt
COLLISION type [,statement]
type Type of interrupt, as follows:
1 = Sprite-to-sprite collision
2 = Sprite-to-display data collision
3 = Light pen (VIC screen only)
statement BASIC line number of a subroutine
E X A M P LE :
Collision 1, 5000 Enables a sprite-to-sprite collision and program control sent to
subroutine at line 5000.
Collision 1 Stops interrupt action which was initiated in above example.
Collision 2, 1000 Enables a sprite-to-data collision and program control directed

to subroutine in line 1000.
COLOR
Define colors for each screen area
COLOR source number, color number
This statement assigns a color to one of the seven color areas:
AREA SOURCE
0 40-column (VIC) background

1 40-column (VIC) foreground
2 multicolor 1
3 multicolor 2
4 40-column (VIC) border
5 character color (40- or 80-column screen)
6 80-column background color
Colors that are usable are in the range 1-16.
CO LOR CODE COLOR CO LOR CODE COLOR
1 Black 9 Orange
2 White 10 Brown
3 Red 11 Light Red
4 Cyan 12 Dark Gray
5 Purple 13 Medium Gray
6 Green 14 Light Green
7 Blue 15 Light Blue
8 Yellow 16 Light Gray
C o lo r N u m b ers in 40-C olum n O u tp u t
1 Black 9 Dark Purple

2 White 10 Dark Yellow
3 Dark Red 11 Light Red
4 Light Cyan 12 Dark Cyan
5 Light Purple 13 Medium Gray
6 Dark Green 14 Light Green
7 Dark Blue 15 Light Blue
8 Light Yellow 16 Light Gray
C o lo r N u m b ers in 80-C olum n O u tp u t
EXA M PLES:
COLOR 0, 1: Changes background color of 40-column screen to black.
COLOR 5 ,8 : Changes character color to yellow.
CONCAT
Concatenate two data files
CONCAT “ file 2” [,Ddrive number] TO “ file 1”

[,Ddrive num ber][<O N |,>U device]
EX A M P LE :
Concat “ File B ” to “ File A ” FILE B is attached to FILE A, and the combined

file is designated FILE A.
Concat (A$) to (B$), D l, U9 The file named by B$ becomes a new file with
the same name with the file named by A$ at
tached to the end of B$. This is performed on
Unit 9, drive 1 (a dual disk drive).
36 COMMODORE 128
Whenever a variable is used as a filename, as in the last example, the filename variable
must be within parentheses.
CONT
Continue program execution
CONT
COPY
Copy a file from one drive to another within a dual disk drive. Copy one file to
another with a different name within a single drive
COPY [Ddrive num ber,]“ source fiIename” TO[Ddrive num ber,]“ destination
filename” [<O N |,>U device]
NOTE: Copying between two single or double disk drive units cannot be
done. This command does not support unit-to-unit copying.
EX A M P LE S:
COPY DO, “ TEST” TO D1, “ TEST PROG” Copies “ test” from drive 0 to drive
1, renaming it “ test prog” on drive 1.
COPY DO. “ STUFF” TO D l, “ STUFF” Copies “ STUFF” from drive 0 to
drive 1.
COPY DO TO Dl Copies all files from drive 0 to drive

1.
COPY “ W ORK.PROG” TO “ BACKUP” Copies “ W ORK.PROG” as a file

called “ BACKUP” on the same disk
(drive 0).
DATA
Define data to be used by a program
DATA list of constants
EX A M P LE :
DATA 100, 200, FRED, “ HELLO, M OM ” ,, 3, 14, ABC123

DCLEAR
Clear all open channels on disk drive
DCLEAR [Ddrive num ber][<O N |,>U device]
EXA M P LE S:
DCLEAR DO Clears all open files on drive 0, device number 8.
DCLEAR D1,U9 Clears all open files (channels) on drive 1, device number 9.
DCLOSE
Close disk file
DCLOSE [#logical file num ber][<O N |,>U device]
EXA M PLES:
DCLOSE Closes all channels currently open on unit 8.

DCLOSE # 2 Closes the channel associated with the logical file number 2 on
unit 8.
DCLOSE ON U9 Closes all channels currently open on unit 9.
DEF FN
Define a user function
DEF FN name (variable) = expression
E X A M P LE :
10 DEF FNA(X) = 12*(34.75-X/.3) + X

20 PRINT FNA(7)
The number 7 is inserted each place X is located in the formula given in the DEF
statement. In the example above, the answer returned is 144.
NOTE: If you plan to define a function in a program that will use BASIC
7.0 graphics commands, invoke the GRAPHIC command before defining
your function. The portion of memory where functions are defined and
where the graphics screen is located is shared. Once you allocate your
graphics area, the function definitions are safely placed somewhere else
in memory. If you don’t take this precaution and you invoke the GRAPHIC
command after you define a function, the function definition (between
$1C00 and $4000) is destroyed.
38 COMMODORE 128
DELETE
Delete lines of a BASIC program in the specified range
DELETE [first line] [-last line]
EXA M PLES:
DELETE 75 Deletes line 75.
DELETE 10-50 Deletes lines 10 through 50, inclusive.
DELETE-50 Deletes all lines from the beginning of the program up to and
including line 50.
DELETE 75- Deletes all lines from 75 to the end of the program, inclusive.
DIM
Declare number of elements in an array
DIM variable (subscripts) [,variable(subscripts)] . . .
E X A M P LE :
10 DIM A$(40),B7(15),CC% (4,4,4)
Dimension three arrays where arrays A$, B7 and CC% have 41 elements, 16 elements
and 125 elements respectively.
DIRECTORY
Display the contents of the disk directory on the screen
DIRECTORY [Ddrive num ber][<O N|,>Udevice][,wildcard]
E XA M PLES:
DIRECTORY Lists all files on the disk in unit 8.

DIRECTORY D l, U9, “ WORK” Lists the file named “ W O R K ,” on drive 1 of
unit 9.
DIRECTORY “ AB*” Lists all files starting with the letters “ A B”
like ABOVE, ABOARD, etc. on unit 8. The
asterisk specifies a wild card, where all files
starting with “ A B” are displayed.
DIRECTORY DO, “ 7.BAK” The ? is a wild card that matches any single
character in that position. For example: FILE
1.BAK, FILE 2.BAK, FILE 3.BAK all match
the string.
DIRECTORY D1,U9,(A$) LISTS the filename stored in the variable A$

on device number 9, drive 1. Remember, when
ever a variable is used as a filename, put the
variable in parentheses.
NOTE: To print the DIRECTORY of the disk in drive 0, unit 8, use the
following example:
LOAD“ $0” ,8
OPEN4,4:CMD4:LIST
PRINT#4:CLOSE4
D LO A D
Load a BASIC program from the disk drive, device 8.
DLOAD “ filenam e” [,Ddrive num ber][<O N |,>U device number]
EXA M PLES:
DLOAD “ BANKRECS” Searches the disk for the program “ BANKRECS”

and LOADS it.
DLOAD (A$) LOADS a program from disk in which the name is

stored in the variable A$. An error message is given if
A$ is null. Remember, when a variable is used as a
filename, it must be enclosed in parentheses.
DO / LOOP / W HILE / U N T IL / EXIT

Define and control a program loop
DO [UNTIL condition | WHILE condition]

statements [EXIT]
LOOP [UNTIL condition | WHILE condition]
This loop structure performs the statements between the DO statement and the LOOP
statement. If no UNTIL or WHILE modifies either the DO or the LOOP statement,
execution of the statements in between continues indefinitely. If an EXIT statement is
encountered in the body of a DO loop, execution is transferred to the first statement
following the LOOP statement. DO loops may be nested, following the rules defined by
the FOR-NEXT structure. If the UNTIL parameter is specified, the program continues
looping until the condition is satisfied (becomes true). The WHILE parameter is the
opposite of the UNTIL parameter: the program continues looping as long as the
condition is TRUE. As soon as the condition is no longer true, program control resumes
with the statement immediately following the LOOP statement. An example of a
condition (boolean argument) is A = 1, or G > 65.
40 COMMODORE 128
EX A M P LE S:
10 X = 25 This example performs the statements X = X -l

20 DO UNTIL X = 0 and PRINT “ X = ” ;X until X = 0 . When X = Othe
30 X = X -l program resumes with the PRINT “ End of Loop”
40 PRINT “ X = ” ;X statement immediately following LOOP.
50 LOOP
60 PRINT “ End of Loop”
10 DO WHILE A $ < > CHR$ (13):GETKEY A$:PRINT A$:LOOP

20 PRINT “ THE RETURN KEY HAS BEEN PRESSED”
This DO loop waits for a key to be pressed,
receives input from the keyboard one character at
a time and prints the letter of the key which is
pressed. If the RETURN key is pressed, control is
transferred out of the loop and line 20 is executed.
10 DOPEN # 8 , “ SEQFILE” This program opens file “ SEQFILE” and gets
20 DO data until the ST system variable indicates all data
30 GET # 8 ,A $ is input.
40 PRINT AS;
50 LOOP UNTIL ST
60 DCLOSE # 8
DOPEN
Open a disk file for a read and/or write operation
DOPEN # logical file num ber,“ filen am e[,< typ e> ]” [,Lrecord length]
t,Ddrive num ber][<O N |,>U device number][,W]
where type is:
S = Sequential File Type

P = Program File Type
U = User File Type
R = Relative File Type
L = Record Length = the length of records in a relative file only
W = Write Operation (if not specified a read operation occurs)
EX A M P LE S:
D O PEN # 1, “ ADDRESS” ,W Create the sequential file number 1 (ADDRESS)

for a write operation
D O PEN #2 “ RECIPES” ,D1,U9 Open the sequential file number 2 (RECIPES)
for a read operation on device number 9, drive 1
DRAW
Draw dots, lines and shapes at specified positions on the screen
DRAW [color source] [,X1, Y l][T O X2, Y2] . . .
where:
Color source 0 = Bit map background

1 = Bit map foreground
2 = Multi-color 1
3 = Multi-color 2
X I,Y 1 Starting coordinate (0,0 through 319,199)
X2,Y2 Ending coordinate (0,0 through 319,199)
EX A M P LE S:
DRAW 1 ,1 0 0 , 50 Draw a dot.
DRAW , 10,10 TO 100,60 Draw a line.
DRAW , 10,10 TO 10,60 TO 100,60 TO 10,10 Draw a triangle.

DRAW 1, 120;45 Draw a dot 45° relative and 120 pixels
away from the current pixel cursor
position.
DRAW Draw a dot at the present pixel cursor

position. Use LOCATE to position the
pixel cursor.
You may omit a parameter but you still must include the comma that would have
followed the unspecified parameter. Omitted parameters take on the default values.
DSAVE
Save a BASIC program file to disk
DSAVE “ filename” [,Ddrive num ber][<O N |,>U device number]
EX A M P LE S:
DSAVE "B A N K R ECS” Saves the program “ BANKRECS” to disk.

DSAVE (A$) Saves the disk program named in the variable A$.
DSAVE “ PROG 3” ,D1,U9 Saves the program “ PROG3” to disk on unitnum

ber 9, drive 1.
42 COMMODORE 128
□VERIFY
Verify the program in memory against the one on disk
DVERIFY “ filename” [,Ddrive num ber][<O N |,>U device number]
To verify Binary data, see VERIFY “ filename” ,8,1 format, under VERIFY command
description.
EX A M P LE S:
DVERIFY “ C128” Verifies program “ C128” on drive 0, unit 8.
DVERIFY “ SPRITES” ,DO,U9 Verifies program “ SPRITES” on drive 0, de

vice 9.
END
Define the end of program execution
END
ENVELOPE
Define a musical instrument envelope
ENVELOPE n[,atk] [,dec] [,sus] [,r e l][,w f] [,pw]
where:
n Envelope number (0-9)
a tk Attack rate (0-15)
dec Decay rate (0-15)
sus Sustain (0-15)
rel Release rate (0-15)
wf Waveform: 0 = triangle
1 = sawtooth
2 = variable pulse (square)
3 = noise
4 = ring modulation
pw Pulse width (0-4095)
See the “ T ” option in the PLAY command to select an envelope in a PLAY string.
E X A M P LE :
ENVELOPE 1, 10, 5, 10, 0, 2, 2048 This command sets envelope 1 to Attack

= 10, Decay = 5, Sustain = 10, Release
= 0, waveform = variable pulse (2), and
the pulse width = 2048
FAST
Sets the 8502 microprocessor at a speed of 2MHz.
FAST
This command initiates 2MHz mode, causing the VIC 40-column screen to be turned off.
All operations are speeded up considerably. Graphics may be used, but will not be visible
until a SLOW command is issued. The Commodore 128 powers up in 1MHz mode. The
DMA operations (FETCH, SWAP, STASH) must be performed at 1MHz (slow) speed.
FETCH
Get data from expansion (RAM module) memory
FETCH # b ytes, intsa, expsa, expb
where bytes = Number of bytes to get from expansion memory (0-65535) where 0 =
64K (65535 bytes)
intsa = Starting address of host RAM (0-65535)
expb = 64K expansion RAM bank number (0-7) where expb = 0-1 for 128K
and expb = 0-7 for up to 512K.
expsa = Starting address of expansion RAM (0-65535)
The host BANK for the ROM and I/O configuration is selected with the BANK
command. The DMA(VIC) RAM bank is selected by bits 6 and 7 of the RAM
configuration register within the MMU($D506).
FILTER
Define sound (SID chip) filter parameters
FILTER [freq][,lp] [,bp] [,hp] [,res]
where:
freq Filter cut-off frequency (0-2047)
lp Low-pass filter on (1), off (0)
bp Band-pass filter on (1), off (0)
hp High-pass filter on (1), off (0)
res Resonance (0-15)
Unspecified parameters result in no change to the current value.
EXA M PLES:
FILTER 1024,0,1,0,2 Set the cutoff frequency at 1024, select the band pass
filter and a resonance level of 2.
FILTER 2000,1,0,1,10 Set the cutoff frequency at 2000, select both the low
pass and high pass filters (to form a notch reject) and set
the resonance level at 10.
44 COMMODORE 128
FOR / T O / STEP / N E X T
Define a repetitive program loop structure.
FOR variable = start value TO end value [STEP increment] NEXT variable
The logic of the FOR/NEXT statement is as follows. First, the loop variable is set to the
start value. When the program reaches a program line containing the NEXT statement, it
adds the STEP increment (default = 1) to the value of the loop variable and checks to
see if it is higher than the end value of the loop. If the loop variable is less than or equal
to the end value, the loop is executed again, starting with the statement immediately
following the FOR statement. If the loop variable is greater than the end value, the loop
terminates and the program resumes immediately following the NEXT statement. The
opposite is true if the step size is negative. See also the NEXT statement.
E X A M P LE :
10 FOR L = 1 TO 10
20 PRINT L
30 NEXT L
40 PRINT “ I ’M DONE! L = ” L
This program prints the numbers from one to 10 followed by the message I ’M DONE!
L = 11.
E X A M P LE :
10 FOR L = 1 TO 100
20 FOR A = 5 TO 11 STEP .5
30 NEXT A
40 NEXT L
The FOR . . . NEXT loop in lines 20 and30 are nested inside the one in line 10 and 40.
Using a STEP increment of .5 is used to illustrate thefact thatfloating point indices are
valid. The inner rested loop must lie completely within the outer rested loop (lines 10
and 40).
GET
Receive input data from the keyboard, one character at a time, without waiting for a key
to be pressed.
GET variable list
E X A M P LE :
10 DO:GETA$:LOOP UNTIL A $ = ” A ” This line waits for the A key to be

pressed to continue.
20 GET B, C, D GET numeric variables B,C and D from the keyboard without
waiting for a key to be pressed.
GETKEY
Receive input data from the keyboard, one character at a time and wait for a key to be
pressed.
GETKEY variable list
E X A M P LE :
10 GETKEY A$
This line waits for a key to be pressed. Typing any key continues the program.
10 GETKEY A$,B$,C$
This line waits for three alphanumeric characters to be entered from the keyboard.
G ET#
Receive input data from a tape, disk or RS232
G E T # logical file number, variable list
E X A M P LE :
10 G E T #1,A $ This example receives one character, which is stored in the

variable A$, from logical file number 1. This example assumes
that file 1 was previously opened. See the OPEN statement.
G 064
Switch to C64 mode
G 064
To return to C128 mode, press the reset button, or turn off the computer power and
turn it on again.
GOSUB
Call a subroutine from the specified line number
GOSUB line number
E X A M P LE :
20 GOSUB 800 This example calls the subroutine beginning at line 800 and executes
it. All subroutines must terminate with a RETURN statement.
46 COMMODORE 128
800 PRINT “ THE C l 28 WAS WORTH THE W AIT!” : RETURN
G O TO / GO TO
Transfer program execution to the specified line number
GOTO line number
E XA M PLES:
10 PRINT“ COMMODORE” The GOTO in line 20 makes line 10 repeat continu-
20 GOTO 10 ously until RUN/STOP is pressed.
GOTO 100 Starts (RUNs) the program starting at line 100,
without clearing the variable storage area.
GRAPHIC
Select a graphic mode
1) GRAPHIC mode [,clear][,s]
2) GRAPHIC CLR
This statement puts the Commodore 128 in one of the six graphic modes:
M ODE DESCRIPTION
0 40-column text (default)

1 standard bit-map graphics
2 standard bit-map graphics (split screen)
3 multi-color bit-map graphics
4 multi-color bit-map graphics (split screen)
5 80-column text
E XA M PLES:
GRAPHIC 1,1 Select standard bit map mode and clear thebit map.
GRAPHIC 4,0,10 Select split screen multi-color bitmap mode, do not clear the
bit map and start the split screen at line 10.
GRAPHIC 0 Select 40-column text.
GRAPHIC CLR Clear and deallocate the bit map screen.
GSHAPE
See SSHAPE.
HEADER
Format a diskette
HEADER “ diskname” [,1 i.d.] [,Ddrive number]

[< O N |,> U d evice number]
Before a new disk can be used for the first time, it must be formatted with the HEADER
command. The HEADER command can also be used to erase a previously formatted
disk, which can then be reused.
When you enter a HEADER command in direct mode, the prompt ARE YOU
SURE? appears. In program mode, the prompt does not appear.
The HEADER command is analogous to the BASIC 2.0 command:
OPEN l,8 ,1 5 ,“ N0:disknam e,i.d.”
EX A M P LE S:
HEADER “ M YDISK” ,I23, DO This headers “ MYDISK” using i.d. 23

on drive 0, (default) device number 8.
HEADER “ RECS” , 145, D1 ON U9 This headers “ RECS” using i.d. 45, on

drive 1, device number 9.
HEADER “ C l 28 PROGRAMS” , DO This is a quick header on drive 0, device

number 8, assuming the disk in the drive
was already formatted. The old i.d. is
used.
HEADER (A$),I76,D0,U9 This example headers the diskette with

the name specified by the variable A$,
and the i.d. 76 on drive 0, device num
ber 9.
HELP
Highlight the line where the error occurred
HELP
The HELP command is used after an error has been reported in a program. When HELP
is typed in 40-column format, the line where the error occurs is listed, with the portion
containing the error displayed in reverse field. In 80-column format, the portion of the
line where the error occurs is underlined.
IF / TH EN / ELSE
Evaluate a conditional expression and execute portions of a program depending on the
outcome of the expression
48 COMMODORE 128
IF expression THEN statements [:ELSE else-clause]

THE IF . . . THEN statement evaluates a BASIC expression and takes one of two
possible courses of action depending upon the outcome of the expression. If the
expression is true, the statement(s) following THEN is executed. This can be any
BASIC statement or a line number. If the expression is false, the program resumes with
the program line immediately following the program line containing the IF statement,
unless an ELSE clause is present. The entire IF . . . THEN statement must be contained
within 160 characters. Also see BEGIN/BEND.
The ELSE clause, if present, must be on the same line as the IF . . . THEN
portion of the statement, and separated from the THEN clause by a colon. W hen an
ELSE clause is present, it is executed only when the expression is false. The expression
being evaluated may be a variable or formula, in which case it is considered true if
nonzero, and false if zero. In most cases, there is an expression involving relational
operators ( = , < , > , < = ,> = , < > ) .
EX A M P LE :
50 IF X > 0 THEN PRINT “ O K ” : ELSE END
This line checks the value of X. If X is greater than 0, the statement immediately
following the keyword THEN (PRINT “ O K ” ) is executed and the ELSE clause is
ignored. If X is less than or equal to 0, the ELSE clause is executed and the statement
immediately following THEN is ignored.
10 IF X = 10 THEN 100 This example evaluates the value of X.

20 PRINT “ X DOES NOT EQUAL 10” IF X equals 10, the program control is
transferred to line 100 and the message
99 STOP “ X EQUALS 10” is printed. IF X
100 PRINT “ X EQUALS 10” does not equal 10, the program resu
mes with line 20, the C l28 prints the
prompt “ X DOES NOT EQUAL 10”
and the program stops.
IN P U T
Receive a data string or a number from the keyboard and wait for the user to press
RETURN
INPUT [“ prompt string” ;] variable list
EX A M P LE :
10 INPUT “ PLEASE TYPE A NUMBER” ;A

20 INPUT “ AND YOUR NAM E” ;A$
30 PRINT A$ “ YOU TYPED THE NUMBER” ;A
IN P U T #
Input data from an I/O channel into a string or numeric variable
IN PU T # file number, variable list
E X A M P LE :
10 OPEN 2,8,2
20 IN PU T #2. A$, C, D$
This statement INPUTs the data stored in variables A$, C and D$ from the disk file
number 2, which was OPENed in line 10.
KEY
Define or list function key assignments
KEY [key number, string]
The maximum length for all the definitions together is 241 characters, (p. 3-41)
E X A M P LE :
KEY 7, “ GRAPHICO” + CHR$(13) + “ LIST” + CHR$(13)
This tells the computer to select the (VIC) text screen and list the program whenever the
F7 key is pressed (in direct mode). CHR$(13) is the ASCII character for RETURN and
performs the same action as pressing the RETURN key. Use CHR$(27) for ESCape.
Use CHR$(34) to incorporate the double quote character into a KEY string. The keys
may be redefined in a program. For example:
10 KEY 2 ,“ PRINT DS$” + CHR$(13)
This tells the computer to check and display the disk drive error channel variables
(PRINT DS$) each time the F2 function key is pressed.
10 FOR 1= 1 to 7 STEP 2
20 KEY I, CHR$(I + 132):NEXT
30 FOR I = 2 to 8 STEP 2
40 KEY I, CHR$(I + 132):NEXT
This defines the function keys as they are defined on the Commodore 64.
LET
Assigns a value to a variable
[LET] variable = expression
E X A M P LE :
10 LET A = 5 Assign the value 5 to numeric variable A.
50 COMMODORE 128
20 B = 6 Assign the value 6 to numeric variable B.

30C = A * B + 3 Assign the numeric variable C, the value resulting from 5
times 6 plus 3.
40 D$ = “ HELLO” Assign the string “ HELLO” to string variable D$.
LIST
List the BASIC program currently in memory
LIST [first line] [- last line]
In C128 mode, LIST can be used within a program without terminating program execution.
EXA M P LE S:
LIST Shows entire program.
LIST 1 0 0 - Shows from line 100 until the end of theprogram.
LIST 10 Shows only line 10.
LIST - 1 0 0 Shows all lines from the beginning through line 100.
LIST 10-200 Shows lines from 10 to 200, inclusive.
LOAD
Load a program from a peripheral device such as the disk drive or Datassette
LOAD “ filename” [,device number] [,relocate flag]
This is the command used to recall a program stored on disk or cassette tape. Here, the
filename is a program name up to 16 characters long, in quotes. The name must be
followed by a comma (outside the quotes) and a number which acts as a device number
to determine where the program is stored (disk or tape). If no number is supplied, the
Commodore 128 assumes device number 1 (the Datassette tape recorder).
EXA M PLES:
LOAD Reads in the next program from tape.
LOAD “ HELLO” Searches tape for a program called HELLO, and
LOADs it if found.
LOAD (A$),8 LOADs the program from disk whose name is
stored in the variable A$.
LOAD“ HELLO” ,8 Looks for the program called HELLO on disk drive
number 8, drive 0. (This is equivalent to DLOAD
“ HELLO” ).
LOAD“ MACHLANG” ,8 ,l LOADs the m achine language program called
“ MACHLANG” into the location from which it
was SAVEd.
LOCATE
Position the bit map pixel cursor on the screen
LOCATE X,Y
The LOCATE statement places the pixel cursor (PC) at any specified pixel coordinate on
the screen.
The pixel cursor (PC) is the coordinate on the bit map screen where drawing of
circles, boxes, lines and points and where PAINTing begins.
E X A M P LE :
LOCATE 160,100 Positions the PC in the center of the bit map screen. Noth
ing will be seen until something is drawn.
LOCATE + 2 0 ,1 0 0 Move the pixel cursor 20 pixels to the right of the last PC
position and place it at Y coordinate 100.
L O C A T E - 3 0 ,+ 20 Move the PC 30 pixels to the right and 20 down from the

previous PC position.
The PC can be found by using the RDOT(O) function to get the X-coordinate and
R D O T(l) to get the Y-coordinate. The color source of the pixel at the PC can be found
by PRINTing RDOT(2).
M O N ITO R
Enter the Commodore 128 machine language monitor
MONITOR
See Chapter 6 for details on the Commodore 128 Machine Language Monitor.
MOVSPR
Position or move sprite on the screen
1) MOVSPR number,X,Y Place the specified sprite at absolute

sprite coordinate X,Y.
2) MOVSPR number, + /- X , + /-Y Move sprite relative to the position
of the sprite’s current position.
3) MOVSPR number,X;Y Move sprite distance X at angle Y
relative to the sprite’s current position.
52 COMMODORE 128
4) MOVSPR number, angle # speed Move sprite at an angle relative to

its current coordinate, in the clock
wise direction and at the specified
speed.
where:
number is sprite’s number (1 through 8)

X,Y is coordinate of the sprite location.
angle is the angle (0-360) of motion in the clockwise direction relative to the
sprite’s original coordinate.
speed is a speed (0-15) in which the sprite moves.
This statement moves a sprite to a specific location on the screen according to

the SPRITE coordinate plane (not the bit map plane) or initiates sprite motion at a
specified rate. See MOVSPR in Chapter 9 for a diagram of the sprite coordinate
system.
EXA M PLES:
MOVSPR 1,150,150 Position sprite 1 near the center of the screen, x,y
coordinate 150,150.
MOVSPR l, + 20,-30 Move sprite 1 to the right 20 coordinates and up 30

coordinates.
MOVSPR 4, -5 0 , + 100 Move sprite 4 to the left 50 coordinates and down 100
coordinates.
MOVSPR 5, 45 # 1 5 Move sprite 5 at a 45 degree angle in the clockwise

direction, relative to its original x and y coordinates.
The sprite moves at the fastest rate (15).
NOTE: Once you specify an angle and a speed as in the fourth example
of the MOVSPR statement, the sprite continues on its path (even if the
sprite display is disabled) after the program stops, until you set the speed
to 0 or press RUN/STOP and RESTORE. Also, keep in mind that the
SCALE command affects the MOVSPR coordinates. If you add SCALing
to your programs, you also must adjust the sprites' new coordinates so
they appear correctly on the screen.
NEW
Clear (erase) BASIC program and variable storage
NEW
ON
Conditionally branch to a specified program line number according to the results of the
specified expression
ON expression <G O T O /G O SU B> line # 1 [, line # 2 , . . .]
E X A M P LE :
10 INPUT X:IF X < 0 THEN 10

20 ON X GOTO 30, 40, 50, 60 When X = 1, ON sends control to the first line
25 STOP number in the list (30). When X = 2, ON sends
30 PRINT “ X = 1” control to the second line (40), etc.
40 PRINT “ X = 2 ”
50 PRINT “ X = 3 ”
60 PRINT “ X = 4 ”
OPEN
Open files for input or output
OPEN logical file number, device number [,secondary address] [ < , “ filename
[,filetype[, [mode” ]]|< ,cm d string>]
EXA M P LE S:
10 OPEN 3,3 OPEN the screen as file number

3.
20 OPEN 1,0 OPEN the keyboard as file num

ber 1.
30 OPEN 1 ,1 ,0 ,“ DOT” OPEN the cassette for reading, as

file number 1, using “ DOT” as
the filename.
OPEN 4,4 OPEN the printer as file number 4.
OPEN 15,8,15 OPEN the command channel on

the disk as file 15, with secondary
address 15. Secondary address 15
is reserved for the disk drive error
channel.
5 OPEN 8 ,8 ,1 2 ,“ TESTFILE,S,W ” OPEN a sequential disk file for

writing called TESTFILE as file
number 8, with secondary address
12.
54 COMMODORE 128
See also: CLOSE, CMD, G E T # , IN P U T #, and PR IN T# statements and system

variables ST, DS, and DS$.
PA IN T
Fill area with color
PAINT [color source],X,Y[,mode]
where:
color source 0 = bit map foreground

1 = bit map background (default)
2 = multi-color 1
3 = multi-color 2
X ,Y starting coordinate, scaled (default at pixel cursor (PC))
mode 0 = paint an area defined by the color source selected

1 = paint an area defined by any nonbackground source
The PAINT command fills an area with color. It fills in the area around the specified
point until a boundary of the same specified color source is encountered. For example, if
you draw a circle in the foreground color source, start PAINTing the circle where the
coordinate assumes the background source. The Commodore 128 will only PAINT
where the specified source in the PAINT statement is different from the source of the x
and y pixel coordinate. It cannot PAINT points where the sources are the same in the
PAINT statement and the specified coordinate. The X and Y coordinate must lie
completely within the boundary of the shape you intend to PAINT, and the source of the
starting pixel coordinate and the specified color source must be different.
E X A M P LE :
10 CIRCLE 1, 160,100,65,50 Draws an outline of a circle.
20 PAINT 1, 160,100 Fills in the circle with color from source 1 (VIC
foreground), assuming point 160,100 is colored in
the background color (source 0).
10 BOX 1, 10, 10, 20, 20 Draws an outline of a box.
20 PAINT 1, 15, 15 Fills the box with color from source 1, assuming
point 15,15 is colored in the background source
( 0 ).
30 PAINT 1, + 10, + 10 PAINT the screen in the foreground color source

at the coordinate relative to the pixel cursor’s
previous position plus 10 in both the vertical and
horizontal positions.
100 PAINT 1, 100;90 Paint the screen area 90° relative to and 100
pixels away from the current pixel cursor co
ordinate.
PLAY
Defines and plays musical notes and elements within a string or string variable.
PLAY “ V n,O n,Tn,U n,X n,elem ents, notes”
where the string or string variable is composed of the following
Vn = Voice (n = 1-3)
On = Octave (n = 0-6)
Tn = Tune Envelope Defaults (n = 0-9)
0 = piano
1 = accordion
2 = calliope
3 = drum
4 = flute
5 = guitar
6 = harpsichord
7 = organ
8 = trumpet
9 = xylophone
Un = Volume (n = 0-8)
Xn = Filter on (n = 1), off(n = 0)
notes: A ,B ,C ,D ,E ,F,G
elements: # ...........................Sharp
$ .............................Flat
W ...........................Whole note
H ...........................Half note
Q ...........................Quarter note
I .............................Eighth note
S .............................Sixteenth note
................................ Dotted
R .............................Rest
M ...........................Wait for all voices currently playing to end
the current “ measure”
The PLAY statement gives you the power to select voice, octave and tune envelope
(including ten predefined musical instrument envelopes), the volume, the filter, and the
notes you want to PLAY. All these controls are enclosed in quotes. You may include
spaces in a PLAY string for readability.
All elements except R and M precede the musical notes in a PLAY string.
56 COMMODORE 128
EX A M P LE S:
PLAY “ VIO4T0U5X0CDEFGAB ’ Play the notes C ,D ,E ,F,G ,A and B

in voice 1, octave 4, tune envelope
0 (piano), at volume 5, with the
filter off.
PLAY “ V 305T6U 7X 1#B $A W .C H D Q EIF Play the notes B-sharp, A-flat, a
whole dotted-C note, a half D-note,
a quarter E-note and an eighth
F-note.
A$ = “ V305T6U 3A BCD E” : PLAY A$ PLAY the notes and elements within
A$.
PLAY “ V1CV2EV3G” Plays a chord in the default setting.
POKE
Change the contents of a RAM memory location
POKE address, value
EX A M P LE :
10 POKE 53280,1 Changes VIC border color
PRINT
Output to the text screen
PRINT [print list]

The word PRINT can be followed by any of the following:
Characters inside quotes (“ text” )

Variable names (A, B, A$, X$)
Functions (SIN(23), ABS(33))
Expressions (2 + 2),A + 3,A = B)
Punctuation marks (;,)
EX A M P LE S: RESULTS
10 PRINT “ HELLO” HELLO

20 A$ = “ THERE” :PRINT “ HELLO” ;A$ HELLO THERE
30 A = 4:B = 2:?A + B 6
40 J = 4LPR IN T J;:PRINT J - 1 41 40
50 PRINT A;B;:D = A + B:PRINTD;A-B 4 2 6 2
See also POS, SPC, TAB and CHAR functions.

PR IN T#
Output data to files
P R IN T # file number[, print list]
P R IN T # is followed by a number which refers to the data file previously OPENed.
E X A M P LE :
10 OPEN 4,4 Outputs the data “ HELLO THERE”

20 P R IN T # 4 ,“ HELLO THERE!” ,A$,B$ and the variables A$ and B$ to the
printer.
10 OPEN 2,8,2 Outputs the data variables A. B$, C
20 PR IN T #2,A ,B $,C ,D and D to the disk file number 2.
NOTE: The PR IN T# command is used by itself to close the channel to

the printer before closing the file, as follows:
10 OPEN 4,4
30 PR IN T #4, “ PRINT W ORDS”
40 PR IN T#4
50 CLOSE 4
PRINT USING
Output using format
PRINT [# file number,] USING“ format list” ; print list
This statement defines the format of string and numeric items for printing to the text
screen, printer or other device.
E X A M P LE :
5 X = 32: Y = 100.23: A$ = “ CAT”

10 PRINT USING “ $ # # . # # # ” ;13.25,X,Y
20 PRINT USING “ # # # > # ” ;“ CBM ” ,A$
When this is RUN, line 10 prints:

$13.25 $32.00 $***** p jve asterisks (*****) are printed instead of a Y
value because Y has five digits, and this condition
does not conform to format list (as explained below).
Line 20 prints this:
CBM CAT Leaves two spaces before printing “ CBM ” as de

fined in format list.
58 COMMODORE 128
The pound sign ( # ) reserves room for a single character in the output field. If the data
item contains more characters than there are # signs in the format field, the entire field
is filled with asterisks (*): no characters are printed.
E X A M P LE :
10 PRINT USING “ # # # # ” ;X
For these values of X, this format displays:
A = 12.34 12
A = 567.89 568
A = 123456 ****
For a STRING item, the string data is truncated at the bounds of the field. Only as many
characters are printed as there are pound signs (# ) in the format item. Truncation occurs
on the right.
EXA M PLES:
FIELD EXPRESSION RESULT COMMENT
# # .# -.1 -0 .1 Leading zero added.
# # .# 1 1 .0 Trailing zero added.
#### -1 0 0 .5 -1 0 1 Rounded to no decimal places.
#### -1 0 0 0 Overflow because four digits and a minus sign

cannot fit in field.
###. 10 10. Decimal point added.
#$## 1 $1 Floating dollar sign.
PUDEF
Redefine symbols in PRINT USING statement
PUDEF “ nnnn”
Where “ nnnn” is any combination of characters, up to four in all. PUDEF allows you to
redefine any of the following four symbols in the PRINT USING statement: blanks, commas,
decimal points and dollar signs. These four symbols can be changed into some other char
acter by placing the new character in the correct position in the PUDEF control string.
Position 1 is the filler character. The default is a blank. Place anew character here
for another character to appear in place of blanks.
Position 2 is the comma character. Default is a comma.
Position 3 is the decimal point. Default is a decimal point.
Position 4 is the dollar sign. Default is a dollar sign.
EXA M PLES:
10 PUDEF“ *” PRINT * in the place of blanks.
20 PUDEF“ < ” PRINT < in the place of commas.
READ
Read data from DATA statements and input it into a numeric or string variable)
READ variable list
This statement inputs information from DATA statements and stores it in variables,
where the data can be used by the RUNning program.
In a program, you can READ the data and then re-read it by issuing the
RESTORE statement. The RESTORE sets the sequential data pointer back to the
beginning, where the data can be read again. See the RESTORE and DATA statements.
EXA M P LE S:
10 READ A, B, C READ the first three numeric variables from
20 DATA 3, 4, 5 the closest data statement.
10 READ A$, B$, C$ READ the first three string variables from
20 DATA JOHN, PAUL, GEORGE the nearest data statement.
10 READ A, B$, C READ (and input into the C l28 memory) a
20 DATA 1200, NANCY, 345 numeric variable, a string variable and an
other numeric variable.
RECORD
Position relative file pointers
R EC O R D # logical file number, record number [,byte number]
This statement positions a relative file pointer to select any byte (character) of any
record in the relative file.
When the record number value is set higher than the last record number in the file,
the following occurs:
For a write (PR IN T#) operation, additional records are created to expand the file
to the desired record number.
For a read (IN PU T#) operation, a null record is returned and a “ RECORD NOT
PRESENT ERROR occurs” . See your disk drive manual for details about relative
files.
EX A M P LE S:
10 DOPEN# 2 , “ FILE ’’
20 R E C O R D # 2 ,10,1
30 PR IN T#2,A $
40 D CLO SE#2
60 COMMODORE 128
This example opens an existing relative file called “ FILE” as file number 2 in
line 10. Line 20 positions the relative file pointer at the first byte in record number 10.
Line 30 actually writes the data, A$, to file number 2.
REM
Comments or remarks about the operation of a program line
REM message
E X A M P LE :
10 NEXT X:REM THIS LINE INCREMENTS X.
RENAME
Change the name of a file on disk
RENAME “ old filenam e” TO “ new filename” [,Ddrive num ber][<O N I,>

Udevice number]
EXA M PLES:
RENAME “ TEST” TO “ FINALTEST” ,D0 Change the name of the file

“ TEST” to “ FINAL TEST” .
RENAME (A$) TO (B$),D0,U9 Change the filename specified in

A$ to the filename specified in B$
on drive 0, device number 9. Re
member, whenever a variable name
is used as a filename, it must be
enclosed in parentheses.
RENUMBER
Renumber lines of a BASIC program
R ENUM BER [new starting line n u m b er][,in crem en ts,old starting line
number]
EXA M PLES:
RENUMBER Renumbers the program starting at 10, and increments

each additional line by 10.
RENUMBER 20, 20, 1 Starting at line 1, renumbers the program. Line 1 be
comes line 20, and other lines are numbered in incre
ments of 20.
RENUM BER,, 65 Starting at line 65, renumbers in increments of 10. Line

65 becomes line 10. If you omit a parameter, you must
still enter a comma in its place.
RESTORE
Reset READ pointer so the DATA can be reREAD
RESTORE [line#]
If a line number follows the RESTORE statement, the READ pointer is set to the first
data item in the specified program line. Otherwise the pointer is reset to the beginning of
the first DATA statement in the BASIC program.
EXA M P LE S:
10 FOR I = 1 TO 3 This example READs the data in line 70 and stores it in

20 READ X numeric variable X. It adds the total of all the numeric
30 ALL = X + ALL data items. Once all the data has been READ, three
40 NEXT cycles through the loop, the READ pointer is RE-
50 RESTORE STOREd to the beginning of the program and it returns
60 GOTO 10 to line 10 and performs repetitively.
70 DATA 10,20,30
10 READ A ,B,C Line 50 of this example RESTORES the DATA pointer
20 DATA 100,500,750 to the beginning data item in line 40. When line 60 is
30 READ X ,Y ,Z executed, it will READ the DATA 36,24,38 from line
40 DATA 36,24,38 40, and store it in numeric variables S, P, and Q,
50 RESTORE 40 respectively.
60 READ S,P,Q
RESUME
Define where the program will continue (RESUME) after an error has been trapped
RESUME [line number | NEXT]
This statement is used to restart program execution after TRAPping an error. With no
parameters, RESUME attempts to re-execute the statement in which the error occurred.
RESUME NEXT resumes execution at the statement immediately following the one indi
cating the error. RESUME followed by a line number will GOTO the specific line and
resume execution from that line number. RESUME can only be used in program mode.
E X A M P LE :
10 TRAP 100
15 INPUT “ ENTER A NUMBER’';A
20 B = 100/A
40 PRINT“ THE RESULT = ” ;B
62 COMMODORE 128
50 INPUT “ DO YOU WANT TO RUN IT AGAIN (Y/N)” ;Z$:IF Z$ = “ Y ”

THEN 10
60 STOP
100 INPUT“ ENTER ANOTHER NUMBER (NOT ZERO)” ;A
110 RESUME 20
This example traps a “ DIVISION BY ZERO ERROR” in line 20 if 0 is entered in line

15. If zero is entered, the program goes to line 100, where you are asked to input another
number besides 0. Line 110 returns to line 20 to complete the calculation. Line 50 asks
if you want to repeat the program again. If you do, press the Y key.
RETURN
Return from subroutine
RETURN
E X A M P LE :
10 PRINT “ ENTER MAIN PROGRAM”

20 GOSUB 100
30 PRINT “ END OF PROGRAM”
90 STOP
100 PRINT “ SUBROUTINE 1”
110 RETURN
This example calls the subroutine at line 100 which prints the message “ SUBROU
TINE 1” and RETURNS to line 30, the rest of the program.
RUN
Execute BASIC program
1) RUN [line number]
2) RUN “ filenam e” [,Ddrive num ber][<O N |,>U device number]
EXA M PLES:
RUN Starts execution from the beginning of the program.

RUN 100 Starts program execution at line 100.
RUN“ PRG 1” DLOADs “ PRG1” from disk drive 8, and runs it from the
starting line number.
RUN(A$) DLOADs the program named in the variable A$.
SAVE
Store the program in memory to disk or tape
SAVE [“ filenam e” ][,device number][,EOT flag]
EXA M PLES:
SAVE Stores program on tape, without a name.
SAVE “ HELLO” Stores a program on tape, under the name HELLO.
SAVE A$,8 Stores on disk, with the name stored in variable A$.
SAVE “ HELLO” ,£ Stores on disk, with name HELLO (equivalent to

DSAVE “ HELLO” ).
SAVE “ H ELLO” , 1, 2 Stores on tape, with name HELLO, and places an END
OF TAPE marker after the program.
SCALE
Alter scaling in graphics mode
SCALE n [,Xmax,Ymax]
where:
n = 1 (on) or 0 (off)
Coordinates may be scaled from 0 to 32767 (default = 1023) in both X and Y (in either
standard or multicolor bit map mode), rather than the normal scale values, which are:
multi-color mode X = 0 to 159 Y = 0 to 199

bit map mode X = 0 to 319 Y = 0 to 199
EXA M P LE S:
10 GRAPHIC 1,1 Enter standard bit map, turn scaling

20 SCALE LCIRCLE 1,180,100,100,100 on to default size (1023, 1023) and
draw a circle.
10 GRAPHIC 1,3 Enter multi-color mode, turn scaling

20 SCALE 1,1000,5000 on to size (1000, 5000) and draw a
30 CIRCLE 1,180,100,100,100 circle.
The SCALE command affects the sprite coordinates in the MOVSPR command. If
you add scaling to a program that contains sprites, adjust the MOVSPR coordinates
accordingly.
64 COMMODORE 128
SCNCLR
Clear screen
SCNCLR mode number
The modes are as follows:
MODE NUMBER MODE
0 40 column (VIC) text

1 bit map
2 split screen bit map
3 multi-color bit map
4 split screen multi-color bit map
5 80 column (8563) text
This statement with no argument clears the graphic screen, if it is present, otherwise the
current text screen is cleared.
EXA M PLES:
SCNCLR 5 Clears 80 column text screen.

SCNCLR I Clears the (VIC) bit map screen.
SCNCLR 4 Clears the (VIC) multicolor bit map and 40-column text split screen.
SCRATCH
Delete file from the disk directory
SCRATCH “ filenam e” [,Ddrive num ber][<O N |,>U device number]
E X A M P LE :
SCRATCH “ MY BACK” , DO
This erases the file MY BACK from the disk in drive 0.
SLEEP
Delay program for a specific period of time
SLEEP N
where N is seconds 0 < N < = 65535.
SLOW
Return the Commodore 128 to 1MHz operation
SLOW
SO UND
Output sound effects and musical notes
SOUND v,f,d[,dir][,m ][,s][,w ][,p]
where: v = voice (1..3)

f = frequency value (0..65535)
d = duration (0..32767)
dir = step direction (0(up), l(down) or 2(oscillate)) default = 0
m = minimum frequency (if sweep is used) (0..65535) default = 0
s = step value for sweep (0..32767) default = 0
w = waveform (0 = triangle, 1 = sawtooth, 2 = variable, 3 = noise)
default = 2
p = pulse width (0..4095) default = 2048
EXA M PLES:
SOUND 1,40960,60 Play a SOUND at frequency 40960 in voice 1

for 1 second.
SOUND 2.20000,50,0,2000,100 Output a sound by sweeping through frequen

cies starting at 2000 and incrementing upward
in units of 100 up to 20,000. Each frequency is
played for 50 jiffies.
SOUND3,5000,90,2,3000,500,1 This example outputs a range of sounds start
ing at a minimum frequency of 3000, through
5000, in increments of 500. The direction of
the sweep is back and forth (oscillating). The
selected waveform is sawtooth and the voice
selected is 3.
SPRCOLOR
Set multi-color 1 and/or multi-color 2 colors for all sprites
SPRCOLOR [sm cr-l][,sm cr-2]
where:
smcr-1 Sets multi-color 1 for all sprites.
smcr-2 Sets multi-color 2 for all sprites.
66 COMMODORE 128
Either of these parameters may be any color from 1 through 16.
EXA M PLES:
SPRCOLOR 3,7 Sets sprite multi-color 1 to red and multi-color 2 to blue.
SPRCOLOR 1,2 Sets sprite multi-color 1 to black and multi-color 2 to white.
SPRDEF
Enter the SPRite DEFinition mode to create and edit sprite images.
SPRDEF
The SPRDEF command defines sprites interactively

Entering the SPRDEF command displays a sprite work area on the screen which
is 24 characters wide by 21 characters tall. Each character position in the grid corre
sponds to a sprite pixel in the sprite displayed to the right of the work area. Here
is a summary of the SPRite DEFinition mode operations and the keys that perform
them:
USER INPUT DESCRIPTION
1-8 Selects a sprite number at the SPRITE NUMBER? prompt only.

A Turns on and off automatic cursor movement.
CRSR keys Moves cursor in work/area.
RETURN KEY M oves cursor to start o f next line.
RETURN key Exits sprite designer mode at the SPRITE NUMBER? prompt
only.
HOME key Moves cursor to top left corner of sprite work area.
CLR key Erases entire grid.
1-4 Selects color source (enables/disables pixels).
CTRL key, 1-8 Selects sprite foreground color (1-8).
Commodore key, 1-8 Selects sprite foreground color (9-16).
STOP key Cancels changes and returns to prompt.
SHIFT RETURN Saves sprite and returns to SPRITE NUMBER? prompt.
X Expands sprite in X (horizontal) direction.
Y Expands sprite in Y (vertical) direction.
M M ulti-color sprite mode.
C Copies sprite data from one sprite to another.
SPRITE
Turn on and off, color, expand and set screen priorities for a sprite
SPRITE < n u m b er> [,on/off][,fgnd][,priority][,x-exp] [,y-exp][,mode]
The SPRITE statement controls most of the characteristics of a sprite.

PARAMETER DESCRIPTION
number Sprite number (1-8)

on/off Turn sprite on (1) or off (0)
foreground Sprite foreground color (1-16) (default = sprite number)
priority Priority is 0 if sprites appear in front o f objects on the screen. Priority
is 1 if sprites appear in back o f objects on the screen.
x-exp Horizontal EXPansion on (1) or off (0)
y-exp Vertical EXPansion on (1) or off (0)
mode Select standard sprite (0) or multi-color sprite (1)
Unspecified parameters in subsequent sprite statements take on the characteristics of the

previous SPRITE statement. You may check the characteristics of a SPRITE with the
RSPRITE function.
E XA M PLES:
SPRITE 1,1,3 Turn on sprite number 1 and color it red.
SPRITE 2,1,7,1,1,1 Turn on sprite number 2, color it blue, make it pass

behind objects on the screen and expand it in the vertical
and horizontal directions.
SPRITE 6,1,1,0,0,1,1 Turn on SPRITE number 6, color it black. The first 0

tells the computer to display the sprites in front of objects
on the screen. The second 0 and the 1 following tell the
C l 28 to expand the sprite vertically only. The last 1
specifies multi-color mode. Use the SPRCOLOR com
mand to select the sprite’s multi-colors.
SPRSAV
Copy sprite data from a text string variable into a sprite or vice versa, or copy data from
one sprite to another.
SPRSAV < o rig in > ,< d estin a tio n >
Either the origin or the destination can be a sprite number or a string variable but they
both cannot be string variables. They can both be sprite numbers. If you are storing a
string into a sprite, only the first 63 bytes of data are used. The rest are ignored since a
sprite can only hold 63 data bytes.
EX A M P LE S:
SPRSAV 1,A$ Transfers the image (data) from sprite 1 to the string named A$.
SPRSAV B$,2 Transfers the data from string variable B$ into sprite 2.
SPRSAV 2,3 Transfers the data from sprite 2 to sprite 3.

68 COMMODORE 128
S S H A P E / GSHAPE
Save/retrieve shapes to/from string variables
SSHAPE and GSHAPE are used to save and load rectangular areas of bit map
screens to/from BASIC string variables. The command to save an area of the bit map
screen into a string variable is:
SSHAPE string variable, X I, Y1 [,X2,Y2]
where:
string variable String name to save data in

X I,Y 1 Corner coordinate (0,0 through 319,199) (scaled)
X2,Y2 Corner coordinate opposite (X I,Y l) (default is the PC)
The command to retrieve (load) the data from a string variable and display it on
specified screen coordinates is:
GSHAPE string variable [X,Y][,mode]
where:
string Contains shape to be drawn

X ,Y Top left coordinate (0,0 through 319,199) telling where to draw the shape
(scaled— the default is the pixel cursor)
mode Replacement mode:
0 = place shape as is (default)
1 = invert shape
2 = OR shape with area
3 = AND shape with area
4 = XOR shape with area
The replacement mode allows you to change the data in the string variable so you can
invert it, perform a logical OR, exclusive OR (turn off bytes that are on) or AND
operation on the image.
EXA M PLES:
SSHAPE A $,10,10 Saves a rectangular area from the coordinates 10,10

to the location of the pixel cursor, into string vari
able A$.
SSHAPE B$,20,30,43,50 Saves a rectangular area from top left coordinate
(20,30) through bottom right coordinate (43,50) into
string variable B$.
SSHAPE D$, + 10, + 10 Saves a rectangular area 10 pixels to the right and
10 pixels down from the current position of the pixel
cursor.
GSHAPE A$, 120,20 Retrieves shape contained in string variable A$ and

displays it at top left coordinate (120,20).
GSHAPE B$,30,30,1 Retrieves shape contained in string variable B$ and
displays it at top left coordinate 30,30. The shape is
inverted due to the replacement mode being selected
by the 1.
GSHAPE C$, + 20, + 30 Retrieves shape from string variable C$ and displays
it 20 pixels to the right and 30 pixels down from the
current position of the pixel cursor.
NOTE: Beware using modes 1-4 with multi-color shapes. You may
obtain unpredictable results.
STASH
Move contents of host memory to expansion RAM
STASH #b ytes, intsa, expsa, expb

Refer to FETCH command for description of parameters.
STOP
Halt program execution
STOP
SWAP
Swap contents of host RAM with contents of expansion RAM
SWAP # b y tes, intsa, expsa, expb

Refer to FETCH command for description of parameters.
SYS
Call and execute a machine language subroutine at the specified address
SYS address [,a][,x][,y][,s]
This statement calls a subroutine at a given address in a memory configuration previously
set up according to the BANK command. Optionally, arguments a,x,y and s are loaded into
the accumulator, x, y and status registers, respectively, before the subroutine is called.
The address range is 0 to 65535. The 8502 microprocessor begins executing the
machine-language program starting at the specified memory location. Also see the
BANK command.
70 COMMODORE 128
EXA M PLES:
SYS 32768 Calls and executes the machine-language routine at location 32768
($8000).
SYS 6144,0 Calls and executes the machine-language routine at location 6144
($1800) and loads zero into the accumulator.
TEMPO
Define the speed of the song being played
TEMPO n
where n is a relative duration between (1 and 255)
The default value is 8, and note duration increases with n.
EXA M PLES:
TEMPO 16 Defines the Tempo at 16.
TEMPO 1 Defines the TEMPO at the slowest speed.
TEMPO 250 Defines the TEMPO at 250.
TRAP
Detect and correct program errors while a BASIC program is RUNning
TRAP [line number]

The RESUME statement can be used to resume program execution. TRAP with no line
number turns off error trapping. An error in a TRAP routine cannot be trapped. Also see
system variables ST, EL, DS and DS$.
EXA M P LE S:
.100 TRAP 1000 If an error occurs, GOTO line 1000.
. 1000?ERR$(ER);EL Print the error message, and the error number.

.1010 RESUME Resume with program execution.
TROFF
Turn off error tracing mode
TROFF
TRON
Turn on error tracing
TRON
TRON is used in program debugging. This statement begins trace mode. When you
RUN the program, the line numbers of the program appear in brackets before any action
for that line occurs.
VERIFY
Verify program in memory against one saved to disk or tape
VERIFY “ filename” f,device number][,relocate flag]
Issue the VERIFY command immediately after you SAVE a program.
EXA M P LE S:
VERIFY Checks the next program on the tape.
VERIFY “ H ELLO " Searches for HELLO on tape, checks it against memory.
VERIFY “ HELLO” ,8,1 Searches for HELLO on disk, then checks it against
memory.
VOL
Define output level of sound for SOUND and PLAY statements
VOL volume level
EX A M P LE S:
VOL 0 Sets volume to its lowest level.
VOL 15 Sets volume for SOUND and PLAY statements to its highest output.
WAST
Pause program execution until a data condition is satisfied
WAIT < lo ca tio n > , < m a sk -l> [,m ask-2>]
where:
location = 0-65535
masks = 0-255
The WAIT statement causes program execution to be suspended until a given memory
address recognizes a specified bit pattern or value.
The first example below WAITs until a key is pressed on the tape unit to
continue with the program. The second example will WAIT until a sprite collides with
the screen background.
72 COMMODORE 128
EXA M PLES:
WAIT 1, 32, 32
WAIT 53273, 2
WAIT 36868, 144, 16
W ID TH
Set the width of drawn lines
WIDTH n
EXA M PLES:
WIDTH 1 Set single width for graphic commands

WIDTH 2 Set double width for drawn lines
W IN D O W
Define a screen window
WINDOW top left col,top left row,bot right col.bot right row[,clear]
This command defines a logical window within the 40 or 80 column text screen. The
coordinates must be in the range 0-39/79 for 40- and 80-column values respectively and
0-24 for row values. The clear flag, if provided (1), causes a screen-clear to be
performed (but only within the limits of the newly described window).
EXA M P LE S:
WINDOW 5,5,35,20 Defines a window with top left corner coordinate as

5,5 and bottom right corner coordinate as 35,20.
WINDOW 10,2,33,24,1 Defines a window with upper left corner coordinate

10,2 and lower right corner coordinate 33,24. Also
clears the portion of the screen within the window as
specified by the 1.
BASIC FU NC TIO N S
The format of the function description is:
FUNCTION (argument)
where the argument can be a numeric value, variable or string.
Each function description is followed by an EXAMPLE. The first line appearing
below the word “ EXAM PLE” is the function you type. The second line without bold is
the computer’s response.
ABS
Return absolute value of argument X
A B S (X )
E X A M P LE :
PRINT ABS (7*(-5))
35
ASC
Return CBM ASCII code for the first character in X$
ASC(X$)
This function returns the CBM ASCII code of the first character of X$.
E X A M P LE :
X$ = “ C128” :PRINT ASC (X$)
67
A TN
Return the arctangent of X in radians
ATN (X)
The value returned is in the range - tt/2 through tt/2.
E X A M P LE :
PRINT ATN (3)

1.24904577
BUMP
Return sprite collision information
BUM P (N)
To determine which sprites have collided since the last check, use the BUMP function.
BU M P(l) records which sprites have collided with each other, and BUMP(2) records
which sprites have collided with other objects on the screen. COLLISION need not be
active to use BUMP. The bit positions (0-7) in the BUMP value correspond to sprites 1
through 8 respectively. BUMP(n) is reset to zero after each call.
74 COMMODORE 128
H ere’s how the sprite numbers and BUMP values that are returned correspond:
BUMP Value: 128 64 32 16 8 4 2 1
Sprite Number: 8 7 6 5 4 3 2 1
EXA M P LE S:
PRINT BUMP (1) 12 indicates that sprites 3 and 4 have collided.
PRINT BUMP (2) 32 indicates the sprite 6 has collided with an object on the screen.
CHR$
Return character for specified CBM ASCII code X
CHR$(X)
The argument (X) must be in the range 0-255. This is the opposite of ASC and returns the
string character whose CBM ASCII code is X. Refer to Appendix E for a table of CHR$ codes.
EXA M P LE S:
PRINT CHR$ (65) Prints the A character.

A
PRINT CHR$ (147) Clears the text screen.
COS
Return cosine for angle of X in radians
COS(X)
E X A M P LE :
PRINT COS (-tt/3)
.5
FNxx
Return value from user defined function xx
FNxx(X)
This function returns the value from the user defined function xx created in a DEF
FNxx statement
E X A M P LE :
10 DEF FNAA(X) = (X-32)*5/9
20 INPUT X
30 PRINT FNAA(X)
RUN
?40 (? is input prompt)
4.44444445
NOTE: If GRAPHIC is used in a program that defines a function, issue

the GRAPHIC command before defining the function or the function
definition is destroyed.
FRE
Return number of available bytes in memory
FRE (X)
where X is the RAM bank number. X = 0 for BASIC program storage and X = 1 to
check for available BASIC variable storage.
EX A M P LE S:
PRINT FRE (0) Returns the number of free bytes for BASIC programs.
58109
PRINT FRE (1) Returns the number of free bytes for BASIC variable storage.
64256
HEX$
Return hexadecimal string equivalent to decimal number X
HEX$(X)
E X A M P LE :
PRINT HEX$(53280)
D020
INSTR
Return starting position of string 2 within string 1
INSTR (string 1, string 2 [,starting position])
E X A M P LE :
PRINT INSTR (“ COMMODORE 128” ,“ 128” )

11
IN T
Return integer form (truncated) of a floating point value
INT(X)
This function returns the integer value of the expression. If the expression is positive,
the fractional part is left out. If the expression is negative, any fraction causes the next
lower integer to be returned.
76 COMMODORE 128
EX A M P LE S:
PRINT IN T (3.14)
3
PRINT IN T (-3 .14)

-4
JOY
Return position of joystick and the status of the fire button
JOY(N)
when N equals:
1 JOY returns position of joystick 1.

2 JOY returns postion of joystick 2.
Any value of 128 or more means that the fire button is also pressed. To find the joystick
position if the fire button is pressed subtract 128 from the JOY value. The direction is
indicated as follows.
1
8
7 0
6
5
EX A M P LE S:
PRINT JOY (2) Joystick 2 fires to the left.

135
IF (JOY (1) > 127) THEN PRINT “ FIRE” Determines whether the fire button
is pressed.
DIR = JO Y (l) AND 15 Returns direction (only) of joystick 1.
LEFT$
Return the leftmost characters of string
LEFT$ (string,integer)
E X A M P LE :
PRINT LEFT$(“ COMMODORE” ,5)

COMMO
LEN
Return the length of a string
LEN (string)
The returned integer value is in the range 0-255.
EXAMPLE:
PRINT LEN (“ COMMODORE 128” )
12
LOG
Return natural log of X
LOG(X)
The argument X must be greater than 0.
EXAMPLE:
PRINT LOG (37/5)
2.00148
MID$
Return a substring from a larger string
MID$ (string,starting position!,length])
This function extracts the number of characters specified by length (0-255), from string,
starting with the character specified by starting position (1-255).
EXAMPLE:
PRINT M ID$(“ COMMODORE 128” ,3,5)
MMODO
PEEK
Return contents of a specified memory location
PEEK(X)
The data will be returned from the bank selected by the most recent BANK command.
See the BANK command.
EXAMPLE:
10 BANK 15:VIC = DEC(“ D 000” )
20 FOR I = 1 TO 47
30 PRINT PEEK(VIC + I)
40 NEXT
This example displays the contents of the registers of the VIC chip (some of which
are ever-changing).
78 COMMODORE 128
PEN
Return X and Y coordinates of the light pen
PEN(n)
where n = 0 PEN returns the X coordinate of light pen position on any VIC screen,
n = 1 PEN returns the Y coordinate of light pen position on any VIC screen,
n = 2 PEN returns the character column position of the 80 column display,
n = 3 PEN returns the character row position of the 80 column display,
n = 4 PEN returns the (80-column) light pen trigger value.
The VIC PEN values are not sealed and are taken from the same coordinate plane as
sprites use. Unlike the 40 column (VIC) screen, the 80 column (8563) coordinates are
character row and column positions and not pixel coordinates like the VIC screen. Both
the 40 and 80 column screen coordinate values are approximate and vary, due to the
nature of light pens. The 80-column read values are not valid until PEN(4) is true.
Light pens are always plugged in to control port 1.
EXA M P LE S:
10 PRINT PEN(0);PEN(1) Displays the X and Y coordinates of the light
' pen (for the 40 column screen).
10 DO UNTIL PEN(4):LOOP Ensures that the read values are valid (for the
80 column screen).
20 X = PEN(2)
30 Y = PEN(3)
40 REM:REST OF PROGRAM
TT
Return the value of pi (3.14159265)
TT
E X A M P LE :
PRINT it This returns the result 3.14159265.
POINTER
Return the address of a variable
POINTER (variable name)
This function returns a zero if the variable is not defined.
E X A M P LE :
A = POINTER (Z) This example returns the address of variable Z.

NOTE: Address returned is in RAM BANK 1.
POS
Return the current cursor column position within the current screen window
POS(X)
The POS function indicates where the cursor is within the defined screen window. X is a
dummy argument, which must be specified, but the value is ignored. The values
returned range from 0-39 on the VIC screen and 0-79 on the 80-column screen.
EXAMPLE:
FOR I = 1 to 10 : ?SPC(I); POS(O): NEXT
This displays the current cursor position within the defined text window.
POT
Returns the value of the game-paddle potentiometer
POT (n)
when:
n = 1, POT returnstheposition of paddle # 1(control port 1)

n = 2, POT returns theposition of paddle # 2 (control port 1)
The values for POT range from 0 to 255. Any value of 256 or more means that the fire
button is also depressed.
EXAMPLE:
10 PRINT PO T(l)
20 IF PO T (l) > 256 THEN PRINT “ FIRE”
This example displays the value of game paddle 1.
RCLR
Return color of color source
RCLR(N)
This function returns the color (1 through 16) assigned to the color source N (0< = N = <
6), where the following N values apply:
80 COMMODORE 128
SO U R CE DESCRIPTION
0 40-column background
1 bit map foreground
2 multi-color 1
3 multi-color 2
4 40-column border
5 40- or 80-column character color
The counterpart to the RCLR function is the COLOR command.
EX A M P LE :
10 FOR I = 0 TO 6
20 PRINT “ SOURCE” ;I;“ IS COLOR CODE” ;RCLR(I)
30 NEXT
This example prints the color codes for all six color sources.
RDOT
Return current position or color source of pixel cursor
RDOT (N)
where:
N = 0 returns the X coordinate of the pixel cursor
N = 1 returns the Y coordinate of the pixel cursor
N = 2 returns the color source (0-3) of the pixel cursor
This function returns the location of the current position of the pixel cursor or the
current color source of the pixel cursor.
EXA M PLES:
PRINT RDOT(O) Returns X position of pixel cursor
PRINT RD O T(l) Returns Y position of pixel cursor
PRINT RDOT(2) Returns color source of pixel cursor
RGR
Return current graphic mode
RGR(X)
This function returns the current graphic mode. X is a dummy argument, which must be
specified. The counterpart of the RGR function is the GRAPHIC command. The value
returned by RGR(X) pertains to the following modes:
BASIC BUILDING BLOCKS AN D BASIC 7.0 ENCYCLOPEDIA 81
VALUE GRAPHIC MODE
0 40 column (VIC) text

1 Standard bit map
2 Split screen bit map
3 M ulti-color bit map
4 Split screen M ulti-color bit map
5 80 column (8563) text
E X A M P LE :
PRINT RGR(O) Displays the current graphic mode; in this case, standard bit
1 map mode.
PRINT RGR(O) Both multi-color bit map and 80-column text modes are enabled.
RIGHT$
Return sub-string from rightmost end of string
RIGHT$(string, numeric)
EX A M P LE :
PRINT RIGHT$(‘'BASEBALL” ,5)

EBALL
RND
Return a random number
RND (X)
If X = 0 RND returns a random number based on the hardware clock.

If X > 0 RND generates a reproducible random number based on the seed
value below.
If X < 0 produces a random number which is used as a base called a seed.
EXA M PLES:
PRINT RND(O) Displays a random number between 0 and 1.

.507824123
PRINT INT(RND(1)*100 + 1) Displays a random number between 1 and 100.
89
82 COMMODORE 128
RSPCOLOR
Return sprite multicolor values
RSPCOLOR (X)
When:
X = 1 RSPCOLOR returns the sprite multi-color 1.
X = 2 RSPCOLOR returns the sprite multi-color 2.
The returned color value is a value between 1 and 16. The counterpart of the RSPCOLOR
function is the SPRCOLOR statement. Also see the SPRCOLOR statement.
E X A M P LE :
10 SPRITE 1,1,2,0,1,1,1
20 SPRCOLOR 5,7
30 PRINT “ SPRITE MULTI-COLOR 1 1S” ;R SPC 0L0R (1)
40 PRINT “ SPRITE MULTI-COLOR 2 IS” ;RSPCOLOR(2)
RUN
SPRITE MULTI-COLOR 1 IS 5
SPRITE MULTI-COLOR 2 IS 7
In this example line 10 turns on sprite 1, colors it white, expands it in both the X and Y
directions and displays it in multi-color mode. Line 20 selects sprite multi-colors 1 and 2
(5 and 7 respectively). Lines 30 and 40 print the RSPCOLOR values for multi-color 1 and 2.
RSPPOS
Return the speed and position values of a sprite
RSPPOS (sprite number,position|speed)
where sprite number identifies which sprite is being checked, and position and speed
specifies X and Y coordinates or the sprite’s speed.
When position equals:
0 RSPPOS returns the current X position of the specified sprite.
1 RSPPOS returns the current Y position of the specified sprite.
When speed equals:
2 RSPPOS returns the speed (0-15) of the specified sprite.
E X A M P LE :
10 SPRITE 1,1,2
20 MOVSPR 1,45# 13
30 PRINT RSPPOS (1,0);RSPPOS (l,l);R SPPO S (1,2)
This example returns the current X and Y sprite coordinates and the speed (13).
RSPRITE
Return sprite characteristics
RSPRITE (sprite number,characteristic)
RSPRITE returns sprite characteristics that were specified in the SPRITE command.
Sprite number specifies the sprite (1-8) you are checking and the characteristic specifies
the sprite’s display qualities as follows:
RSPRITE RETURNS
CH ARA CTERISTIC THESE VALUES:
0 Enabled(l) / disabled(O)
1 Sprite color (1-16)
2 Sprites are displayed in front of (0) or behind
(1) objects on the screen
3 Expand in X direction yes = 1, no = 0
4 Expand in Y direction yes = 1, no = 0
5 M ulti-color yes = 1, no = 0
E X A M P LE :
10 FOR 1 = 0 TO 5 This example prints all 6 characteristics of sprite 1.

20 PRINT RSPRITE (1,1)
30 NEXT
RW INDO W
Returns the size of the current window or the number of columns of the current
screen
RWINDOW (n)
When n equals:
0 RWINDOW returns the number of lines in the current window.
1 RWINDOW returns the number of rows in the current window.
2 RWINDOW returns either of the values 40 or 80, depending on the current
screen output format you are using.
The counterpart of the RWINDOW function is the WINDOW command.
E X A M P LE :
10 WINDOW 1,1,10,10
20 PRINT RWINDOW(0);RWINDOW( 1);RWINDOW(2)
RUN
9 9 40
This example returns the lines (10) and columns (10) in the current window. This
example assumes you are displaying the window in 40 column format.
84 COMMODORE 128
SGN
Return sign of argument X
SGN(X)
E X A M P LE :
PRINT SGN(4.5);SGN(0);SGN(-2.3)
1 0-1
SIN
Return sine of argument
SIN(X)
E X A M P LE :
PRINT SIN ( it/ 3 )

.866025404
SPC
Skip spaces on printed output
SPC (X)
E X A M P LE :
PRINT “ COMMODORE” ;SPC(3);“ 128”

COMMODORE 128
SQR
Return square root of argument
SQR (X)
E X A M P LE :
PRINT SQR(25)
5
STR$
Return string representation of number
S T R $ (X )
E X A M P LE :
PRINT STR$(123.45)
123.45
PRINT STR$(-89.03)
-89.03
PRINT STR$(1E20)
IE + 20
TAB
Moves cursor to tab position in present statement
TAB (X)
E X A M P LE :
10 PRINT“ COM M ODORE” TAB(25)“ 128”

COMMODORE 128
TA N
Return tangent of argument in radians
TAN(X)
This function returns the tangent of X, where X is an angle in radians
E X A M P LE :
PRINT TA N (.785398163)
1
USR
Call user-defined subprogram
USR(X)
When this function is used, the BASIC program jumps to a machine language program
whose starting point is contained in memory locations 4633($1219) and 4634($121A), (or
785($0311) and 786($0312) in C64 mode). The parameter X is passed to the machine-
language program in the floating-point accumulator ($63-$68 in C l 28 mode). A value is
returned to the BASIC program through the calling variable. You must direct the value
into a variable in your program in order to receive the value back from the floating-point
accumulator. An ILLEGAL QUANTITY ERROR results if you don’t specify this
variable. This allows the user to exchange a variable between machine code and
BASIC.
86 COMMODORE 128
E X A M P LE :
10 POKE 4633,0
20 POKE 4634,48
30 A = USR(X)
40 PRINT A
Place starting location ($3000 = 12288:$00 = 0:$30) = 48 of machine language
routine in location 4633 and 4634. Line 30 stores the returning value from the floating
point accumulator. The USER vector is assumed to be in BANK 15. Your machine
language routine M UST be in RAM bank 0 below address $4000.
VAL
Return the numeric value of a number string
VAL(X$)
E X A M P LE :
10 A$ = “ 120”
20 B$ = “ 365”
30 PRINT VAL (A$ + B$)
RUN
485
XO R
Return exclusive OR value
XOR (n l,n 2)
This function returns the exclusive OR of the numeric argument values n l and n2.
X = XOR (nl,n2)
where n l, n2, are 2 unsigned values (0-65535)
E X A M P LE :
PRINT XOR( 128,64)
192
RESERVED SYSTEM WORDS

(KEYWORDS)
This section lists the words used to make up the BASIC 7.0 language. These words
cannot be used within a program as other than a component of the BASIC language. The
only exception is that they may be used within quotes (in a PRINT statement, for example).
ABS DELETE HELP POT SPRCOLOR

AND DIM HEX$ PRINT SPRDEF
APPEND DIRECTORY IF PRINT# SPRITE
ASC DLOAD INPUT PUDEF SPRSAV
ATN DO INPUT# (QUIT) SQR
AUTO DOPEN INSTR RCLR SSHAPE
BACKUP DRAW INT RDOT ST
BANK DS JOY READ STASH
BEGIN DS$ KEY RECORD STEP
BEND DSAVE LEFTS REM STOP
BLOAD DVERIFY LEN RENAME STR$
BOOT EL LET RENUMBER SWAP
BOX ELSE LIST RESTORE SYS
BSAVE END LOAD RESUME TAB
BUMP ENVELOPE LOCATE RETURN TAN
CATALOG ER LOG RGR TEMPO
CHAR ERR$ LOOP RIGHTS THEN
CHR$ EXIT MID$ RND TI
CIRCLE EXP MONITOR RREG TI$
CLOSE FAST MOVSPR RSPCOLOR TO
CLR FETCH NEW RSPPOS TRAP
CMD FILTER NEXT RSPRITE TROFF
COLLECT FN NOT RUN TRON
COLLISION FOR (OFF) RWINDOW UNTIL
COLOR FRE ON SAVE USING
CONCAT GET OPEN SCALE USR
CONT GET# OR SCNCLR VAL
COPY G064 PAINT SCRATCH VERIFY
COS GOSUB PEEK SGN VOL
DATA GOTO PEN SIN WAIT
DCLEAR GO TO PLAY SLEEP WHILE
DCLOSE GRAPHIC POINTER SLOW WIDTH
DEC GSHAPE POKE SOUND WINDOW
DEF FN HEADER POS SPC XOR
N O T E : Keywords shown in parentheses are not implemented in C128

BASIC 7.0.
Reserved variable names are names reserved for the variables DS, DS$, ER, EL,
ST, TI and TI$, and the function ERR$. Keywords such as TO and IF or any other
names that contain keywords, such as RUN, NEW or LOAD cannot be used.
ST is a status variable for input and output (except normal screen/keyboard
operations). The value of ST depends on the results of the last I/O operation. In general,
if the value of ST is 0, then the operation was successful.
88 COMMODORE 128
TI and TI$ are variables that relate to the real time clock built into the Commodore
128. The system clock is updated every 1/60th of a second. It starts at 0 when the
Commodore 128 is turned on, and is reset only by changing the value of TI$. The
variable TI gives the current value of the clock in l/60th of a second. TI$ is a string that
reads the value of the real time clock as a 24-hour clock. The first two characters of TI$
contain the hour, the third and fourth characters are minutes and the fifth and sixth
characters are seconds. This variable can be set to any value (so long as all characters
are numbers) and will be updated automatically as a 24-hour clock.
E X A M P LE :
TI$ = “ 101530” Sets the clock to 10:15 and 30 seconds (AM).
The value of the clock is lost when the Commodore 128 is turned off. It starts at
zero when the Commodore 128 is turned on, and is reset to zero when the value of the
clock exceeds 235959 (23 hours, 59 minutes and 59 seconds).
The variable DS reads the disk drive command channel and returns the current
status of the drive. To get this information in words, PRINT DS$. These status variables
are used after a disk operation, like DLOAD or DSAVE, to find out why the error light
on the disk drive is blinking.
ER, EL and the ERR$ function are variables used in error trapping routines. They
are usually only useful within a program. ER returns the last error number encountered
since the program was RUN. EL is the line where the error occurred. ERR$ is a
function that allows the program to print one of the BASIC error messages. PRINT
ERR$(ER) prints out the proper error message.
RESERVED SYSTEM SYMBOLS

The following characters are reserved system symbols.
SYMBOL USE(S)
+ Plus sign Arithmetic addition; string concatenation; relative pixel

cursor/sprite movement; declare decimal number in ma
chine language monitor
— Minus sign Arithmetic subtraction; negative number; unary minus;
relative pixel cursor/ sprite movement
* Asterisk Arithmetic multiplication
/ Slash Arithmetic division
t Up arrow Arithmetic exponentiation
Blank space Separate keywords and variable names
= Equal sign Value assignment; relationship testing
< Less than Relationship testing
> Greater than Relationship testing
) Comma Format output in variable lists; command/statement func
tion parameters
SYM BOL USE(S)
. Period Decimal point in floating-point constants

> Semicolon Format output in variable lists; delimiter
Colon Separate multiple BASIC statements on a program line
66 91
Quotation mark Enclose string constants
9 Question mark Abbreviation for the keyword PRINT
( Left parenthesis Expression evaluation and functions
) Right parenthesis Expression evaluation and functions
% Percent Declare a variable name as integer; declare binary num
ber in machine language monitor
# Number Precede the logical fde number in input/output statements
$ Dollar sign Declare a variable name as a string and declare hexadeci
mal number in machine language monitor
& And sign Declare octal number in machine language monitor
TT Pi Declare the numeric constant 3.141592654
1
ONE STEP
BEYOND
SIMPLE B A S I C
91
92 COMMODORE 128
■
This chapter takes you one step beyond simple BASIC and presents a collection of useful
routines. You can incorporate these routines into your own programs as needed. In most
cases the routines will require only line number changes to be fitted into your programs.
CREATING A MENU
A menu is a list of choices you select to perform a specific operation within an
application program. A menu directs the computer to a particular part of a program.
Here is a general example of a menu program:
5 R E M ME NU S K E L E T O N
10 SCNCLR 0
20 P R I N T " 1 . F I R S T ITEM"
3 0 P R I N T " 2 . SECOND IT EM "
40 P R I N T " 3 . THIRD ITEM"
50 P R I N T " 4 . FOURTH IT EM "
1 0 0 P R I N T : P R I N T " S E L E C T AN I T E M F R O M A B O V E "
1 1 0 G E T K E Y A5
1 2 0 A=VAL ( A S ) : I F A > 4 THEN 10
1 3 0 ON A G O S U B 1 0 0 0 , 2 0 0 0 , 3 0 0 0 , 4 0 0 0
1 4 0 G O T O 1 0 : REM R E T U R N T O MENU
999 STOP
1 0 0 0 REM S T A R T F I R S T R O U T I N E F O R I T E M ONE H E R E
1 9 9 9 RETURN
2 0 0 0 REM S T A R T S E C O N D R O U T I N E HE R E
2 9 9 9 RETURN
3 0 0 0 REM S T A R T T H I R D R O U T I N E H E R E
3 99 9 RETURN
4 0 0 0 REM S T A R T F O U R T H R O U T I N E HE R E
4 9 9 9 RETURN
Pro g ram 3—1. Menu Skeleton
The SCNCLR 0 command in line 10 clears the 40-column screen. (Use SCNCLR
5 if you are using the 80-column screen. The easiest selection is by a number. You may
use as many selections as can fit on the screen. Line 100 displays a message to the user.
The GETKEY command in line 110 forces the computer to wait for a key to be pressed.
Since a key represents a character symbol, A$ is a string variable. So that it can be
interpreted as a numeric value in an ON GOTO statement, the string variable is
converted to a number with the VAL function in line 120. The IF . . . THEN statement
in line 120 screens user errors by preventing the user from selecting a number that is not
in the range of numbers used for choices (4). Line 130 directs control to the appropriate
section (i.e., line number) in your program. Since four selections are offered in this
example, you must include at least four line numbers. Line 1999 returns to the menu at
the end of each subroutine that you add at lines 1000, 2000, 3000 and 4000 in the menu
skeleton.
ONE STEP BEYOND SIMPLE BASIC 93
BUFFER ROUTINE
The C l 28 keyboard buffer can hold and dispense up to ten characters from the
keyboard. This is useful in a word processing program where it is possible at certain
moments to type faster than the software can actually process. The characters that
haven’t been displayed yet are temporarily stored in the keyboard buffer. The computer
can hold the next instruction in the buffer for use when the program is ready. This buffer
allows a maximum of ten characters in queue. To see the buffer in action, enter the
command SLEEP 5 and immediately press ten different letter keys. After five seconds,
all ten characters are displayed on the screen.
Here is a buffer routine that allows you to put items in the keyboard buffer
from within a program so they are dispensed automatically as the computer is able to act
upon them.
In line 10, memory location 208 (198 in C64 mode) is filled with a number
between 0 and 10— the number of keyboard characters in the keyboard buffer. In line
20, memory locations 842 through 851 (631-640 in C64 mode) are filled with any ten
characters you want placed there. In this example, seven characters are in the buffer,
each a carriage RETURN character. CHR$(13) is the character string code for the
carriage return character.
Line 40 places the text “ ?CHR$(156)” on the screen, but does not execute the
instruction. Line 50 displays the word “ LIST” on the screen. Neither command is
executed until the program ends. In the C128, the keyboard buffer automatically empties
when a program ends. In this case, the characters in the buffer (carriage return) are
emptied and act as though you are pressing the RETURN key manually. When this occurs
on a line where the commands from lines 40 and 50 are displayed, they are executed
as though you typed them in direct mode and pressed the RETURN key yourself. When
this program ends, the character color is changed to purple and the program is LISTED
to the screen. This technique is handy in restarting programs (with RUN or GOTO).
The next section gives a practical example of using the buffer routine.
10 P O K E 2 0 8 , 7 : RE M S P E C I F Y # O F C H A R S I N B U F F E R
20 FO R 1 = 8 4 2 TO 8 4 9 : POKE1 , 1 3 : N E X T : REM P L A C E CHARS I N BUFFER
30 S L E E P 2 : REM DE LAY
40 SCNCLR : P R I N T : P R I N T : P R I N T : P R I N T : P R I N T : P R I N T : P R I N T " ? CHR$ ( 1 5 6 ) ' '
50 P R I N T : P R I N T : P R I N T : P R I N T " L I S T " : RE M P L A C E L I S T ON S C R E E N
60 P R I N T C H R $ ( 1 9 ) : P R I N T : P R I N T : RE M GO HOME AND C U R S O R DOWN T W I C E
70 R E M WHEN P R O G R A M E N D S , B U F F E R E M P T I E S AN D E X E C U T E S 7 R E T U R N S .
80 R E M T H I S C H A N G E S C H A R C O L O R T O P U R P L E AN D L I S T S T H E P R O G R A M AU TO MA TI CAL LY
90 R E M A S I F Y O U P R E S S E D T H E R E T U R N K E Y MA N U A L L Y
P ro g ram 3 -2 . B uffer Return
LOADING ROUTINE
The buffer can be used in automatic loader routines. Many programs often involve the
loading of several machine code routines as well as a BASIC program. The results of
the following loader are similar to many found on commercial software packages.
94 COMMODORE 128
2 COLOR 4 , 1 : COLOR 0 , 1 : COLOR 5 , 1

5 A$="PICTURE"
10 SCNCLR: P R I N T : P R I N T : P R I N T : P R IN T " L O A D " C H R $ ( 3 4 ) A $ C H R $ ( 3 4 ) " , 8 , 1 "
15 P R I N T : P R I N T : PRINT"NEW"
25 B $ = " F I L E 3 . B I N "
30 P R IN T : P R IN T : PRINT"LOAD"CHR$( 3 4 ) B$CHR$( 3 4 ) " , 8 , 1 "
4 5 PR IN T: PRINT: PR IN T:PRIN T:PRINT:PR IN T"SY S12*256"
90 PRINT CHR$( 5 ) : PRINT" G R E E T I N G S F R O M C O M MO D O R E "
100 PRINT" P L E A S E S T A N D BY - L O A D I N G " : P R I N T C H R $ ( 1 4 4 )
200 PRINT CHR$(1 9 )
3 0 0 PO K E 20 8 , 7 : F O R I = 8 4 2 T 0 8 5 1 : P O K E I , 1 3 :NEXT
P ro g ram 3 -3 . Loading Routine
Line 2 colors the border, screen and characters black. Line 5 assigns A$ the
filename “ PICTURE” , which in this example assumes that it is an 8K binary file of a
bit-mapped screen. Line 10 places the LOAD instruction for the picture file on the
screen, but does not execute it. A carriage return from the keyboard buffer executes the
load instruction once the program ends and the keyboard buffer empties. Line 15 prints
the word “ N EW ” on the screen. Again, this operation is not carried out until a carriage
return is executed on the same line once the keyboard buffer empties. After loading a
machine language program, a NEW is performed to set pointers to their original
positions and clear the variable storage area. Line 30 displays the second load instruc
tion for the machine language program “ FILE3.BIN ” . This hypothetical program
enables the bit mapped PICTURE, and anything else you want to supply in the program.
Line 45 initiates (SYS12*256), the “ FILE3.BIN” program starting at 3072 ($OCOO)
once the keyboard buffer empties. This is only a template sample for you to follow.
“ PICTURE” and “ FILE3.BIN” are programs you supply and are only used to illustrate
one technique of automatic loading. Since the previous character color was black, all the
loading instructions are displayed in black on a black background, so they can’t be seen.
The CHR$(5) in line 90 changes the character color to white, so the only visible
messages are the ones in white in lines 90 and 100, while the disk drive is loading
“ PICTURE” and “ F1LE3.BIN” . Line 300 is the buffer routine.
If you were to do each step manually it would require seven “ RETURNS” . This
program places seven carriage return characters in the keyboard buffer, and they are
dispensed automatically when the program ends. As each RETURN is accepted, the
corresponding screen instruction is enacted automatically as if you had pressed the
RETURN key manually.
PROGRAMMING THE
C l 28 FU N C TIO N KEYS
As each of the function keys (FI through F8) is pressed, the computer displays a BASIC
command on the screen and in some cases acts immediately. These are known as the
default values of the function keys. Enter a KEY command to get a list of function key
values at any time.
C H A N G IN G FU N C T IO N KEYS
You can change the value assigned to any function key by entering the KEY command
followed by the number (1 through 8) of the key you want changed, a comma, and the
new key instruction in a string format. For example:
KEY1, “ DLOAD” + C H R $(34)+ “ PROGRAM NAME”

+ CHR$(34) + CHR$( 13) + “ LIST” + CHR$( 13)
This tells the computer to automatically load the BASIC program called “ program
name” and list it immediately (whenever FI is pressed). The character string code value
for the quote character is 34. It is necessary for LOAD and SAVE operations. The
character string code value for RETURN is 13 and provides immediate execution.
Otherwise, the instruction is only displayed on the screen and requires you to supply the
additional response and press the RETURN key.
The following example uses the ASCII value for the ESCape key to assign the F3
key to cause a downward scroll:
KEY 3,CHR$(27) + “ W ”
N O TE: All eight KEY definitions in total must not exceed 246 characters.
USING C64 FU N C T IO N KEY VALUES

IN C l 28 MODE
Programs previously written for the C64 which incorporate the function keys may still
be used in C l 28 mode by first assigning the C64 ASCII values to them with this
instruction:
10J = 132:FORA = 1T02:F0R K = AT08STEP2:J = J + 1:KEYK,CHR$(J):NEXT:

NEXT
H O W TO CRUNCH B A S I C PROGRAMS
Several techniques known collectively as memory crunching allow you to get the most
out of your computer’s memory. These techniques include eliminating spaces, using
multiple instructions, having syntax relief, removing remark statements, using variables,
and in general using BASIC intelligently.
E LIM IN A TIN G SPACES

In most BASIC commands, spacing is unnecessary, except inside quotes when you want
the spaces to appear on the screen. Although spaces improve readability, the extra space
consumes additional memory. Here is an instructional line presented both ways:
10INPUT“ FIRST NAME” ;N$:FOR T = A TO M:PR1NT “ O K ” :

10INPUT1‘FIRST NAME’’;N$:FORT = ATOM:PRINT‘‘OK ” :
96 COMMODORE 128
USING MULTIPLE INSTRUCTIO NS

Colons allow you to place several instructions within a single program line. Each
program line consumes additional memory. Be careful, however, crunching IF state
ments. Any instruction after the IF statement with the same line number can be bypassed
along with the IF . . . THEN condition. The following line is the equivalent of five
lines:
(A)
10 PRINTX:INPUTY:PRINTY:SCNCLR0:?J
(B)
10 PRINTX
20 INPUTY
30 PRINTY
40 SCNCLRO
50 PRINTJ
Example A requires less space in memory and on disk. Example B requires 16

additional bytes; 2 bytes for each additional line number and 2 bytes for the link to the
next line number.
S Y N T A X RELIEF
Some BASIC syntax is very flexible and this flexibility can be used to your advantage.
The LET statement, for example, can be written without LET. LET Y = 2 is the same as
Y = 2. Although it is good practice to initialize all variables to zero, it is not necessary
since the computer automatically sets all variables to zero, including subscripted vari
ables. DIMension all arrays (subscripted variables) to have twelve or more elements. The
C l 28 automatically dimensions each variable to have eleven subscripted elements if no
dimension is specified following DIM and the variable names. Often semicolons are not
required in PRINT statements. Both of these perform the same results:
10 PRINT” A ” ;Z$;“ W ORD” ;CHR$(65);“ NOW $ ”

20 P R IN T '‘A ” Z $‘‘W ORD” CHR$(65)“ NOW $ ”
REMOVING R EM STATEMENTS
Although REM statements are useful to the programmer, removing them makes a
considerable amount of memory available again. It might be a good idea to create a
separate listing with REM statements.
USING VARIABLES
Replace repeated numbers with a variable. This is especially important with large
numbers such as memory addresses. POKEing several numbers in sequence conserves
memory if a variable is used, such as POKE 54273+ V, etc. Of course, single-letter
variable names require the least memory. Reuse old variables such as those used in FOR
. . . NEXT loops. Whenever possible, make use of integer variables since they consume
far less memory than floating-point variables.
USING B A S I C INTELLIGENTLY
If information is used repeatedly, store the data in integer arrays, if possible. Use DATA
statements where feasible. Where a similar line is used repeatedly, create a single line
with variables and access it with GOSUBs. Use TAB and SPC functions in place of
extensive cursor controls.
MASKING BITS
Any of the bits within a byte can be controlled individually, using the Boolean operators
AND and OR. Calculations with AND and OR are based on a truth table (Table 3-1)
showing the results given all possible true and false combinations of the arguments X and Y.
X Y X AND Y X OR Y
0 0 0 0
0 1 0 1
1 0 0 1
I 1 1 1
T a b le 3 -1 . A N D and O R T ru th T ab le
With “ 0 ” representing False and “ 1” Truth, Table 3-1 shows how the operators
AND and OR work. The result of an AND operation between two bits is only true if
both bits are true (1). Otherwise the combination is false. Any bit combination with a
zero yields a zero in an AND operation. The result of an AND operation is only true
(equal to 1) if both bits are true (equal to 1).
The result of an OR operation is only false if each bit is false. Otherwise the result
is true. Any bit combination with a one yields a one in an OR operation. ONLY two
zeros result in a zero.
Observe the following example with the numbers 5 and 6 in binary form. When
you type the command PRINT 5 AND 6, the result is 4. Here’s why:
5= 0000 0101
6= 0000 0110
ANDed 4= 0000 0100
Instead of adding, ANDing performs a bit-by-bit comparison in accordance with

the rules of the AND truth table. Compare column-for-column from the right: 1 AND
0 = 0, 0 AND 1 = 0 , 1 AND 1 = 1, 0 AND 0 = 0. The result “ 0100” converted to
decimal is the number 4.
What is the effect of ORing 5 and 6? Again comparing bit-by-bit, using the rules
from the OR truth table:
98 COMMODORE 128
5= 0000 0101
6 = 0000 0110
ORing 7= 0000 0111
The result 0111 is decimal 7. Notice from the right that 1 OR 0 = 1 , 0 OR 1 = 1, 1 OR

1 = 1 and 0 OR 0 = 0.
Understanding how these OR and AND combinations work gives you the power to
control individual bits within your computer’s memory. Many of the 8-bit bytes utilize
each bit for separate control parameters.
USING OR A N D A N D T O M O DIFY
TH E BIT VALUES IN A BYTE
A byte is a group of eight binary digits labeled, from right to left, 0 to 7. Each binary
digit position represents a decimal value equal to two raised to the power of the position
number. For example, the decimal value of position 6 is 2**6 or 64. From left to right
the positions are:
7 6 5 4 3 2 1 0
and the corresponding values in decimal are:
128 64 32 16 8 4 2 1
To turn on a bit, place a “ 1” in its position. To turn it off, enter a “ 0 ” . Hence the
binary 10010000 has bits 4 and 7 on. Their values are 128 and 16. So if a particular byte
is POKED with 144 (1 2 8 + 1 6 ), these two bits are turned on. To turn bits on, store
(POKE) a new value to that byte— a value equal to the sum of all the decimal
equivalents of all the bits that are enabled (on). O f course, you do not always know
which bits are already on. You may only want to turn on specific bits without
affecting the others. That’s the purpose of the logical operations AND and OR.
First, obtain the decimal value of the byte by PEEKing. Then add the decimal
value of the bit you wish to turn on. The following command turns on bit 2 of memory
address “ V ” :
POKEV, PEEK(V) + 4
This assumes bit 2 (third bit from the right) had a value of 0. Had it already been
“ o n ,” it would have no effect. To prevent such confusion, the C l 28 uses the power of
Boolean Logic.
Ideally you want to read (PEEK) each bit. The proper approach is to OR the byte
with an operand byte which will yield the desired binary value. Suppose we want to
turn on bit 5; the operand byte becomes 00100000. By ORing this with any byte it will
affect only bit 5, because any combination involving 1 in an OR operation results in 1.
Thus no bit already ON can be inadvertently turned off.
POKEV,PEEK(V) OR 32
Just as OR turns a switch on, AND can turn a switch off—with a slight difference.
AND results in a “ 1” only if both bits compared are “ 1.” The trick is to compare the
byte in question with an operand byte of all ON bits except the bit you want turned off.
Bits to remain on will not be affected. To turn off bit 5, AND the byte in question with
the mirror image of 00100000 or the operand byte 11011111. In decimal this value is
always 255 minus the value of the bit(s) you want to turn off. Thus:
POKEV,PEEK(V) AND (255-32)
turns off bit 5.

Use OR to turn bits ON
Use AND to turn bits OFF
EXA M PLES:
POKEW,PEEK(W) OR 129 Turns ON bits 0 and 7 of memory address W.
POKES,PEEK(S) AND 126 Turns OFF bits 0 and 7 of memory register S

(Remember 255-129 = 126)
POKEC,PEEK(C)AND254 Turns OFF bit 0
POKEC,PEEK( V)OR63 Turns ON all bits except 6 and 7
DEBUGGING PROGRAMS
No program when first written is free of “ bugs” or errors. The process of finding errors
and removing them, debugging, combines editing with problem solving.
S Y N T A X ERRORS
Syntax errors result from misspelling or misusing the guidelines and formats of BASIC
commands. An error message is displayed on the screen defining the line where the
error occurs. Typing HELP < R E T U R N > or pressing the HELP key also highlights the
line with the error. Common syntax errors include misspelled BASIC terms, misplaced
punctuation, unpaired parentheses, reserved variable names such as TI$, use of line
numbers that do not exist, etc.
LOGIC ERRORS
Sometimes errors exist in the program logic, with the result that the program doesn’t do
exactly what you think it is supposed to do. Some logic errors are caused by the order of
instructions. One common fault occurs when you forget that anything on a line after an
IF statement is affected by the IF condition.
Some errors in logic require a trial-and-error investigation. This is best initiated by
asking the computer for help in BASIC.
USING A DELAY
Where the computer responds rapidly, it often helps to see a response by inserting a
SLEEP command for a temporary time delay. This gives you a chance to see exactly
what is happening in the program.
100 COMMODORE 128
USING PRINT A N D STOP

Insert STOP statements within your program prior to the suspect instruction line. Good
locations are at the end of specific tasks. Run the program. After the STOP statement
puts you into direct mode, use the PRINT command to identify clues to the problem by
determining the values of the various variables, especially those within loops. Check
these with what you expect. Continue the program with CONT to the next STOP
statement until you modify your program.
TRAPPING A N ERROR
Debugging is the art of detecting the source of problem. The following program is
perfectly valid; however, it produces an error when B equals zero.
10 INPUT A.B
20 PRINT A/B
30 GOTO 10
Although in this case the computer defines the error as a DIVISION BY ZERO
error, it is not always obvious how the variable B became a zero. It could have been
derived from a complex formula embedded in your program, or directly inputting the
value zero into a variable.
The BASIC TRAP command has a technique of trapping such an error in a
program without crashing. Since you can’t always foresee all the possible values of the
variable B, you can screen the probable error of division of zero by including a TRAP at
the beginning of the program.
5 TRAP 50
10 INPUT A,B
20 PRINT A/B
30 GOTO 10
50 PRINT“ DIVISION BY ZERO IS NOT POSSIBLE”
60 PR1NT“ ENTER ANOTHER NUMBER FOR B BESIDES ZERO’
70 RESUME
RESUME is required after the TRAP response in order to reactivate the TRAP. If
you include the option to enter a replacement for B, RESUME without a line number
returns to the cause of the error— line 20— and executes it as follows:
65 INPUT B
The use of RESUME NEXT proceeds with the next line after the TRAP command,
i.e., line 10.
TRAP tells the computer to go to a specific line number whenever an error occurs.
Do NOT use TRAP until you have removed all syntax errors first. TRAP can only catch
the error condition it is looking for. An error in the syntax or the logic of your TRAP
routine may cause an error, or may not catch the error that you are looking for. In other
words, TRAP routines are sensitive to errors, too.
ERROR FU N C TIO N S
Several reserved variables inherent in the system store information about program
errors. ER stores the error number. EL stores the relevant program line number.
ERR$(N) returns the string representing ER or EL. In the example of division by zero,
ERR$(ER) returns “ DIVISION BY ZERO” and ERR$(EL) returns “ BREAK” . Add
this to the program in the previous section. See Appendix A for a complete listing of
errors.
DOS ERRORS
Information on disk errors is determined from the variables DS and DS$ where DS is the
error number (See Appendix B) and DS$ provides the error number, error message, and
track and sector of the error. DS$ reads the disk error channel and is used during a disk
operation to determine why the disk drive error light is blinking.
Trying to read a directory without a disk in place results in the following error
when the PRINT DS$ command is issued:
74, DRIVE NOT READY, 00, 00

Appendix B highlights specific causes of errors. To convert a function key to read
the disk-drive error channel automatically, use:
KEY 1, “ PRINT DS$ + CHR$(13)
TR A C IN G A N ERROR
Some programs have many complex loops that are tedious to follow. A methodical
step-by-step trace is useful. The BASIC TRON and TROFF commands can be used
within a program as a debugging tool to trace specific routines.
Some errors can only be found by acting like the computer and methodically
following each instruction step-by-step, and then doing all the calculations until you
discover something wrong. Fortunately the Commodore 128 can trace errors for you.
Enter the direct command TRON prior to running a program. The program displays each
line number as they occur in brackets, followed by each result. (To slow down the
display, hold the Commodore ( O ) key down.)
Try it with this double loop:
10 FOR A = 1T05
20 FOR B = 2T06
30 C = B* A:K = K + C:PRINTK
40 NEXTB:NEXTA
50 PRINTK
The results will start off like this:
[10] [20] [30] [30] [30]2
[40] [30] [30] [30]5
meaning the first printed result is the number 2 after operations in lines 10, 20, 30 are
performed. Then lines 40 and 30 result in 5, etc. Notice three activities were performed
in line 30. The Trace function is turned off with the direct command TROFF.
102 COMMODORE 128
W IN D O W IN G
The standard screen display size is 40- or 80-columns by 25 lines. It is often convenient
to have a portion of the screen available for other work. The process of producing and
isolating small segments of your screen is called “ windowing.”
DEFIN IN G A W IN D O W
There are two ways to create a window— either directly or within a program using the
WINDOW command. Using the ESCape key followed by a T or B is all that is
necessary to describe and set a window.
Here’s how to define a window in direct mode:
1. Move the cursor to the upper-left comer position of the proposed window.
Press the (ESC) escape key, then press the letter T key.
2. Move the cursor to the bottom right corner and press the escape key (ESC)
then press the letter B key.
Now that your window is in effect, all commands and listings remain in the
window until you exit by pressing the HOME key twice. This is useful if you have a
listing on the main screen and wish to keep it while you display something else in a
window. See Chapter 13, the Commodore 128 Operating System, under the screen
editor for special ESCape controls within a window.
Although it is possible to define several windows simultaneously on the screen, only
one window can be used at a time. The other windows remain on the display, but they are
inactive. To re-enter a window you have exited, define the top and bottom corners of the
window with the ESC T and ESC B commands, respectively, as you did originally.
The second way to define a window is with the BASIC window command. The
command:
WINDOW 20,12,39,24,1
establishes a window with the top-left corner at column 20, row 12, and the bottom-
right corner at column 39, row 24. The 1 signifies the area to be cleared. Once this
command is specified, all activities are restricted to this window.
Use the window command within a program whenever you want to perform an
activity in an isolated area on the screen.
ADVANCED B A S I C
PROGRAMMING TECHNIQUES
FOR COMMODORE MODEMS
The following information tells you how to:
1. Generate Touch Tone™ frequencies

2. Detect telephone ringing
3. Program the telephone to be on or off the hook
4. Detect carrier
The programming procedures operate in C128 or C64 modes with the Modem/300.
In C128 mode, select a bank configuration which contains BASIC, I/O, and the Kernal.
GENERATING T O U C H TONE
(DTM F) FREQUENCIES
Each button on the face of a Touch Tone telephone generates a different pair of tones
(frequencies). You can simulate these tones with your Commodore 128 computer. Each
button has a row and column value in which you must store the appropriate memory
location in order to output the correct frequency. Here are the row and column
frequency values that apply to each button on the face of your Touch Tone telephone:
TOUCH TONE FREQUENCY TABLE
COLUMN 1 (1029 HZ) COLUMN 2 (1336 HZ) COLUMN 3 (1477 HZ)
Row 1 (697 Hz) 1 2 3

Row 2 (770 Hz) 4 5 6
Row 3 (852 Hz) 7 8 9
Row 4 (941 Hz) * 0 #
To generate these tones in BASIC with your Commodore 128, follow this procedure:
1. Initialize the sound (SID) chip with the following BASIC statements:
SID = 54272
POKE SID + 2 4 .15:POKE SID + 4,16
POKE SID + 11 ,16:POKE SID + 5,0:POKE SID + 12,0
POKE SID + 6,15* 16:POKE SID + 13,15*16:POKE SID + 23,0
2. Next, select one row and one column value for each digit in the telephone
number. The POKE statement for each row and column are as follows:
104 COMMODORE 128
Column 1: POKE SID, 117:POKESID + 1,77

Column 2: POKE SID,152:POKE SID + 1,85
Column 3: POKE SID, 161, POKE SID + 1,94
Row 1: POKE SID + 7,168:POKE SID + 8,44
Row 2: POKE SID + 7,85,:POKE SID + 8,49
Row 3: POKE SID + 7,150:POKE SID + 8,54
Row 4: POKE SID + 7,74 :POKE SID + 8,60
For example, to generate a tone for the number 1, POKE the values for row
1, column 1 as follows
POKE SID + 7 , 168:POKE SID + 8,44:REM ROW 1

POKE SID, 117:POKE SID + 1,77:REM COLUMN 1
3. Turn on the tones and add a time delay with these statements:
POKE SID + 4,17:POKE SID + 11,17:REM ENABLE TONES
FOR 1= 1 TO 50:NEXT:REM TIME DELAY
4. Turn off the tones and add a time delay with the following statements:
POKE SID + 4,16:POKE SID + 11,16:REM DISABLE TONES
FOR 1= 1 TO 50:NEXT:REM TIME DELAY
5. Now repeat steps 2 through 4 for each digit in the telephone number you are
dialing.
6. Finally, disable the sound chip with this statement:
POKE SID + 24,0
DETECTING TELEPHONE RINGING

To detect whether your telephone is ringing using a Commodore 128, use the following
statement:
IF (PEEK(56577) AND 8) = 0 THEN PRINT “ RINGING”
If bit 3 of location 56577 contains a value other than 0, the phone is not ringing.
PROGRAMMING THE TELEPHONE

T O BE O N OR OFF THE H O O K
To program the phone to be off the hook using a Commodore 128, enter the following
statements in a program:
OH = 56577:HI = 32:LO = 255 - 32
POKE (OH + 2),(PEEK(OH + 2) OR HI)
POKE OH,(PEEK(OH) AND LO)
To hang up the phone with a Commodore 128, enter this statement in a program:
POKE OH,(PEEK(OH) OR HI)
Here is the procedure to dial and originate a communication link:
1. Set the modem’s answer/originate switch to the “ O ” for originate.

2. Program the telephone to be OFF the hook.
3. Wait 2 seconds (FOR I = 1 to 500:NEXT:REM 2-SECOND DELAY)
4. Dial each digit and follow it with a delay (FOR I = 1 TO 50:NEXT)
5. When a carrier (high pitched tone) is detected, the Modem/300 automatically
goes on-line with the computer you are connecting with.
6. Program the phone to hang up when you are finished.
Here is the procedure to answer a call:
1. Set the modem’s answer/originate switch to “ A ” for answer.

2. To manually answer, program the telephone to be OFF the hook.
3. To automatically answer, detect if the phone is ringing then program the
phone to be OFF the hook.
4. The Modem/300 automatically answers the call.
5. Program the phone to hang up when you are finished.
DETECTING CARRIER
Your Commodore M odem /1200 and Modem/300 are shipped from the factory with the
ability to detect a carrier on the Commodore 128.
That ability is useful in an unattended auto-answer mode. By monitoring the
carrier detect line, the computer can be programmed to hang up after loss of carrier.
Since a caller may forget to hang up, your program should monitor the transmit and
receive data lines. If there is no activity for five minutes or so, the modem itself should
hang up.
To detect carrier on the Commodore 128, the following statement can be used in a
BASIC program:
OH = 56577:
IF ((PEEK (OH) AND 16) = 0) THEN PRINT “ CARRIER DETECTED”
If bit 4 of location 56577 contains a value other than 0, then no carrier is detected.
ROTARY (PULSE) DIALING

In order to dial a number with a modem, the software in the computer must generate
pulses at a prescribed rate. In the United States and Canada, the rate is between 8 and 10
pulses per second with a 58% to 64% break duty cycle. Most people, however, use 10
pulses per second with a 60% break duty cycle.
So to make a call, your software must first take the phone “ off-hook” (the
equivalent of you picking up the receiver). Then to dial the first digit, a 3 for instance,
106 COMMODORE 128
the software must put the phone on-hook for 60 milliseconds and off-hook for 40
milliseconds. Repeat this process three times to dial a 3.
The same method is used to dial other digits, except 0, which is pulsed ten times.
Pause at least 600 milliseconds between each digit.
USING E S C APE CODES
To perform any of the escape capabilities within a program, use a line such as:
10150 PRINT CHR$(27) + “ U ”
to create an underline cursor (in 80-column only). For example, to clear from the cursor
to the end of a window:
10160 PRINT CHR$(27) + “ @ ”
(See the Screen Editor section of Chapter 13 for all the escape and control codes
available on the Commodore 128.)
RELOCATING BASIC
To relocate the beginning or ending of BASIC (in C128 mode) for additional memory or
to protect machine-language programs from being overwritten by BASIC text, it is
necessary to redefine the starting and ending pointers in required memory addresses.
The Start of BASIC pointer is located at address 45($2D) and 46($2E). The Top
of BASIC pointer is at addresses 4626($ 1212) and 4627($ 1213). The following instruc
tion displays the default locations of the beginning and end of BASIC text, respectively
(when a VIC bit-mapped screen is not allocated):
PRINT PEEK(45) ,PEEK(46) ,PEEK(4626), PEEK(4627)
1 28 0 255
Since the second number in each case is the high byte value, the default start of
basic is 28*256 plus 1 or 7169 while the top is 255*256 or 65280.
The following command reduces the size of BASIC text (program) area by 4K by
lowering the top of BASIC to address 61184 (239*256):
POKE4626,239:POKE4627,0:NEW
To move the beginning of BASIC up in memory by IK, from 7168 to 8192, use
this command line:
POKE 46,32: POKE45,1: NEW
This is the case only when a bit-mapped graphics screen is not allocated. Remem
ber, the beginning of BASIC starts at 16384($4000) when a bit-mapped screen is
allocated, and other parts of memory are shifted around.
MERGING PROGRAM A N D FILES

Files can be merged (combined) by opening an existing file and locating the pointer to
the end of the file so subsequent data can be written to the disk file. C l28 BASIC has
included the APPEND command to accomplish this:
A PPEN D #5, “ FILE NAM E”
opens channel 5 to a previously stored file named “ FILE N A M E.” Subsequent write
(PR IN T#5) statements will add further information to the file. APPEND is primarily
used for data files.
The command CONCAT allows the concatenation (combine in sequence) of two
files or programs while maintaining the name of the first.
CONCAT“ PART2B” TO “ PART2”
creates a new file called Part 2, consisting of the old Part 2, plus the new Part 2b in
sequence. Concatenated BASIC program files must be renumbered before they can
work. Other corrections may also be necessary.
The BASIC routines described in this chapter can greatly enchance the capabilities
of your programs. So far, BASIC has been discussed in detail. The machine language
programming introduced in Chapter 5 can extend program capabilities even further.
And, as shown in Chapter 7, for still greater flexibility and power, you can combine
BASIC and machine language in your programs.
4
COMMODORE 128
GRAPHICS
PROGRAMMING
HOW TO USE
T H E GRAPHICS SYSTEM
NO COMMODORE 128
COMMODORE 128 VIDEO FEATURES

In C128 M ode, Commodore BASIC 7.0 offers fourteen high-level graphics commands
that make difficult programming jobs easy. You can now draw circles, boxes, lines,
points and other geometric shapes, with ten high level commands such as DRAW, BOX
and CIRCLE, and with four sprite commands. (The sprite commands are described in
Chapter 9.) You no longer have to be a machine language programmer, or purchase
additional graphics software packages to display intricate and visually pleasing graphics
displays— the Commodore 128 BASIC graphics capabilities take care of this for you. Of
course, if you are a machine language programmer or a software developer, the
exceptional C l 28 video hardware features offer high price/performance value for any
microcomputer application.
The C l 28 graphics features include:

■ Specialized graphics and sprite commands
■ 16 colors
■ 6 display modes, including:
Standard character mode
M ulti-color character mode
Extended background color mode
Standard bit map mode
Multi-color bit map mode
Combined bit map and character modes (split-screen)
■ 8 programmable, movable graphic objects called SPRITES which make anima
tion possible
■ Custom programmable characters
■ Vertical and horizontal scrolling
The Commodore 128 is capable of producing two types of video signals: 40-
column composite video, and 80-column RGBI video. The composite video signal,
channeled through a VIC II (Video Interface Controller) chip (8564)— similar to that
used in the Commodore 64— mixes all of the colors of the spectrum in a single signal to
the video monitor. The 8563 separates the colors red, green and blue to drive separate
cathode ray guns within the video monitor for a cleaner, crisper and sharper image than
composite video.
The VIC II chip supports all of the Commodore BASIC 7.0 graphics commands,
SPRITES, sixteen colors, and the graphic display modes mentioned before. The 80-
column chip, primarily designed for business applications, also supports sixteen colors
(a few of which are different from those of the VIC chip), standard text mode, and bit
map mode. Sprites are not available in 80-column output. Bit map mode is not
supported by the Commodore BASIC 7.0 language in 80-column output. The 80-column
screen can be bit mapped through programming the 8563 video chip with machine
language programs. See Chapter 10, Programming the 80-Column (8563) Chip, for
information on bit mapping the 80-column screen.
COMMODORE 128 GRAPHICS PROGRAMMING 111
This chapter discusses how to use the Commodore 128 graphics features through
BASIC using the VIC (40-column) screen. Except for the sprite commands, each
graphic command is listed in alphabetical order. The sprite commands are covered in
Chapter 9. Following the format of each command are example programs that illustrate
the features of that command. Wherever possible, machine language routines are
included to show how the machine language equivalent of a BASIC graphics command
operates.
Chapter 8, The Power Behind Commodore 128 Graphics, is a description of the
inner workings of the Commodore 128 graphics capabilities. It explains how screen,
color and character memory are used and how these memory components store and
address data in each display mode. Chapter 9 then explains how to use sprites with the
new BASIC commands. Chapter 9 also discusses the inner workings of sprites, their
storage and addressing requirements, color assignments, and describes how to control
sprites through machine language.
TYPES OF SCREEN DISPLAY

Your C l 28 displays information several different ways on the screen; the parameter
“ source” in the command pertains to three different modes of screen display.
T E X T DISPLAY
Text display shows only text or characters, such as letters, numbers, special symbols
and the graphics characters on the front faces of most C l28 keys. The C l 28 can display
text in both 40-column and 80-column screen formats. Text display includes standard
character mode, multi-color character mode and extended background color mode.
The Commodore 128 normally operates in standard character mode. When you
first turn on the Commodore 128, you are automatically in standard character mode. In
addition, when you write programs, the C128 is in standard character mode. Standard
character mode displays characters in one of sixteen colors on a background of one of
sixteen colors.
M ulti-color character mode gives you more control over color than the standard
graphics modes. Each screen dot, a pixel, within an 8-by-8 character grid can have one
of four colors, compared with the standard mode which has only one of two colors.
Multi-color mode uses two additional background color registers. The three background
color registers and the character color register together give you a choice of four colors
for each dot within an 8-by-8 dot character grid.
Each pixel in multi-color mode is twice as wide as a pixel in standard character
mode and standard bit map mode. As a result, multi-color mode has only half the
horizontal resolution (160 x 200) of the standard graphics modes. However, the
increased control of color more than compensates for the reduced horizontal resolution.
Extended background color mode allows you to control the background color and
foreground color of each character. Extended background color mode uses all four
background color registers. In extended color mode, however, you can only use the first
sixty-four characters of the screen code character set. The second set of sixty-four
characters is the same as the first, but they are displayed in the color assigned to
112 COMMODORE 128
background color register two. The same holds true for the third set of sixty-four
characters and background color register three, and the fourth set of sixty-four characters
and background color register four. The character color is controlled by color memory.
For example, in extended color mode, you can display a purple character with a yellow
background on a black screen.
Each of the character display modes receives character information from one of
two places in the Commodore 128 memory. Normally, character information is taken
from character memory stored in a separate chip called ROM (Read Only Memory).
However, the Commodore 128 gives you the option of designing your own characters
and replacing the original characters with your own. Your own programmable characters
are stored in RAM.
BIT MAP DISPLAY

Bit map mode allows you to display highly detailed graphics, pictures and intricate
drawings. This type of display mode includes standard bit map mode and multi-color bit
map mode. Bit map modes allow you to control each individual screen dot or pixel
(picture element) which provides for considerable detail in drawing pictures and other
computer art. These graphic displays are only supported in BASIC by the VIC chip.
The 80-column chip is designed primarily for character display, but you can bit
map it through your own programs. See Chapter 10, Programming the 80-Column (8563)
Chip, for detailed information.
The difference between text and bit map modes lies in the way in which each
screen addresses and stores information. The text screen can only manipulate entire
characters, each of which covers an area of 8 by 8 pixels on the screen. The more
powerful bit map mode exercises control over each pixel on your screen.
Standard bit map mode allows you to assign each screen dot one of two colors. Bit
mapping is a technique that stores a bit in memory for each dot on the screen. In
standard bit map mode, if the bit in memory is turned off, the corresponding dot on the
screen becomes the color of the background. If the bit in memory is turned on, the
corresponding dot on the screen becomes the color of the foreground image. The series
of 64,000 dots on the screen and 64,000 corresponding bits in memory control the
image you see on the screen. Most of the finely detailed computer graphics you see in
demonstrations and video games are bit mapped high-resolution graphics.
Multi-color bit map mode is a combination of standard bit map mode and
multi-color character mode. You can display each screen dot in one of four colors within
an 8 x 8 character grid. Again, as in multi-color character mode, there is a tradeoff
between the horizontal resolution and color control.
SPLIT SCREEN DISPLAY

The third type of screen display, split screen, is a combination of the first two types.
The split-screen display outputs part of the screen as text and part in bit map mode
(either standard or multi-color). The C l 28 is capable of this since it depends on two
parts of its memory to store the two screens: one part for text, and the other for
graphics.
CO M M AND SUMMARY
Following is abrief explanation of each graphics command available in BASIC 7.0:
B O X : Draws rectangles on the bit-map screen

C H A R : Displays characters on the bit-map screen
C IR C LE: Draws circles, ellipses and other geometric shapes
C O L O R : Selects colors for screen border, foreground, background and characters
D R A W : Displays lines and points on the bit-map screen
G R A P H IC : Selects a screen display (text, bit map or split-screen bit map)
G S H A P E: Gets data from a string variable and places it at a specified position on the
bit-map screen
L O C A T E : Positions the bit-map pixel cursor on the screen
P A IN T : Fills area on the bit-map screen with color
SCALE: Sets the relative size of the images on the bit-map screen
SSHAPE: Stores the image of a portion of the bit-map screen into a text-string variable
W ID T H : Sets the width of lines drawn
The following paragraphs give the format and examples for each of the non-sprite
BASIC 7.0 graphic commands. For a full explanation of each of these commands, see
the BASIC 7.0 Encyclopedia in Chapter 2.
BOX
Draw a box at a specified position on the screen.
BOX [color source], X I, Y l[,X 2,Y 2][,angle][,paint]

where:
color source 0 = Background color

1 = Foreground color
2 = Multi-color 1
3 = Multi-color 2
X I, Y l Top left corner coordinate (scaled)
X2, Y2 Bottom right corner opposite X I, Y l, is the pixel cursor
location (scaled)
angle Rotation in clockwise degrees; default is 0 degrees
paint Paint shape with color
0 = Do not paint
1 = Paint
114 COMMODORE 128
EXA M P LE S:

2 0 G R A P H I C 1 , 1 : R E M S E L E C T BMM
3 0 B O X 1 , 1 0 , 1 0 , 7 0 , 7 0 , 9 0 , 1 : RE M DRAW F I L L E D G R E E N B O X
40 FOR 1 = 2 0 TO 1 4 0 S T E P 3
5 0 B O X 1 , 1 , 1 , 1 + 6 0 , 1 + 6 0 , 1 + 8 0 : RE M DRAW AN D R O T A T E B O X E S
6 0 NEXT
7 0 B O X 1 , 1 4 0 , 1 4 0 , 2 0 0 , 2 0 0 , 2 2 0 , 1 : R E M DRAW 2 N D F I L L E D G R E E N B O X
8 0 CO L O R 1 , 3 : REM S W I T C H TO RED
9 0 B O X 1 , 1 5 0 , 2 0 , 2 1 0 , 8 0 , 9 0 , 1 : RE M DRAW F I L L E D R E D BOX
1 0 0 FOR 1 = 2 0 TO 1 4 0 S T E P 3
1 1 0 B O X 1 , 1 + 1 3 0 , I , 1 + 1 9 0 , 1 + 6 0 , I + 7 0 : R E M DRAW AND R O T A T E R E D B O X E S
1 2 0 NEXT
1 3 0 B O X 1 , 2 7 0 , 1 4 0 , 3 3 0 , 2 0 0 , 2 1 0 , 1 : RE M DRAW 2 N D F I L L E D R E D BOX
140 SLEEP 5 : REM DELAY
1 5 0 G R A P H I C 0 , 1 : R E M S W I T C H T O T E X T MODE
10 COLOR 0 , 1 : COLOR 4 , 1 : COLOR 1,6

20 GRAPHIC 1 , 1
30 BOX 1 , 0 , 0 , 3 1 9 , 1 9 9
40 FOR X = 1 0 TO 1 6 0 S T E P 1 0
50 C=X/10
60 COLOR 1 ,C
70 BOX 1 , X , X , 3 2 0 - X , 3 2 0 - X
80 NEXT
90 SLEEP 5
100 GRAPHIC 0 , 1

20 GRAPHIC 1 , 1
30 BOX 1 , 5 0 , 5 0 , 1 5 0 , 1 2 0
40 BOX 1 , 7 0 , 7 0 , 1 7 0 , 1 4 0
50 DRAW 1 , 5 0 , 5 0 T O 7 0 , 7 0
60 DRAW 1 , 1 5 0 , 1 2 0 T O 1 7 0 , 1 4 0
70 DRAW 1 , 5 0 , 1 2 0 T O 7 0 , 1 4 0
80 DRAW 1 , 1 5 0 , 5 0 T O 1 7 0 , 7 0
90 CHAR 1 , 2 0 , 2 0 , " CU B E E XAM PLE"
100 SLEEP 5
110 GRAPHIC 0 , 1
10 COLOR 1 , 6 : COLOR 4 , l : C O L O R 0 , 1
2 0 G R A P H I C 1 , 1 : R E M S E L E C T B I T MAP MODE
3 0 DO : R E M C A L C U L A T E RA ND OM P O I N T S
4 0 X 1 = 1 NT ( RND ( 1 ) * 3 1 9 + 1 )
50 X 2 = I N T (RND( 1 ) * 3 1 9 + 1)
60 X 3= IN T (RND(1 )* 3 1 9 + 1 }
70 X 4=IN T(R N D (1 )* 3 1 9 + 1 )
8 0 Y1 = I N T ( R N D ( 1 ) * 1 9 9 + 1)
90 Y 2=IN T(R N D (1 )* 1 9 9 + 1 )
100 Y3 = 1 NT(RND( 1 ) * 199 +1)
1 1 0 Y 4 = I N T ( RND( 1 ) * 19 9+ 1)
1 2 0 B O X 1 , X 1 , Y 1 , X 2 , Y 2 : R E M DRAW T H E RA ND OM B O X E S
1 3 0 BOX 1 , X 3 , Y 3 , X 4 , Y 4
1 4 0 DRAW 1 , X I , Y 1 T O X 3 , Y 3 : R E M C O N N E C T T H E P O I N T S
1 5 0 DRAW 1 , X 2 , Y 2 T O X 4 , Y 4
1 6 0 DRAW 1 , X 1 , Y 2 T O X 3 , Y 4
1 7 0 DRAW 1 , X 2 , Y 1 T O X 4 , Y 3
1 8 0 S L E E P 2 : R E M DELAY
1 9 0 SCNCLR
2 0 0 L O O P : REM L O O P C O N T I N U O U S L Y
CHAR
Display characters at the specified position on the screen.
CHAR [color source],X,Y[,string][,RVS]
This is primarily designed to display characters on a bit mapped screen, but it can also
be used on a text screen. Here’s what the parameters mean:
color source 0 = Background

1 = Foreground
X Character column (0-79) (wraps around to the next line
in 40-column mode)
Y Character row (0-24)
string String to print
RVS Reverse field flag (0 = off, 1 = o n )
E X A M P LE :
10 COLOR 2,3: REM multi-color 1 = R ed

20 COLOR 3,7: REM multi-color 2 = Blue
30 GRAPHIC 3,1
40 CHAR 0,10,10, “ TEX T” ,0
CIRCLE
Draw circles, ellipses, arcs, etc. at specified positions on the screen.
CIRCLE [color source],X,Y[,Xr][,Yr]
[,sa][,ea][,angle][,inc]
where:
color source 0 = background color

1 = foreground color
2 = multi-color 1
3 = multi-color 2
X,Y Center coordinate of the CIRCLE
Xr X radius (scaled)
Yr Y radius (default is xr)
sa Starting arc angle (default 0 degrees)
ea Ending arc angle (default 360 degrees)
angle Rotation in clockwise degrees (default is 0 degrees)
inc Degrees between segments (default is 2 degrees)
EX A M P LE S:
CIRCLE1, 160,100,65,10 Draws an ellipse.

CIRCLE1, 160,100,65,50 Draws a circle.
CIRCLE1, 60,40,20,18,,,,45 Draws an octagon.
116 COMMODORE 128
CIRCLE1, 260,40,2 0 ,,,,,9 0 Draws a diamond.

CIRCLE 1, 60,140,20,18,,,, 120 Draws a triangle.
C IR C L E 1,+ 2 ,+ 2,50,50 Draws a circle (two pixels down and two to
the right) relative to the original coordinates of
the pixel cursor.
SAM PLE PROG RAM S:
1 0 REM S U B M A R I N E T R A C K I N G S Y S T E M
2 0 C O LO R 0 , 1 : C O LO R 4 , 1 : C O LO R 1 , 2 : REM S E L E C T B K G R N D , B R D R , S C R E E N C O L OR S
3 0 G R A P H I C 1 , 1 : R E M E N T E R B I T MAP MODE
4 0 BOX 1 , 0 , 0 , 3 1 9 , 1 9 9
5 0 C H A R 1 , 7 , 2 4 , " S U B M A R I N E T R A C K I N G S Y S T E M " : R E M D I S P L A Y C H A R S O N B I T MAP
6 0 C O L O R 1 , 3 : REM S E L E C T R E D
7 0 X R = 0 : Y R = 0 : R E M I N I T X AND Y R A D I U S
8 0 DO
9 0 C I R C L E 1 , 1 6 0 , 1 0 0 , X R , Y R , 0 , 3 6 0 , 0 , 2 : R E M DRAW C I R C L E S
1 0 0 X R = X R + 1 0 : Y R = Y R + 1 0 : R E M UPDATE RA DIU S
1 1 0 LOOP U N T I L XR=9 0
1 2 0 DO
1 3 0 XR= 0 : YR= 0
1 4 0 DO
150 CIRCLE 0 , 1 6 0 , 1 0 0 , X R ,Y R ,0 , 3 6 0 , 0 , 2 : RE M E R A S E C I R C L E
1 6 0 COLOR 1 , 2 : REM S W I T C H TO W H I T E
1 7 0 DRAW 1 , 1 6 0 , 1 0 0 + X R : D R A W 0 , 1 6 0 , 1 0 0 + X R : RE M DRAW S U B M A R I N E B L I P
1 8 0 COLOR 1 , 3 : REM S W I T C H BACK T O RE D
1 9 0 SOUND 1 , 1 6 0 0 0 , 1 5 : R E M B E E P
200 CIRCLE 1 , 1 6 0 , 1 0 0 , X R , Y R , 0 , 3 6 0 , 0 , 2 : R E M DRAW C I R C L E
2 1 0 X R = X R + 1 0 : Y R = Y R + 1 0 : REM U P D A T E R A D I U S
2 2 0 LOOP U N T I L X R = 9 0 :REM LOOP
2 3 0 LOOP

20 G R A P H I C 1 , 1 : R E M S E L E C T BMM
30 X = 1 5 0 : Y= 1 5 0 : X R = 1 5 0 : Y R = 1 5 0
40 DO
50 C I R C L E 1 , X , Y , X R , YR
60 X = X + 7 : Y = Y - 5 : REM I N C R E M E N T X AND Y C O O R D I N A T E S
70 X R = X R - 5 : Y R = Y R - 5 : R E M D E CRE ME NT X AND Y R A D I I
80 LOOP U N T I L XR =0
90 G R A P H I C 0 , 1 : R E M S E L E C T T E X T MODE
COLOR
Define colors for each screen area.
COLOR source number, color number
This statement assigns a color to one of the seven color areas:
AREA SOURCE
0 40-column (VIC) background

1 40-column (VIC) foreground
2 multi-color 1
3 multi-color 2
4 40-column (VIC) border
5 character color (40- or 80-column screen)
Colors codes are in th e range 1-16.

CO LO R CODE COLOR COLOR CODE COLOR
1 Black 9 Orange
2 White 10 Brown
3 Red 11 Light Red
4 Cyan 12 Dark Gray
C o lo r Codes in 40-C olum n (V IC ) O u tp u t
EX A M P LE :
COLOR 0, 1: Changes background color of 40-column screen to black.

COLOR 5, 8: Changes character color to yellow.
SAM PLE PR O G RA M :
10 REM CHANGE F O R E G R OU ND B I T MAP COLOR

20 GRAPHIC 1 , 1
30 1= 1
40 DO
50 COLOR 1 , 1
60 BOX 1 , 1 0 0 , 1 0 0 , 2 1 9 , 1 5 9
70 1 = 1 + 1 :SLEEP 1
80 LOOP U N T I L 1 = 1 7
90 GRAPHIC 0 , 1
100 REM CHANGE B O R D E R COLOR
110 1=1
120 DO
130 COLOR 4 , 1
140 1 = 1 + 1 : SLEEP 1
150 LOOP U N T IL 1 = 1 7
160 REM CH AN GE C H A R A C T E R CO LOR
170 1=1
180 DO
190 COLOR 5 , 1
200 PRINT"COLOR C O D E ";I
210 1 = 1 + 1 : SLEEP 1
220 LOOP U N T I L 1 = 1 7
230 REM CHANGE BACKGROUND C O L O R
240 1=1
250 DO
260 COLOR 0 , 1
270 1 = 1 + 1 : SLEEP 1
280 LOOP U N T I L 1 = 1 7
290 COLOR 0 , 1 : COLOR 4 , l : C O L O R 5 , 2
DRAW
Draw dots, lines and shapes at specified positions on screen.
DRAW [color source], [XI, Y l][TO X2, Y2] . . .
Here are the parameter values:

118 COMMODORE 128
where:
color source 0 Bit map background
1 Bit map foreground
2 Multi-color 1
3 Multi-color 2
X I,Y l Starting coordinate (0,0 through 319,199) (scaled)
X2,Y2 Ending coordinate (0,0 through 319,199) (scaled)
EX A M P LE S:
DRAW 1, 100, 50 Draw a dot.

DRAW , 10, 10 TO 100,60 Draw a line.
DRAW , 10, 10 TO 10,60 TO 100,60 TO 10,10 Draw a triangle.
SAM PLE PROG RAM S:
10 R E M DRAW E X A M P L E S
30 GRAPHIC 1 , 1
40 C H A R 1 , 1 0 , 1 , " T H E DRAW COMMAND"
50 X=10
60 DO
70 DRAW 1 , X , 5 0 : R E M DRAW P O I N T S
80 X=X+10
90 LOOP U N T I L X = 3 2 0
100 CHAR 1 , 1 2 , 7 /'DRAWS P O IN T S"
110 Y=70
120 DO
130 Y=Y+5
140 DRAW 1 , 1 , Y T O Y , Y : R E M DRAW L I N E S
150 LOOP U N T I L Y = 1 3 0
160 CHAR 1 , 1 8 , 1 1 , " L I N E S "
170 DRAW 1 , 1 0 , 1 4 0 T O 1 0 , 1 9 9 T O 9 0 , 1 6 5 T O 4 0 , 1 6 0 T O 1 0 , 1 4 0 : R E M DRAW S H A P E 1
180 DRAW 1 , 1 2 0 , 1 4 5 TO 1 4 0 , 1 9 5 TO 1 9 5 , 1 9 5 TO 2 2 5 , 1 4 5 T O 1 2 0 , 1 4 5 : R E M DRAW S H A P E
190 DRAW 1 , 2 5 0 , 1 9 9 T O 3 1 9 , 1 9 9 T O 3 1 9 , 6 0 T O 2 5 0 , 1 9 9 : R E M DRAW SHAPE 3
200 CHAR 1 , 2 2 , 1 5 , "AND S H A P E S "
210 SLEEP 5 : GRAPHIC 0 , 1

20 G R A P H I C 1 , 1 : R E M S E L E C T BMM
30 Y=1
40 DO
50 DRAW 1 , 1 , Y T O 3 2 0 , Y : R E M DRAW H O R I Z O N T A L L I N E S
60 Y=Y+10
70 LOOP W HI LE Y < 2 0 0
75 X=1
80 DO
90 DRAW 1 , X , 1 T O X , 2 0 0 : R E M DRAW V E R T I C A L L I N E S
95 X=X+10
97 LOOP WHI LE X < 3 2 0
100 CO L O R 1 , 3 : REM S W I T C H T O RED
110 DRAW 1 , 1 6 0 , 0 T O 1 6 0 , 2 0 0 : R E M DRAW X A X I S I N R E D
120 DRAW 1 , 0 , 1 0 0 T O 3 2 0 , 1 0 0 : RE M DRAW Y A X I S I N R E D
130 C O L O R 1 , 6 : RE M S W I T C H T O G R E E N
140 DRAW 1 , 0 , 1 9 9 T O 5 0 , 1 0 0 T O 9 0 , 5 0 T O 1 1 0 , 3 0 T O 1 5 0 , 2 0 T O 1 8 0 , 3 0
150 DRAW 1 , 1 8 0 , 3 0 T O 2 2 0 , 1 0 T O 2 6 0 , 8 0 T O 3 2 0 , 0 : R E M DRAW GR OWTH C U R V E
160 CHAR 1 , 7 , 2 3 , " P R O J E C T E D S A L E S THROUGH 1 9 9 0 "
170 CHAR 1 , 1 , 2 1 , " 1 9 7 0 1975 1980 1985 1990"
180 S L E E P 1 0 : G R A P H I C 0 , 1 : R E M D E L A Y A N D S W I T C H T O T E X T MODE
GRAPHIC
Select a graphic mode.
1) GRAPHIC mode [,clear][,s]
2) GRAPHIC CLR
This statement puts the Commodore 128 in one of the six graphic modes:
MODE DESCRIPTION
0 40-coluinn text
1 standard bit map graphics
2 standard bit map graphics (split screen)
3 multi-color bit map graphics
4 multi-color bit map graphics (split screen)
5 80-column text
EXAMPLES:
GRAPHIC 1,1 Select standard bit map modeand clear the bit map.
GRAPHIC 4,0,10 Select split screen multi-color bit map mode, do not clear
the bit map and start the split screen at line 10.
GRAPHIC CLR Clear and deallocate the bit map screen.
SAMPLE PROGRAM:
10 REM G R A P H I C MODES EX AM P L E
30 G R A P H I C 1 , 1 : R E M E N T E R S T N D B I T MAP
40 CIRCLE 1 , 1 6 0 , 1 0 0 , 6 0 , 6 0
50 CIRCLE 1 , 1 6 0 , 1 0 0 , 3 0 , 3 0
60 C H A R 1 , 9 , 2 4 , " S T A N D A R D B I T MA P M O D E "
70 SLEEP 4
80 G R A P H I C 0 , 1 : R E M E N T E R S T N D C H A R MODE
90 CO L O R 1 , 6 : REM S W I T C H TO G R E E N
100 F O R 1 = 1 TO 2 5
110 P R I N T " S T A N D A R D C H AR ACT ER MODE"
120 NEXT
130 SLEEP 4
140 G R A P H I C 2 , 1 : REM S E L E C T S P L I T S C R E E N
150 CIRCLE 1 , 1 6 0 , 7 0 , 5 0 , 5 0
160 CHAR 1 , 1 4 , 1 , " S P L I T S C R E E N "
170 C H A R 1 , 8 , 1 6 , " S T A N D A R D B I T MAP MODE ON TOP"
180 FOR 1 = 1 TO 25
190 P R I N T " S T A N D A R D C H A R A C T E R MODE ON T H E BOTTOM"
200 NEXT
210 S L E E P 3 : RE M D E L A Y
220 S C N C L R : RE M C L E A R S C R E E N
230 G R A P H I C C L R : R E M D E - A L L O C A T E B I T MAP
GSHAPE
Retrieve (load) the data from a string variable and display it on a specified coordinate.
GSHAPE string variable [X,Y][,mode]
120 COMMODORE 128
where:
string Contains shape to be drawn
X,Y Top left coordinate (0,0 through 319,199) telling where to
draw the shape (scaled— the default is the pixel cursor)
mode Replacement mode:
0 = place shape as is (default)
1 = invert shape
2 = OR shape with area
3 = AND shape with area
4 = XOR shape with area
SAM PLE PRO G RAM :
1 0 R E M D R A W , S A V E AND G E T T H E COMMO DO RE S Y M B O L
2 0 COLOR 0 , 1 : COLOR 4 , 1 : COLOR 1 , 7
3 0 G R A P H I C 1 , 1 : R E M S E L E C T BMM
4 0 C I R C L E 1 , 1 6 0 , 1 0 0 , 2 0 , 1 5 : R E M OUTER C I R C L E
50 CIR C LE 1 , 1 6 0 , 1 0 0 , 1 0 , 9:REM INNER CIRC LE
6 0 BOX 1 , 1 6 5 , 8 5 , 1 8 5 , 1 1 5 : R E M I S O L A T E A R E A TO B E E R A S E D
7 0 S S H A P E A 5 , 1 6 6 , 8 5 , 1 8 5 , 1 1 5 : REM S A V E T H E A R E A I N T O A $
8 0 G SHAP E A $ , 1 6 6 , 8 5 , 4 : REM E X C L U S I V E OR T HE A R E A - T H I S ( E R A S E S ) TURNS O F F P I X E L S
9 0 DRAW 0 , 1 6 5 , 9 4 T O 1 6 5 , 1 0 6 : R E M T U R N O F F ( DR A W I N B K G R N D C O L O R ) P I X E L S I N " C = "
1 0 0 DRAW 1 , 1 6 6 , 9 4 T O 1 6 6 , 9 9 T O 1 8 0 , 9 9 TO 1 8 5 , 9 4 TO 1 6 6 , 9 4 : REM U P P E R F LAG
1 1 0 DRAW 1 , 1 6 6 , 1 0 6 T O 1 6 6 , 1 0 1 T O 1 8 0 , 1 0 1 T O 1 8 5 , 1 0 6 T O 1 6 6 , 1 0 6 : RE M L O WE R F L A G
120 PAINT 1 , 1 6 0 , 110:REM PAINT "C"
130 PAINT 1 , 1 6 8 , 9 8 : REM U P P E R F L A G
1 4 0 S L E E P 5 : REM DE LAY
1 5 0 S S H A P E B $ , 1 3 7 , 8 4 , 1 8 7 , 1 1 6 : R E M SAVE S H A P E I N T O B$
1 6 0 DO
1 7 0 SCNCLR
180 Y=10
1 9 0 DO
200 X=10
2 1 0 DO
2 2 0 G S H A P E B $ , X , Y : REM G E T AND D I S P L A Y S H A P E
2 3 0 X = X + 5 0 : RE M U P D A T E X
2 4 0 LOOP W H I L E X < 2 80
2 5 0 Y = Y + 4 0 : REM U P D A T E Y
2 6 0 LOOP W H I L E Y < 1 6 0
270 SLEEP 3
2 8 0 LOOP
LOCATE
Position the bit map pixel cursor (PC) on the screen.
LOCATE X, Y
E X A M P LE :
LOCATE 160,100 Position the PC in the center of the bit map screen.
Nothing will be seen until something is drawn.
LOCATE + 2 0 ,1 0 0 Move the PC 20 pixels to the right of the last PC position
and place it at Y coordinate 100.
LOCATE + 30, + 20 Move the PC 30 pixels to the right and 20 down from
the previous PC position.
P A IN T
Fill area with color.
PAINT [color source],X,Y[,m ode]
where:
color source 0 Bit map background
1 Bit map foreground (default)
2 Multi-color 1
3 Multi-color 2
X ,Y starting coordinate, scaled (default at pixel cursor (PC))
mode 0 = paint an area defined by the color source selected
1= paint an area defined by any non-background source
EX A M P LE :
10 CIRCLE 1, 160,100,65,50 Draws an outline of a circle.

20 PAINT 1, 160,100 Fills in the circle with color from source 1 (VIC
foreground), assuming point 160,100 is colored
in the background color (source 0).
10 BOX 1, 10, 10, 20, 20 Draws an outline of a box.
20 PAINT 1, 15, 15 Fills the box with color from source 1, assuming
point 15,15 is colored in the background source
(0).
30 PAINT 1, + 1 0 , + 1 0 PAINT the screen in the foreground color source
at the coordinate relative to the pixel cursor’s
previous position plus 10 in both the vertical and
horizontal positions.
SCALE
Alter scaling in graphics mode.
SCALE n [,Xmax,Ymax]
where:
n = 1 (on) or 0 (off)
X max = 320-32767
(default = 1023)
Y max = 200-32767
(default = 1023)
The default scale values are:
Multi-color mode X = 0 to 159 Y = 0 to 199
Bit map mode X = 0 to 319 Y = 0 to 199
EX A M P LE S:
10 GRAPHIC 1,1 Enter standard bit map, turn scaling
20 SCALE 1:CIRCLE 1,180,100,100,100 on to default size (1023, 1023) and
draw a circle.
122 COMMODORE 128
10 GRAPHIC 1,3 Enter multi-color mode, turn scaling

20 SCALE 1,1000,5000 on to size (1000,5000) and draw a
30 CIRCLE 1,180,100,100,100 circle.
SSHAPE
Save shapes to string variables.
SSHAPE and GSHAPE are used to save and load rectangular areas of multi-color or bit
mapped screens to/from BASIC string variables. The command to save an area of the
screen into a string variable is:
SSHAPE string variable, X I, Y l [,X2,Y2]
where:

X I ,Y l Corner coordinate (0,0 through 319,199) (scaled)
X2,Y2 Corner coordinate opposite (X1,Y1) (default is the PC)
EXA M PLES:
SSHAPE A$,10,10 Saves a rectangular area from the coordinate 10,10

to the location of the pixel cursor, into string vari
able A$.
SSHAPE B$, 20,30,47,51 Saves a rectangular area from top left coordinate
string variable B$.
SSHAPE D$, + 10, + 10 Saves a rectangular area 10 pixels to the right and 10
pixels down from the current position of the pixel
cursor.
Also, see the example program under GSHAPE for another example.
W ID T H
Set the width of drawn lines.
WIDTH n
This command sets the width of lines drawn using BASIC’s graphic commands to either
single or double width. Giving n a value of 1 defines a single width line; a value of 2
defines a double width line.
EXA M PLES:
WIDTH 1 Set Single width for graphic commands

WIDTH 2 Set double width for drawn lines
5______
MACHINE
LANGUAGE
ON THE
COMMODORE 128
123
124 COMMODORE 128
This chapter introduces you to 6502-based machine language programming. Read this
section if you are a beginner or novice machine language programmer. This section
explains the elementary principles behind programming your Commodore 128 in machine
language. It also introduces you to the 8502 machine language instruction set and how
to use each instruction. If you are already an experienced machine language program
mer, skip this section and continue to the 8502 Instruction and Addressing Table at the
end of the chapter for reference material on machine language instructions. The 8502
instruction set is exactly the same as the 6502 microprocessor instruction set.
W H A T IS MACHINE LANGUAGE?
Every computer has its own machine language. The type of machine language depends
on which processor is built into the computer. Your Commodore 128 understands 8502
machine language, which is based on 6502 machine language, to carry out its opera
tions. Think of the microprocessor as the brain of the computer and the instructions as
the thoughts of the brain.
Machine language is the most elementary level of code that the computer actually
interprets. True machine language is composed of binary strings of zeroes and ones.
These zeroes and ones act as switches to the hardware, and tell the circuit where to apply
voltage levels.
The machine language discussed in this chapter is symbolic 6502 Assembly
language as it appears in the C l28 Machine Language Monitor. This is not the
full-blown symbolic assembly language as it appears in an Assembler package, since
symbolic addresses or other higher level utilities that an Assembler software package
would provide are not implemented.
Machine language is the lowest level language in which you can instruct your
computer. BASIC is considered a high-level language. Although your Commodore 128
has BASIC built in, the computer must first interpret and translate it to a lower level that
it can understand, before the computer can act upon BASIC instructions.
With each microinstruction, you give the computer a specific detail to perform.
The computer takes nothing for granted in machine language, unlike BASIC, where
many unnoticed machine-level functions are performed by one statement. One BASIC
statement requires several machine language instructions to perform the same operation.
Actually, when you issue a BASIC command, you are really calling a machine language
subroutine that performs a computer operation.
W H Y USE MACHINE LANGUAGE?

If machine language is more intricate and complicated than BASIC, why use it? Certain
applications, such as graphics and telecommunications, require machine language be
cause of its speed. Since the computer does not have to translate from a higher-level
language, it runs many times faster than BASIC.
MACHINE LANGUAGE O N THE COMMODORE 128 125
Programs such as those used in arcade games cannot operate in the relatively slow
speed of BASIC, so they are written in machine language. Other instances dictate the
use of machine language simply because those programming operations are handled
better than in a high-level language like BASIC. But some programming functions such
as string operations are easier in BASIC than in machine language. In these cases,
BASIC and machine language can be used together. You can find information on how to
mix machine language with BASIC in Chapter 7.
Inside your computer is a perpetually running program called the operating
system. The operating system program controls every function of your computer. It
performs functions at lightning speeds you are not even aware of.
The operating system program is written entirely in machine language and is
stored in a portion of the computer called the Kernal ROM. (Chapter 13 describes how
to take advantage of the machine language programs within the Kernal, and how to use
parts of the operating system in your own machine-language programs.)
Though machine language programming may seem more complicated and difficult
than BASIC at first, think back to when you didn’t know BASIC or your first
programming language. That seemed difficult at first, too. If you learned BASIC or
another programming language, you can learn machine language. Although it’s a good
idea to learn a higher-level language such as BASIC before you start machine language,
it’s not absolutely necessary.
W HAT DOES MACHINE LANGUAGE

LOOK LIKE?
Chapter 2 describes the C128 BASIC 7.0 language. Most statements in BASIC start
with a BASIC verb or keyword, followed by an operand. The BASIC keywords
resemble English verbs. The operands are variables, or constants, that are part of an
expression. For example, A + B = 2, is an expression where A, B, and 2 are operands
in the expression. Machine-language instructions are similar, though they have a uniform
format. Here’s the format for an 8502 symbolic machine language instruction as it
appears in the C128 Machine Language Monitor:
OP-CODE FIELD OPERAND FIELD
O PERATIO N CODE (OP-CODE) FIELD

The first part of a machine-language instruction is called the operation code or op-code.
The op-code is comparable to a BASIC verb, in that it is the part of the instruction that
performs an action. A machine language op-code is also referred to in an assembly
language as a mnemonic. All 8502 (6502) machine language assembler mnemonics are
three-letter abbreviations for the functions they perform. For example, the first and most
common instruction you will learn is LDA, which stands for LoaD the Accumulator.
This chapter defines all of the mnemonics.
126 COMMODORE 128
OPERAND FIELD
The second portion of a machine-language instruction is the OPERAND field. In the
C128 Machine Language Monitor, the operand is separated from the op-code with at
least one space and preceded by a ($) dollar sign, ( + ) plus sign (decimal), (&)
ampersand (octal), or a (%) percent (binary) sign to signify that the operand is a
hexadecimal, decimal, octal or binary number. An ADDRESS is the name of or
reference to a specific memory location within the computer.
The number of a memory location is its address, just like houses on your street are
numbered. Addresses in your computer are necessary so they can receive, store and send
(LOAD) data back and forth to the microprocessor.
When you use the Commodore 128’s built-in machine-language monitor, all
numbers and addresses default to hexadecimal numbers, but they can be represented in
decimal, octal or binary. The address is the hexadecimal number of the specified
memory location. When you use an ASSEMBLER, the addresses are referred to as
symbolic addresses. Symbolic addresses allow you to use variable names, instead of
absolute addresses that specify the actual memory location. You declare the symbolic
address to be the numeric address in the beginning of your machine language program or
allow the assembler to assign the address.
When you refer to that address later in the program, you can refer to the symbolic
address rather than to the absolute address as does the Machine Language Monitor.
Using an assembler and symbolic addresses make programming in machine language
easier than using the machine-language monitor and absolute addresses. You will
learn about the eleven addressing modes later in this chapter.
As you know, the second part of a machine-language instruction is the OPER
AND. A machine language operand can be a constant; it does not necessarily have to be
an address reference. When a constant in machine language appears in place of an
address as the second part of an instruction, an operation is performed on a data value
rather than a memory location.
A pound sign (# ) in front of the operand signifies immediate addressing, which
you will learn more about later in the chapter. The pound sign is only used as an aid for
the symbolic language programmer. The pound sign tells the computer to perform
machine-language instruction on a constant, and not an address. In the case of the
Machine Language Monitor, variable names are not allowed. To represent variables in
the monitor, you must reference a memory location where your variable data value is
stored.
EXAMPLES OF
M ACHINE-LANGUAGE INSTRUCTIONS
LDA $100 ; Absolute addressing
LDA $10 ; Zero page absolute addressing
LDA ($FA),Y ; Indirect indexed addressing
LDA $2000,X ; Indexed addressing (absolute)
LDA #$10 ; Immediate addressing (constants)
M ACHINE LANGUAGE ON THE COMMODORE 128 127
TH E SIMILARITIES A N D DIFFERENCES BETWEEN A N

ASSEMBLER A N D A M A CH IN E LANGUAGE M O N IT O R
An assembler and machine-language monitor both provide for symbolic op-codes.
Assemblers typically allow symbolic operands as well, whereas the C l 28 machine-
language monitor refers to addresses and operands literally (absolutely).
An assembler typically has two forms of a file: source code and object code.
Source code is the file you create when you are writing the program including symbolic
start addresses and comments.
The source code file is not executable. It must be assembled (in an intermediate
process) into object code, which is executable code.
The machine language monitor start address is determined by where you place the
actual instructions in memory. The monitor does not provide for comments. The resulting
program, once it is input, is executed immediately as a binary file. No intermediate
assembly step is needed.
THE 8502 MICROPROCESSOR REGISTERS
You have learned that an address is a reference to a specific memory location among the
2 banks of RAM within the Commodore 128. Separate and independent of those RAM
locations are special purpose work and storage areas within the microprocessor chip
itself, called registers. These registers are where the values are manipulated. The
manipulation of the microprocessor registers and their communication with the comput
e r’s memory (RAM and ROM) accomplishes all the functions of machine language and
your computer’s operating system.
Figure 5-1 shows a block diagram of the 8502 microprocessor. As shown in the
figure, the 8502 microprocessor registers are:
Accumulator
X index register
Y index register
Status register
Program counter
Stack pointer
Following are descriptions o f these registers.
TH E A C C U M U LA TO R
The accumulator is one of the most important registers within the 8502 microprocessor.
As the name implies, it accumulates the results of specific operations. Think of
the accumulator as the doorway to your microprocessor. All information that enters
your computer must first pass through the accumulator (or the X or Y register).
128 COMMODORE 128
For example, if you want to store a value within one of the RAM locations, you must
first load the value into the accumulator (or the X or Y register) and then store it into the
specified RAM location. You cannot store a value directly into RAM, without placing it
into the accumulator or the index registers first. (The index registers are described in the
following section.)
Ac
A,
A;
Aa
A,
A,
A.
A7
A.
A,
A,o
A„
A,3
A,3
A,.
A,s
Figure S - l. 8502 Block D iag ram

All mathematical operations are performed within the arithmetic logic unit (ALU)
and stored in the accumulator. It is considered a temporary mathematical work area. For
example, you want to add two numbers, 2 + 3. First, load the accumulator with the 2.
Next add 3 with the ADC mnemonic. Now, you want to perform another operation. You
must store the answer from the accumulator into a RAM location before you perform the
next math operation. If you don’t, your original answer is erased.
The accumulator is so important that it has an addressing mode of its own. All the
instructions using this mode pertain specifically to the accumulator. The following three
sample instructions pertain solely to the accumulator in its own addressing mode:
LDA - LOAD accumulator with memory

STA - STORE the accumulator in memory
ADC - ADD contents of memory to the accumulator
Details on all of the accumulator addressing commands are given later in this chapter.
TH E X A N D Y IN D E X REGISTERS
The second most used registers are the X and Y index registers. These index registers are
used primarily to modify an address by adding an index within a machine-language
instruction. They also can be used as temporary storage locations or to load values and
store them in RAM like the accumulator.
When modifying an address, the contents of the index registers are added to an
original address, called the base address, to find an address relative to the base address.
The resulting address yields the effective address— i.e., the location where a data value
is stored or retrieved. The effective address is acted upon by machine-language instruc
tions. For example, you want to place the value 0 in locations 1024 through 1034. In
BASIC, here’s how you do it:
10 FOR I = 1024 to 1034
20 POKE 1,0
30 NEXT
Here’s how you do it in symbolic machine language by using the X or Y index
register. NOTE: Don’t worry if you don’t understand all of the following instructions. They
are discussed fully in the TYPES OF INSTRUCTIONS section, later in this chapter.
LDA # $ 0 0 Load the Accumulator with 0

TAX Transfer the contents of Accumulator (0) to X
Register.
START STA $0400,X Store contents of Accumulator in address $0400 + X
INX Increment the X register
CPX #$0B Compare the X register with $0B (11 decimal)
BNE START If X register does not equal 11 branch to START.
BRK Stop
* = In the machine-language monitor the symbolic label START is not allowed, so it

would appear as an absolute address reference (eg; $183B).
130 COMMODORE 128
The BASIC example above places a 0 in locations (addresses) 1024 through 1034.
Line 10 sets up a loop from memory locations 1024 to 1034. Line 20 POKEs the value 0
into the location specified by I. The first time through the loop, I equals 1024. The
second time through the loop, I equals 1025 and so on. Line 30 increments the index
variable I by 1 each time it is encountered.
The previous machine-language example accomplishes the same task as the BA
SIC example. LDA # $ 0 0 loads a 0 into the accumulator. TAX transfers the contents of
the accumulator into the X-index register. The following machine-language instructions
form a loop:
START STA $0400,X

INX
CPX #$0B
BN ESTA R T
Here’s what happens within the loop. STA $0400,X stores a 0 in location $0400
(hex) the first time through the loop. Location $0400 is location 1024 decimal. INX
increments the X register by 1, each cycle through the loop. CPX #$0B compares the
contents of the X register with the constant 11 ($0B). If the contents of the X register do
not equal 11, the program branches back to START STA $0400,X and the loop is
repeated.
The second time through the loop, 0 is stored in address $0401 (1025 decimal) and
the X register is incremented again. The program continues to branch until the contents
of the X register equal 11.
The effective address for the first cycle through the loop is $0400 which is 1024
decimal. For the second cycle through the loop the effective address is $0400 + 1, and
so on. Now you can see how the index registers modify the address within machine-
language instruction.
TH E STATUS REGISTER
The microprocessor’s status register indicates the status of certain conditions within the
8502. The status register is controlled by seven programming states of the microproces
sor, and indicates the conditions with flags. The status register is one byte, so each flag
is represented by a single bit. Bit 5 is not implemented.
Branching instructions check (4 of the 7 bits in) the status register to determine
whether a condition has occurred. The conditions for branching pertain to the value of
the bits in the status register. If a condition is true, meaning the FLAG bit corresponding
to one of the four conditions is high (equal to a 1), the computer branches. If the
condition you are testing is not true, the computer does not branch and the program
resumes with the instruction immediately following the branch.
Figure 5 -2 shows the layout of the 8502 status register and lists the conditions
the status register flags.
MACHINE LANGUAGE ON THE COMMODORE 128 131
7 0
[ N 1V j 1B j D 1 I | Z~|C] PROCESSOR STATUS REG "P"
(-► CARRY 1 = TRUE

----- ZERO 1 = RESULT ZERO
-----------► IR Q DISABLE 1 = DISABLE
--------------- ► DECIMAL MODE 1 = TRUE
..................— ► BRK COMMAND
-----------------------------► OVERFLOW 1 = TRUE

--------------------------------* - NEGATIVE 1 = NEG
Figure 5 -2 , 8502 Status Register
The Carry bit (0) is set if an addition operation carries a bit into the next position
to the left of the leftmost bit. The Carry bit is set by other conditions, of which this is
one. The SEC instruction sets the Carry bit. CLear the Carry bit with the CLC
instruction.
The Zero bit (1) is set if the result of an operation equals zero. The command BEQ
stands for Branch on result EQual to Zero. The command BNE stands for Branch on
Result Not Equal to zero. If the zero bit in the status register is set, the program
branches to the address relative to the current program counter value (for a BEQ
instruction). Otherwise, the BEQ command is skipped and the program resumes with the
instruction immediately following the BEQ statement.
The IRQ Disabled bit (2) is set if your program requests interrupts to be dis
abled with the SEI command (Set Interrupt Disable Status). The Disable Interrupt
Status bit is cleared with the CLI command (CLear Interrupt Disable bit) to permit
interrupts to occur. You will learn more about programming interrupts in the section
entitled TYPES OF INSTRUCTIONS and in the Raster Interrupt program explanation in
Chapter 8.
The microprocessor sets the Decimal Mode bit (3) if you instruct the microproces
sor to SEt Decimal Mode with the SED instruction. CLear the Decimal Mode bit with
the CLD instruction, CLear Decimal Mode.
The BRK flag (bit 4) operates similar to the IRQ disable flag. If a BRK instruction
occurs, it is set to 1. Like an IRQ interrupt, the BRK causes the contents of the
program counter to be pushed onto the stack. The contents of the status register is
pushed on top of the stack and evaluated. If the BRK flag is set, the operating system
or your application program must evaluate whether or not a BRK or interrupt has
occurred.
If the BRK flag is cleared once the status register is pushed onto the stack, the
processor handles this as an interrupt and services it. Unlike an interrupt, the BRK flag
causes the address of the program counter plus two to be saved. The microprocessor
expects this to be the address of the next instruction to be executed. You may have to
132 COMMODORE 128
adjust this address since it may not be the actual address of the next instruction within
your program.
The Overflow flag (bit 6) is set by a signed operation overflowing into the sign
bit (bit 7) of the status register. You can clear the Overflow bit in the status register with
the CLV instruction (CLear Overflow flag). You can conditionally branch if the
Overflow bit is set with the BVS (Branch Overflow Set) instruction. Similarly, you can
conditionally branch if the overflow bit is clear with the BVC (Branch Overflow Clear)
instruction. The BIT instruction can be used to intentionally set the overflow flag.
The microprocessor sets the negative flag (bit 7) if the result of an arithmetic
operation is less than 0. You can conditionally branch if the result of an arithmetic
operation is negative, using the BMI instruction, (Branch on result Minus) or positive
using the BPL instruction, (Branch on Result Positive).
The status register indicates seven important conditions within the microprocessor
while your machine language program is executing. Your program can test for certain
conditions, and act upon the results. It gives you a way to conditionally control certain
machine level functions depending on the value of the status flags.
TH E PROGRAM C O UN TER
So far all of the registers within the 8502 are 8 bits, or one byte. The program counter
is twice as wide (16 bits) as the accumulator, X or Y registers or the status register. The
program counter is a 16-bit register because it holds the current address of the next
instruction to be executed. The addresses used in an 8502-based microprocessor are all 16
bits wide. They have to be in order to address all locations within each 64K RAM
bank.
The program counter holds the address of the next instruction to be executed. It
fetches the addresses of the instructions sequentially (usually) and places them on the
16-bit address bus. The processor obtains the data or instructions at the specified 16-bit
address from the data bus. Then they are decoded and executed.
TH E STACK POINTER
Within the RAM of the Commodore 128 is a temporary work area called the stack. It
starts at location decimal 256 and ends at location 511 (hex $100 to $1FF). This area of
computer RAM is referred to as page 1. Paging is explained in the next section.
The stack is used for three purposes in your computer: temporary storage, control
of subroutines, and interrupts. The stack is a LIFO (Last In, First Out) structure which
means the last value placed on the stack is the first one taken off. When you place a
value on the stack, it is referred to as pushing. When you take a value off the stack, it is
considered pulling or popping.
Think of the structure as a stack of lunch trays in a cafeteria. The first tray used is
the one that is pulled off the top. The last one used is the one on the bottom, and it is
used only if all the others are pulled off before it.
The stack pointer is the address of the top stack value (plus 1). W hen a value
is pulled from the stack, the stack pointer then indicates the new address of the
next item on the stack. When a subroutine is called or an interrupt occurs, the
address where the interrupt or subroutine occurs is pushed on top of the stack. Once
the interrupt or subroutine is serviced, the address where it occurred is popped off
the stack and the computer continues where it left off when the interrupt or subroutine
occurred.
16-BIT ADDRESSING:
THE CONCEPT OF PAGING
The Commodore 128 contains 128K of Random Access Memory (RAM). This means
you have two banks of 65536 (64K) RAM memory locations (minus two for locations 0
and 1, which are always present in a RAM bank). Since the 8502 is an 8-bit micropro
cessor, it needs two 8-bit bytes to represent any number between 0 and 65535. One
eight-bit byte can only represent numbers between 0 and 255. Your computer needs a
way to represent numbers as large as 65535 in order to address all the memory
locations.
H ere’s how your computer represents the largest number in one 8-bit byte. The
computer stores it as a binary number. You usually represent it as a hexadecimal number
in your machine-language programs. Figure 5-3 shows the relationship between binary,
hexadecimal and decimal numbers.
BINARY HEXADECIMAL DECIMAL
1 eight-bit Byte = 1 1 1 1 1 1 1 1 $FF 255
Figure 5 -3 . Com parison o f N u m b e r Systems
A byte contains eight binary digits (bits). Each bit can have a value of 0 or 1. The
largest number your computer can represent in eight binary digits is 1 1 1 1 1 1 1 1 ,
which equals 255 in decimal. This means all eight bits are set, or equal to 1. A
bit is considered off if it is equal to 0. In converting binary to decimal, all the binary
digits that are set are equal to 2 raised to the power of the bit position. The bit
positions are labeled 0 through 7 from right to left. Figure 5^4 provides a visual
representation of converting binary to decimal.
27 26 25 24 23 22 21 2°
One binary byte = 1 1 1 1 1 1 1 1

The byte in decimal = 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255
Figure 5-4 . B inary/D ecim al Conversion

134 COMMODORE 128
The top of each column represents the value of 2 raised to the power of the bit
position. Since each bit is turned on when you represent the largest number in one byte,
add all the values at the bottom of each to obtain the decimal equivalent. Figure 5-5
shows another example that converts the binary number 1 1 0 0 1 0 1 0 to decimal.
27 26 2s 24 23 22 21 2°
One binary byte = 1 1 0 0 1 0 1 0

The byte in decimal = 128 + 64 + 0 + 0 + 8 + 0 + 2 + 0 = 202
Figure 5 -5 . B in ary/D ecim al Conversion
Remember, only add the values of two raised to the bit position if the bit is set.
If a bit is off, it equals zero.
Now that you can convert one byte from binary to decimal, you are probably
wondering what this has to do with 16-bit addressing. We mentioned before that the
program counter— the register responsible for storing the address of the next instruction
to be executed— is 16 bits wide. This means it uses two bytes side-by-side to calculate
the address.
You just learned about the low byte, the lower half of the 16 bits used to represent
an address. The upper half of the 16-bit address is called the high byte. The high byte
calculates the upper half of the address the same way as the low byte, except the bit
position numbers are labeled from 8 on the right to 15 on the left. When you raise 2 to
the power of these bit positions, and add the resulting values to the low byte of the
address, you arrive at addresses that go up to 65535. This allows your computer
to represent any number between 0 and 65535, and address any memory location
within each 64K RAM bank. Figure 5 -6 is an illustration of a 16-bit address in
decimal:
High Byte 2 15 2 14 2 13 2 12 2 11 2 10 29 28
One binary byte = 1 1 1 1 1 1 1 1
The byte in
decimal = 3 2 7 6 8 + 1 6 3 8 4 + 8192 + 4096 + 2048 + 1024 + 512 + 256 = 65280
Low Byte 27 26 2s 24 23 22 2' 2°
One binary byte = 1 1 1 1 1 1 1 1
The byte in
decimal = 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1
Figure 5 -6 . E xam p le of 16-Bit Address in Decim al

You can see that the highest number of the high byte of the 16-bit address is
65280. And you know that the highest number of the low byte of the 16-bit address
equals 255. Add the highest high-byte and the highest low-byte number (65280 + 255),
to arrive at 65535, the highest address within each of the two 64K RAM banks.
When the microprocessor calculates the address of the next instruction, it looks at
the high byte of the 16-bit program counter. Try to think of the high byte of the address
as just another 8-bit byte. If this was the case, the bit positions would be labeled from 0
on the right through 7 on the left, just like the low byte of the address. Therefore, the
largest number this 8-bit byte can represent again is 255 decimal.
The value in the high byte determines which 256-byte block is accessed. These
256-byte blocks are referred to as pages. The high byte determines the page boundary of
the address, so the high byte is calculated in increments of 256 bytes. The high byte of
the program counter determines which of the possible 256 pages is being addressed. If
you multiply the number of possible pages, 255 by 256 bytes, you realize the highest
page starts at location 65280, decimal, the same number as in the high byte in Figure
5-6 . Location 65280 is the highest page boundary addressable.
What if you want to address a memory location that does not lie on a page
boundary? That’s where the low byte of the 16-bit address comes in.
The high byte of the program counter represents the 256-byte page boundary.
All addresses between boundaries are represented by the low byte. For example, to
address location 65380 decimal represent the high byte as 255, since 255 times
256 equals 65280. You still have to move 100 addresses higher in memory to location
65380.
The low byte contains the value 100 decimal. The low byte value is added to the
high byte to find the actual, or effective address.
When you look at the memory map of your Commodore 128, you will see
references to the low byte and high byte pointers or vectors to certain machine-language
routines within the operating system or to important system memory locations, like the
start of BASIC.
You can find out the contents of these addresses and where the routines reside in
your Commodore 128’s memory by using the PEEK command in BASIC, or the
Memory command in the Machine Language Monitor. To find the effective address
using BASIC, look in the memory map for the reference to a specific routine or system
function, sometimes called a vector. PEEK the high byte, the page number of the
routine. Multiply by 256 to find the page boundary. Then PEEK the low byte and add it
to the page boundary to arrive at the effective decimal address.
Keep in mind that all the address calculations are performed in binary. They are
explained in decimal so they’re easier to understand. In your machine language pro
grams, you will usually reference memory in hexadecimal notation, explained in the
next section.
136 COMMODORE 128
HEXADECIMAL N O TA TIO N
Your 8502 microprocessor only understands the binary digits 0 and 1. Although machine
language usually requires hexadecimal notation and BASIC processes decimal numbers,
those numbers are translated and processed as binary numbers. Your computer uses
three different number systems, binary (base 2), hexadecimal (base 16) and decimal
(base 10). The machine-language monitor also uses the octal number base. A number
base is also referred to as a radix; therefore, the C l28 uses four radices, but the
microprocessor only understands binary at machine level.
BASIC understands decimal numbers because they are easiest for people to use.
Although BASIC doesn’t process as fast as machine language, the ease of use makes up
for the loss of speed.
Machine language uses hexadecimal notation because it is closer to the binary
number system and easier to translate than decimal. Hexadecimal representation is also
used usually by machine-level programmers because it is easier for people to think of a
group of eight binary digits (a whole byte) than it is to think of them as separate digits
by themselves. How do you find it easier to represent this value:
3A (hexadecimal), or as 00111010 (binary)?
Once values are translated from the higher level language into a form that the
microprocessor can understand (binary digits or bits), they are interpreted as electronic
switches by the internal circuitry. The switches determine if an electronic impulse will
be transmitted by the integrated circuit (I.C.) to perform a specific function, such as
addressing a memory location. If the bit equals 1, the switch is interpreted as on, which
sends a voltage level (approximately 3 to 5 volts) through the I.C. If the binary digit is
equal to 0, no voltage is transmitted. Though this is a simplified illustration, you get an
idea of how the microcomputer system can translate, process and perform the instruc
tions you give to your computer. The hardware and software merge here, at machine
level.
UNDERSTANDING HEXADECIMAL
(H E X ) N O T A T IO N
The key behind understanding hexadecimal (base 16) numbers is to forget about decimal
(base 10). Hexadecimal digits are labeled from 0 through 9 and continuing with A
through F, where F equals 15 in decimal. By convention, hexadecimal numbers are
written with a dollar sign preceding the value so that they can be distinguished from
decimal values. Figure 5-7 provides a table of the hexadecimal digits and their decimal
and binary equivalents:
HEXADECIMAL DECIMAL BINARY
$0 0 0000
$1 1 0001
$2 2 0010
$3 3 0011
$4 4 0100
$5 5 0101
$6 6 0110
$7 7 0111
$8 8 1000
$9 9 1001
$A 10 1010
$B 11 1011
SC 12 1100
$D 13 1101
$E 14 1110
$F 15 1111
Figure 5-7 . H e xa d ecim al/D ecim al/B in ary Conversion
Each hex digit represents four bits. The highest number you can represent with
four bits is 15 decimal. In machine language, you usually represent operands and
addresses as two or four hex digits. Since each hex digit of a four-digit hexadecimal
address takes up four bits, four of them represent 16 bits for addressing.
At first you’ll find yourself converting decimal addresses and operands into
hexadecimal. Then you’ll want to convert the other way. See the HEX$ and DEC
functions for quick and easy decimal to HEX conversions. In the machine language
monitor, use the ( + ) plus sign to represent decimal numbers. Use the conversions for
now, but eventually you should find yourself thinking hexadecimal notation instead of
always converting from decimal to hexadecimal.
ADDRESSING MODES IN
THE COMMODORE 128
Addressing is the process by which the microprocessor references memory. The 8502
microprocessor has many ways to address the internal locations in memory. The
different addressing modes require either one, two or three bytes of storage depending
on the instruction. Each instruction has a different version and op-code. For example,
LDA (LoaD the Accumulator) has eight versions, each with a different op-code to
specify the various addressing modes. See the 8502 Instruction and Addressing Table
section for the different versions of all the 8502 machine-language instructions.
138 COMMODORE 128
A CCU M ULA TO R ADDRESSING

Accumulator addressing implies that the specified operation code operates on the
accumulator. The operand field is omitted since the instruction can only perform the
operation on the accumulator. Accumulator instructions require only one byte of stor
age. Here are some examples of accumulator addressing instructions:
INSTRUCTION HEX OPCODE MEANING
ASL $0A Shift one bit left

LSR $4A Shift one bit right
ROR $6A Rotate one bit right
IMMEDIATE ADDRESSING
Immediate addressing specifies that the operand be a constant value rather than the
contents of a particular address. The operand is the data, not a pointer to the data. At
machine level, the microprocessor actually interprets an operand field constant
and an address in the operand field as two different op-codes, so the pound sign gives
the programmer a way to distinguish between the data and a pointer to the data.
Immediate addressing instructions require two bytes of storage. Here are some immedi
ate addressing instruction examples:
LDA # $ 0 F $A9 Load the accumulator with 15 ($0F)
CMP # $ F F $C9 Compare the accumulator with 255 ($FF)
SBC #$E 0 $E9 Subtract 224 ($E0) from accumulator
ABSOLUTE ADDRESSING
Absolute addressing allows you to access any of the memory locations within either 64K
RAM bank. Absolute addressing requires three bytes of storage; the first byte for the
op-code, the second for the low byte of the address and the third for the high byte. Here
are some examples of absolute addressing instructions:
INC $4FFC $EE Increment the contents of address $4FFC by 1
LDX $200C $AE Load the X register with the contents of address
$200C
JSR SFFC3 $20 Jump to location $FFC3 and save the return address
ZERO-PAGE ADDRESSING
Zero-page addressing requires two bytes of storage; the first byte is used for the opcode
and the second for the zero-page address. Since zero page ranges from addresses 0
through 255, the computer only needs the low byte to represent the actual address. The
high byte is assumed to be 0; therefore, it is not specified. When addressing a zero-page
location, you can still use absolute addressing; however, the execution time is not as fast
as zero-page addressing. Here are some examples:
LDA $FF $A5 Load the accumulator with the contents of zero-page
location $FF (255)
ORA $E4 $05 OR the accumulator with the contents of location

$E4
ROR $0F $66 Rotate the contents o f location $0F one bit to the
right
IMPLIED ADDRESSING
In implied addressing mode, no operand is specified because the op-code suggests the
action to be taken. Since no address or operand is specified, an implied instruction
requires only one byte for the op-code. Some examples are:
DEX $CA Decrement the contents of the X register
INY $C8 Increment the contents of the Y register
RTS $60 Return from Subroutine

140 COMMODORE 128
RELATIVE ADDRESSING
Relative addressing is used exclusively with branch instructions. The branch instructions
(BEQ, BNE, BCC, etc.) allow you to alter the execution path depending on a particular
condition. Branch instructions are similar to IF . . . THEN statements in BASIC since
they both conditionally perform a specified set of instructions.
The operand in the branch instruction determines the destination of the conditional
branch. For example, the op-code BEQ stands for Branch on result EQual to zero. If the
zero flag in the status register is equal to 1 add the operand to the program counter and
continue execution at this new address. Figure 5-8 provides an example in symbolic
assembly language.
LDA #$01 .01800 A9 01 LDA #$ 01

STA TEMP .01802 85 FA STA $FA
DEC TEMP .01804 C6 FA DEC $FA
START BEQ START .01806 F0 FC BEQ $1804
LDX #$01 .01808 A2 01 LDX #$ 01
STA COUNT .0180A 85 FB STA $FB
(A) (B) (C)
*NOTE: The machine language monitor does not provide for

symbolic addresses and labels like TEM P and START.
Figure 5 -8 . Relative Addressing
Figure 5-8 lists the (A) code on the left as it appears in symbolic assembly
language. The code (B) in the middle is the actual machine-level machine code
as it appears in the machine language monitor. The (C) code to the right is the symbolic
machine language as it appears in the monitor as executable code.
In this program segment, the first instruction LoaDs the Accumulator with 1.
STA is the op-code for STore the contents of the Accumulator in the variable
TEMP. The third instruction, DEC, decrements the contents of the variable TEMP. In
the third instruction, START is a label which marks the beginning of the conditional
loop. The branch instruction (BEQ) checks to see if the value stored in TEMP equals
0 as a result of the DECrement instruction. The instruction marks the end of the
loop.
The first time through this loop, the result in TEMP equals 0 so program control
branches back to the instruction specified by the label START.
The second time through the loop, TEMP is less than zero; therefore, the zero
flag in the status register is cleared, the program does not branch to START and
continues with the statement directly following the branch instruction (LDX #$01).
Because of the way this program segment is written, a branch can occur only once, the
first time through the loop.
Under relative addressing, the first byte of the instruction is the op-code and the
second is the operand, representing an offset of a number of memory locations. The
location to branch back to is not interpreted as an absolute address but an offset relative to
the location of the branch instruction in memory.
The offset ranges from -128 through 127. If the condition of the branch is met,
the offset is added to the program counter and the program branches to the address in
memory.
In the example in Figure 5-8, notice that the operand in the branch instruction is
only one instruction past the label START. The operand START is interpreted by the
computer as an offset of three bytes backward in memory since the DEC instruction use
2 bytes and the BEQ op-code uses one byte. The 8502 can only branch forward 127
bytes and branch backward by 128 bytes.
If you enter the machine-language monitor and disassemble the machine-language
code, you’ll see how the computer represents a branch instruction operand as in part (B)
of Figure 5-8. The symbolic code in part (C) operand field represents the operands as
absolute addresses but the assembled hexadecimal code to the left in part (B) of the op
code stores the operand using one byte, a number plus or minus the address of the
branch instruction. The largest number for a forward branch is $7F. A backward branch
is represented by hex numbers greater than $80. When you are within the machine-
language monitor, subtract the operand offset from 255 ($FF) to find the actual value of
the negative offset. In this case $FF minus 3 equals $FC, which is the operand in the
branch instruction in part (B) of Figure 5-8.
Here are some examples of relative addressing branch instructions:
INSTRUCTION HEX OP-CODE MEANING
BEQ $F0 Branch on result Equal to 0
BNE $D0 Branch on result Not Equal to 0
BCC $90 Branch on Carry Clear
IN D EXED ADDRESSING MODES

The Commodore 128 has two special-purpose registers: the X and Y index registers.
In indexing addressing modes, index registers modify an address by adding their
contents to a base address to arrive at the actual or effective address. For example,
here’s a program segment that illustrates the importance of address modification, using
the X and Y index registers:
142 COMMODORE 128
LDA #$0F
LDX #$00
LOOP STA $2000,X
INX
BNE LOOP
The first instruction in this program loads the accumulator with $0F(15 decimal).
The second instruction loads the X register with 0. The third instruction stores the
contents of the accumulator into the address $2000 added to the contents of the X index
register. The first time the loop cycles, $0F is stored in address $2000 ($2000 + 0 =
$2000). The next instruction (INX) increments the contents of the X register. The last
instruction in the loop branches to the statement specified by the label LOOP, which is
the STA $2000, X instruction. The second time through the loop, $0F is stored in
location $2001 ($2000 + 1). The third cycle of the loop stores $0F in location $2002, etc.
The loop continues to cycle and stores $0F in consecutive locations until the X
register equals 0. In other words, the loop circulates 256 times until the X register
equals 0, since 255 plus 1 is represented as 0. This is because the extra bit is carried
over to the ninth bit position, which doesn’t exist in an eight-bit number, so the register
is reset to zero. This is similar to when your car odometer is set at 99,999 miles. When
you travel another mile the dial resets to 00,000.
This example shows just one way to modify addresses with the index registers.
The Commodore 128 has four indexed addressing modes: (1) indexed absolute address
ing (illustrated in the example just shown), (2) indexed zero-page addressing, (3)
indexed indirect addressing, and (4) indirect indexed addressing.
IN D E XE D ZERO-PAGE ADDRESSING
This type of addressing is similar to zero-page addressing except that the index registers
(X or Y) are used to modify addresses within page zero ($00 to $FF) of memory. Since
zero-page addressing requires no high byte to represent the page number, this type of
instruction requires only two bytes of memory. The effective (actual) address is calcu
lated by adding the contents of the index register to the low byte of the address in the
program counter. This addressing mode is faster and more efficient than using indexed
absolute addressing in zero page.
Here are some examples of indexed zero-page addressing instructions:
INSTRUCTION HEX OP CODE MEANING
INC operand, X $F6 Increment the contents of memory by 1.

The base address (the operand) is added to the
contents of the index register (X)).
CMP operand,X $D5 Compare the contents of the accumulator with

memory. The memory base address (the oper
and) is added to the contents o f the index register
(X)).
IN D E XE D ABSOLUTE ADDRESSING
Indexed absolute addressing allows you to access and modify any of the memory
locations in each of the two 64K banks. The effective address is calculated by adding
the contents of the index register (X or Y) to the high and low byte base address
determined by the operand. Since absolute addressing can access any of the available
memory locations, high and low bytes are required to form the 16-bit address. There
fore, this type of addressing requires three bytes.
Here are some examples of indexed absolute addressing instructions:
INSTRUCTION HEX OP CODE MEANING
AND operand,Y $39 Perform the logical AND operation on the

accumulator and the contents of memory base
address plus the contents of the register (Y).
ASL operand,X $1E Shift the contents o f the memory (the memory
is the base address (the operand) added to the
contents of the index register (X)) one bit to the
left.
THE INDIRECT ADDRESSING CONCEPT

So far you’ve learned that the computer calculates the effective address as the
base address (in the program counter) plus the offset from the contents of the index
registers if indexed addressing is used. Indirect addressing calculates the effective
address differently.
Think of indirect addressing as the address of an address. Here’s an illustration
using absolute indirect addressing:
JMP ($0326)
The above JuMP instruction is an example of absolute indirect addressing. This

type of instruction requires three bytes: one for the op-code, one for the low byte and
one byte for the high byte of the 16-bit address. The parentheses indicate that indirect
addressing is used. The second and third bytes of the JMP instruction specify the low and
high byte of the address. The address in the operand field is only the low byte of the effective
address. The contents of the byte immediately following the address specified in the JMP
instruction is automatically placed into the program counter as the high byte of the effective
address. In this example, the contents of location $0326 and $0327 represent the address of
the actual instructions to be executed. For example, location $0326, the low byte of
the effective address, contains the value $65 and location $0327, the high byte of the
effective address, contains the value $F2. The high- and low-byte values are placed in
144 COMMODORE 128
the program counter as the address $F265, the actual address of the next instruction
the computer executes then is $F265.
If the parentheses were not present, the assembler interprets the instruction
as an absolute addressing instruction. The computer would understand the low
byte to be $26 and the high byte to be $03 and would JuMP to the instruction
located at $0326 instead of the intended address of $F265. Since this is not the
case, the high byte is automatically presumed to be the low byte address plus 1
(the contents of $0327).
The last two addressing modes, indirect indexed and indexed indirect, use the
same principle as absolute indirect addressing. Here’s an explanation of each.
IN D E XE D INDIRECT ADDRESSING
Indexed indirect addressing is similar to absolute indirect, although it uses index
registers to modify an address. This type of addressing, sometimes called indirect X
addressing, requires two bytes of storage: the first byte is for the op-code and the second
is for the operand which is used in the effective address calculation. The address
specified in the second byte is added to the contents of the X register and the carry, if
any, is ignored. The results point to an address in page zero in which its contents
contain the low byte of the effective address. The zero page address plus 1 indicates the
high byte of the effective address. Both locations in which the low and high bytes of the
effective address are contained must be located in page zero, locations $00 through $FF.
H ere’s an example:
LDX # $ 0 4
LDA # $ 0 0
STA ($DF,X)
The first line loads the X register with $04. Next, the accumulator is loaded with
0. The third instruction stores zero in the effective address. Calculate the effective
address by taking the base address $DF (not the contents of it) and add the contents of
the X register ($04) to it, which equals $E3. The contents of location $E3 is the low
byte of the effective address and the contents of $E4 is the high byte of the effective
address. For example, the contents of address $E3 contain $56 and the contents of
address $E4 contain $F3. Since the contents of $E3 is the low byte and the contents of
$E4 is the high byte, the effective address is $F356. Indexed indirect addressing is
referred to as pre-indexing because the indexing occurs before the effective address is
actually obtained. Indirect X addressing is useful in addressing a series of pointers such
as the zero-page memory of the Commodore 128.
IND IRECT IN D E XE D ADDRESSING

This mode, also called indirect Y addressing, is post-indexed, which means the adding
o f the index itself obtains the effective address. This mode operates on the principle of a
base address and a displacement. Here’s how it works.
The first of two bytes is the op-code, the second is the operand, a pointer to a
MACHINE LANGUAGE O N THE COMMODORE 128 MS
zero-page memory address. The contents of the pointer and the contents of the Y
register are added to arrive at the low byte of the effective address. The contents of the
pointer act as the base address and the contents of the Y register act as the displacement.
The carry, if any, is added to the memory location directly following the low-byte
address which becomes the high byte of the effective address. This is true indexing,
designed specifically for manipulating tables of data. In order to access different table
values, just change the contents of the Y register since the base address is already
established. H ere’s an example:
LDY # $ 0 8
LDA # $ 0 0
STA ($EA),Y
The first instruction loads the Y register with $08. The second instruction loads
the accumulator with 0. The third instruction stores the contents of the accumulator in
the effective address.
To find the effective address, add the contents of the zero page memory location
(base address) specified in the instruction to the contents of the Y register (displace
ment). In this example, the contents of the address $EA equals $F0. Add $F0 to the
contents of the Y register ($08) to arrive at $F8, the low byte of the effective address of
the next instruction. The high byte of the effective address is obtained by adding the
carry (none in this case) to the zero-page memory location immediately following the
low-byte address. For example, location $F9 contains the value $3F. Since the low byte
is $F8 and the high byte equals $3F, the effective address is $3FF8.
Notice the difference between indirect indexed and indexed indirect addressing
modes as they can be confusing. Remember, the most important difference between the
two addressing modes is the way the effective address is calculated. Indexed indirect is
X indexing, which is indexed prior to the arrival of the effective address. Indirect
indexed is post-indexed with the Y register.
You have just covered all the addressing modes in the Commodore 128. Each calls
for different circumstances and you should use the correct mode whenever circum
stances dictate it to obtain optimal performance from the microprocessor. For example,
use indexed zero-page addressing when you are manipulating zero-page locations in
stead of using indexed absolute.
TYPES OF INSTRUCTIONS
This section explains all the types of machine-language instructions available in the
Commodore 128. They are first covered by type of instruction, such as REGISTER TO
MEMORY and COMPARE instructions; then they are listed alphabetically by op-code
mnemonic with all the different addressing options. This section provides important
information on programming in machine language on the Commodore 128 (or any
6502-based microcomputer).
146 COMMODORE 128
Use this information as a reference for background on each instruction. Figure 5-9
provides an alphabetized list of the 8502 microprocessor op-code mnemonics. For
detailed, quick-reference information, see the following section for an alphabetic list of
instructions, their hexadecimal op-codes, the different versions of the instructions for
each addressing mode and the way they affect the flags in the status register.
8502 M ICROPROCESSOR INSTRUCTION S E T -

ALPHABETIC SEQUENCE
ADC Add Memory to Accumulator with Carry

AND “ A ND” Memory with Accumulator
ASL Shift Left One Bit (Memory or Accumulator)
BCC Branch on Carry Clear

BCS Branch on Carry Set
BEQ Branch on Result Zero
BIT Test Bits in Memory with Accumulator
BMI Branch on Result Minus
BNE Branch on Result not Zero
BPL Branch on Result Plus
BRK Force Break
BVC Branch on Overflow Clear
BVS Branch on Overflow Set
CLC Clear Carry Flag

CLD Clear Decimal Mode
CLI Clear Interrupt Disable Bit
CLV Clear Overflow Flag
CMP Compare Memory and Accumulator
CPX Compare Memory and Index X
CPY Compare Memory and Index Y
DEC Decrement Memory by One

DEX Decrement Index X by One
DEY Decrement Index Y by One
EOR “ Exclusive-Or” Memory with Accumulator
INC Increment Memory by One

INX Increment Index X by One
INY Increment Index Y by One
JM P Jump to New Location

JSR Jump to New Location Saving Return Address
LDA Load Accumulator with Memory

LDX Load Index X with Memory
LDY Load Index Y with Memory
LSR Shift Right One Bit (Memory or Accumulator)
NOP No Operation
8502 M ICROPROCESSOR INSTRUCTION S E T -

ALPHABETIC SEQUENCE (cont’d)
ORA “ O R ” Memory with Accumulator
PHA Push Accumulator on Stack

PHP Push Processor Status on Stack
PLA Pull Accumulator from Stack
PLP Pull Processor Status from Stack
ROL Rotate One Bit Left (Memory or Accumulator)

ROR Rotate One Bit Right (Memory or Accumulator)
RTI Return from Interrupt
RTS Return from Subroutine
SBC Subtract M emory from Accumulator with Borrow

SEC Set Carry Flag
SED Set Decimal Mode
SEI Set Interrupt Disable Status
STA Store Accumulator in Memory
STX Store Index X in Memory
STY Store Index Y in Memory
TAX Transfer Accumulator to Index X

TAY Transfer Accumulator to Index Y
TSX Transfer Stack Pointer to Index X
TXA Transfer Index X to Accumulator
TXS Transfer Index X to Stack Pointer
TYA Transfer Index Y to Accumulator
Figure 5 -9 . 8502 O p -C o d e Mnemonics
REGISTER T O MEMORY
INSTRUCTIONS
The REGISTER TO MEMORY instructions are:
LDA STA
LDX STX
LDY STY
The register to memory instructions either place a value into the accumulator, X
register or Y register from memory, or store a value from a register (A, X, or Y) into a
memory address.
LO A D IN G THE A C C U M U LA TO R
The first and most common instruction is LDA, LoaD the Accumulator. This places a value
into the accumulator, the most powerful and active register in the microprocessor. The
value is derived from the contents of a memory location or a constant. Here’s an example:
148 COMMODORE 128
LDA $2000
This instruction loads the contents of the memory location $2000 (8192 decimal)
into the accumulator. The value in the memory location $2000 remains the same. The
value also remains in the accumulator until another value is placed there or another
operation acts upon it.
The previous example is just one of the addressing modes for loading the accumu
lator. Another form the LDA instruction can take is to load a constant. To load a
constant into the accumulator, you must precede the dollar sign ($) with a pound sign
(# ). For readability, it’s a good idea to place at least one space between the op-code and
the operand but it is not necessary. Here’s an example of loading a constant into the
accumulator:
LDA #$0A
This loads the constant $0A (10 decimal) into the accumulator. Remember,
precede a constant with a pound sign, or else the assembler interprets the instruction as
the contents of a memory address.
The LDX and LDY instructions work the same way as the LDA instruction.
Again, you can load a constant or the contents of a memory address into the X and Y
registers. Examples:
LDX # $ 0 A
LDX $2000
LDX #$FB
STORE: TH E OPPOSITE OF LOAD

You know how to place a value into a register, but how do you do the opposite? The
STORE instruction performs the opposite of a load. It places a value from the A
(accumulator), X, or Y registers into a specified memory address. As you learned in the
addressing section, the load, store and most other machine-language instructions have
several versions, depending on the type of addressing used. Here’s an example:
STA $FC3E
This stores the contents of the accumulator into memory location $FC3E. The
contents of the accumulator remain the same until another instruction modifies it. The
STX and STY instructions work the same way;they store the contents of the register
into a specified memory address. There is no immediate version or pound sign version
of the store command.
CO UNTER INSTRUCTIONS
The COUNTER instructions are
INC DEC
INX DEX
INY DEY
Counter instructions can be used to keep track of or count the number of times an
event occurs. These instructions are used for mathematical manipulations or indexing a
series of addresses. The counter instruction, INC, increments the contents of a memory
address by a value of 1 each time it is encountered. These instructions are used primarily
within a program loop and in conjunction with a branch instruction. Here’s an example
o f a loop and how INC keeps track of a number of occurrences of an event:
LDX # $ 0 0
TXA
START STA $2000,X
INX
BNE START
The first instruction loads a 0 into the X register. The second instruction transfers
the contents of the X register into the accumulator (without erasing the X register).
Instruction three stores the contents of the accumulator (0) into location $2000 the first
time through the loop. The fourth instruction increments the contents of the X register.
The last instruction branches to the instruction specified by the label START, until the
value of the X register equals 0.
This program segment stores 0 ’s in an entire page (256 locations) starting at $2000
and ending at $20FF. When the contents of the X register equals 255 and it is
incremented again, it is reset to 0, since it can only hold an eight-bit number. When this
occurs, the branch is skipped and the program continues with the instruction directly
following the branch instruction.
The INY instruction operates in the same way as INX, since it also only uses
implied addressing. The INC instruction, on the other hand, uses several different
addressing modes including absolute, which uses 16-bit addresses. With the INC
instruction, you can count past the capacity of an 8-bit number, though you must
separate the counter into a high byte and a low byte. For example, the low byte counts
the increments of less than a page and the high byte keeps track of the number of pages.
When low-byte counter is at 255 and is incremented, it is set back to 0. When this
occurs, increment the high-byte counter. To count up to 260 (decimal), the high-
byte value equals 1 and the low byte equals 4. H ere’s an equation to illustrate the point:
(1 * 256) + 4 = 260
H ere’s the machine-language code that does this:
LDA #$00
STA HIGH
STA LOW
LOOP INC LOW
BNE LOOP
INC HIGH
LOOP 2 INC LOW
LDA LOW
CMP #$04
BNE LOOP2
The DECrement instructions operate the same way as the increment instructions.
They are the negative number counterparts of the increment counters.
150 COMMODORE 128
COMPARE INSTRUCTIONS
The Commodore 128 has three compare instructions that check the contents of a register
with the contents of memory. A compare operation can be used to determine which
instructions to execute as a result of a conditioned value. The compare instructions are:
CMP
CPX
CPY
The CMP instruction compares the contents of the accumulator with the contents
of the specified address in the instruction. Compare instructions essentially subtract
memory from a register value but change neither— they just set status flags. CPX
compares the contents of the X register with the specified address. CPY compares the
contents of the Y register with the specified memory location.
All three instructions have versions that will operate in immediate, zero-page and
absolute addressing modes. This means you can compare the contents of a register
(A,X, or Y) with the contents of a zero-page location, any other address above zero page,
or against a constant. H ere’s an example:
LDX # $ 0 0
LDA # $ 0 0
ONE STA $DF,X
INX
CPX #$0A
BNE ONE
The preceding program segment stores 0 ’s in 10 consecutive memory addresses
starting at $DF. The first instruction loads the X register with 0, the second loads 0 into
the accumulator. The third instruction stores 0 in location $DF plus the contents of the X
register. The fourth instruction increments the X register. The fifth instruction compares
the contents of the X index register with the constant $0A (10 decimal). If the contents
of the X register does not equal $0A, the program segment branches back to the store
instruction specified by the label ONE. After the loop cycles ten times, the X register
and the constant $0A are equal. Therefore the processor does not take the branch
and the program continues with the instruction immediately following BNE.
You can compare the value of a register with the contents of an absolute memory
address. H ere’s the same example as above using the contents of a memory address
instead of a constant:
LDA #$0A
STA $FB
LDX # $ 0 0
LDA # $ 0 0
ONE STA $DF,X
INX
CPX $FB
BNE ONE
Remember, if you want to compare numbers larger than eight bits can represent
(greater than 255 decimal), you must separate the number into a low byte and a
high byte.
The BIT instruction can also be used for comparisons. See the logical instructions
next.
ARITHM ETIC A N D
LOGICAL INSTRUCTIONS
The accumulator is responsible for all mathematical and logical operations performed in
your computer. The mathematical and logical instructions available in machine language are:
ADC EOR
AND ORA
BIT SBC
Here’s what each instruction means:
A D C — Add the contents of the specified memory address to the contents of the
accumulator with a carry. It is considered a good programming practice to clear
the carry bit with the CLC instruction before performing any addition. This avoids
adding the carry into the result.
A N D — Perform the logical AND operation with the contents of the accumulator and the
contents of the specified memory address.
BIT— Compare the bits in the specified memory address with those in the accu
mulator. Bits 6 and 7 are transferred to the status register flags. Bit 7 is trans
ferred to the negative status flag bit and bit 6 is sent to the overflow status flag bit.
EO R — Perform the exclusive OR operation with the contents of the specified memory
address and the contents of the accumulator.
O R A — Perform the logical OR operation with the contents of the specified memory
address and the contents of the accumulator.
SB C — Subtract the contents of the specified memory address from the contents of the
accumulator with a borrow. (It is a good practice to set the carry flag before
performing subtraction. This avoids subtracting the borrowed bit from the result.)
A R ITH M E TIC INSTRUCTIONS

(ADC, SBC)
The addition and subtraction instructions are easy to understand. Here’s an example:
CLC
LDA #$0A
STA $FB
ADC #$04
SEC
SBC # $ 0 6
ADC $FB
STA $FD
152 COMMODORE 128
This program segment essentially performs the following mathematical operation:

(10 + 4 ) - 6 + 10= 18.
The first instruction clears the carry bit. The second instruction loads the accumu
lator with $0A (10 decimal). The third instruction stores the value in address $FB for
later use. The fourth instruction adds the constant $04 to the value already in the
accumulator. The SBC instruction subtracts the constant $06 from the contents of the
accumulator. The next instruction, ADC $FB, adds the contents of memory location
$FB to the contents of the accumulator. The resulting value (18($ 12)) of all the
mathematical operations is stored in address $FD.
LO GICAL IN STRUCTIO NS
(A N D , EOR, A N D ORA)
These instructions operate on the contents of a memory address and a register. The AND
operation is a binary (Boolean) algebra operation having two operands that can result in
one of two values, 0 or 1. The only way an AND operation can result in a 1 is if both
the operands equal 1; otherwise the result is 0. For example, the two operands are the
contents of a specified memory address and the contents of the accumulator. Here’s an
illustration of this concept:
Memory address = 10001010

Accumulator = 11110010
Result of AND = 10000010
As noted, the result of an AND operation is (true) 1, only if the two operands are
equal to 1; otherwise the result is 0. Notice bit 7 (high-order bit) equals 1 because both
bit 7 ’s in the operands are 1. The only other resulting bit equal to 1 is bit 1, since both bit
l ’s are equal to 1. The rest of the bits are equal to zero since no other bit positions in
both operands are equal to 1. A 1 and a 0 equals 0, as does a 0 and a 0.
The Boolean OR works differently. The general rule is:
If one of the operands equals 1, the resulting Boolean value equals 1.
For example, the two operands are the contents of a specified memory address and
the contents of the accumulator. Each individual bit can be treated as an operand. H ere’s
an illustration.
Contents of Memory Address = 10101001

Contents of Accumulator = 10000011
Result of the OR operation = 10101011
For all the bit positions that equal one in either operand, the resulting value of that
bit position equals 1. The result is 1 if either operand or both operands are equal to 1.
The exclusive OR works similarly to the OR operation, except if both operands
equal 1, the result is zero. This suggests the following general rule:
M ACHINE LANGUAGE O N THE COMMODORE 128 153
If either of the operands equals 1, the resulting Boolean value is 1, except if both
operands are 1, then the result equals 0.
Here’s an example using this rule:
Contents of Memory Address = 10101001

Contents of Accumulator = 10000011
Result of the exclusive OR = 00101010
In this example, the operands are the same as in the previous OR example. Notice
bits 0 and 7 are now equal to 0 since both operands are equal to 1. All other bit values
remain the same.
BIT
The BIT instruction performs a logical AND operation on the contents of the specified
memory address and the contents of the accumulator, but the resulting value is not
stored in the accumulator. Instead, the zero flag in the status register is set by the result
of the operation. The BIT instruction compares the contents of the accumulator and the
contents of the memory address, bit-for-bit. If the result of the operation of the
accumulator being ANDed by a memory location is 0, then the zero flag (in the status
register) is set to a 1. Otherwise the zero flag is 0.
Your machine language program can then act conditionally depending on the
result of the zero flag in the status register. In addition, bits 7 and 6 from the specified
memory address are moved into the negative-flag and overflow-flag bit positions in the
status register, respectively. These flags can also be used to perform conditional
instructions depending on the value of the flag. For example, the BIT instruction
performs the following:
7 0
________ NV B DI ZC
Contents of Memory Address = 10101001 _______________
Contents of Accumulator = 11001101 -h> 1 0 0
Result of BIT instruction = 10001001 Status Register

(Not stored in accumulator)
Since the resulting bit pattern is not 0, the zero flag in the status register is 0.
In addition, bits 7 and 6 are placed in the bit positions of the negative and overflow
flags, respectively, in the status register. Notice the result of the BIT instruction’s AND
operation is not stored in the accumulator. The original contents of the accumulator
remain intact. See the following example of 2-bit pattern operands that result in 0 when
ANDed:
154 COMMODORE 128
7 0
________ NV BDIZC
Contents of Memory Address = 01111010 _______________
Contents of Accumulator = 10000100 —> 0 1 1
Result of BIT instruction = 00001000 Status Register
This time the bit patterns result in 0. Therefore, the zero flag in the status register
is set to 1. Bits 7 and 6 are also placed into their respective negative and overflow status
register bit positions from their positions in the memory location.
Now you know how each of the arithmetic and logical instructions operate. The
next section discusses branching instructions. Branching instructions are designed so
you can conditionally execute a certain set of instructions, depending on the result of a
condition. Many times the conditions are contingent on the results of an arithmetic or
logical operation, which affects the flags in the status register. The branching instruc
tions then act according to the flags in the status register.
BRANCHING INSTRUCTIONS
The 8502 microprocessor has many conditional branching instructions. By definition, a
branch temporarily redirects the otherwise sequential execution of program instructions.
It transfers control to a location of a machine-language instruction other than the one
immediately following the branch instruction in memory.
The conditional branch instructions cause the microprocessor to examine a particu
lar flag in the status register. The processor, depending on the value of the tested flag,
either takes the branch and transfers control of the program to another location or skips
the branch and resumes with the instruction immediately following the branch.
Think of a conditional branch as a test. For example, if the condition passes the
test, the program branches or shifts control to an instruction that is not the next
sequential instruction in the computer’s memory. If it fails the test, the branch is skipped
and program control resumes with the instruction immediately following the branch
instruction in memory. Remember that program control can also be shifted to an
instruction that is out of sequential order if it fails a test. This means you can transfer
control of the execution of your program depending on the conditions you create. You
may set a condition that branches if the value of a certain flag (operand) is zero.
In another instance, you may set a condition to branch if a specific flag is set
to 1.
The conditional branch instructions available in the 8502 microprocessor are:
BCC BNE
BCS BPL
BEQ BVC
BMI BVS
Here’s what the conditional branch instructions mean. The phrases in parentheses
are the literal translations of the op-code mnemonics. The remainder explains the
meaning behind the op-codes.
B C C — (Branch on Carry Clear) Branch if the Carry flag in the status register equals 0.
BCS— (Branch on Carry Set) Branch if the Carry flag in the status register equals 1.
B EQ — (Branch on result EQual zero) Branch if the zero flag in the status register equals 1.
BMI— (Branch on result Minus) Branch if the negative flag in the status register equals 1.
BN E— (Branch on result Not Equal to zero) Branch if the zero flag in the status register
equals 0.
BPL— (Branch on result PLus) Branch if the negative flag in the status register equals 0.
BVC— (Branch on oVerflow Clear) Branch if the overflow flag in the status register
equals 0.
BVS— (Branch on oVerflow Set) Branch if the overflow flag in the status register
equals 1.
As you can see, all branching instructions depend on the value of a flag in the
status register.
Here are some branching examples.
READY.
MONITOR
PC SR AC XR YR S P
; FBOOO 00 00 00 00 F8
01828 E6 FA INC $FA

0182A A5 FA LDA $FA
0182C DO 02 BNE $1830
0182E E6 FB INC $FB
01830 C8 INY
This program segment keeps track of the low and high pointers in $FA and $FB
respectively. The first instruction (INC $FA) increments the low byte address pointer.
Next, the contents of $FA is loaded into the accumulator.The branch instruction (BNE
$1830) evaluates the value of theaccumulator. If thevalue is notequal tozero, the
branch is taken to the instruction located at address $1830 (INY). In this case the high
byte pointer is not yet ready to be incremented, so the INC $FB instruction is skipped. If
the value in the accumulator is equal to zero, the branch is skipped and the high byte
address pointer is incremented.
This is an example of the BPL (Branch on Result Plus) instruction.
READY.
MONITOR
PC SR AC XR YR S P
; FBOOO 0 0 0 0 00 00 F8
. 01858 8E 0 0 D6 STX $D600

. 0185B 2C 0 0 D6 BIT $D600
. 0185E 10 FB BPL $ 1 85B
. 01860 8D 0 1 D6 STA $D601
This example is a routine that checks the update ready status bit for the 8563
address register, and ensures that data is valid before writing a value to an 8563 register.
The first instruction stores the contents of the X register, which was previously loaded
156 COMMODORE 128
with an 8563 register number, into the 8563 address register. The BIT instruction places
bit 7 of location $D600 into the negative flag in the 8502 status register. The BPL
instruction branches to the BIT instruction in location $ 185B as long as the value of the
negative flag is equal to 0. To the 8563 chip, this means the data is not yet valid and
cannot be written to or read from until bit 7 is set. This loop continues until the value of
bit 7 is 1, then it is transferred to the negative flag. The result now becomes negative
so the branch is skipped and control is passed to the next instruction inmemory, which
stores the data into the 8563 data register. Refer to Chapter 10, Writing to an 8563
Register for an expanded version of this program.
REGISTER TRANSFER INSTRUCTIONS

Register transfer instructions move a value from one register (A, X, or Y) to another.
This instruction is useful since it only requires one byte of memory and saves the
programmer the trouble of loading the value from one register and storing it in another.
The 8502 microprocessor has the following six register transfer instructions:
T A X — Transfer contents of accumulator to X index register

T A Y — Transfer contents of accumulator to Y index register
T S X — Transfer the contents of the stack pointer to X index register
T X A — Transfer the contents of X index register to the accumulator
T Y A — Transfer the contents of the Y index register to the accumulator
T X S — Transfer the contents of the X register to the stack pointer
The TXS and TSX instructions transfer values from the X index register to the
stack pointer and vice versa. This is useful if you need to take a value off the stack
temporarily, in a mathematical operation (for example, to operate on it and then replace
it on the stack). Another use is to take a value off the stack, place it in the X register for
temporary storage, add a new value on the stack, and then place the old value back on
top. This could be the case when you need to sort values in ascending order.
SHIFT A N D ROTATE INSTRUCTIONS

The shift and rotate instructions manipulate the bits of the accumulator or memory.
Following are the shift and rotate instructions used by the 8502 family of microprocessors:
A SL— Shift the whole byte one bit to the left

LSR— Shift the whole byte one bit to the right
ROL— Rotate the whole byte one bit to the left
RO R— Rotate the whole byte one bit to the right
SHIFT INSTRU C TIO NS

The shift instructions are useful when evaluating the value of a single bit at a time in a
series of bits that control your program. For example, a joystick read routine is an
example that calls for the shift instruction. Locations $DC00 and $DC01 control the
joystick direction (bits 0-3 ), and the joystick fire button (bit 4). One way to evaluate
these values is to shift them to the right. This causes the value to be passed to the carry
flag. If the carry flag is enabled (1), then the joystick is being pushed in the direction
corresponding to that bit. Here is a joystick read routine that uses the LSR instruction to
evaluate the direction of the joystick:
READY.
MONITOR
PC SR AC XR YR SP
; FB000 00 00 00 00 F8
01800 AD 00 DC LDA $DC00

01803 A0 00 LDY #$00
01805 A2 00 LD X #$00
01807 4A LSR
01808 B0 01 BCS $ 1 80B
0180A 88 DEY
0180B 4A LSR
0180C B0 01 BCS $180F
0180E C8 INY
0180F 4A LSR
01810 B0 01 BCS $1813
01812 CA DEX
01813 4A LSR
01814 B0 01 BCS $1817
01816 E8 INX
01817 4A LSR
01818 86 FA STX $FA
0181A 84 FB STY $FB
0181C 60 RTS
ROTATE IN STR U CTIO N S

The rotate instructions operate a little differently. Instead of the shifted bit falling into
the carry flag, the bit “ falling off the edge” is placed in the carry bit, then the carry bit
is placed at the opposite end of the byte. For example, if the ROR (rotate right)
instruction is specified, each bit is moved one position to the right. Now bit 7 is placed in
the carry bit and the carry bit is rotated around to the left and placed in the bit 7 bit
position. The ROL instruction operates in the same manner, except the rotation is
leftward rather than to the right. See Figure 5-10 to visualize the rotation concept of the
ROR (rotate right) instruction:
Bit Position
Figure 5-1 0. C oncept o f RO R (R o tate Right) Instruction

158 COMMODORE 128
SET A N D CLEAR INSTRUCTIONS

The set and clear instructions are designed to manipulate the bits (flags) within the status
register and control certain conditions within the microprocessor. These are the set and
clear instructions available in 8502 machine language:
SEC Set the Carry Flag

SED Set Decimal Mode
SEI Set the Interrupt Disable Bit
CLC Clear the Carry Flag
CLD Clear Decimal Mode
C LI Clear the Interrupt Disable Bit
CLV Clear the Overflow Flag
Each of these instructions applies to a flag in the status register that controls a
particular microprocessor condition. Notice that each clear instruction has a counterpart
which sets the condition, except for CLV (Clear Overflow Flag). The overflow flag can
be set by the BIT instruction or from the result of a signed mathematical operation
overflowing into the sign bit.
Figure 5-11 shows the 8502 status register:
7 o
| N 1V 1 j B | D 1 I 12 | C] PROCESSOR STATUS REG "P"
L».CARflY 1 x TRUE
------►ZERO 1 = RESULT ZERO
----------► IRQ DISABLE 1 * DISABLE
-------------- ► DECIMAL MODE 1 = TRUE
------------------ ► BRK COMMAND
---------------------------►OVERFLOW 1 * TRUE
-------------------------------► NEGATIVE 1 * NEG
Figure 5-1 1. 8502 Status Register
The flags of the status register are set for various reasons. For example, set
decimal mode when you want to perform calculations in binary coded decimal (BCD)
notation rather than hexadecimal. Set the carry flag when you are performing subtrac
tion. Set the interrupt disable bit when you want to prevent interrupts from occurring.
An example of a split screen, smooth scrolling raster interrupt routine is given at the end
of Chapter 8.
The clear instructions operate in the reverse of the set instructions. To make sure
that a carry does not occur during an addition operation, clear the carry flag before
adding two numbers in the accumulator. To perform mathematical operations in hexa

decimal or binary numbers, clear the decimal mode flag so that your calculations are not
mistakenly performed in binary coded decimal. W henever the result of a signed mathe
matical operation overflows into the sign bit an overflow error occurs. To correct this,
clear the overflow flag with the CLV op-code.
When a program requires interrupts, first set the interrupt disable bit (SEI) to
prevent interrupts from occurring. At the end of the interrupt initialization routine, issue
the CLI (Clear Interrupt Disable bit) instruction to enable (allow) interrupts to occur.
JUMP A N D RETURN INSTRUCTIONS

JUMP IN STR U CTIO N S
The 8502 processor makes use of two jum p instructions:
JMP — Jump to new location

JSR— Jump to new location Saving the Return address
These instructions both redirect control of the microprocessor to a location other than
the one immediately following it in memory. The first instruction, JMP,is a one-way trip
to the location specified in the operand field, orthecontentsof it (indirect).Forexample:
JMP $1800
jumps to location $1800 and executes the instruction contained in that location. This is a
direct jump.
You can also jump indirectly. For example:
JMP ($1800)
jumps to the address specified in the contents of location $1800. For instance, location
$1800 contains the value $FE and location $1801 contains the value $C0. Therefore, the
above instruction jumps to location $C0FE, and not location $1800. Jumping indirectly
is always denoted by parentheses around the address in the operand field, and it means
to jum p to the location specified by the CONTENTS OF the address in the operand field.
The JSR instruction calls subroutines and saves the return address to the stack, so
when an RTS instruction is encountered at the end of the subroutine, the microprocessor
knows where to resume processing in the main (calling) program. Program control
resumes with the instruction in memory immediately following the JSR instruction. In
short, JSR is a round trip, while JMP is one way. For example:
. 01804 20 58 18 JSR S1858

. 01807 A2 0C LDX #$0C
jumps to the subroutine starting at location $1858. The return address is saved on the
stack, so when the RTS instruction is encountered in this subroutine:
160 COMMODORE 128
01858 8E 00 D6 STX $D600

0185B 2C 00 D6 BIT $D600
0185E 10 FB BPL $ 1 8 5B
01860 8D 01 D6 STA $D601
01863 60 RTS
the processor resumes with the main program instruction (LDX #$0C ) in location $1807.
RETURN INSTR U C TIO N S

The 8502 instruction set has two return instructions:
RTI— Return from Interrupt

RTS— Return from Subroutine
The first instruction returns from your interrupt service routine after the interrupt
disable bit is cleared (CLI) and the interrupt occurs. The RTI is the last instruction in the
interrupt service routine. The interrupt service routine is the series of instructions which
are performed on the occurrence of an interrupt. Refer to Chapter 8, Raster Interrupt
Split Screen Program with Horizontal Scrolling for a working example of an interrupt
service routine.
The RTS instruction is the last instruction in a machine language subroutine called
from BASIC or by the machine language JSR instruction. See the Jump instructions
above for an example.
STACK INSTRUCTIONS
Four stack instructions are included in the 8502 instruction set to manipulate the values
on the stack. These instructions are as follows:
P H A — Push accumulator on the stack

P H P — Push processor status on the stack
PLA— Pull accumulator from the stack
PL P— Pull processor status from the stack
The term push means to place a value on the stack, while pull means to remove
a value from the stack. The only values pushed or pulled on to or off the stack are the
contents of the status register or the accumulator. The manipulation of the stack values
is important to the programmer when processing interrupts. The Raster Interrupt Split
Screen Program with Horizontal Scrolling section in Chapter 8 illustrates the manipula
tion of the stack values prior to returning from the interrupt.
THE NOP INSTRUCTION

The NOP instruction stands for no operation. It is often used to add space between
program segments for readability. This instruction is not executable.
8502 INSTRUCTION AND

ADDRESSING TABLE
The next 16 pages contain the 8502 Instruction and Addressing Table. These are the
conventions used in the table:
1. OP-CODE
2. Brief definition
3. Operation notation
4. Status flags
5. Flags affected
6. Addressing Modes
7. Assembly language form
8. OP-CODE (in hex)
9. Number of bytes
10. Number of instruction cycles
The following notation applies to this summary:
A Accumulator
X,Y Index Registers
M Memory
P Processor Status Register
S Stack Pointer
J Change
- No Change
+ Add
A Logical AND
- Subtract
-V- Logical Exclusive Or
T Transfer from Stack
4 Transfer to Stack
— » Transfer to
<- Transfer from
V Logical OR
PC Program Counter
PCH Program Counter High
PCL Program Counter Low
OPER OPERAND
# IMMEDIATE ADDRESSING MODE
162 COMMODORE 128
ADC Add memory to accumulator with carry ADC
Operation: A + M + C - ^ A , C N E C I E V
y y y - - y
ADDRESSING ASSEMBLY OP NO. NO.

MODE LANGUAGE FORM CODE BYTES CYCLES
Immediate ADC # Oper 69 2 2

Zero Page ADC Oper 65 2 3
Zero Page, X ADC Oper, X 75 2 4
Absolute ADC Oper 6D 3 4
Absolute, X ADC Oper, X 7D 3 4*
Absolute, Y ADC Oper, Y 79 3 4*
(Indirect, X) ADC (Oper, X) 61 2 6
(Indirect), Y ADC (Oper), Y 71 2 5*
* Add 1 if page boundary is crossed.
AND “ A N D ” m e m o ry w ith a c c u m u la to r AND
Logical AND to the accumulator

Operation: A A M —> A N E C
y y -

Immediate AND # Oper 29 2 2

Zero Page AND Oper 25 2 3
Zero Page, X AND Oper, X 35 2 4
Absolute AND Oper 2D 3 4
Absolute, X AND Oper, X 3D 3 4*
Absolute, Y AND Oper, Y 39 3 4*
(Indirect, X) AND (Oper, X) 21 2 6
(Indirect), Y AND (Oper), Y 31 2 5

ASL ASL S h ift L e ft O n e B it (M e m o r y o r A c c u m u la to r) ASL
Operation: C [7 |6 |5 |4 |3 |2 lH o 1 < - 0 N Z C I D V
/ / / - -

Accumulator ASL A OA 1 2
Zero Page ASL Oper 06 2 5
Zero Page, X ASL Oper, X 16 2 6
Absolute ASL Oper 0E 3 6
Absolute, X ASL Oper, X IE 3 7
BCC BCC B ranch on C a rry C le a r BC C
Operation: Branch on C = 0 N Z C V

Relative BCC Oper 90 2 2*
* Add 1 if branch occurs to same page.

* Add 2 if branch occurs to different page.
BCS BCS B ranch on c a rry set BCS
Operation: Branch on C = 1 N Z C V

Relative BCS Oper BO 2 2*

* Add 2 if branch occurs to next page.
164 COMMODORE 128
BEQ BEQ B ranch on result ze ro BEQ
Operation: Branch on Z = 1 N Z C I D V

Relative BEQ Oper FO 2 2*

* Add 2 if branch occurs to next page.
B IT B IT Test bits in m e m o ry w ith a c c u m u la to r B IT
Operation: A A M , M 7 —> N , M6 —» V
Bit 6 and 7 are transferred to the status register. N Z C I D V
If the result of AAM is zero then Z = 1, otherwise M7 / - - M,6
Z =0

Zero Page BIT Oper 24 2 3

Absolute BIT Oper 2C 3 4
BMI B M I B ranch on result minus BMI
Operation: Branch on N = 1 N z c I D V

Relative BMI Oper 30 2 2*

BNE BNE B ranch on result n o t ze ro BN E
Operation: Branch on Z = 0 N Z C I D V

Relative BNE Oper DO 2 2*

BPL BPL B ranch on result plus BPL
Operation: Branch on N = 0 N Z C I D V

Relative BPL Oper 10 2 2*

BRK BRK Force B re a k BRK
Operation: Forced Interrupt PC + 2 J, P | N Z C I D V

Implied BRK 00 1 7
1. A BRK com mand cannot be masked by setting I.
BVC BVC B ranch on o v e rflo w c le a r BVC
Operation: Branch on V = 0 N Z C I D V

Relative BVC Oper 50 2 2*

BVS BVS B ranch on o ve rflo w set BVS
Operation: Branch on V = 1 N Z C I D V

Relative BVS Oper 70 2 2*

166 COMMODORE 128
CLC CLC C le a r c a rry fla g C LC
Operation: O ^ C N Z C I D V
- - 0 - - -

Implied CLC 18 1 2
CLD CLD C le a r d ec im al m ode CLD
Operation: 0 D N Z C I D V
- - - - 0 -

Implied CLD D8 1 2
CLI C LI C le a r in te rru p t disable b it CLI
Operation: 0 -» I N Z C I D V

Implied CLI 58 1 2
CLV CLV C le a r o ve rflo w fla g CLV
Operation: 0 -> V N Z C I D V
- - - - - 0

Implied CLV B8 1 2
CM P C M P C o m p a re m e m o ry and a c c u m u la to r CMP
Operation: A - M N Z C I E V

Immediate CMP #O p er C9 2 2
Zero Page CMP Oper C5 2 3
Zero Page, X CMP Oper, X D5 2 4
Absolute CMP Oper CD 3 4
Absolute, X CMP Oper, X DD 3 4*
Absolute, Y CMP Oper, Y D9 3 4*
(Indirect, X) CMP (Oper, X) Cl 2 6
(Indirect), Y CMP (Oper), Y D1 2 5*
Add 1 if page boundary is crossed.
CPX C PX C o m p a re M e m o r y and In d e x X CPX
Operation: X - M N Z C I E V
/

Immediate CPX #O p er EO 2 2
Zero Page CPX Oper E4 2 3
Absolute CPX Oper EC 3 4
CPY CPY C o m p a re m e m o ry and index Y CPY
Operation: Y - M N Z C I E V
J J
Immediate CPY #O p er CO 2 2
Zero Page CPY Oper C4 2 3
Absolute CPY Oper CC 3 4
168 COMMODORE 128
DEC D E C D e c re m e n t m e m o ry by one DEC

Operation: M - 1 —> M N Z C I V

MODE LANGUAGE FORM CODE; BYTES CYCLES
Zero Page DEC Oper C6 2 5

Zero Page, X DEC Oper, X D6 2 6
Absolute DEC Oper CE 3 6
Absolute, X DEC Oper, X DE 3 7
DEX D E X D e c re m e n t index X by one DEX

Operation: X - 1 —» X N Z C V
/ / -

Implied DEX CA 1 2
DEY D E Y D e c re m e n t index Y by one DEY

Operation: Y - 1 —» Y N Z C V
y y -
Implied DEY 88 1 2
EO R E O R “ Exclusive — O r ” m em o ry w ith a c c u m u la to r EO R
Operation : A - * M -V" A N Z C V
y y -
Immediate EOR #O p er 49 2 2
Zero Page EOR Oper 45 2 3
Zero Page, X EOR Oper, X 55 2 4
Absolute EOR Oper 4D 3 4
Absolute, X EOR Oper, X 5D 3 4*
Absolute, Y EOR Oper, Y 59 3 4*
(Indirect, X) EOR (Oper, X) 41 2 6
(Indirect), Y EOR (Oper), Y 51 2 5*
* A d d 1 if p age b o u ndary is crossed.

IN C IN C In c re m e n t m e m o ry by one IN C
Operation: M + 1 —> M N Z C V
J y -
Zero Page INC Oper E6 2 5

Zero Page, X INC Oper, X F6 2 6
Absolute INC Oper EE 3 6
Absolute, X INC Oper, X FE 3 7
IN X I N X In c re m e n t In d e x X by one IN X
Operation: X + 1 —> X N Z C V
y y -
Implied INX E8 1 2
IN Y IN Y In c re m e n t In d e x Y by one IN Y
Operation: Y + 1 —» Y N Z C V
y y -
Implied INY C8 1 2
JMP J M P Jum p to n ew location JMP
Operation: (PC + 1) —» PCL N Z C

(PC + 2) PCH

Absolute JM P Oper 4C 3
Indirect JM P (Oper) 6C 3 5
170 COMMODORE 128
JSR JSR Jump to new location saving return address JSR
Operation: PC + 2 j , (PC + 1) -> PCL N Z C I D V

(PC + 2) -> PCH _ _ _ _ _ _

Absolute JSR Oper 20 3 6
LDA LDA Load accumulator with memory LDA
Operation: M ^ A N Z C I D V

Immediate LDA #O p er A9 2 2
Zero Page LDA Oper A5 2 3
Zero Page, X LDA Oper, X B5 2 4
Absolute LDA Oper AD 3 4
Absolute, X LDA Oper, X BD 3 4 *
Absolute, Y LDA Oper, Y B9 3 4 *
(Indirect, X) LDA (Oper, X) A1 2 6
(Indirect), Y LDA (Oper), Y B1 2 5*
LDX L D X Load index X with memory LDX
Operation: M - ^ X N Z C I D V

Immediate LDX #O p er A2 2 2
Zero Page LDX Oper A6 2 3
Zero Page, Y LDX Oper, Y B6 2 4
Absolute LDX Oper AE 3 4
Absolute, Y LDX Oper, Y BE 3 4*
* A dd 1 w hen p age bou n d ary is crossed.

LDY L D Y Load in d e x Y w ith m e m o ry LDY
Operation: M —> Y N Z C I V
J J ~ ~ ~

Immediate LDY #O p er AO 2 2
Zero Page LDY Oper A4 2 3
Zero Page, X LDY Oper, X B4 2 4
Absolute LDY Oper AC 3 4
Absolute, X LDY Oper, X BC 3 4*
* Add 1 when page boundary is crossed.
LSR LSR S h ift rig h t one b it (m e m o ry o r a c c u m u la to r) LSR
Operation: 0 -> |716]5[413121 llo l -* C N Z C I D V

o y y - -
Accumulator LSR A 4A I 2
Zero Page LSR Oper 46 2 5

Zero Page, X LSR Oper, X 56 2 6
Absolute LSR Oper 4E 3 6
Absolute, X LSR Oper, X 5E 3 7
NOP N O P N o o p e ra tio n NOP
Operation: No Operation (2 cycles) N Z C V

Implied NOP EA 1 2
172 COMMODORE 128
ORA O RA " O R ” m e m o ry w ith a c c u m u la to r ORA
Operation: A V M - » A N Z C V
/ y -
MODE LANGUAGE FORM CODE: BYTES CYCLES
Immediate ORA #O p er 09 2 2
Zero Page ORA Oper 05 2 3

Zero Page, X ORA Oper, X 15 2 4
Absolute ORA Oper 0D 3 4
Absolute, X ORA Oper, X ID 3 4*
Absolute, Y ORA Oper, Y 19 3 4*
(Indirect, X) ORA (Oper, X) 01 2 6

(Indirect), Y ORA (Oper), Y 11 2 5
* Add 1 on page crossing
PHA P H A Push a c c u m u la to r on stack PHA
Operation : A | N Z C V

Implied PHA 48 1 3
PHP P H P Push processor status on stack PHP
Operation: P J, N Z C V

Implied PHP 08 1 3
PLA PLA P ull a c c u m u la to r fro m stack PLA
Operation: A f N Z C V
y y -
Implied PLA 68 1 4
PLP PLP P u ll processor status fro m stack PLP
Operation: P f N Z C I D
From Stack

Implied PLP 28 1 4
ROL ROL R o ta te one b it le ft (m e m o ry o r a c c u m u la to r) ROL
M or A
Operation: ----- |7 |6 |5 |4 |3 |2 |l |0 | <-|c h— N Z C I D
y y y - -
Accumulator ROL A 2A 1 2
Zero Page ROL Oper 26 2 5
Zero Page, X ROL Oper, X 36 2 6
Absolute ROL Oper 2E 3 6
Absolute, X ROL Oper, X 3E 3 7
RO R R O R R o ta te one b it rig h t (m e m o ry o r a c c u m u la to r) ROR
Operation: |C k |71 6 15 14131211 |o| 0 N Z C I D
1 y y y - -
Accumulator ROR A 6 A 1 2
Zero Page ROR Oper 66 2 5
Zero Page, X ROR Oper, X 76 2 6
Absolute ROR Oper 6E 3 6
Absolute, X ROR Oper,X 7E 3 7
174 COMMODORE 128
RTI R T I R eturn fro m in te rru p t RTI
Operation: P f P C f N Z C I D V
From Stack

Implied RTI 40 1 6
RTS RTS R eturn fro m subroutine RTS
Operation: PC f , PC + 1 —» PC N Z C I D V

Implied RTS 60 1 6
SBC SBC S u b tra c t m em o ry fro m a c c u m u la to r w ith b o rro w SBC
Operation: A - M - C —> A N Z C I D V
Note: C = Borrow J / / - - /

Immediate SBC #O p er E9 2 2
Zero Page SBC Oper E5 2 3

Zero Page, X SBC Oper, X F5 2 4
Absolute SBC Oper ED 3 4
Absolute, X SBC Oper, X FD 3 4*
Absolute, Y SBC Oper, Y F9 3 4*
(Indirect, X) SBC (Oper, X) El 2 6
(Indirect), Y SBC (Oper), Y FI 2 5*
* A dd 1 w hen p age bou n d ary is crossed.

SEC SEC S et c a rry flag SEC
Operation: 1 —» C N Z C I E V
_ _ 1 _ _

Implied SEC 38 1 2
SED SED Set d ec im al mode SED
Operation: 1 —* D N X C V

Implied SED F8 1 2
SEI SEI S et in te rru p t disable status SEI
Operation: 1 —» I N Z C V

Implied SEI 78 1 2
STA STA S to re a c c u m u la to r in m e m o ry STA
Operation: A —» M N Z C V

Zero Page STA Oper 85 2 3

Zero Page, X STA Oper, X 95 2 4
Absolute STA Oper 8D 3 4
Absolute, X STA Oper, X 9D 3 5
Absolute, Y STA Oper, Y 99 3 5
(Indirect, X) STA (Oper, X) 81 2 6
(Indirect), Y STA (Oper), Y 91 2 6
176 COMMODORE 128
STX STX Store ind ex X in m e m o ry STX
Operation: X —> M N Z C I V

Zero Page STX Oper 86 2 3

Zero Page, Y STX Oper, Y 96 2 4
Absolute STX Oper 8E 3 4
STY STY Store ind ex Y in m em o ry STY
Operation: Y —> M N Z C V

Zero Page STY Oper 84 2 3

Zero Page, X STY Oper, X 94 2 4
Absolute STY Oper 8C 3 4
TAX T A X T ra n s fe r a c c u m u la to r to ind ex X TAX
Operation: A -> X N Z C I V

Implied TAX AA 1 2
TAY TAY T ra n s fe r a c c u m u la to r to ind ex Y TAY
Operation: A N E V

Implied TAY A8 1 2
TSX TSX Transfer stack pointer to index X T SX
Operation: S X r V

Implied TSX BA 1 2
TXA TXA Transfer index X to accumulator TXA
Operation: X N Z C I E V
y
Implied TXA 8A 1 2
TXS TXS Transfer index X to stack pointer TXS
Operation: X —» S N Z C I V

Implied TXS 9A 1 2
TYA TYA Transfer index Y to accumulator TYA
Operation: Y N Z C I D V
J J - -

Implied TYA 98 1 2
178 COMMODORE 128
INSTRUCTION ADDRESSING MODES AND

RELATED EXECUTION TIMES (in clock cycles)
H
U
W
os 2
O x >* 3
H W X z
< H w w u
-
< o o U3
H w H
U
H
U H
< £J3 Q > W a -
s a* Cn
a h j
w O O O —J < 2 ss oEfl
os c*
w w
N N <
y:
CQ 0,
5
a
-
W
os
gsw sz'—- ea
<
ADC . 2 3 4 4 4* 4* 6 5* .
AND . 3 4 4 4* 4* .
2 6 5* .
ASL 2 . 5 6 6 7 ,
BCC 2 **
BCS 2 **
BEQ 2 **
BIT 3 4 ,
BMI 2 **
BNE 2 **
BPL 2 **
BRK ,
BVC 2 **
BVS 2 **
CLC 2
CLD 2
CLI 2
CLV 2
CMP . 2 3 4 4 4* 4* , 5* .
CPX . 2 3 4
CPY . 2 3 4
DEC 5 6
DEX 2
DEY . 2
EOR . 2 3 4 4 4* 4* . 5* .
INC 5 6 6 7
INX , . . 2
INY . 2
JMP 3 .
JSR . . 6
LDA . 2 3 4 4 4* 4* 5*
LDX . 2 3 4 4*
LDY . 2 3 4 4 4*
LSR 2 5 6 6 7
NOP . . . 2
ORA . 2 3 4 4 4* 4* 5*
PHA , . 3
PHP . . . , . 3
H
U
w
&
cc
o X >
X >- z
H X >
< w w w" W
- H o e> a w
H
w
H
w
H
w H p w
H
< < < Q > U u
a. CL. a. P 3 W w w S3
s Q
w o o
J J H fie; os -J
o
P
U S
O
CS2 c*
O
OS C/5
O
c/3
o
C/3
J
a.
<
-J 3 o C/j
u w w w ca SQ CQ W Z z CO
< N N SI < c C OS w <
PLA . 4 .
PLP 4 .
ROL 2 . 5 6 . 6 7 .
ROR 2 . 5 6 . 6 7 .
RTI . . . . , . 6 .
RTS . 6 .
SBC 2 3 4 . 4 4* 4* 6 5* .
SEC 2 .
SED . . 2 .
SEI . . . . 2 .
STA . 3 4 . 4 5 6 6 .
STX . 3 . 4 4 .
STY . 3 4 . 4 .
TAX 2 .
TAY . # 2 .
TSX . , . 2 .
TXA , , 2 .
TXS . . . 2 .
TYA , # . 2 .
* Add one cycle if indexing across page boundary

** Add one cycle if branch is taken. Add one additional if branching
operation crosses page boundary
A clock cycle is the speed at which the processor operates as determined by the
number of bytes transferred from one internal logic component to another. The 8502
operates at a default speed of 1 MHz, which is equivalent to 1,000,000 cycles per
second.
6________
HOW TO ENTER
MACHINE LANGUAGE
PROGRAMS INTO THE
COMMODORE 128
181
182 COMMODORE 128
Now that you know about addressing modes, types of instructions and opcodes, you
need to know how to actually enter machine language instructions into the Commodore
128 memory. The C l 28 offers three methods of inputting instructions so that they may
be operated on by the microprocessor. You can enter machine language instructions by:
1. Using the built-in machine language monitor (available in C l 28 mode only).

2. PO K E ing the translated decimal opcode values into memory with a BASIC
program (C l28 and C64 modes).
3. Using an additional software program called an assembler.
All three methods have advantages and disadvantages. For instance, the built-in
machine language monitor is easy to use and allows you to program in machine
language without any additional aids such as an assembler. It makes merging BASIC
and machine language easy. In addition, you can save machine language programs as
binary files with the monitor SAVE command. Since you are already working in an
object code, there is no need to compile from source code into an object code, as is
necessary with an assembler.
Though these are powerful features, the monitor does not allow the use of symbolic
operand names or commented code. The monitor produces executable (object) code;
hence, no source files are produced. The resulting coded program contains actual
(absolute) address references, whereas an assembler source code file allows the use of
symbolic addresses and arguments as well as comments. When you display a machine
language program in the monitor, you do not have the luxury of comments or symbolic
address variables, so you really have to know what you are looking for when reading
other people’s code. On the other hand, an assembler source file must be compiled into
executable object code, then used often with an additional program called a loader. This
requires three steps, whereas the monitor’s machine language is ready to run as soon as
you finish writing the program.
The second method, POKEing translated decimal opcode data into memory with a
BASIC program, is an alternative usually implemented only when the first two options
are not available. This is the case if you have no assembler and are writing a machine
language routine in Commodore 64 mode, which does not make the built-in monitor
available to you. However, it is sometimes handy to POKE small routines from
BASIC if the application program you are writing is more suited for BASIC and you
need the speed of machine language for only a small portion of the program (though for
the most part, this method is tedious, bulky and time-consuming). Use it only if you
have no alternative, since once it is POKED into memory, you cannot display a listing
of the machine language routine as in the monitor or the assembler.
This chapter explains how to enter machine language programs in the first two
methods described above. The third method, using an assembler, requires an additional
software package similar to the Commodore 64 Assembler Development System. For
specific details on how to enter machine language programs with the assembler, refer to
the manual that is packed with the assembler software package you buy.
HO W TO ENTER MACHINE LANGUAGE PROGRAMS IN TO THE COMMODORE 128 183
ENTERING MACHINE LANGUAGE

INSTRUCTIONS IN THE M O NITO R
Begin entering machine language instructions by entering the monitor from BASIC with
the following command:
MONITOR RETURN
The Commodore 128 responds with the following display:
MONITOR
PC SR AC XR YR SP
; FB000 00 00 00 00 F8
These values indicate the contents of the microprocessor registers upon entering
the monitor. The abbreviations and definitions of the register names are as follows:
PC— Program Counter Marks the address of the current machine language
instruction
SR— Status Register Flags that alert the microprocessor of certain conditions
AC— Accumulator Register for all mathematical operations
XR— X Index Register Used for effective address modification
YR— Y Index Register Same as X register
SP— Stack Pointer Indicates the address of the first available memory
location on the stack
Now you can begin to enter machine language instructions. The ASSEMBLE
command within the monitor enters the instructions into the specified memory location.
To enter instructions, follow the format of this example:
A 01800 LDA # $ 0 0
Make sure to leave at least one space between each of the fields. Here’s what each
part of the instruction means:
<A ssem ble> <Address in memory where opcode is stored> < O pcode> <O perand>
The A stands for ASSEMBLE an opcode. The second part (field) is the address
where the opcode in the instruction is placed in the Commodore 128 memory. Notice
the 5-digit hexadecimal number specifying the address. The leftmost digit (0-F) speci
fies the configuration of the Commodore 128 memory layout. This is the same as the
BANK command in BASIC.
Once the entire machine language program is entered, reference the address that is
contained in the first instruction you entered to start execution of the program. Execute
the program with the GO command in the monitor, or exit the monitor with the X
(EXIT) command and issue the SYS command from BASIC. If you SYS to the start of
the program, you must use the decimal equivalent of the hexadecimal address, which
184 COMMODORE 128
appears in the first instruction you entered. You must have an RTS instruction at the end
of the routine if you want to return to BASIC. Often, the Kernal must be resident in the
current configuration in context in order to obtain results.
The opcode is the 8502 instruction that is carried out by the microprocessor when
your program is running. See the 8502 Instruction Set Table in Chapter 5 for allowable
instructions.
The operand is the address or value that is acted upon by the opcode in the
instruction. If the operand field is preceded by a pound sign ( # ), the opcode will act
upon a constant value. If no pound sign is specified, the microprocessor assumes the
opcode will act upon an address.
Remember to separate each field in the instruction with at least one space. If you
don’t, the computer indicates that an error has occurred by displaying a question mark at
the end of the instruction.
Once a routine is displayed on the screen, the monitor allows shortcuts in entering
instructions. To display a listing of a machine language program, issue the DISASSEM
BLE command as follows:
D 04000 04010 RETURN
The “ D ” stands for disassemble. The first number (04000) specifies the starting
memory location in which you want the contents displayed. The second number
specifies the end address in which to display.
Now for the shortcut. Since the address where the opcodes are stored is already on
the screen, you can simply move the cursor to the opcode field, type over the exist
ing opcode and operand on the screen, erase any unwanted characters and press
RETURN . The computer registers the instruction in memory by displaying the
hexadecimal values for the opcode and operand directly to the left of the opcode
mnemonic you just entered. This is a faster and easier way of entering machine-
language routines, rather than typing the ASSEMBLE command and the address each
time you enter an instruction.
EXECUTING (RUNNING)
YOUR MACHINE-LANGUAGE PROGRAM
Once you have finished entering your machine language routine, you may execute it in
three different ways. Within the monitor, issue the GO or JUMP to Subroutine com
mand as follows:
G FI 800 (JMP)
J F I 800 (JSR)
The G stands for G O , or go to the start address of the machine language program
in memory, and begin executing it at the specified address. The value following the
letter G refers to the start address of your routine. The J stands for Jump to Subrou
tine, similar to the JSR mnemonic in machine language.
The third way to invoke a machine language routine is to exit the monitor by
H O W T O ENTER MACHINE LANGUAGE PROGRAMS IN TO THE COMMODORE 128 185
pressing the X key and R E T U R N . This places you back within the control of the
BASIC language. Next, issue the SYS command and reference the starting address in
decimal as follows:
BANK 15
SYS 6144
This SYS command is the same as the GO command (G F I 800) example above.
The BANK 15 command and the leading F in the 5-digit hexadecimal number F I 800
specify memory configuration 15. The Kernal, BASIC and other ROM code are
resident in this configuration. The only difference is that it executes the machine
language routine from BASIC, instead of within the monitor.
The machine language routine given below clears the text screen. Starting at
location 1024 ($0400), the value 32 ($20) is stored in each screen location. The
character string value 32 is the space character, which blanks out each character position
on the screen. When finished, an RTS instruction returns control to BASIC. H ere’s the
main BASIC program and the machine language screen-clear subroutine as it appears in
the machine language monitor.
10 FOR 1 = 1 TO 25
20 P R I N T " F I L L THE SCREEN WITH CHARACTERS"
30 NEXT
40 PR IN T: PRINT
50 PR IN T" NO W CALL THE MACHINE LANGUAGE"
60 P R I N T " R O UT IN E TO CLEAR THE SC REE N"
70 SLEEP 5
80 SYS D E C ( " 1 8 0 0 " )
90 P R I N T " T H E S C R E E N I S NOW C L E A R E D "
READY.
MONITOR
PC SR AC XR YR SP
; FBOOO 00 00 00 00 F8
01800 A2 00 LDX #$00

01802 A9 20 LDA #$20
01804 9D 00 04 STA $ 0 4 0 0 ,X
01807 9D 00 05 STA $ 0 5 0 0 ,X
0180A 9D 00 06 STA $0600,X
0180D 9D E7 06 STA $06E7,X
01810 E8 INX
01811 DO FI BNE $1804
01813 60 RTS
In this sample program, the SYS command executes the subroutine to clear the
text screen. Once the text screen is cleared, control of the microprocessor is returned to
BASIC by the RTS instruction, and the READY prompt is displayed.
186 COMMODORE 128
M ACHINE LANGUAGE
M O N ITO R COMMANDS
The C128’s built-in machine language monitor has several additional commands that
manipulate your machine language routines once they are entered into memory. Figure
6-1 is a summary of all the commands available to you in the machine language
MONITOR.
KEYW ORD FUNCTION FORMAT
ASSEMBLE Assembles a line of 8502 code A < start ad d ress> < o p co d e>
[operand]
COMPARE Compares two sections o f mem C <start address> < en d address>
ory and reports differences < n ew start address>
DISASSEMBLE Disassembles a line or lines of 8502 D [<start address> < end address> ]
code
FILL Fills a range of memory with spec F <start address> < end address>
ified byte 
GO Starts execution at the specified G [address]
address
HUNT Hunts through memory within a H <start address> < en d address>
specified range for all occurrences [ . . .]
of a set of bytes H <start address> < en d address>
< ascii string>
GOSUB Jumps to the subroutine J [address]
LOAD Loads a file from tape or disk L “ < file n a m e > ” [,< d ev ice # >
[,< lo a d address>]]
MEMORY Displays the hexadecimal values M [< start address>
of memory locations [< en d address>]]
REGISTERS Displays the 8502 registers R
SAVE Saves to tape or disk S “ < filen a in e > ” ,< d ev ice # > ,
< start address> c la s t address
+ 1>
TRANSFER Transfers code from one section T <start address> < en d address>
of memory to another <n ew start address>
VERIFY Compares memory with tape or V “ < file n a n ie> ” [,< d evice # > [ ,
disk c lo a d address>]]
EXIT Exits Commodore 128 MONITOR X
(period) Assembles a line of 8502 code .
(greater than) Modifies memory > [address]

(semicolon) Modifies 8502 register displays
H O W TO ENTER MACHINE LANGUAGE PROGRAMS IN TO THE COMMODORE 128 187
KEYW ORD FUNCTION FORMAT
(at sign) Displays disk status, sends disk @

command, displays directory
disk status © [device # ]
disk command © [device # ],< co m m a n d
string>]
disk catalog @ [device # ],$ [[< d r iv e > :< file
spec>]]
NOTES < > enclose required parameters

[] enclose optional parameters
Figure 6 -1 . S u m m a ry o f C o m m o d o re 128 M o n ito r Com m ands
N O TE: 5-Digit Addresses

The Commodore 128 displays 5-digit hexadecimal addresses within the
machine language monitor. Normally, a hexadecimal number is only four
digits, representing the allowable address range. The extra left-most (high
order) digit specifies the BANK configuration (at the time the given
command is executed) according to the following memory configuration
table:
0— RAM 0 only 8 — EXTROM , RAM 0, I/O

1— RAM 1 only 9— EXT ROM , RAM 1, I/O
2— RAM 2 only A— EXT ROM , RAM 2, I/O
3— RAM 3 only B— EXT ROM , RAM 3, I/O
4— INT ROM , RAM 0, I/O C— KERNAL + INT (lo). RAM 0, I/O
5— INT ROM , RAM 1, I/O D— KERNAL + EXT (lo), RAM 1, I/O
6 — INT ROM , RAM 2, I/O E— KERNAL + BASIC, RAM 0, CHARROM
7— INT ROM , RAM 3, I/O F— KERNAL + BASIC, RAM 0, I/O
SUMMARY OF M O N ITO R
FIELD DESCRIPTORS
The following designators precede monitor data fields (e.g., memory dumps). When
encountered as a command, these designators instruct the monitor to alter memory or
register contents using the given data.
 precedes lines of disassembled code.

> < rig h t angle> precedes lines of a memory dump.
; < sem icolon> precedes line of a register dump.
The following designators precede number fields (e.g., address) and specify the radix
(number base) of the value. Entered as commands, these designators instruct the monitor
simply to display the given value in each of the four radices.
188 COMMODORE 128
< n u ll> (default) precedes hexadecimal values.

$ < d o llar> precedes hexadecimal (base-16) values.
+ precedes decimal (base-10) values.
& < am persand> precedes octal (base-8) values.
% precedes binary (base-2) values.
The following characters are used by the monitor as field delimiters or line terminators
(unless encountered within an ASCII string).
< sp ac e > delimiter— separates two fields.

< co m m a> delimiter— separates two fields.
: < c o lo n > terminator— logical end of line.
? < q u estio n > terminator—logical end of line.
M O N ITO R C O M M A N D DESCRIPTIONS
The following are descriptions of each of the C l28 Machine Language Monitor commands.
COMMAND:
PURPOSE: Enter a line of assembly code.
SYNTAX: A < ad d ress> <opcode m nem onic> < operand>
< ad d ress> A number indicating the location in memory to
place the opcode. (See 5-digit address note on
previous page.)
< o p co d e> A standard MOS technology assembly language
mnemonic, e.g., LDA, STX, ROR.
< o p eran d > The operand, when required, can be any of the
legal addresses or constants.
A RETURN is used to indicate the end of the assembly line. If there are
any errors on the line, a question mark is displayed to indicate an error, and the
cursor moves to the next line. The screen editor can be used to correct the error(s) on
that line.
E X A M P LE :
.A 01200 LDX #$00

.A 01202
N O TE: A period (.) is equal to the ASSEMBLE command.
E X A M P LE :
.02000 LDA #$23

HO W T O ENTER MACHINE LANGUAGE PROGRAMS IN TO THE COMMODORE 128 189
COMMAND:
PURPOSE: Compare two areas of memory.
SYNTAX: C <address 1> <address 2 > <address 3 >
<address 1> A number indicating the start address of the area
of memory to compare against.
<address 2 > A number indicating the end address of the area
o f memory to compare against.
<address 3 > A number indicating the start address of the other
area of memory to compare with. Addresses that
do not agree are printed on the screen.
COMMAND: D
PURPOSE: Disassemble machine code into assembly language mnemonics and
operands.
SYNTAX: D [< address> ] |< address 2 > ]
< ad d ress> A number setting the address to start the dis
assembly.
< address 2 > An optional ending address of code to be dis
assembled.
The format of the disassembly differs slightly from the input format of an assembly. The
difference is that the first character of a disassembly is a period rather than an A (for
readability), and the hexadecimal value of the op-code is listed as well.
A disassembly listing can be modified using the screen editor. Make any changes
to the mnemonic or operand on the screen, then hit the carriage return. This enters the
line and calls the assembler for further modifications.
A disassembly can be paged. Typing a D R E T U R N causes the next page
of disassembly to be displayed.
E X A M P LE :
D3000 3003
.03000 A9 00 LDA # $ 0 0
.03002 FF ???
.03003 DO 2B BNE $3030
COMMAND:
PURPOSE: Fill a range of locations with a specified byte.
SYNTAX: F <address 1> <address 2 > 
<address 1> The first location to fill with the .
<address 2 > The last location to fill with the .
<byte value> A 1- or 2-digit hexadecimal number to be written.
This command is useful for initializing data structures or any other RAM area.
190 COMMODORE 128
E X A M P LE :
F0400 0518 EA
Fill memory locations from $0400 to $0518 with $EA (a NOP instruction).
COMMAND: G
PURPOSE: Begin execution of a program at a specified address.
SYNTAX: G [< address> ]
< ad d ress> An address where execution is to start. When
address is left out, execution begins at the current
PC. (The current PC can be viewed using the R
command.)
The GO command restores all registers (displayable by using the R command) and
begins execution at the specified starting address. Caution is recommended in using the
GO command. To return to the Commodore 128 MONITOR after executing a machine
language program, use the BRK instruction at the end of the program.
E X A M P LE :
G 140C
Execution begins at location $140C in configuration (BANK)0. Certain applica
tions may require that Kernal and/or I/O be present when execution begins.
Precede the four-digit hexadecimal number with the hex configuration number
which contains those appropriate portions of memory.)
COMMAND: H
PURPOSE: Hunt through memory within a specified range for all occurrences of a
set of bytes.
SYNTAX: H <address 1> <address 2 > < d ata >
<address 1> Beginning address of hunt procedure.
<address 2> Ending address of hunt procedure.
< d a ta > Data set to search for data may be hexadecimal
for an ASCII string.
E X A M P LE :
H A000 A101 A9
Search for data $A9 from A000 to A 101.
H2000 9800 ‘CASH’
Search for the alpha string “ CASH” .
COMMAND: J
PURPOSE: Jump to a machine language subroutine.
SYNTAX: J < ad d ress>
The JUMP to SUBROUTINE command directs program control to the machine language
subroutine located at the specified address. This command saves the return address as
does the 8502 instruction JSR (Jump to Subroutine). In other words, the JUMP
command is a two-way instruction, where the application gains control of the computer.
Only after the subroutine encounters an RTS instruction does the machine language
monitor regain control.
E X A M P LE :
J 2000
Jump to the subroutine starting at $2000 in configuration 0.
COMMAND: L
PURPOSE: Load a file from cassette or disk.
SYNTAX: L < “ file nam e” > [,< d e v ic e > [,a lt load address]]
< “ file nam e” > Any legal Commodore 128 file name.
< d ev ic e > A number indicating the device to load from. 1 is
cassette. 8 is disk (or 9, A, etc.).
[alt load address] Option to load a file to a specified address.
The LOAD command causes a file to be loaded into memory. The starting address is
contained in the first two bytes of the disk file (a program file). In other words, the
LOAD command always loads a file into the same place it was saved from. This is very
important in machine language work, since few programs are completely relocatable.
The file is loaded into memory until the end of file (EOF) is found.
E X A M P LE :
L “ PROGRAM ” ,8 Loads the file name PROGRAM from the disk.
COMMAND: M
PURPOSE:To display memory as a hexadecimal and ASCII dump within the
specified address range.
SYNTAX: M [<address 1>] [<address 2> ]
<address 1> First address of memory dump. Optional. If omit
ted, one page is displayed. The first digit is the
bank number to be displayed, the next four digits
are the first address to be displayed.
<address 2 > Last address of memory dump. Optional. If omit
ted, one page is displayed. The first digit is the
bank number to be displayed; the next four digits
are the ending address to be displayed.
Memory is displayed in the following format:
>1A 048 41 42 43 44 45 46 47 48:ABCDEFGH

192 COMMODORE 128
Memory contents may be edited using the screen editor. Move the cursor to the data to be
modified, type the desired correction and hit R E T U R N . If a syntax error
or an attempt to modify ROM has occurred, an error flag (?) is displayed. An
ASCII dump of the data is displayed in reverse (to contrast with other data displayed on
the screen) to the right of the hex data. When a character is not printable, it is displayed
as a reverse period. As with the disassembly command, paging down is accom
plished by typing M and R E T U R N
E X A M P LE :
M 21 COO
>21 COO 41 4A 4B 4C 4D 4E 4F 50 :AJKLMNOP
N O TE: The above display is produced by the 40-column editor.
COMMAND: R
PURPOSE: Show important 8502 registers. The status register, the program counter,
the accumulator, the X and Y index registers and the stack pointer are
displayed. The data in these registers is copied into the microprocessor
registers when a “ G ” or “ J ” command is issued.
SYNTAX: R
EX A M P LE :
R
PC SR AC XR YR SP
; 01002 01 02 03 04 F6
N O TE: ; (semicolon) can be used to modify register displays in the same

fashion as > can be used to modify memory registers.
COMMAND:
PURPOSE: Save the contents of memory onto tape or disk.
SYNTAX: S < “ filenam e” > ,< d e v ic e > ,< a d d ress 1 > , <address 2 >
< “ filename ” > Any legal Commodore 128 filename. To save
the data, the file name must be enclosed in dou
ble quotes. Single quotes cannot be used.
< d ev ice> A number indicating on which device the file is
to be placed. Cassette is 01; disk is 08, 09, etc.
<address 1> Starting address of memory to be saved.
<address 2 > Ending address of memory to be saved + 1. All
data up to, but not including, the byte of data at
this address is saved.
HO W T O ENTER MACHINE LANGUAGE PROGRAMS IN T O THE COMMODORE 128 193
The file created by this command is a program file. The first two bytes contain the
starting address <address 1> of the data. The file may be recalled, using the L
command.
E X A M P LE :
S “ GAM E” ,8,0400,0C00
Saves memory from $0400 to $0BFF onto disk.
COMMAND: T
PURPOSE: Transfer segments of memory from one memory area to another.
SYNTAX: T <address 1> <address 2 > <address 3 >
<address 1> Starting address of data to be moved.
< address 2 > Ending address of data to be moved.
<address 3 > Starting address of new location where data will
be moved.
Data can be moved from low memory to high memory and vice versa. Additional
memory segments of any length can be moved forward or backward. An automatic
“ com pare” is performed as each byte is transferred, and any differences are listed by
address.
E X A M P LE :
T1400 1600 1401

Shifts data from $1400 up to and including $1600 one byte higher in memory.
COMMAND: V
PURPOSE: Verify a file on cassette or disk with the memory contents.
SYNTAX: V < “ filename” > [,< d e v ic e > ][,a lt start address]
< “ filename” > Any legal Commodore 128 file name.
< d ev ic e > A number indicating which device the file is on.
Cassette is 01; disk is 08, 09, etc.
[alt start address] Option to start vertification at this address.
The VERIFY command compares a file to memory contents. If an error is found, the
words VERIFY ERROR are displayed; if the file is successfully verified, the cursor
reappears without any message.
E X A M P LE :
V “ W ORKLOAD” ,08
194 COMMODORE 128
COMMAND: X
PURPOSE: Exit to BASIC.
SYNTAX: X
COMMAND: > (greater than)
PURPOSE:Can be used to assign values for one to eight memory locations at a time
(in 40-column mode; up to 16 in 80-column mode).
SYNTAX: > < ad d ress> < data byte 1> < data byte 2 . . . 8>
< ad d ress> First memory address to set.
< d ata byte 1> Data to be put at address.
< d ata byte 2 . . . 8> D ata to be placed in the successive memory
locations following the first address (optional)
with a space preceding each data byte.
COMMAND: @ (at sign)

PURPOSE: Can be used to send commands to the disk drive.
SYNTAX: @ [<device num ber>], < disk cmd string>
<device num ber> Device unit number (optional).
<diskcm d string>String command to disk.
N O TE: @ alone gives the status of the disk drive.
EXAM PLES:
@ checks disk status

00, OK, 00, 00
@ ,I initializes drive 8
@ ,$ displays disk directory on unit 8.
@ ,$0:F* display all files on Drive 0,unit 8 starting with the letter F.
As a further aid to programmers, the Kernal error message facility has been automati
cally enabled, while in the Monitor. This means the Kernal will display ‘I/O E R R O R # ’
and the error code, should there be any failed I/O attempt from the MONITOR. The
message facility is turned off when exiting the MONITOR.
M A NIPULATING T E X T W ITH IN
THE MACHINE LANGUAGE M O NITO R
Certain machine language application programs require the manipulation of strings of
characters. If you are using an assembler package, it contains provisions for handling
strings of characters. However, within the monitor, strings of characters must be placed
in memory, either (1) through modifying a memory dump using the screen editor, or (2)
by placing the ASCII values of the characters in memory locations within a program.
To modify a memory dump using the screen editor, issue the M EM O R Y com
mand with the address range in which you want to place the character string informa
tion. For example, suppose you want to place the word “ TEXT” in memory starting at
location $2000. First, enter the machine language monitor with the MONITOR com
mand. Next, issue the memory command containing the address $2000 as follows:
M 2000
The 128 responds with this display:
02000 FF 00 FF 00 FF 00 FF 00: iT.TT.Tr.Tr.
The entire screen is filled with the contents of the memory (dump) locations $2000
through $205F. For illustrative purposes, only one line of the memory dump is shown.
This line pertains to the address range $2000 through $2007. At the right of the screen is
an area that displays the corresponding ASCII character for each value within a memory
location in that line of the memory dump. The left character in the display area
corresponds to location $2000, the second character position to the right pertains to
address $2001, and so on. To place the word “ TEXT” in memory starting at location
$2000, move the cursor up to the first line of the memory dump, move the cursor right
to the memory address that pertains to address $2000, and place the ASCII character
string code for the letter T in this position. To do this, type over the characters that are
there and replace them with the hexadecimal equivalent of decimal 84 ($54) and press
RETURN . Notice that the letter T is now displayed at the right of the screen.
Refer to Appendix E, ASCII and CHR$ Codes, for a list of the Commodore ASCII
codes for each character available in the Commodore 128.
Now do the same procedure for the letters E, X and T. When you are through, the
word “ TEX T” is displayed in the display area. The first line of the memory dump now
looks like this:
02000 54 45 58 54 FF 00 FF 00: TEXT tt . tt.
Now the character string you wish to manipulate is in memory, starting at address
$2000. Your machine language routine can now act upon the characters of the word
“ TEX T” in order to display them on the screen. An efficient way of manipulating
entire words is to use the start address in memory where the text begins, in this case
$2000. Determine the length of the string, and use an index register as an offset to the
end of the word. See the section Raster Interrupt Split Screen Program with Horizontal
Scrolling in Chapter 8, for a working example of manipulating text. This chapter has
described the use of machine language. For additional information on machine language
topics, see Chapter 5, 7, 8, 9, 10, 11, and 13.
7
MIXING
MACHINE
LANGUAGE
AND B A S I C
198 COMMODORE 128
W H Y M IX B A S I C AND
M ACHINE LANGUAGE?
Certain application programs are better suited for a high-level language such as BASIC
rather than low-level machine language. In other cases, however, certain portions of a
program, such as displaying graphics, may require the speed of machine language while
the rest of the program lends itself to the use of BASIC. This is the main reason for
mixing BASIC programs with machine language subroutines. Another reason may be
the lack of an alternative in programming machine language. For example, in C64
mode, a machine language monitor is not ordinarily available to the user. In addition,
you may not have an assembler package, so the only alternative is to enter machine
language programs through the BASIC language. This method has disadvantages; it can
be tedious and time-consuming, and once the routine is entered into memory, you have
no way of listing it to the screen for editing. This method is recommended only if no
alternative is available.
ENTERING MACHINE LANGUAGE

SUBROUTINES THROUGH B A S I C
In Chapter 6, you saw an example of how to use the SYS command to go from BASIC
to a machine language routine that cleared the text screen. The SYS command invoked
the subroutine, cleared the screen, and returned to BASIC with the RTS instruction.
This example illustrated the SYS command within a program. You can also SYS to a
machine language subroutine outside a BASIC program, as follows:
SYS 8192
This example assumes you entered the machine language subroutine through the
monitor (and it is still in memory). What if you don’t have the monitor available to you,
as in C64 mode, and you want to mix a machine language subroutine with a BASIC
program?
The answer to this is to POKE decimal data that represents the hexadecimal
opcodes and operands into memory. To activate the subroutine, you SYS to it as you did
before. This method of entering machine language programs requires these steps:
1. Write your machine language program on a piece of paper.

2. Translate the hexadecimal op-code values into decimal. Some instructions
require 3 bytes of memory, while others only use 1 or 2 bytes of memory. For
a list of hexadecimal opcodes for the 8502 machine language instruction set,
see the 8502 Instruction and Addressing Table, in Chapter 5.
3. Enter the decimal equivalents of the opcodes into DATA statements in such a
way that a 2-byte instruction, for example, is entered as follows:
M IXIN G MACHINE LANGUAGE A N D BASIC 199
1000 DATA 162,0: REM = LDX # $ 0 0 = $A2, $00

The hexadecimal number $A2 represents the 8502 instruction for LDX, which
equals 162 in decimal. The 0 (zero) represents the operand 0, which is
loaded into the X register in the instruction. In hex, 0 is the same as in
decimal, so the second byte of the instruction is the operand value 0. The
hexadecimal opcodes are translated into strings of binary digits (bits), so the
microprocessor can interpret and operate them. This is the true machine
language of the computer at its lowest level.
4. Once you have translated all the opcodes and operands into decimal, you
must place them in memory. Accomplish this by READing a DATA value in
BASIC and POKEing it into memory in an appropriate address range. For
example, to enter the LDX # $ 0 0 instruction into memory, in C128 mode,
perform the following routine:
10 ALPHA = 8192
20 I = 0
30 DO
40 READ A
45 IF A = -999 THEN EXIT
50 POKE ALPHA + 1,A
60 I =1+1
70 LOOP
80 PRINT “ ALL DATA IS NOW IN M EM ORY”
1000 DATA 162,0,-999

For C64 mode use this routine:
10 ALPHA = 8192
20 FOR 1 = 0 TO 1
40 READ A
50 POKE ALPHA + I,A
60 NEXT
80 PRINT “ ALL DATA IS NOW IN M EM ORY”
1000 DATA 162,0

The first example, in C l 28 mode, READs all DATA and POKEs it into
memory, starting at location 8192 ($2000), until a data item is equal to -999.
When the data item equals -9 9 9 , EXIT the loop and print a message saying
all the data is read into memory at the specified locations. This DO . . .
L O O P allows you to enter as many data items as you please without the need
to know how many data items are in the data list, since it checks for the
terminator value -999. This programming approach makes it easy to modify
the number of data entries without having to change any of the code. This
program assumes that the bit map screen is not being used ($2000-$3FFF).
200 COMMODORE 128
The second example, in C64 mode, uses a FO R . . . NEXT loop to

enter the data. This requires you to know how many data items are in the data
list. You can add an IF . . . TH E N statement to check for a terminator value
like -9 9 9 , as in the first example. However, this method illustrates a different
way of accomplishing the same thing.
5. The final step in entering machine language subroutines through BASIC is
executing the subroutine with the SYS command. Add line 90 to your BASIC
routines above as follows:
90 SYS 8192
This command executes the machine language routine that you POKEd into
memory starting at location 8192.
Although the DATA statement in the example in Step 4 does not show it, all
machine language subroutines must end with an RTS instruction so you can return to
BASIC. The decimal code for an RTS machine language instruction is 96. Your last
decimal data item in the final data statement in your program must be 96, unless you use
a terminator like -999; then -999 will be your last decimal data item.
Figure 7-1 shows a step-by-step translation from the machine language screen-
clear routine as it appears in the monitor and a complete program that mixes the clear
screen routine with the BASIC program that POKEs in the data and executes the
machine language subroutine. It only operates in the 40-column (VIC) screen.
Symbolic
A ddress Hex O pcode Instruction Decimal Equivalent
1st Byte 2nd Byte 3rd Byte
(O pcode) (Operand)
02000 A2 00 LDX #$00 162 0
02002 A9 20 LDA #$20 169 32
02004 9D 00 04 STA $0400 ,X 157 0 4
02007 9D 00 05 STA $ 0 5 0 0 ,X 157 0 5
0200A 9D 00 06 STA $ 0 6 0 0 ,X 157 0 6
0200D 9D E7 06 STA $ 0 6E7,X 157 231 6
02010 E8 INX 232
02011 DO FI BNE $2004 208 241
02013 60 RTS 96
Figure 7 -1 . Step-by-step Translation i Decim al
To find the hexadecimal opcodes, refer to the 8502 Instruction and Address
ing Table in Chapter 5. Notice in Figure 7-1 that the hexadecimal opcodes are displayed
within the monitor, directly to the left of the symbolic instruction. These hexadecimal
numbers are the codes that you translate into decimal data items in a BASIC program.
Notice that the second byte in the BNE instruction is the value 241. In a branch
instruction, the operand is not an absolute address, but is instead an offset to the
instruction to which it will branch. In this case, the BNE instruction branches backward
to location $2004; therefore, it branches backward 15 locations in memory to the first
store (STA) instruction.
Y ou’re probably wondering how the code 241 tells the computer to branch
backward by 15. The number 241 (decimal) is the 2 ’s complement of the value 15.
When bit 7 is enabled, the microprocessor branches backward. The number 241
signifies to branch backward by 15 memory locations and execute the instruction in that
location. To find the root value of a 2 ’s complement number, do this:
1. Translate the value into binary: 241 = 1111 0001

2. Subtract 1: = 1111 0000 = 240
3. Convert each bit to its complement; in other words,
change each bit to the opposite value: = 0000 1111 = 15
To find the tw o’s complement of a number, perform steps 1 and 3 above;

then add 1 to the value instead of subtracting.
H ere’s the complete C128 mode program, including all the decimal DATA
equivalents of the instructions in the machine language, clear-screen subroutine above:
10 A L PH A = 8192
20 1=0
30 DO
40 : READ A
45 : I F A = - 9 9 9 THEN EX IT
5 0 : POKE A L P H A + I, A
60 : 1 = 1 + 1
7 0 LOOP
80 P R IN T "A L L DATA I S NOW IN MEMORY"
8 5 SL EE P 1
90 SYS 8 1 9 2
1 0 0 0 DATA 1 6 2 , 0 , 1 6 9 , 3 2 , 1 5 7 , 0 , 4 , 1 5 7 , 0 , 5 , 1 5 7 , 0 , 6 , 1 5 7 , 2 3 1 , 6
2 0 0 0 DATA 2 3 2 , 2 0 8 , 2 4 1 , 9 6 , - 9 9 9
202 COMMODORE 128
Here is the corresponding program in C64:
10 ALPHA=8192
2 0 FOR 1 = 0 TO 19
40 : READ A
50 : POKE A L P H A + I , A
6 0 NEXT
8 0 P R I N T " A L L D A T A I S NOW I N ME MO R Y "
8 5 FOR 1 = 1 TO 2 5 0 0 : N E X T
90 SYS 8 1 9 2
1 0 0 0 DATA 1 6 2 , 0 , 1 6 9 , 3 2 , 1 5 7 , 0 , 4 , 1 5 7 , 0 , 5 , 1 5 7 , 0 , 6 , 1 5 7 , 2 3 1 , 6
2 0 0 0 DATA 2 3 2 , 2 0 8 , 2 4 1 , 9 6
When you run this program, the computer READs the DATA, POKEs it into
memory, and executes the machine language, clear-screen subroutine with the SYS
command. After you RUN the program, enter the machine language monitor (assuming
you are currently in C l 28 mode) and disassemble the code in the range $2000 through
$2015 with this command:
D 2000 2015
Notice that the subroutine you POKEd in through BASIC is the same as the
subroutine that appears in Figure 7-1. The two different methods accomplish the same
goal— programming in 8502 machine language.
WHERE TO PLACE MACHINE

LANGUAGE PROGRAMS IN MEMORY
The Commodore 128 has 128K of RAM memory, divided into two 64K RAM banks.
Much of the 128K of RAM is overlaid by ROM, but not at the same time. The
Commodore 128 memory is layered, so RAM is beneath the overlaid ROM. The
designers of the Commodore 128 have managed to squeeze 28K of ROM and 128K of
RAM into 128K of address space. Only one bank is available or mapped in at a time,
since the highest address an 8-bit microprocessor can address is 65535 ($FFFF).
However, because the C l 28 is capable of banking RAM and ROM in and out so fast, it
may seem as though 128K is always available.
In the portions of memory shared by RAM and ROM, a read operation returns a
ROM data value and a write operation “ bleeds through” to the RAM beneath the
layered ROM. The data is stored in the RAM memory location. If the data in RAM
beneath the ROM is a program, the ROM on top must be switched out before the
program in RAM can be executed. The RAM and ROM layout in memory is all
regulated and controlled through the Configuration Register (CR) of the Memory-
Management Unit (MMU). For detailed information, refer to the sections on the
Registers of the Memory Management Unit (specifically, the discussion of the Configu
ration Register) in Chapter 13.
WHERE TO PLACE MACHINE LANGUAGE ROUTINES

IN C O N JU N C TIO N W ITH B A S I C
Within BASIC, the operating system takes care of the mapping in and out of ROM and
RAM. The C128 operating system provides sixteen (default) memory configurations.
They each contain different values in the Configuration Register; therefore, each has a
different configuration. This gives you sixteen different configurations of memory to
choose from. Figure 7-2 lists the sixteen default memory configurations available under
the control of the BASIC language.
BANK CONFIGURATION
0 RAM(O) only
1 R A M (l) only
2 RAM(2) only
3 RAM(3) only
6 Internal ROM, RAM(2), I/O
7 Internal ROM, RAM(3), I/O
10 External ROM , RAM (2), I/O
11 External ROM , RAM (3), I/O
14 Kernal and BASIC ROM, RAM(O), Character ROM
Figure 7 -2 . Bank C onfiguration T ab le
If you want to place a machine language subroutine in memory while the BASIC
language is running, put the subroutine in a bank that contains RAM, preferably bank 0
since this bank is composed entirely of RAM. If you place the machine language
subroutine in a bank other than 0, not all of the RAM is available for pro
grams, since ROM overlays some of the RAM. You must check the value of the
Configuration Register within that bank to see which addresses within these banks
contain ROM. Check the value of the Configuration Register within each of the sixteen
configurations and compare the value with the table in Figure 13-5 to see exactly where
ROM maps in.
Follow this procedure when calling machine language subroutines from BASIC:
204 COMMODORE 128
1. Place the subroutine, preferably in bank 0, either through the monitor or by

POKEing in the code through BASIC. If the Kernal, BASIC, and I/O are
required, execute your program from configuration (BASIC Bank) 15. If you
enter the subroutine through the monitor, place the routine in bank 0 by
placing the digit 0 before the 4-digit hexadecimal address where the instruc
tions are stored. If you are POKEing the codes in through BASIC (not the
preferred method), issue the BANK 0 command within your program, assum
ing you are placing the routine into BANK 0; then POKE in the decimal data
for the opcodes and operands. The recommended memory range to place
machine language routines in conjunction with BASIC is between $1300 and
$1BFF. The upper part of BASIC text is also available, provided your BASIC
program is small and does not overwrite your routine.
2. Now, to process the rest of your BASIC program, return to bank 15 (the
default bank) with this command:
BANK 15
3. Now call the subroutine with the SYS command. SYS to the start address
where the first machine language instruction of your program is stored in
memory. In this case, assume the subroutine starts at hex location $2000
(assuming the VIC bit map screen is not used) and enter:
SYS 8192
The RAM in configuration 0 in Figure 7 -2 is the same RAM that appears in configura
tions (BANKS) 4, 8, 12, 13, 14, and 15. For example, you can enter programs into
BANK 15, but you must make sure that no ROM overlays your program area.
N O T E : If you plan to return to BASIC, make sure your subroutine ends

with an RTS instruction.
WHERE T O PLACE MACHINE LANGUAGE

ROUTINES W HEN B A S I C IS DISABLED
When you are programming in machine language and you don’t require the services of
the BASIC ROM, you can disable BASIC by mapping out the BASIC ROMs. Do this
by placing certain values into the Configuration Register in your own machine language
routines as follows:
LDA # $ O E ;Set up the Configuration Register value
STA $D501 ;Write to Preconfiguration Register A
STA $FF01 ;Write to LCR A to change value of CR
You can use this sequence:
LDA # $ O E
STA $FF00
When you switch out BASIC, the sixteen default configurations no longer exist
via the BASIC command, so it becomes your responsibility to manage the memory
configurations by manipulating the Configuration Register in your application program.
Figure 13-5, on page 462, defines the values to place in the configuration register to
arrive at the different memory configurations.
When you switch out the BASIC ROMs, the address range where BASIC usually
resides ($4000 through $7FFF for BASIC low and $8000 through $BFFF for BASIC
high), is available for your machine language programs. Be careful when switching out
the Kernal, since the Kernal controls the entire operation of the C128, including routines
that seem transparent to the user (i.e., routines that you may take for granted).
At certain points within your machine language programs, you may need to
disable the I/O operation of the C l 28 temporarily. For instance, if you want to copy
portions of the character ROM into RAM when programming your own characters, you
must switch out the I/O registers $D000 through $DFFF of the C l28, transfer the
character data into RAM, and then switch the I/O back in.
See the section discussing the Configuration Register, in Chapter 13, for a full
explanation of how the C128 RAM and ROM memory is configured.
This chapter has described the use of BASIC and machine language together. For
material on using BASIC alone, see Chapters 2, 3 and 4. For material on using machine
language see Chapters 5, 6, 8, 9, 10, 11 and 13.
•»
•*
*>
«■
8
THE POWER
BEHIND
COMMODORE
GRAPHICS
208 COMMODORE 128
THE RELATIONSHIP BETWEEN

VIDEO BANKS, RAM BANKS
A N D MEMORY CONFIGURATIONS
Many of you are familiar with how the Commodore 64 manages memory. This section
explains how the Commodore 128 manages video memory, and how the video banks
relate to the currently selected memory configuration.
M A N A G IN G BANKED MEMORY
Banking is a process in which a section of memory is addressed by the microprocessor.
The memory is said to be banked in when it is available to the microprocessor in the
current memory configuration.
The Commodore 128 is programmable in its memory configuration. BASIC and
the Machine Language Monitor give you 16 pre-programmed default configurations of
memory (referred to in BASIC as banks). For the puiposes of this discussion, BASIC
banks are referred to simply as default memory configurations which are combinations
of ROM and RAM in various ranges of memory. The current configuration, whether in
BASIC or machine language, is determined by the value in the configuration register of
the C l 28 Memory Management Unit (MMU) chip.
The sixteen different configurations in BASIC and the Machine Language Monitor
require different values to be placed in the configuration register so that particular
combinations of ROM and RAM can be banked into memory simultaneously. For
example, the character ROM is only available in memory configuration 14 (Bank 14 in
BASIC; the fifth digit hexadecimal prefix “ E ” in the Machine Language Monitor),
since this configuration tells the C l 28 MMU to swap out the I/O registers between
$D000 and $DFFF, and replace them with the character ROM. To swap the I/O
capabilities back in, change to any configuration number that contains I/O. Figure 8-1
lists the sixteen default memory configurations available in BASIC and the Machine
Language Monitor. Information on programming the MMU is contained in Chapter 13,
The Commodore 128 Operating System.
TH E T W O 64K RAM BANKS

The Commodore 128 memory is composed of two RAM banks (labeled 0 and 1), each
having 64K of RAM, giving a total of 128K. The 8502 microprocessor address bus is 16
bits wide, allowing it to address 65536 (64K) memory locations at a time. Figure 8-2
illustrates the two separate 64K RAM banks.
Although only 64K can be accessed at one time, the MMU has provisions for
sharing up to 16K of common RAM between the two RAM banks. Common RAM is
discussed in Chapter 13.
The 8502 microprocessor and the VIC chip can each access a different 64K RAM
bank. The 8502 RAM bank is selected by the configuration register (bits 6 and 7) and
THE POWER BEHIND COMMODORE 128 GRAPHICS 209
BANK CONFIGURATION
0 RAM(O) only
1 R A M (l) only
10 External ROM , RAM (2), I/O (same as 8 )
14 Kernal and BASIC ROM , RAM(O), Character ROM
Figure 8 - i . C l 28 D e fa u lt M em o ry Configurations
-64K
■N
-64K
>
Figure 8 -2 . C l 28 64K RA M Banks
the VIC RAM bank is selected by the RAM configuration register (bits 6 and 7). This
also is covered in detail in Chapter 13.
The configuration determined by the configuration register can be composed of
RAM and ROM, where the ROM portion overlays the RAM layer underneath, as
illustrated in Figure 8-3.
210 COMMODORE 128
Write
Operation
(POKE, STA $ ------ ) READ Operation (PEEK, LDA $ - - - - )
ROM ROM
RAM
0 65535
memory
Figure 8 -3 . R O M O ve rla y
A read (PEEK) operation returns a ROM value, and a write (POKE) operation
bypasses the ROM and stores the value in the RAM underneath.
Many different combinations of memory can be constructed to comprise a 64K
configuration of accessible memory. Bits six and seven of the configuration register
specify which RAM bank lies beneath the ROM layers specified by bits zero through
five. The underlying RAM bank can be switched independently of any ROM layers on
top. For instance, you may switch from RAM Bank 0 to RAM Bank 1, while
maintaining the Kernal, BASIC and I/O.
I6K VIDEO BANKS

In C l 28 graphics programming, a video bank is a 16K block of memory that contains
the essential portions of memory controlling the C128 graphics system: screen and
character memory. These two types of memory, which are discussed in the following
section, must lie within the 16K range of memory referred to as a (VIC) video bank.
The VIC chip is capable of addressing 16K of memory at any one time, so all graphics
memory must be present in that 16K. Since the Commodore 128 microprocessor
addresses 64K at a time, there are four video banks in each 64K RAM bank, or a total
of eight video banks. (See Figure 8-4.)
Where you place the VIC video bank depends on your application program. The
Commodore 128 ROM operating system expects this bank in default video Bank 0, in
the bottom of RAM Bank 0. Screen and character memory may be located at different
positions within each 16K video bank, though in order to successfully program the VIC
chip, the current 16K bank must contain screen and character memory in their entirety.
Y ou’ll understand this after reading the next few pages.
3 video bank 3
2 video bank 2
1 video bank 1
0 video bank 0
\
3 video bank 3
2 video bank 2
;
1 video bank 1
0 video bank 0
Figure 8 -4 . V ideo Banks w ith in RA M Banks
The four video banks in each 64K RAM bank are set up in the memory ranges
specified in Figure 8-5:
BANK ADDRESS RANGE VALUE OF BITS 1 & 0 IN $DD00
BINARY DECIM AL
0 $0-$3FFF 11 = 3 (DEFAULT)
1 $4000-$7FFF 10 = 2
2 $8000-$BFFF 01 = 1
3 $C 000-$FFFF 00 = 0
Figure 8 -5 . V id e o Banks M em o ry Ranges
Each RAM bank (0 and 1) has this memory layout.

Bits 0 and 1 of location $DD00 select the video bank. To select a video bank in
BASIC, type this command:
P O K E 5 6 5 7 6 , (P E E K (5 6 5 7 6 ) A N D 2 5 2 ) O R X
Where X is the decimal value of bits 1 and 0 in Figure 8-5.

In machine language, run the following program segment to select video banks:
L D A $D D 0 0 ; load the accumulator with contents of $DD00
A N D # $ F C ; preserve the upper 6 bits of $DD00
O RA # $ X ; where X is the hex value of bits 1 and 0 from Figure 8-5 above
STA $ D D 0 0 ; Place the value in $DD00
212 COMMODORE 128
In the third instruction, replace X with the hexadecimal value of bits 1 and 0 in
Figure 8-5. The default value is $03, which selects video bank zero.
Whenever you change video banks, you must add $4000 to the address of your
starting screen memory (video matrix) and character memory (bit map in bit map mode)
for each bank above 0. To change to video Bank 1, add $4000 to your starting screen
and character address; for Bank 2 add $8000; for Bank 3 add $C000. You must always
add an offset of $4000 to the start of your screen and character memory for each video
bank that is greater than zero.
SUMMARY OF BANKING CONCEPT

The major features of the banking concept can be summarized as follows:
1. BASIC and the Machine Language Monitor have sixteen 64K memory con
figurations that give you sixteen different combinations of memory layouts.
The MMU chip, particularly the value in the configuration register, controls
most of the memory management in the Commodore 128. In order to PEEK
(read) from or POKE (write) to a particular portion of memory, you must
choose a BASIC or monitor configuration that contains the desired section of
memory. Figure 8-1 lists the sixteen default memory configurations available
in BASIC and the Machine Language Monitor.
2. The 128K of memory is divided into two 64K RAM banks. Only one bank is
addressable at a time by the microprocessor. RAM bank selection is con
trolled by the MMU configuration register (bits 6 and 7), which is part of the
C l 28 I/O memory. The VIC chip and 8502 microprocessor can each access a
different 64K RAM bank. Figure 8-2 illustrates the two separate and indepen
dent 64K RAM banks.
3. Each 64K RAM bank is divided into four 16K video segments. The screen
and character memory must both lie within the selected 16K video segment in
order to successfully display graphics and characters on the screen. For each
16K video bank higher than zero, remember to add $4000 (16384 decimal) to
the start address of screen and character memory. Figure 8-4 shows how four
16K video banks fit into each of the two 64K RAM banks.
Here’s how the banks fit together and operate within the Commodore 128. One
64K RAM bank is always mapped into memory. Within BASIC or the Machine
Language Monitor, sixteen different memory configurations are available in a 64K bank.
To change the configuration, issue the BASIC BANK command, or precede the four
digit hexadecimal address in the Machine Language Monitor with an additional hexadec
imal digit 0 through F. Outside of BASIC or the monitor, you can select other
configurations, by changing the value in the configuration register at location $FF00
(or $D500). See Chapter 13 for details.
Within the selected configuration, and part of the current 64K RAM bank, is a
16K range reserved for a video bank. The 16K video bank must encompass IK of screen
memory, and either 4K of character ROM or an 8K block of memory for the bit map
data. All these components must be present in order for graphics to operate.
In essence, the bank concept can be thought of in this way: The C l 28 has a 16K
(VIC) video bank within a selected memory configuration within a 64K RAM bank.
Figures 8-28 through 8-32 at the end of this chapter provides a graphics program
ming summary.
SH ADO W REGISTERS: INTERMEDIATE STORAGE

LO C A TIO N S USED BY THE C l28 SCREEN EDITOR
Users who are experienced in programming the Commodore 64 VIC chip will find that
most of the graphics operations of the Commodore 128 are performed in the same way
as the C64. The main difference between the Commodore 64 and the Commodore 128
graphics systems is the hardware implementation of split-screen modes.
C l 28 mode provides two types of split-screen displays:
1. Standard character mode and bit map mode

2. Standard character mode and multi-color bit map mode
Because the split-screens switch from one display mode to another at a given
time, the screen editor must be interrupt-driven. The interrupt indicates at what point
the mode is to be switched. At that point, the VIC chip is loaded with presetvalues
already contained in RAM. These preset values are known as shadow registers. Each
time an interrupt occurs, certain video-chip registers are cleared and refreshed with the
values in the shadow registers. These shadow registers add a variation in programming
the VIC chip compared to the way the Commodore 64 handles it.
The primary intermediate storage locations for VIC chip programming are.
NAM E IN D IR E C T LO C A TIO N ACTU AL LO CA TIO N D ESCRIPTIO N
GRAPHM BIT 7 - 216 ($00D8) BIT 4 - 53270 ($D016) M ulticolor Mode Bit
GRAPHM Bit 6 - 216 ($00D8) Split Screen Bit
GRAPHM BIT 5 - 216 ($00D8) BIT 5 - 53265 ($D011) Bit M ap Mode Bit
VM1* BITS 7-4 - 2604 <$0A2C) BITS 7-4 - 53272 ($D018) Video M atrix (screen memory) Pointer
VM1 BITS 3-0 - 2604 ($0A2C) BITS 3-0 - 53272 ($D018) C haracter Base Pointer
VM2** BITS 7-4 - 2605 ($0A2D) BITS 7-4 - 53272 ($D018) Video M atrix (screen memory) Pointer
VM2 BITS 3-0 - 2605 ($0A2D) BITS 3-0 - 53272 (SD018) Bit M ap Pointer
* VM1 appl i es onl y t o s t andar d a nd mul t i - col or char act er (text) modes.
* * V M2 appl i es o nl y t o st andar d and mul t i - col or bit ma p modes .

214 COMMODORE 128
You must store to and load from the indirect locations when accessing the above
features of the VIC (8564) chip. For example, in C64 mode, this is how you set up the
video matrix and bit map mode:
10 POKE 53272, 120: REM Select bit map @ 8192, video matrix @ 7168
20 POKE 53265, PEEK(53265) OR 32: REM Enter bit map mode
Line 10 sets the video matrix at 7168 ($1C00) and the bit map at 8192 ($2000).
Line 20 enables bit map mode.
Normally, you would perform this operation with the high-level, 7.0 BASIC
command:
GRAPHIC 1
The comparable way to accomplish this with POKE commands in C l 28 mode is
as follows:
10 POKE 2605,120
20 POKE 216,PEEK(216) OR 32
In C128 machine language, use these instructions:
LDA # $ 7 8 ; set bit map @ $2000

STA $0A2D; set video matrix @ 7 1 6 8
LDA $00D8
ORA # $ 2 0
STA $00D8; select bit map mode
Although these examples do more than just select bit map mode and set up the
video matrix and bit map pointer (such as wait on a video retrace when the raster is off
the visible coordinate plane), these examples give you an idea of how to perform these
programming steps.
As you can see, C l 28 mode requires a slight variation in programming the VIC
chip. You must keep this in mind when programming graphics in C128 mode. Usually,
the high-level BASIC 7.0 commands take care of these variations. However, if you are
programming in machine language, remember to address these indirect storage locations
and not the actual ones. If you store values directly to the actual registers, the value will
be cleared in a jiffy and no apparent action occurs.
DISABLING THE INTERRUPT DRIVEN

SCREEN EDITOR
You can disable the interrupt-driven C l 28 screen editor by storing the value 255
($FF) in location 216 ($00D8). The actual VIC registers are not affected, and you
can program the VIC chip the same way as the C64. This makes it unnecessary to
address the indirect shadow registers. In BASIC, enter:
POKE 216,255
In m ach in e la n g u ag e , y o u enter:
LDA # $ F F
STA $00D8
Since disabling the interrupt allows you to program the VIC chip in the same way
as the Commodore 64, you can store values directly to the actual registers. You do not
have to address the indirect storage locations for VIC chip programming. However, if
you don’t disable the interrupt, it is still active and your values will be cleared upon the
first occurrence of the raster interrupt.
Remember, you must either disable the interrupt or address the indirect storage
locations. Failure to do one or the other can cause serious problems in your program.
The 80-column chip indirect memory locations are discussed in Chapter 10,
Programming the 80-Column 8563 Chip. Certain other I/O functions require the use of
indirect locations also. These are covered in Chapter 12, Input/Output Guide.
THE COMMODORE 128

GRAPHICS SYSTEM
This section describes where the SCREEN, COLOR and CHARACTER memory com
ponents in the graphics system are located in character modes and bit map modes.
Screen and character memory are addressed and stored differently in the character
modes than in the bit-map modes. The split-screen modes use a section of both the
character screen storage and the bit map screen storage.
In graphics operations, the C l 28 can operate in either BASIC or machine language
in both C l 28 and C64 modes.
This section tells you where the graphics locations and screen color character
memory are stored under each graphic mode. The next section details the inner workings
of each graphic display mode including how color and data are assigned and how
screen, color and character memory are interpreted.
SCREEN MEMORY (RAM)

The location in which screen RAM is stored in memory and the way the data are stored
within it depends on the current graphics mode and operational mode of the C l 28.
C l 28 BASIC
In Commodore 128 BASIC, the character screen memory is located in the default address
range 1024 ($0400) through 2023 ($07E7). The text screen memory can be moved.
Remember, certain addresses use indirect memory locations to change the value of the
actual address. The shadow register for the pointer to the text screen memory is location
2604 ($0A2C). The actual location is 53272, but the screen editor uses a shadow since
the VIC screen is interrupt-driven. A direct poke to 53272 ($D018) is changed back to
its original value every sixtieth of a second. Here’s how to change the location of screen
memory in C128 BASIC:
216 COMMODORE 128
POKE, 2604 (PEEK(2604) AND 15) OR X
where X is a value in Figure 8-6.

If you move the screen memory, make sure that the screen and character memory
do not overlap. In addition, make sure to add an offset of $4000 to the start address of
screen and character memory for each bank above 0. Additional commands are required
to make the program work. Details follow in the discussion of each graphic mode, as
well as program examples.
LOCATION*
X BITS DECIMAL HEX
0 OOOOXXXX 0 $0000
16 0001XXXX 1024 $0400 (DEFAULT)

32 0010XXXX 2048 $0800
48 0011XXXX 3072 $0C00
64 0100XXXX 4096 $1000
80 0101XXXX 5120 $1400

96 0110XXXX 6144 $1800
112 0111XXXX 7168 $1C00
128 1000XXXX 8192 $2000
144 1001XXXX 9216 $2400

160 1010XXXX 10240 $2800
176 1011XXXX 11264 $2C00
192 1100XXXX 12288 $3000
208 1101XXXX 13312 $3400
224 1110XXXX 14336 $3800
240 1111XXXX 15360 $3C00
*Remember that the BANK ADDRESS offset of $4000 per

video bank m ust be added if changing to a higher video
bank above 0.
Figure ft-6. Screen M em o ry Locations
This register also controls where character memory is placed in memory. The
upper four bits control the screen, the lower four control character memory. The “ AND
15” in the POKE 2604 statement ensures that the lower nybble is not upset. (If it had
been, you would not see the correct character data.)
In Commodore 128 bit map mode (standard or multi-color), the default bit map
screen memory (video matrix) is located between 7168 ($1C00) and 8167 ($1FFF).
Screen memory is interpreted differently in bit map mode than in text mode. The video
matrix in bit map mode actually supplies color information to the bit map. This is
explained in detail in the Standard Bit Map Mode section elsewhere in this chapter. To
change the location of the bit map screen memory (video matrix), use this command:
POKE 2605, (PEEK(2605) AND 15) OR X
where X is a value in Figure 8-6. Location 2605 is also a shadow register for 53272, but
only for bit map mode. When you move the video matrix you must ensure that it does
not overlap the bit map (data). In addition, be sure to add an offset of $4000 to the start
address of the video matrix and the bit map for each video bank above zero.
C64 BASIC
In C64 mode, the text screen defaults to locations 1024 ($0400) through 2023
($07E7). In bit map mode, the video matrix (screen memory) also defaults to this range
though the screen memory is interpreted differently in either mode. Commodore 64
BASIC allows you to move the location of the video matrix to any one of the sixteen
locations specified in Figure 8-6. The upper four bits of location 53272 ($D018) control
the location of the screen memory. To change the location of screen memory, use the
following command:
POKE 53272, (PEEK(53272) AND 15) OR X
where X is equal to one of the values in Figure 8-6
NOTE: The following paragraph pertains to both C l28 and C64 modes.
Bits zero and one of location 56576 ($DD00) control which of the four video
banks is selected. The default bank is 0. If you change to another video bank (from 0 to
1, for example), then for each bank higher than bank zero, you must add an offset of
$4000 to the starting video matrix (screen memory) address in Figure 8-6. This yields
the actual address of the video matrix. For example, if you’re changing from bank 0 to
bank 1, add $4000. If you are going to bank 2, add $8000; if you are changing to bank
3, add $C000. Remember, this is true for both C128 and C64 modes.
M A C H IN E LANGUAGE
In machine language, use the commands listed under A in Figure 8-7 to move the C128
(VIC) text screen. Use the commands under B to move the C128 bit map screen memory
(video matrix). Use the commands under C to move the C64 text or bit map screen
memory (video matrix).
(A) (B) (C)

M OVE C64
M OVE C l28 MOVE C128 TEXT OR BJT MAP
TEXT SCREEN BIT M AP SCREEN SCREEN MEMORY
LDA $0A2C LDA $0A2D LDA $D018

AND # $ 0 F AND # $ 0 F AND # $ 0 F
O R A #$X ORA # $ X ORA # $ X
STA $0A2C STA $0A2D STA $D018
Figure 8—7. M oving Screen M em o ry in Machine Language

218 COMMODORE 128
In Figure 8-7, X is the hexadecimal equivalent of the decimal value X in the left
column in Figure 8-6. The second and third instructions in each example in Figure
8-7 make sure not to upset the lower four bits of location 53272 or its shadow registers,
2604 ($0A2C) and 2605 ($0A2D), since they control the character data for text and bit
map modes.
COLOR RAM
C l 28 BASIC
Color RAM within the Commodore 128 is always stationary in memory. It occupies the
address range 55296 ($D800) through 56295 (SDBE7). In standard character mode,
screen RAM and color RAM correspond to one another on a one-to-one basis. Location
1024 gets color data from 55296, 1025 gets color from 55297 and so on. Multi-color
character mode utilizes color RAM also, but in a different manner. Additional explana
tions and examples are provided in the Standard Character Mode section of this chapter.
COLOR RAM BANKING

In C l 28 mode, the LORAM and HIRAM signal lines allow the graphics system to
make use of an additional Color RAM bank, which is not available in C64 mode. This
allows fast and clean switching of colors for the character or multi-color bit map screen.
The LORAM signal line allows the 8502 microprocessor to access one color RAM
bank, while the HIRAM control line allows the VIC chip to access either Color RAM
bank independently of the microprocessor. Bit 0 of location 1 controls the LORAM
signal line. LORAM selects color RAM bank 0 or 1 as seen by the 8502 microprocessor
depending on the value of the bit. If the bit value is low, the color RAM bank 0 is
accessed by the 8502. If the value of the bit is high, the upper color RAM bank is
accessed by the 8502 microprocessor.
Bit 1 of location 1 controls the HIRAM signal line. HIRAM selects color RAM
bank 0 or 1 as seen by the VIC chip, depending on the value of the bit. If the bit value is
low, the color RAM bank 0 is accessed by the VIC chip. If the value of the bit is high,
the upper color RAM bank 1 is accessed by the VIC chip.
These control lines add flexibility to the already powerful C l 28 graphics system.
This allows you to change colors of the multi-color bit map or character screen on the fly,
without any time delay. It allows you to swap color RAM banks instantly.
In standard bit map mode, color information is obtained from the bit map
screen memory (the video matrix, $1C00 through $1FFF), not color memory. Bit
map mode interprets screen memory differently than character mode. Color RAM
is used in standard character mode, multi-color bit map mode, multi-color character
mode and the split-screen mode.
C64 BASIC
In standard character mode, color RAM is located in the same place as in C128 mode:
55296 ($D800) through 56295 ($DBE7).
In bit map mode, C64 BASIC receives color information from screen memory (the
video matrix) as does C l 28 mode, though the default location for screen memory is
1024 ($0400) through 2023 ($07E7).
In machine language or in BASIC, standard character mode color data always comes
from the same place. Color RAM is used for multi-color character mode. In
standard bit map mode, however, color data originates from screen memory, so wher
ever you place screen memory, the color data for the bit map comes from the specified
screen memory (video matrix) range. Multi-color bit map mode receives color from three
places: color RAM, screen memory and background color register 0. This is explained in
depth in the sections on the multi-color character and multi-color bit map modes.
CHARACTER MEMORY (ROM)

C l 28 BASIC-CHARACTER MODES
In standard character mode, character information is stored in the character ROM in the
memory range 53248 (SD000) through 57343 ($DFFF). In location 1 of the Com
modore 128 memory map, the CHAREN (CHARacter ENable) signal determines whether
the character set is available in any given video bank (0-3). Bit 2 of location 1 is the
CHAREN bit. If the CHAREN bit is high (1), the Commodore character set is not
available within the currently selected video bank in context. If the value of bit 2 in
location 1 is low, equal to zero, then the C l 28 character set is available in the currently
selected video bank. This is true in any of the four video banks in both 64K RAM
banks. This feature allows the Commodore 128 character set to be available in any video
bank at any time. To read the character ROM, enter BANK 14 either in BASIC or the
MONITOR, and read the ROM, starting at location 53248. This configuration switches
out I/O, and maps in character ROM in the range $D000 through $DFFF. Figure 8-8
shows how the character sets are stored in the character ROM:
ADDRESS
VIC*
BLOCK DECIM AL HEX IMAGE CONTENTS
0 53248 DOOO-DIFF 1000-1IFF Upper case characters

53760 D200-D3FF 1200-13FF Graphics characters
54272 D400-D5FF 1400-15FF Reversed upper case characters
54784 D600-D7FF 1600-17FF Reversed graphics characters
1 55296 D 800-D9FF 1800-19FF Lower case characters
55808 DA00-DBFF 1A00-1BFF Upper case & graphics characters
56320 DC00-DDFF 1C00-1DFF Reversed lower case characters
56832 DE00-DFFF 1E00-1FFF Reversed upper case & graphics
characters
* = in C 6 4 m o d e only
Figure 8 -8 . B reakdow n o f C h arac ter Set Storage in C h arac ter R O M

220 COMMODORE 128
The character memory is relocatable as is screen memory. To move standard

character memory in C128 BASIC, alter the lower four bits (nybble) of location 2604
($0A2C). Location 2604 is a shadow register for 53272 for the text screen memory
(upper four bits) and character memory (lower four bits). To move the standard
character memory use the following command:
POKE 2604, (PEEK(2604) AND 240) OR Z.
where Z is a value in Figure 8-9.
LOCATION OF C H A R A C T E R MEMORY
VALUE
O FZ BITS DECIM AL HEX
0 XXXX000X 0 $0000-$07FF
2 XXXX001X 2048 $0800-$0FFF
4 XXXX010X 4096 $1000-$17FF ROM IMAGE in BANK 0 & 2
(default)*
6 XXXX011X 6144 $1800-$1FF ROM IMAGE in BANK 0 & 2*
8 XXXX100X 8192 $2000-$27FF
10 XXXX101X 10240 $2800-$2FFF
12 XXXX110X 12288 $3000-$37FF
14 XXXX111X 14336 $3800-$3FFF
= in C64 mode only.
Figure 8 -9 . C h a ra c te r M em o ry Locations
As with the other graphic system components, character data behaves differently in
bit map mode than in text mode.
Remember, the upper nybble controls where the screen memory maps in, so make
sure not to upset those bits. The AND 240 in the POKE statement above takes care of
preserving the upper four bits.
In C l 28 mode the character sets are available in all video banks depending on the
value of CHAR ENable
NOTE: Remember to add an offset of $4000 to the start address of

character m em ory, for each bank above 0; i.e ., for bank 3 add
3*$4000 = $C000
C l 28 BASIC-BIT MAP MODES

In bit map mode, the character memory data, also referred to as the bit map, defaults to
the range 8192 ($2000) to 16191 ($3F3F). The bit patterns of these 8000 bytes tell the
computer which pixels to turn on. This block of memory “ maps out” the picture on the
screen, according to the data in this 8000-byte block. Since the standard bit map screen
is 320 X 200, 64000 pixels make up the screen image. Divide 64000 by 8 to arrive at
8000 bytes of memory for the bit map.
Besides the upper four bits for the video matrix, bit 3 of location 2605 ($0A2D) is
the only significant bit in bit map mode. Location 2605 is the indirect memory location
of 53272 for bit map mode only.
When you issue the G R A PH IC 1 command, bit 3 in 2605 is set. This specifies
the bit map (data) to start at location 8192 ($2000). W henever you enter bit map mode
with the GRAPHIC command, bit 3 is always set. Outside of BASIC, you can specify
the bit map to start at location $0000; therefore, the value of bit 3 is zero.
If the C l 28 is running under the control of C l28 BASIC, the bit map always starts
on a boundary of $2000 (since bit 3 is set) within a given video bank. In video bank
zero, the bit map starts at $2000. For banks 1, 2 and 3, the bit map begins at $6000,
$A000 and $E000, respectively, since you must add an offset of $4000 for each bank
number above zero. In machine language, however, bit 3 may have a value of zero or
one. Therefore the bit map may start at $0000 if bit 3 is zero, or at $2000 if bit 3 is one,
in each video bank. This means that the bit map has eight possible starting locations
(per 64K RAM bank) in machine language— two for each of the four video banks. In
C l28 BASIC, the bit map can only start at one of the four locations.
D on’t forget to add the mandatory $4000 offset for each video bank above 0. The
eight possible starting locations of a bit map in a machine language program are shown
in Figure 8-10.
This gives you the choice of eight starting locations in which to place your bit map
data. Though only one bit map can fit in each video bank, you must leave room for the
video matrix. Since the C128 has two RAM banks— each with four video banks per
RAM bank— you may have a total of eight bit maps in memory, one for each video
bank. Each video bank can fit only one bit map, because you need IK for screen RAM,
and 8K for the bit map since a video bank has a maximum of 16K. To access another bit
map, you must switch video banks. To access a bit map in the upper RAM bank (1),
you may have to switch RAM banks and video banks.
VIDEO BANK VALUE O F BIT 3 =

0 1
0 $0000 $2000
1 $4000 $6000
2 $8000 $AOOO
3 $C000 $E000
Figure 8-10. S tartin g Locations fo r B it M ap in M achine Language

222 COMMODORE 128
C64 BASIC-CHARACTER MODES

In standard character mode in C64 BASIC, the lower four bits of location 53272 control
where character memory is placed. As in C128 mode, the character ROM is actually
mapped into memory between 53248 ($D000) and 57343 ($DFFF). The ROM image
appears in RAM in the range 4096-8191 (in video bank 0) and 36864-40959 in bank 2,
since it must be accessible to the VIC chip in a 16K range in video banks 0 and 2. The
character sets are not accessible in video banks 1 and 3. This ROM imaging in RAM
applies only to character data as seen by the VIC chip. These memory ranges are still
usable for data and programs and have no effect on the contents of RAM as far as your
programs are concerned.
In C64 mode, the ROM image overlays the RAM underneath. A write operation
“ bleeds through” to the RAM underneath, while a read returns a ROM value depending
on which memory configuration is currently in context. Since the VIC chip accesses
16K at a time, the character set images must appear in the 16K which the VIC chip is
currently addressing in video banks 0 and 2. Remember, in C128 mode, the character
sets are available in all video banks according to the value of the CHAREN bit in
location 1.
You can change the location of character memory with the following command:
POKE 53272, (PEEK(53272) AND 240) OR Z
where Z is a decimal value in the table in Figure 8-9.

The breakdown of the character sets is the same as in the C l28 for the character
ROM (see Figure 8-8).
C64 BASIC BIT MAP MODE

In bit map mode, bit 3 of location 53272 specifies the start of the bit map either at
$0000 or $2000 depending whether the value of bit 3 is 0 or 1, respectively. Use the
following command:
POKE 53272 (PEEK (53272) AND 240) OR Z
where Z is zero if you want the bit map to start at $0000 in each bank, or Z = 8 if you
want to place the bit map starting at 8192 ($2000) in each video bank.
See Figure 8-10 for the arrangement of the bit map in each of the four 16K video
banks within the two RAM banks. If you switch video banks, don’t forget to add the
$4000 (hex) offset for each bank above 0. See the Character Memory section under
C l28 Bit Map Mode in the last section for more detail on the arrangement of bit maps in
memory.
There are three ways to select the placement of character memory, as shown in Figure
8-11. Example A places character memory using the shadow register $0A2C in place of
the actual $D018 register. Example B specifies the start of the bit map at $2000 (using
shadow register $0A2D). Example C specifies the start of the C64 bit map or character
memory.
LDA $02AC LDA $02AD LDA $D018

AND #$ F 0 AND # $ F 0 AND # $ F 0
ORA # $ Z ORA # $ 0 8 ORA # $ Z
STA $02AC STA $02AD STA $D018
Figure 8 -1 1 . Selected C h a ra c te r M e m o ry Location
In Figure 8-11, Z is a value in the table in Figure 8-9.
STANDARD CHARACTER MODE
H O W T O ENTER STANDARD
CHARACTER MODE
The C l 28 powers up in standard character mode. This mode displays characters on the
default screen. The character is displayed in a single color on a single color background.
This is the mode in which you write (enter) programs. When you press RUN/STOP and
RESTORE, the C l 28 defaults to the text screen.
Location 53265 (and its shadow register $00D8) determine whether the C l28 is
operating in standard character mode. If bit 5 is 0, as it is on power-up, the C128 is in
character mode; otherwise it is in bit map mode.
Location 53270 (and its shadow register $00D8) determine whether the characters
are standard (single color) or multi-color. Bit 4 of 53270, and the shadow bit, bit 7 of
$00D8, specify multi-color mode. If these bits are equal to zero, characters are standard;
otherwise they are multi-color. See the Multi-color Character Mode section for more
details on selecting multi-color character mode.
224 COMMODORE 128
SCREEN LO C A T IO N
In standard character mode, the screen memory defaults to the range 1024 ($0400) through
2023 ($07E7). This is relocatable. See the Screen Memory section in the preceding pages.
Since the screen is 40 columns by 25 lines, the text screen requires 1,000 memory
locations to store all of the screen information in memory. The final twenty-four memory
locations in screen memory do not store displayed data; they are used for other purposes.
Each column of every row you see on the screen has its own screen memory location.
The top-left screen location, referred to as HOME, is stored at address 1024 ($0400).
The second screen location marked by the cursor is 1025 ($0401), and so on. Although
the screen you see is constructed in rows and columns, the screen memory within the
computer is stored linearly, starting at 1024 ($0400) and ending at location 2023 ($07E7).
Figure 8-12 shows a screen memory map, so you can visualize how a screen memory
location corresponds to the location on the physical screen of your video monitor.
SCREEN M E M O R Y M A P
COLUMN
0 10 20 30
1063
I
1024------- -
1064
1104
1144
1184
1224
1264
1304
1344
1384
1424
1464
1504
1544
1584
1624
1664
1704
1744
1784
1824
1864
1904
1944
1984
t
2023
Figure 8 -1 2 . Screen M em o ry Map
H O W THE SCREEN MEMORY D A T A

IS INTERPRETED
This screen memory range stores whole characters only. The characters are not repre
sented as ASCII character string codes (CHR$). Instead, they are stored in memory as
screen codes as shown in Appendix D. The screen codes and character string codes are
different due to the way they are stored in the character ROM. Notice in Appendix D
that the screen code for an at-sign (@) is 0. The @ is numbered 0 because it is the first
character to be stored in the character ROM. The letter “ A ” is the second character
ROM; therefore its code is 1. The letter “ B ” is the third character in the character
ROM. etc. The screen code is actually an index from the starting location of the
character ROM, beginning with zero.
If you want to POKE a character directly into screen memory, use the screen code
rather than the ASCII character string (CHR$) code. The same holds true for the
machine language monitor. For example:
POKE 1024,1
places the letter A in the HOME position on the VIC screen. From the monitor, placing
the value 1 in location $0400 (decimal 1024) also displays the letter A in the HOME
position on the VIC screen.
COLOR DATA
In standard character mode, color information comes from color RAM, in the address
range 55296 ($D800) through 56295 ($DBE7). This memory determines the color of the
characters in each of the 1,000 screen locations. The background color of the screen is
determined by the background color register 0 which is location 53281.
The color RAM and the screen RAM locations correspond on a one-to-one basis.
Screen location 1024 pertains to color RAM location 55296; screen location 1025
corresponds to color location 55297, etc. Figure 8-13 is the color RAM memory map.
The map shows how color RAM corresponds to the locations in screen RAM and the
placement on your video display.
COLOR MEMORY MAP
COLUMN
0 20
55335
I
55296— -
55336
55376
55416
55456
55496
55536
55576
55616
55656
55696 10 o
55736
55776
55816
55856
55896
55936
55976
56016
56056
56096
56136
56176
56216
56256
t
Figure 8 - 13. C o lor M em o ry Map 56295
226 COMMODORE 128
H O W COLOR MEMORY IS
INTERPRETED
The contents of the color RAM locations contains the color codes 0-15. Each color
memory location may have a different color code. The lower four bits (nybble) of
COLOR RAM are significant. Figure 8-14 shows the COLOR RAM color codes:
0 Black 8 Orange
1 White 9 Brown
2 Red 10 Light Red
3 Cyan 11 Dark Gray
Figure 8 -1 4 . C o lor Codes— 40 Colum ns
Notice these color code values are one less than the color codes used by the
keyboard and BASIC. If you want to store a value directly into COLOR RAM, store the
values in the table above, not the color codes used by BASIC and the keyboard. For
example:
P O K E 5 5 2 9 6 ,1
colors the character in the HOME position white. From the monitor, place the value 1 in
location $D800, and the same results occur.
Remember, these color codes only control the color of the foreground character.
The background color is controlled by background color register 0 (53281). The pixels
that make up the character image are enabled by bits in character memory. If the bit is
enabled, the pixel in the foreground is turned on in the foreground color, and is
therefore controlled by color RAM. If the bits making up the character are turned off,
they default to the color in background color register 0. The combination of on and off
bits makes up the image of the character. The value of these bits determines whether the
color data comes from color RAM or background color register 0. You’ll learn more
about character patterns in the next few paragraphs.
CHARACTER MEMORY
In standard character mode, the C l 28 receives character data from the CHARACTER
ROM. The character ROM is stored in the range 53248 (SD000) through 57343
($DFFF). Since the VIC chip is capable of accessing 16K at a time, the C128 needs a
way to have the character ROM available in the 16K VIC range. In C128 mode, the
character ROM is available in any VIC bank in C128 mode, based on the value of
CHAREN. See the chapter set availability in the Character Memory section in the
beginning of this chapter.
In C64 mode the character ROM is available only in banks 0 and 2. This is
accomplished by having a ROM IMAGE of the character ROM (53248-57343) mapped
into memory in place of RAM, in the range 4096-8191 ($1000-$1FFF) in video BANK
0, and 36864-40959 ($9000-$9FFF) in video BANK 2. In banks 1 and 3, the character
ROM is not available to the VIC chip.
Notice that the range where the character ROM is actually stored (53248-57343)
is also occupied by the I/O registers but not at the same time. When the VIC chip
accesses the character ROM, the character ROM is switched into the currently selected
video bank as a ROM image (in C64 mode only). When the character ROM is not
needed, the I/O registers are available in the usual range. It is important to note the
ROM image applies only to the character data as seen by the VIC chip. The RAM loca
tions where the ROM image maps in are still usable for programs and data. The locations
where the VIC chip looks for the character data are relocatable. See the Character
Memory section elsewhere in this chapter for information on moving character memory.
H O W T O INTERPRET CHARACTER MEMORY

IN STANDARD CHARACTER MODE
Typically, a complete character set contains 256 characters. The C l28 contains two sets
of characters, for a total of 2 times 256 or 512 characters. In 40-column (VIC) output,
only one character set is available at a time. Upon power-up, the uppercase/graphics
character set is available through the keyboard. To access the second character set, press
the Commodore key ( O ) and shift key at the same time. The second character set is
composed of the upper- and lowercase/graphics characters.
In character ROM, each character requires eight bytes of storage to make up the
character pattern. Since 256*8 is equal to 2048 bytes or 2K, and since there are two
character sets, the C l 28 has a total of 4K of character ROM. Figure 8-15 shows where
each character set is stored in the character ROM.
ADDRESS
VIC-II
BLOCK DECIM AL HEX IM AGE CONTENTS
0 53248 D 000-D1FF 1000-1IFF Upper case characters

53760 D200-D3FF 1200-13FF Graphics characters
54272 D400-D5FF 1400-15FF Reversed upper case characters
54784 D 600-D7FF 1600-17FF Reversed graphics characters
1 55296 D800-D9FF 1800-19FF Lower case characters
55808 D A00-DBFF 1A00-1BFF Upper case & graphics characters
56320 DC00-D DFF 1C00-1DFF Reversed lower case characters
56832 DE00-DFFF 1E00-1FFF Reversed upper case & graphics
characters
Figure 8 -1 5 . Location of C h arac ter Sets in C h a ra c te r RO M

228 COMMODORE 128
Note that there is really 8K of character ROM— 4K for C64 mode and 4K for
C l 28 mode. The system automatically selects the appropriate character ROM for each
mode of operation.
The bit patterns stored in the character ROM have a direct relationship to the
pixels on the screen, where the character is displayed. In memory, each character
requires eight bytes of storage. On the screen, a character is made up of an 8 by 8 pixel
matrix. Think of a character as eight rows of eight pixels each. Each row of pixels
requires one byte of memory, so each pixel requires one bit.
Since a character is an 8 by 8 pixel matrix, each character requires a total of 64
bits or eight bytes. Within each byte, if a bit is equal to 1, the corresponding pixel in
that character position is turned on. If a bit in a character ROM byte is equal to 0, the
corresponding pixel within the character on that screen position is turned off. The
combination of on and off pixels creates the image of the characters on the screen.
Figure 8-16 demonstrates the correspondence between a character on the screen and the
way it is represented in the character ROM.
60 Byte 1 -> 0 0 1 1 1 1 0 0 = $D000

102 Byte 2 -> 0 1 1 0 0 1 1 0 = $D001
110 Byte 3 -» 0 1 1 0 1 1 1 0 = $D002
110 Byte 4 -> 0 1 1 0 1 1 1 0 = $D003
96 Byte 5 -» 0 1 1 0 0 0 0 0 = $D004
98 Byte 6 -> 0 1 1 0 0 0 1 0 = $D005
60 Byte 7 - 0 0 1 1 1 1 0 0 = $D006
0 Byte 8 -* 0 0 0 0 0 0 0 0 = $D007
Character on Character as
the Screen Represented
in Character ROM
Figure 8-1 6. Relationship of Screen C h a ra c te r to C h arac ter R O M .
In Figure 8-16, the first eight bytes of character ROM, ($D000-$D007) are equal
to 60,102,110,110,96,98,60 and 0. These decimal numbers are calculated from the
binary value of the eight bytes that pertain to each row of pixels in the character. For
each bit that is equal to one, raise two to the bit position (0-7). For example, the first
byte of character ROM ($D000) is equal to 60, which is calculated by raising two to the
following bit positions:
22 + 23 + 24 + 2s = 4 + 8 + 1 6 + 32 = 60
The bits that are set (on) correspond to pixels that are enabled on the screen in the
foreground color. Bits that are clear correspond to pixels that are disabled, which are
displayed in the background color, according to background color register 0 at location
53281.
The second byte (row of pixels) of the at-sign (@ ) character is equal to 102
(decimal) and is obtained by the following:
2 ' + 22 + 25 + 26 = 102
The last byte of the at-sign character is equal to zero, since no bits are set.
Therefore, each pixel on the screen is displayed in the background color. The values of
the binary digits on the right in Figure 8-16 are directly related to the image of the
character as it appears on the screen on the left in Figure 8-16.
ACCESSING CHARACTER ROM
C l 28 BASIC
To access character ROM in C l 28 BASIC, type and run the following program:
10 BA NK 1 4
20 FOR 1 = 5 3 2 4 8 TO 5 3 2 4 8 + 7 : P R I N T P E E K ( I ) ; : NEXT
30 BA NK 15
Enter Bank 14, the only BASIC bank where the character ROM is accessible.
Then print the PEEK value of the first eight bytes of the character ROM. When
finished, return to Bank 15.
M ACHINE LANGUAGE
To access character ROM in C l 28 Machine Language, type and run the following
program:
MONITOR
PC SR AC XR YR SP
; FB000 00 1DO 0 0 00 F6
. 01800 A9 01 LDA #$01

. 01802 8D 00 FF STA $FF00
. 01805 A2 00 LDX #$00
. 01807 BD 00 DO L D A $D000,X
. 0180A 9D 40 1 8 STA $1840,X
. 0180D E8 INX
. 0180E E0 07 CPX #$07
. 01810 DO F5 BNE $1807
. 01812 A9 00 LDA #$00
. 01814 8D 00 F F STA $FF0 0
. 01817 60 RTS
10 SYS 6144
20 FOR 1=6208 TO 6208 + 7 : P R I N T P E E K ( I ) ; : NEXT
These machine language and BASIC routines accomplish the same task as the preceding
four-line BASIC program. The first two machine language instructions switch in the
character ROM, and switch out I/O. The next six instructions transfer the first eight
230 COMMODORE 128
bytes of character ROM into locations 6208 ($1840) through 6215 ($1847). The last
three instructions switch out the character ROM, replace it with the I/O registers and
return from the machine language subroutine to BASIC.
The BASIC routine activates the machine language subroutine, then prints the
values that were temporarily stored in 6208 through 6215. See Chapter 6 , How to
Enter Machine Language Programs, for details on how to input machine language
instructions on the C l 28.
C64 BASIC
To access character ROM in C64 BASIC, enter and run the following program:
40 P O K E 5 6 3 3 4 , P E E K ( 5 6 3 3 4 ) AND 2 5 4
50 P O K E 1 , P E E K ( 1 ) AND 2 5 1
80 FOR 1 = 0 TO 7 : POKE 6 1 4 4 + 1 , P E E K ( 5 3 2 4 8 + 1 ) :NEXT
9 0 P O K E 1 , P E E K { 1 ) OR 4
1 0 5 P O K E 5 6 3 3 4 , P E E K ( 5 6 3 3 4 ) OR 1
1 3 0 FOR 1 = 6 1 4 4 TO 6 1 4 4 + 7 : P R I N T P E E K ( I ) ; :NEXT
Line 40 turns off the interrupt timer. Line 50 switches out I/O and replaces it with
character ROM. Line 80 transfers the first eight bytes of character ROM (53248-53255)
to 6144-6151. Line 90 switches out character ROM, and replaces it with the I/O
registers. Line 105 turns on the interrupt timer. Line 130 prints the first eight character
ROM values that were temporarily stored in 6144 through 6151.
You may need to transfer parts of the character ROM data into RAM if you are
creating your own character set, and you want the remainder to be from the C l 28
character set. This is covered in more detail later in the chapter. These methods of
looking at the character ROM demonstrate how the character ROM is accessed, what the
patterns of the characters look like and why you would want to access the character
ROM.
The next section explains how to program your own custom characters in C l 28
mode.
PROGRAMMABLE CHARACTERS
The Commodore 128 has a feature that allows you to redefine the character set into
custom characters of your own. In most cases, you'll want to redefine only a few
characters at most, while obtaining the rest of the character set from the Commodore
128 character ROM.
With programmable characters, you tell the C l28 to get character information
from RAM. Usually, characters are taken from the character ROM. If you only want
certain characters, you can choose the ones you want, copy the character patterns into
RAM and leave the rest in ROM. You cannot write to the character data in ROM;
however, the character data placed in RAM can be redefined.
The first step in programming your own characters is to define the image. In the
Standard Character Mode section, you saw how a character on the screen is stored in the
character ROM. Each character requires eight bytes of storage. Each byte corresponds to
a row of pixels on the visible screen within the 8 by 8 character matrix; therefore, eight
rows of pixels make up one character.
This section shows how to customize an uppercase cursive (script) character set
for the letters A through H. Figure 8-17 shows the design for the uppercase cursive
letter A. The grid in the figure demonstrates how the character appears on the screen
within the 8 by 8 pixel matrix. Each row of the grid determines which bits are on within
the character bit pattern, and, hence, which corresponding pixels are enabled on the
screen. The eight-bit binary strings to the right of the grid are the bit patterns as stored
in RAM. The numbers to the right of the binary strings are the decimal equivalents of
the binary bit patterns. This decimal value is the data you POKE into RAM in order to
display the character.
7 6 5 4 3 2 1 0
0 • • • = 0 0 0 0 1 1 1 0 = 14
1 • • = 00010001 = 17
2 • = 00100000 = 32
3 • 0 = 01000010 = 66
4 » • = 10 0 0 0 0 1 0 = 130
5 e • = 10 0 0 0 1 0 0 = 132
6 • • • = 10001010 = 138
7 * « • • = 01110001 = 123
Figure 8 -1 7 . Design fo r Cursive le tte r “ A ”
The following program creates and displays the upper-case cursive characters A
through H. Enter it into the computer and RUN it. You’ll see the letters A through H
change from uppercase block letters to uppercase cursive letters. When you press the
newly defined lettered keys, they are displayed in cursive form.
Line 10 selects the uppercase character set, the set being redefined. Line 20
protects the character set from being overwritten by the BASIC program and prepares a
location in RAM in which to place your character set. The end of user BASIC text and
the top of string storage is moved from 65280 to 12288 (decimal), which substantially
cuts down the size of BASIC programming space. The character set will be placed
beginning at location 12288, but it does not have to be located there. The character set
does have to be within the first 16K of memory unless another bank is selected. The
VIC chip can only access 16K at a time so each video bank consists of 16K of memory.
232 COMMODORE 128
10 PRINT CH R $( 1 4 2 ) :REM S E L E C T UP PE R CASE

20 P 0 K E 5 4 , 4 8 : P O K E 5 8 , 4 8 : C L R : R E M P R O T E C T CHAR S E T
30 B A N K 1 4 : RE M S W I T C H T O BA NK 1 4 F O R C H A R A C T E R ROM
40 F O R I = l T O 5 1 1 : P O K E I + 1 2 2 8 8 , P E E K ( 1 + 5 3 2 4 8 ) : N E X T : R E M ROM T O RAM T R A N S F E R
50 B A N K 1 5 : R E M S W I T C H T O D E F A U L T B A NK
60 P O K E 2 6 0 4 , ( P E E K ( 2 6 0 4 ) AND 2 4 0 ) + 1 2 : R E M S T A R T CH AR B A S E A T 1 2 2 8 8
70 F0RJ=12288TO12288+71 : R E M P L A C E C H A R A C T E R D A T A I N RAM
8 0 READ A
90 P O K E J ,A
1 0 0 NEXT J
1 1 0 SCNCLR
1 2 0 DATA 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 : RE M @
1 3 0 DATA 1 4 , 1 7 , 3 2 , 6 6 , 1 3 0 , 1 3 2 , 1 3 8 , 1 2 3 :REM A
1 4 0 DATA 1 2 4 , 6 6 , 6 6 , 1 2 4 , 6 6 , 8 1 , 2 2 5 , 1 2 6 :REM B
1 5 0 DATA 6 2 , 6 7 , 1 3 0 , 1 2 8 , 1 2 8 , 1 2 8 , 1 3 1 , 126:REM C
1 6 0 DATA 1 2 5 , 9 8 , 1 2 5 , 6 5 , 6 5 , 1 9 3 , 1 6 1 , 2 5 4 : RE M D
1 7 0 DATA 2 2 5 , 6 6 , 6 4 , 5 6 , 1 2 0 , 6 5 , 6 6 , 1 2 4 :REM E
1 8 0 DATA 1 2 7 , 1 2 9 , 2 , 4 , 1 4 , 2 2 8 , 6 8 , 5 6 : RE M F
1 9 0 DATA 1 9 3 , 1 6 3 , 2 5 3 , 3 3 , 2 4 9 , 6 5 , 9 9 , 190:REM G
2 0 0 DATA 2 2 7 , 1 6 5 , 3 6 , 3 6 , 1 2 6 , 3 6 , 3 7 , 2 3 1 : REM H
You can leave yourself more BASIC text area, but to do this you must enter a video
bank higher than zero. This sample program operates in Bank 0. If you place your
character set in a higher video bank, remember to add an offset of 16384 ($4000) to the
start of RAM character memory for each bank above video Bank 0.
The CLR in line 20 clears out memory starting at 12288 because, prior to the
placement of the character set there, the memory locations are filled with random bytes.
The random bytes must be cleared before new character information can be stored.
. Line 30 selects BANK configuration 14. This configuration makes the character
ROM visible with a PEEK command or within the machine language monitor, and
temporarily switches out the I/O registers. Both the I/O functions and the 4K character
ROM share the same locations ($D000-$DFFF). Depending on whether bit 0 in the
configuration register (location $FF00) is on or off, the C 128 addresses the I/O registers
or the character ROM. Normally the C128 powers up with bit 1 turned off; therefore,
the I/O registers in locations $D000-$DFFF are addressed. The BANK command in line
30 sets bit 0 in location $FF00; therefore, the character ROM in locations 53248-57343
($D000-$DFFF) is addressed. The character ROM is accessed in order to make the
transfer of characters to RAM.
Line 40 makes the actual transfer from ROM to RAM. Since the character ROM is
accessed in line 30, it now begins at location 53248. The first 512 bytes (the uppercase
character set) from the character ROM are POKEd into the 512 bytes of RAM beginning
at location 12288 and ending at 12800. At this point the character set is ready to be
redefined to custom characters.
Line 50 switches the I/O registers back in, meaning the character ROM is no
longer available.
Line 60 specifies the start of the character set base at location 12288. The character
set base can be stored in locations other than 12288. (See the Character Memory section
earlier in this chapter for more information on moving character memory.) Location
2604 is the intermediate memory location that the interrupt-driven C l 28 screen editor
uses to point to screen and character memory in character mode. You must use this
indirect location to change the value of the actual register that points to the screen and
character memory 53272. If you try to POKE directly to location 53272, the interrupt will
change the value back to the original one within a sixtieth of a second. (You can, how
ever, disable the interrupt-driven screen editor. See the Shadow Register section for details.)
The value (AND 240) or 12 is placed in address 2604 to tell the C l28 to point to
character memory in RAM, starting at address 12288. If the value (AND 240) or 8 is
placed into 2604, the character set will begin at 8192 and your BASIC program must be
less than 6K, but the complete 4K of characters can be redefined. If the value (AND 240)
or 14 is placed into that location, the character set starts at 14336. Your BASIC program
then must be less than 2K, since the programmable character set must reside within a
single 16K block, but the BASIC program that creates the characters can be almost 14K.
If a number other than 12 is POKED into 2604, other program lines must be modified.
Lines 70 through 100 start a loop at the beginning of the character base (12288),
read the values from the data statements which define the new characters (lines 120 and
200), and POKE them (line 90) into the locations allocated for the character set base,
starting at location 12288. The value 12288 + 71 in line 70 sets aside seventy-two
storage locations for the data values in lines 120 through 200 for storage in locations
12288 through 12359. When more data statements are added, the value (12288 + 71)
must be increased to 12288 plus the number of data values in the data statements minus
1. For example, if more characters were defined and there were twenty data statements
with eight values in each, then line 70 would read:
70 F O R J = 12288 T O 12288 + (160-1)
MULTI-COLOR CHARACTER MODE

Standard character mode displays text in two colors: the foreground color of the
character as determined by COLOR RAM, and the background color as determined by
background color register 0 at location 53281 (SD021).
Multi-color character mode gives you the ability to display characters in four colors
within an 8 by 8 character matrix. This substantially increases the freedom of using color.
However, the horizontal resolution is only half the resolution of standard character mode
(160*200), since multi-color mode bits and screen pixels are grouped in pairs. This means
that the color definition and pixel density is twice as wide as standard character mode. The
tradeoff in horizontal resolution is compensated for by the increased freedom of using color.
H O W TO ENTER M ULTI-COLOR MODE

Location 53270 and its shadow register ($00D8) determine whether the C l28 is output
ting standard or multi-color characters on the screen. Bit 4 of 53270 and bit 7 of 216
($00D8) control multi-color mode for character and bit map modes. If bit 4 of 53270
(and bit 7 of 216) is equal to 1, multi-color mode is enabled. Otherwise, standard mode
is enabled. Most of the new 7.0 BASIC graphics commands have provisions for multi
color mode. However, if you want to enter multi-color mode with a POKE command, type:
10 PO K E 216,255: REM Disable IRQ Editor
20 P O K E 53270, P E E K (53270) o r 16: REM Select MCM
This enters multi-color mode, either for character mode or bit map mode.
234 COMMODORE 128
SCREEN LO C A TIO N
The screen location in multi-color character mode defaults to 1024 ($0400) through 2023
($07E7), the same as standard character mode. The screen memory locations can be
relocated. See the Screen Memory section for details.
H O W SCREEN DA TA IS INTERPRETED
In multi-color character mode, the screen data from screen memory is interpreted as
screen codes the same way as in standard character mode. The screen codes are listed in
Appendix D. See the screen data interpretation in the standard character memory
section. The only difference between standard character mode and multi-color character
mode is the way color is assigned to the characters on the screen.
CHARACTER MEMORY LO C A TIO N

The character memory in multi-color character mode, as in standard character mode, is
taken from between 53248 ($D000) and 57343 ($DFF) when I/O is switched out. See
standard character mode for more detailed information on character memory.
H O W CHARACTER MEMORY IS INTERPRETED

IN MULTI-COLOR CHARACTER MODE
Character memory is interpreted virtually the same way in m ulti-color and in
standard character modes, except for one difference: In standard character mode, if
a bit in the character definition image is on, the pixel corresponding to that bit
is colored in the foreground color as specified by color RAM. If a bit in the character
image in the character ROM is equal to zero, the corresponding pixel on the screen is
colored in the background color, as specified by background color register 0 (loca
tion 53281).
In multi-color character mode, color assignments to the pixels that make up the
character on the screen are not in a direct one-to-one relationship to the bits in the
character ROM data patterns. Instead, the bits that make up a character are grouped in
pairs, as shown in figures 8-18, 8-19 and 8-20.
Figure 8-1 8. “ A t ” Sign (@ ) C h a ra c te r as It A ppears on th e Screen

0 0 1 1 1 1 0 0
0 1 1 0 0 1 1 0
0 1 1 0 1 1 1 0
0 1 1 0 1 1 1 0
0 1 1 0 0 0 0 0
0 1 1 0 0 0 1 0
0 0 1 1 1 1 0 0
0 0 0 0 0 0 0 0
Figure 8 -1 9 . B it Patterns of the “ A t ” Sign (@ ) C h arac ter as T h e y A p p e a r in

C h a ra c te r RO M
00 11 11 00
01 10 01 10
01 10 11 10
01 10 11 10
01 10 00 00
01 10 00 10
00 11 11 00
00 00 00 00
Figure 8 -2 0 . B it Patterns o f the “ A t ” Sign (@ ) C h arac ter as T h e y A re G rouped

in Pairs in M u lti-C o lo r C h arac ter Mode
The bits are grouped in pairs, since the horizontal resolution is only half as wide in
multi-color mode. The bit pair determines the color assignments for the pixels within the
character on the screen. The following section describes how the colors are assigned in
multi-color character mode.
COLOR DATA
The color of the pixels in a multi-color character originate from four sources, depending
on the bit pairs. Since the bit pairs have four color possibilities, two bits are needed to
represent four values: 00, 01, 10 and 11. In Figure 8-21, the value of the four bit pair
combinations determines the color assignments for the pixels in a multi-color character.
BIT PAIR COLOR REGISTER LOCATION
00 Background # 0 color (screen color) 53281 ($D021)

01 Background # 1 color 53282 ($D022)
10 Background # 2 color 53283 ($D023)
11 Color specified b y the color RAM
lower 3 bits in color memory
Figure 8 -2 1 . T ru th T a b le fo r C o lor D a ta
236 COMMODORE 128
If the bit pair equals 00 (binary), the color of those two pixels corresponding to the
bit pair are colored by background color register 0 (location 53281 ($D021)). If the bit
pair equals 01 (binary), the pixels are colored by background color register 1 (location
53282 ($D022)). If the bit pair equals 10 (binary), color for those two pixels within the
character are colored from background color register 2 (location 53283 ($D023)).
Finally, if the bit pair from the character pattern equals 11 (binary), those two pixels are
colored from the color specified in the lower three bits (2, 1, 0) of color RAM. Color
RAM is located between 55296 ($D800) and 56295 ($DBE7).
When multi-color character mode is selected, you can still display standard
characters on some screen locations, and display others in multi-color mode. Bit 3 of
each color RAM location determines whether the character is displayed in standard or
multi-color mode. If bit 3 in color RAM is set (1), characters are displayed in
multi-color mode. If bit 3 is clear (0), characters are displayed in standard character mode.
This means that in order to display characters in multi-color mode, you must fill color
RAM with a color code greater than 7. The colors greater than 7 (the ones that are
displayed in multi-color mode) are shown in Figure 8-22.
COLOR CODE COLOR
8 Orange
9 Brown
10 Light Red
11 Dark Gray
12 Medium Gray
13 Light Green
14 Light Blue
15 Light Gray
Figure 8 -2 2 . C o lo r Codes
Remember, the multi-color bit (bit 3) must be set to display multi-color characters.
The following program illustrates multi-color character mode.
10 COLOR 0,1 : RE M BK GN D = B L A C K
20 COLOR 2,1 : RE M MULTCLR 1 = WHITE
30 COLOR 3,2 : REM MULTCLR 2 = RED
31 FOR I = 1 T 0 2 5
32 P R I N T " ABCDEFGHIJKLMNOPQRSTUVWXYZ"
33 NEXT
35 F O R 1 = 5 5 2 9 6 + 5 1 2 TO 5 5 2 9 6 + 1 0 2 3 : P O K E I , 7 : N E X T : R E M PLACE YELLOW IN COLOR RAM
37 POKE 2 1 6 , 2 5 5 : RE M D I S A B L E S C R E E N E D I T O R
40 P O K E 5 3 2 7 0 , P E E K ( 5 3 2 7 0 ) OR 1 6 : R E M S E T M U L T I C O L O R BIT
50 FOR I = 1 T 0 2 5
60 P R I N T "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
70 NEXT
85 F O R 1 = 5 5 2 9 6 T O 5 5 2 9 6 + 1 0 2 3 : P O K E I , 1 4 : N E X T : RE M F I L L COLOR RAM W I T H BLUE
90 G R A P H I C 0 : REM RE TURN TO STANDARD COLOR
Lines 10, 20 and 30 place the color codes for black, white and red into background
color registers 0, 1 and 2, respectively. Lines 31 through 33 print the letters of the
alphabet on the screen twenty-five times. Line 35 fills the last 512 bytes of color RAM
with yellow. Line 37 disables the IRQ VIC screen editor. Line 40 enables multi-color
mode. At this point, all the screen locations corresponding to the color RAM locations
which have a color code greater than or equal to 8 are displayed in multi-color mode.
Since the yellow color code is 7, all the color RAM locations having this code are
placed in standard character mode. The default color for color RAM is code 13 (light
green) for C l 28 mode, and code 14 (light blue) for C64 mode. Line 85 fills color RAM
with the light blue color code. The multi-color characters displayed on the screen are
red, white and blue on a black background.
EXTENDED BACKGROUND COLOR MODE
The third type of character display mode, extended background color mode, allows you
to display three colors at a time on the text screen. For example, you have the character
color, the background color of the screen, and an additional background color within
each 8 by 8 character matrix. This means you can display a white character with a green
background in the 8 by 8 character matrix, on a black screen background. This mode
offers the use of an additional color in an 8 by 8 character matrix, without any loss in
screen resolution.
There is one sacrifice, however. In extended background color mode, only the
first sixty-four characters of the screen code character set are available. The reason for
this is that bits 6 and 7 determine which color will be selected for the background within
the 8 by 8 pixel character matrix. This only leaves five bits for the computer to interpret
which character is currently on the screen. The highest number you can represent with
five bits is 63. This means only the screen code values between 0 and 63 are available
for display on the screen within extended background color mode.
HOW TO ENTER EXTENDED

BACKGROUND COLOR MODE
Enabling bit 6 of location 53265 selects extended background color mode. Use this
POKE in BASIC:
POKE 53265, PEEK(53265) OR 64
To turn it off, use this POKE:
POKE 53265, PEEK(53265) AND 191
SCREEN LOCATION
The screen location in extended background color mode is the same as the standard
character and multi-color character modes, 1024 ($0400) through 2023 ($07E7). This
screen range can be relocated. See the SCREEN MEMORY section for details.
238 COMMODORE 128
H O W T O INTERPRET SCREEN DATA

The data in screen memory is interpreted as screen codes, which are actually the indexes
into the character ROM. Instead of representing the data as ASCII characters, the screen
codes represent the index into the character ROM which provide the ASCII codes. The
first character in character ROM is the at sign (@); therefore the first screen code, 0, is the
code for the at sign.
Remember, since extended background color mode only uses five bits to deter
mine the screen code value, only the first 64 screen code characters (0-63) are available.
COLOR DA TA
The color assignments for the three colors on the screen stem from three sources. Just as
in standard character mode, the foreground color is assigned by COLOR RAM, in the
range 55296 ($D800) through 563295 ($DFE7). As described in the standard character
mode section, each color RAM location has a direct one-to-one correspondence with the
screen memory locations. See the Standard Character Mode section for screen and color
memory maps and an explanation of how the two sections of memory correspond to one
another.
The screen background color is assigned by background color register zero (location
53281 ($D021)). This is the color of the entire screen, on which the foreground and an
additional 8 by 8 character matrix background is placed.
The additional 8 by 8 character matrix background colors are determined by the
value of bits 6 and 7 of the screen code character value. Depending on the value of these
bits, the extended background color (the color within the 8 by 8 character matrix for
each character), comes from one of the four background color registers. Since there are
four choices for the extended background color, the computer needs two bits to
represent the four color choices. Figure 8-23 shows the four-bit combinations and the
corresponding background color registers associated with them.
BACKGROUND COLOR
CHARACTER CODE REGISTER
RANGE BIT 7 BIT 6 NUMBER ADDRESS
0-63 0 0 0 53281 ($D021)

64-127 0 1 1 53282 ($D022)
128-191 1 0 2 53283 ($D023)
192-255 1 1 3 53284 ($D024)
Figure 8 -2 3 . Extended Background C o lo r Registers
For example, POKE the screen code for the letter A (1) into screen location 1024.
Now POKE the screen value 65 into screen location 1025. You might expect the
character to be a reverse A, the second character of the second screen code character set.
However, in this mode, you can only represent the first sixty-four characters, since you
only have five bits to represent screen characters. By trying to represent the screen code
65, bit 6 is enabled, which tells the computer to select the background color register 1
(location 53282 ($D022)), and display the same character, but with the extended
background color specified by background color register 1.
H ere’s a program that illustrates how extended background color mode operates:
5 SCNCLR
1 0 COLOR 0 , 1 : REM B K G R D = B L A C K
2 0 COLOR 2 , 1 : REM M U L T I C O L O R 1
3 0 COLOR 3 , 2 : REM M U L T I C O L O R 2
4 0 P O K E 5 3 2 8 4 , 3 : REM BACKGRD C O L O R 3
4 5 P OKE 5 3 2 6 5 , P E E K ( 5 3 2 6 5 ) OR 6 4 : R E M S E T EXTENDED BKGRND BIT
5 0 F OR 1 = 1 0 2 4 TO 1 2 5 6 : P O K E I , 1 - 1 0 2 3 : NEXT
60 P R IN T :P R IN T :P R IN T :P R IN T
In the program, line 5 clears the screen. Lines 10 through 40 assign colors to the
four background color registers: black, white, red and cyan, respectively. Line 45
enables extended background color mode. Line 50 POKEs 232 characters into screen
memory. Each time sixty-four characters are stored into the screen memory, the same
set of sixty-four characters is POKEd into the next sixty-four screen locations. However,
the next extended background color is displayed. First, the sixty-four characters are
displayed with a black extended background, then a white extended background, then
red, then cyan.
CHARACTER D ATA
Character data is interpreted the same way as in standard character mode, except only
the first 64 characters of the screen character set are available. The character data is also
located in the same range as in standard character mode. See standard character mode
for more information on character data.
STANDARD BIT MAP MODE

Standard Bit Map Mode, also referred to as high-resolution mode, offers the ability to
display detailed graphic images in two colors. The resolution of bit map mode is
320 by 200 pixels. In this mode, the C128 no longer operates in terms of characters,
which are 8 by 8 pixel images stored in those complete units at a time. Bit map mode
allows you to address single pixels at a time, therefore exercising a substantial amount
of control over the detail of images on your screen. The smallest unit addressed in the
character display modes is an 8 by 8 pixel character. Standard bit map mode allows you
to address every individual pixel of the possible 64000 pixels that make up an entire
high-resolution screen image. Figure 8-24 shows how that bit-map coordinate plane is
set up.
240 COMMODORE 128
X Coordinate
0,0 319,0
Y Coordinate
0,199 319,199
Figure 8 -2 4 . B it M ap Screen C oordinates
H O W T O ENTER STANDARD BIT MAP MODE

To enter standard bit map mode, set bit 5 of location 216 ($00D8) (the shadow
register of location 53265 ($D011)). When you issue the GRAPHIC 1,1 command
in C128 BASIC, bit map mode is enabled and the bit map screen is cleared.
You can also use a POKE command as follows, but you should use the highest-
level commands wherever possible:
POKE 216, PEEK (216) OR 32
This command turns on bit 5 of the GRAPHM register, which is the interface between
the VIC chip and the interrupt-driven C l 28 screen editor. This indirectly turns on bit 5
in location 53265 and enters bit map mode.
You can disable the interrupt-driven screen editor and select bit map mode directly
with these commands:
POKE 216, 255
POKE 53265, PEEK (53265) OR 32
In C128 mode machine language, use this program sequence:
LDA $00D8
ORA # $ 2 0
STA $00D8
In C64 mode machine language, try this:
LDA $D011
ORA # $ 2 0
STA $D011
THE VIDEO M ATRIX

(SCREEN MEMORY) LO CATIO N
The default location of the C128 video matrix (i.e., the bit map screen memory) is 7168
($1C00) through 8167 ($1FE7).
The default location of the C64 video matrix is 1024 ($0400) through 2023
($07E7). The video matrix can be moved, however. See the Screen Memory section
elsewhere in this chapter for information on relocating the video matrix.
HOW THE VIDEO MATRIX

IS INTERPRETED
In bit map mode, the video matrix (bit map screen memory) is interpreted differently
than it is in the character display modes. In the character display modes, the screen
memory data is interpreted as screen codes corresponding to the characters in character
ROM. However, in bit map mode, the video matrix data is interpreted as the supplier of
color information for the bit map. The upper four bits (nybble) supply the color
information for the bit map foreground, and the lower nybble supplies the color code for
the bit map background.
The next section explains how pixels on the bit map screen are assigned to either
the foreground color or the background color.
BIT MAP DATA

In bit map mode, the character data, referred to as the bit map, is also interpreted
differently than in the character display modes. The character data is not taken from
character ROM at all. Instead, it is taken from an 8K section of RAM memory, known
as the bit map.
The standard high-resolution screen is composed of 200 rows of 320 pixels, so
that the entire screen is composed of 64,000 pixels. In bit map mode, one bit in memory
is dedicated to an individual pixel on the screen. Therefore 64000 pixels require 64000
bits (or 8000 bytes) of memory to store the entire bit mapped image.
If a bit in memory in the 8000 byte bit map is set, the corresponding pixel on
the screen is enabled, and becomes the color of the foreground as specified in the upper
four bits of the video matrix. If a bit in the bit map is equal to zero, the corresponding
pixel on the screen becomes the color of the background, as specified in the lower four
bits of the video matrix. The combination of on and off bits in the bit map and the
corresponding pixels on the screen define the highly detailed image on the video screen.
The bit map default location in memory ranges from 8192 ($2000) through 16191
($3F3F). This requires 8000 bytes or just under 8K of memory. The spare bytes are used
for other purposes. Location 53272 determines the location of the video matrix and bit
map in memory. Since the screen editor is running on the IRQ, location 2605 ($0A2D)
is an indirect address you must use to place a value in location 53272 ($D018). The
upper four bits of 53272 determine where the video matrix begins and the lower four
bits determine where the bit map begins. In bit map mode, only bit 3 is significant, so
the bit map is either placed starting at location 0 or location 8192 ($2000) in each video
bank.
If you change to a higher bank number, remember to add an offset of $4000 to the
start of the bit map and the video matrix for each bank number above 0 .
The video matrix is relocatable. See the Screen Memory section in the beginning
of the chapter for details on how to move the video matrix.
242 COMMODORE 128
The bit map tells the computer which pixels in the foreground to enable on the
screen. Like a road map, it spells out exactly which pixels to turn on (in the foreground)
and off (in the background color) in order to display a picture on the screen. For
example, if the bit map started at location 8192 (the C l 28 BASIC default) the first
byte of the bit map corresponds to the bit map pixel coordinates 0,0 through 0,7. The
second byte of the bit map, location 8193, corresponds to coordinates 1,0 through 1,7
and so on. See Figure 8-25 to see how the bit map data in locations 8192-16191
correspond to the pixels on the visual screen:
X C oordinate
Figure 8 -2 5 . Relationship o f B it Map D a ta to Screen Pixels
Now you know how bit map mode operates internally within your C l28. However,
you need an easy way to turn on and off pixels in the bit map in order to display
graphics on the screen. The new, high level BASIC 7.0 commands such as DRAW,
CIRCLE, BOX and PAINT allow you to control the turning on and off of bits and their
corresponding screen pixels. The use of the X and Y coordinates on the bit map
coordinate plane easily orient you to displaying graphics. You can display high-
resolution graphics in other ways, outside of the BASIC 7.0 graphics commands.
This includes using commercial software packages that employ graphics tablets or
joysticks to draw on the screen, writing your own draw routines using a joystick or
paddle, or physically entering data into the bit map (which is painfully tedious and not
recommended).
Another way to display graphics, which involves manipulating bits in the bit map,
is through mathematical equations, using geometry. Several books are available which
offer geometrical equations on how to draw three-dimensional objects and to move them.
Refer to the Suggestions for Further Reading the back of the book for sources on graphics.
COLOR RAH
In standard bit map mode, color RAM is not used since the color information for the bit
map is taken from the upper and lower nybble of screen RAM. Color RAM is used,
however, in multi-color bit map mode.
MULTI-COLOR BIT MAP MODE

Multi-color bit map mode is a combination of standard bit map mode and multi-color
character mode. Multi-color bit map mode allows the display of four colors within an 8
by 8 pixel bit map area. Like multi-color character mode, the horizontal resolution is
only half of the standard bit map mode, though the tradeoff in resolution is compensated
for by the use of two additional colors within an 8 by 8 pixel, bit mapped area.
HOW TO ENTER MULTI-COLOR

BIT MAP MODE
To enter multi-color bit map mode from C128 BASIC, issue the following command:
GRAPHIC 3
You can enter this mode with a POKE command as well. But make use of the
highest level commands available for the easiest programming:
POKE 216, PEEK (216) OR 160

This POKEs the value 160 and turns on the multi-color mode bit 7 (value 128) and
the bit map mode bit 5 (value 32) in the GRAPHM register which interfaces to the C l 28
interrupt driven screen editor. This indirectly turns on, respectively, bit 4 (multi-color
mode) of location 53270 ($D 016). and bit 5 (bit map mode) of location 53265 ($D 011).
Bit 5 of location 53265 determines whether the C128 is in bit map mode or
character mode. If bit 5 is equal to 1, bit map mode is enabled. Bit 4 of location 53270
determines whether the C l28 is in standard or multi-color mode. If bit 4 is set, the
C l28 operates in multi-color mode, regardless of whether it is in character or bit map
mode.
In C64 mode, you can store a value directly to these registers. But in C128 mode,
the GRAPHM intermediate register must be used as a gateway to these actual registers.
Again, this is because the C l 28 screen editor is interrupt driven, enabling the split
screen modes for text and simultaneous bit map displays. Since the screen editor is
interrupt driven, an indirect register is used to restore the values that you need to use for
244 COMMODORE 128
specific VIC registers. GRAPHM is one such register; therefore, every sixtieth of a
second, the value in the GRAPHM register is loaded into the appropriate VIC registers.
MACHINE LANGUAGE
To select multi-color bit map mode in C l 28 machine language, perform the following
instructions:
LDA $A0; enables bits 4 and 5 of GRAPHM, the shadow location
STA $00D8
In C64 machine language, enter:
LDA $D011
ORA #$20; select Bit map mode
STA $D011
LDA $D016
ORA #$10; select multi-color mode
STA $D016
In both cases, you must clear the screen, color RAM and the bit map in your own
program.
VIDEO MATRIX LOCATIONS

The video matrix defaults to locations 7168 ($1C00) through 8167 ($1FE7) in C128
mode.
In C64 mode, the video matrix defaults to locations 1024 ($0400) through 2023
($07E7). This is relocatable. See the Screen Memory section earlier in this chapter.
H O W TO INTERPRET THE VIDEO MATRIX

The contents of the video matrix are interpreted the same way as in standard bit map
mode. The upper nybble is the color code for the foreground color; the lower nybble is
the color code for the background color of the bit map.
ADDITIONAL COLOR DATA

The upper and lower nybbles of screen memory supply the multi-color bit map screen
with two of the color sources. This mode offers two additional colors— background
color register 0 (location 53281) and the lower nybble of color RAM.
As in multi-color character mode, the bit patterns of the bytes in the bit map
determine the color assignments for the pixels on the screen. The bits are similarly
grouped in pairs, within 8-bit bytes, so there are 4-bit pairs in each byte. Bits 0 and 1, 2
and 3, 4 and 5, and 6 and 7 are grouped in pairs respectively. Depending on the values
of the bit pairs, the corresponding pixels in the bit map are assigned colors from the
sources in Figure 8-26.
BITS COLOR INFORMATION COMES FROM
00 Background color # 0 (screen color)

01 Upper four bits of video matrix
10 Lower four bits o f video matrix
11 Color RAM
Figure 8 -2 6 . M u lti-C o lo r B it M ap Pixel C o lo r Assignments
THE BIT MAP

Bit patterns determine how color is assigned to the multi-color bit map screen. If the bit
pair is equal to 00 (binary), color is taken from background color register 0 (location
53281). If the bit pair is equal to 01 (binary), the color assigned to these two pixels
comes from the upper nybble of video matrix. If the bit pair in the bit map is equal to
10 (binary), then the color assigned to those two pixels comes from the lower nybble of
video matrix. Finally, if the bit pair in the bit map is equal to 11 (binary), the color is
taken from the lower four bits of color RAM. Unlike multi-color character mode, the
screen is either all standard bit map, or all multi-color bit map, unless you develop a
sophisticated interrupt-driven application program that handles the two separate bit
maps.
COLOR RAM
In multi-color bit map mode, color RAM is used if the bit pair from the bit map equals
11 (binary). Each color RAM location may have one of sixteen color codes, which
means that one 8 by 8 bit map area can have black, red, white and blue colors,
respectively, for the background color register 0 , the upper nybble, the lower nybble, and
the color RAM. The 8 by 8 multi-color bit map area next to it can have black, red, white
and green colors, since each color RAM location is independent of any other. The other
three color sources usually remain constant throughout a bit map screen, though you can
change the upper and lower nybbles of the video matrix. The background color register
is almost always the same throughout a bit map screen.
The C128 has powerful and varied graphics display capabilities. Certain applica
tions call for one type of display over another. Experiment with them all and see which
one meets your needs best. Figures 8-28 through 8-32 provide a graphics programming
summary that should be helpful in understanding graphics on the C l 28.
SPLIT-SCREEN MODES
The Commodore 128 has a split-screen feature that allows you to display the top portion
of the screen in bit map mode and the bottom portion in character mode. This allows you to
enter a BASIC graphics program and RUN it while the BASIC program listing is
246 COMMODORE 128
present and the bit map image is displayed, which saves time switching back and forth
between bit map and character modes.
Before, you would have had to enter the graphics program (in machine language),
RUN it and switch back to the text screen to make a change. Now you can display the
graphic image and have your text screen available to you all at the same time. You can
alter the program while your bit map image is still on the screen, RUN it and see the
immediate results without losing the text screen.
Without the Commodore 128’s split-screen capabilities, you would have to pro
gram a split screen yourself. This involves raster interrupts which utilize either two
screen memories in two different video banks, or a fairly choppy single-screen memory,
usually with a visible raster line. With the C128 split-screen mode, all you have to do to
enter a split-screen mode is to issue the GRAPHIC command in BASIC. For example,
the command:
GRAPHIC 2,1
sets up a standard bit map screen on top and a text screen on the bottom. Similarly, the
command:
GRAPHIC 4,1
constructs a multi-color bit map screen on the top portion of the screen and a text screen
on the bottom portion. The “ 1” in these commands clears the bit map screen. To leave
the bit map screen intact, once you have already displayed an image, replace the “ 1”
with a zero (0).
The GRAPHIC command has an additional parameter that allows you to define
where the split occurs. The split-screen starting location is defined in terms of a
character row, as if the C l 28 were in a character display mode. For example,
GRAPHIC 4,1,15
selects a split screen with multi-color bit map mode on top of the screen and the text
screen on the bottom, starting at character row 15. If the start of the split screen is not
defined, the C l 28 defaults the start to line 19.
H O W SPLIT-SCREEN MODES ARE

ORGANIZED IN MEMORY
SCREEN LO CATIO N S
The split-screen modes, both multi-color and standard, use two independent screen
memories. The bit map video matrix is taken from the address range 7168 ($1C00)
through 8191 ($1FFF), just as in standard and multi-color bit map modes. The text
portion of the screen takes its screen memory from default character mode screen
locations 1024 ($0400) through 2023 ($07E7), just as in standard and multi-color
character mode. The hidden portions of the screen, the bottom portion of the bit map
screen and the upper portion of the text screen, still store data, but it is invisible
since the other screen memory has overlaid it.
INTERPRETING SCREEN D A TA
The text portion of the split screen is interpreted according to the standard character mode
section. The bit map portion, whether standard or multi-color, is interpreted according to the
description in the bit map mode section. Consult the Standard Character Mode, Standard Bit Map
Mode and Multi-color Bit Map Mode sections for information on how screen data is interpreted.
CHARACTER MEMORY LO CA TIO N S

The split-screen modes also take character data from two independent parts of memory.
The bit map data, referred to simply as the bit map, is taken from the default range 8192
($2000) through 16191 ($3FE7) for both the multi-color and standard bit map mode
portions of the screen.
The character memory for the text portion of the split screen is derived from the
character ROM. The actual character ROM occupies the memory locations 53248
($D000) through 57343 ($DFFF) overlaying the I/O registers. The I/O registers must be
switched out to view the actual character ROM, in bank configuration 14, for example.
For information on how character data is interpreted in standard character, stan
dard bit map and multi-color bit map modes, see the sections describing these modes.
See also the Color RAM Banking section.
COLOR DA TA
Each of the standard bit map, multi-color bit map and standard character modes interpret
color differently. See each section for detailed information on color assignments.
M ACHINE LANGUAGE
In machine language, you must program a split screen yourself. This is not the easiest of
programming tasks, since it involves raster interrupt processing, which can be tricky. In
C128 mode, bit 6 in the GRAPHM register is the split screen bit. If bit 6 of $0008
(GRAPHM) is set, a split screen is displayed. Otherwise, bit 6 is clear (0) and a single
screen is displayed.
The C64 mode has no corresponding split screen bit. C64 mode is programmed
differently for split screens. See the Raster Interrupt Split-Screen Program at the end of
the chapter to learn how to program a split screen in machine language.
CAUTION: A system crash may occur if the display mode is changed

while the interrupt-driven screen editor is enabled. See the Shadow
Register section for details.
THE INTERRUPT-DRIVEN SCREEN EDITOR

The intermediate memory locations, sometimes referred to in this guide as shadow
registers, are designed specifically for handling the split-screen modes. In order to
provide split-screen modes, the C128 screen editor has to be wedged into the system’s
interrupt handling routines.
248 COMMODORE 128
Unlike the Commodore 64, the C l 28 handles interrupts exclusively according to

the raster beam. This has made it necessary to merge the C l28 screen editor into the
interrupt request routines (IRQ). The C64 uses interrupt timers which makes interrupt
processing less predictable. By processing the interrupts from the raster beam, the
operating system always knows where and when the interrupt will occur. Timer inter
rupts made catching a raster interrupt less reliable because the operating system never
knew exactly where an interrupt would occur in relation to the raster beam.
The raster interrupt-driven screen editor made it necessary to use indirect storage
locations for certain registers of the VIC chip and 80-column chip. This way, the
intermediate memory locations refresh the actual video chip registers every sixtieth of a
second, each time the raster beam begins a new scan at the top of the screen. The raster
beam scans the entire screen sixty times a second, so on each pass of the raster beam the
intermediate memory locations refresh the actual VIC chip and 8563 chip registers.
RASTER INTERRUPT
SPLIT SCREEN PROGRAM
W ITH H O R IZO N TA L SCROLLING
This section explains how and provides a program to perform split screen raster
interrupts in machine language. The program is explained as it applies to Commodore
64 mode, but it can be modified to run in Commodore 128 mode.
You already have a way to split the screen in C128 mode with the BASIC
GRAPHIC command. The program provided in this section splits the screen in machine
language in C64 mode. See the figure in the shadow register section to see which
addresses must be changed to make this program work in C l 28 Mode. A few differ
ences will occur in the timing of the raster. In Commodore 128 mode, all interrupts
occur according to the position of the raster beam as it scans the screen. This is why
shadow registers are necessary for certain graphics locations, since the C l 28 screen
editor is interrupt driven to allow the split screen modes in BASIC.
The program in this section also scrolls text on the bottom quarter of a standard
character screen, while the top three quarters are displayed in multi-color bit map mode.
The standard character screen resides in video bank 0 ($0400-$07E7) while the multi
color bit map video matrix is stored in video bank 1 starting at $5C00 ($1C00 +
$4000 = $5C00). Every time an interrupt occurs, the program changes video banks,
display modes, the character memory and video matrix pointers. This program supplies
the data that scrolls at the bottom of the screen, but it assumes you have placed an 8000
byte bit map starting at address 8192 ($2000) plus an offset of 16384 ($4000) for the
change of video banks. This makes the absolute start address of the bit map 24576 ($6000).
An 8000 byte bit map is just too large to present in this book. However, the easy
way to place a bit map in this area is as follows:
1. First start in C128 mode, and enter split screen (multi-color) bit map
mode through BASIC with this command:
GRAPHIC 4,1
2. Now draw on the screen with the BOX, CIRCLE, DRAW and PAINT
commands either in a program or in direct mode.
3. When you are finished drawing, enter the machine language monitor either by
pressing the F 8 function key or by typing the MONITOR command.
4. Now transfer the video matrix and bit map from the C l28 default locations of
$1C00 through $1FFF and $2000 through $3FFF respectively, to $5C00 through
$5FFF and $6000 through $7FFF respectively. The new start addresses are
the default locations plus an offset of $4000 for both the video matrix and bit
map pointers. The new start of the video matrix is at $5C00 ($1C00 +
$4000). The new bit map begins at address $6000 ($2000+ $4000). This
transfer can be accomplished with a single transfer command within the
machine language monitor as follows:
T 1C00 3FFF 5C00
This command transfers the contents of memory locations $1C00 through $3FFF
to $5C00 through $7FFF. Since the default locations of the video matrix and bit map are
continuous in memory, the transfer can be done with a single command. If the default
addresses of the video matrix and bit map had not been contiguous, two transfers would
have been necessary. See Chapter 6 for details on using the Machine Language Monitor.
Now that you are still within the control of the Machine Language Monitor, begin
entering the machine language instructions in the listing provided in the next few pages.
Start entering the program at address $0C00. The program, including the scrolled data
occupies memory up to address $0DF9, which means the program is a total of 505 bytes,
or almost half a kilobyte (K).
Now save the program you just painstakingly entered with the Monitor Save (S)
command as follows:
S “ filename” , 08, 0C00, 0DFF
If you have a C128 assembler, create a source file, assemble and load it. If your
assembler allows it, save the program as a binary file.
NOTE: If you have the Commodore 64 Assembler Development System,

create a source file (with start address $0C00), assemble and load it in
C64 mode. Then press the RESET button (not the ON/OFF switch) to enter
C128 mode. D on’t worry, your program will still be in memory, but this
time it’s in C l28 memory. Next enter the Machine Language Monitor
and use the Monitor Save (S) command to make a binary file as described
above.
250 COMMODORE 128
Now you are ready to enter C64 mode and run the program with the following
command:
GO 64
Reply to the question “ ARE YOU SURE?” by pressing the “ Y ” key and
RETURN. You are now placed in C64 mode.
At this point, you may say to yourself, “ After I just did all that work, why am I
going to waste it by changing modes?”
Actually, you are not wasting any effort. When you GO 64 (or press the reset
button), much of the RAM for machine language programs and data is preserved in
RAM bank 0. BASIC programs are erased, however. Specifically, these are the ranges
of RAM that are preserved when changing between C l28 and C64 modes:
C128 M EM ORY LAYOUT
$0C00 - $0DFF RS232 Input and Output Buffers

$1300 - $1BFF Available Memory for Machine Language application programs
$ 1 C 0 0 - $1FFF Bit Map Video Matrix
$2000 - $3FFF Bit Map Data
$4000 - $FF00 BASIC Text Area
Figure 8 -2 7 . Preserved RAM B etw een C l 28 and C 64 Modes
The rest of the RAM memory is allocated for other purposes and the contents
change from mode to mode. There are other particular bytes that are preserved from
mode to mode, but these small chunks of memory are not worth mentioning here. The
blocks of memory mentioned above provide enough of a clue to the RAM used by both
modes. Remember, however, that the RAM is only preserved if you switch from C l 28
mode to C64 mode with the GO 64 command, or you switch from C64 to C l28 mode
with the reset button (warm start). If you perform a cold start, turn the computer power
off, then on again, all RAM is cleared and none is preserved.
Notice that the address ranges where you placed your program, the video
matrix and the bit map are in the portions of RAM that are preserved from mode to
mode.
Now start (run) the program from C64 BASIC with this command:
SYS 12*256
The top three quarters of the screen is the bit map screen you created with the
C128 BASIC graphics commands, the lower quarter is horizontally scrolling text.
The following program is the listing that performs the split screen and scrolling.
1000 ; R A S T E R I N T E R R U P T S P L I T SCRI HORIZONTAL SCROLLING

1010 I VEC= $ 0 3 1 4
1020 TEMP= $ 1 8 0 6
1030 RASTRO=$DO12
1040 BA COL=$D021
1050 POINT=$1802
1060 FLAG= $FC
1070 FLAG 2 = $FD
1080 SCROLL=$FE
1090 SREG= $ D 0 1 6
1100 TXTPTR = $ F A ; POINTER INTO
1110
1120 *=$0C00
1130 ;
1140 LDA # < FRANK ; P O I N T TO TEXT
1150 STA TXTPTR
1160 LDA # > FRANK
1170 STA TXTPTR+1
1180 ;
1190 LDA # $ 0 7 ; SET HI VAL SREG
1200 STA SCROLL
1210 STA FLAG 2
1220 ;
1230 LDA $DD02 ; SET TO OUTPUT
1240 ORA #$03
1250 STA $DD02
1260 j
1270
1280 LDA $D016 ; SET 38 COL.
1290 AND #$F7
1300 STA $D016
1310 ;
1320 •
1330 LDA #$01 ; CLEAR CRAM
1340 LDX #$00
1350 BO NK STA $DB20, X
1360 INX
1370 CPX #200
1380 BNE BONK
1390
1400
1410 INITIALIZE INTERRUPT
1420
1430 LOOPl SEI
1440 LDA IVEC
1450 STA TEMP
1460 LDA IVEC+1
1470 STA TEMP+1
1480 LDA # <MINE
1490 STA IVEC
1500 LDA # >MINE
1510 STA IVEC+1
1520
1530 LD A # $ 0 0 ; DISABLE TIMER IRQ
1540 STA $DC0E
1550 STA FLAG
1560 ;
1570 LDA #49
1580 STA POIN T
1590 STA RASTRO
1600 ;
1610 LDA #201
1620 STA POINT+1
1630
1640 ;
1650 LDA #1
1660 STA $D01A ; ENABLE INT E R R U P T
1670 STA $D019 ; R A S T E R COMPARE
1680 LDA $D 011 ; CLEAR H I B I T
252 COMMODORE 128
1690 AND #$7F

1700 STA $D011 ; DISABLE INTRPT DIS BIT
1710
1720 CL I
1730
1740
1750
1760
1770 CHECK LDA F L A G 2
1780 B P L CHECK
1790 LDA # $ 0 0
1800 STA FLAG 2
1810
1820 LD X # 3 9
1830 LDY # 3 9
1840 S H I F T LDA ( T X T P T R ) , Y
1850 STA $ 0 7 7 0 , X
1860 DEX
1870 DEY
1880 BPL S H IF T
1890 INC TXTPTR
1900 BNE MV T I ME ; T I M E TO R E S E T P O I N T E R
1910 INC TXTPTR+1
1920 M V T I M E L D A T X T P T R ; A R E WE A T T H E E N D O F T H E T E X T
1930 CMP # < E N D T X T
1940 BNE CH ECK
1950 LDA T X T P T R + 1
1960 CMP # > E N D T X T
1970 BNE CHECK
1980 LDA # < F R A N K ; S E T P O I N T E R BACK T O T H E B E G I N N I N G
1990 STA TX TP TR
2000 LDA # > FRANK
2010 STA TXTP TR +1
2020 J MP CHECK
2030
2 0 40 FRANK .BYT
2050 ■B Y T '
2060 ■ BYT ' T H I S I S AN E X A M P L E O F S C R O L L I N G
2070 ■ B YT ' I N THE HORIZONTA L (X) D I R E C T I O N .
2080 ■ B YT ' O N C E T H E DATA HAS B E E N D I S P L A Y E D ,
2090 ■ B YT ' S C R O L L I N G S T A R T S A G A I N FROM T H E
2100 •BYT 'BEGINNING.'
2110
2120 ENDTXT .BYT
2130 • BYT
2140
2150 INTERRUPT SERVICE ROUTINE
2160
2170 MINE LDA $ D 0 1 9
2180 STA $ D 0 19
2190
2200 LDA FLAG ; FLAG = 0 .A=0
2210 EOR # 1 ; FLAG = 0 .A=l
2220 STA FLAG ; FLAG = 1 .A=l
2230 TAX ; . X = 1 . A=1
2240 LDA P O I N T , X ; . X = 1 ,A=200
2250 STA RASTRO ; . X = 1 R A S T R O = 2 0 0
2260 CPX #1
2270 BNE BOTTOM
2280
2290 LDA $ D 0 11 ; BIT MAP MODE
2300 ORA #$2 0
2310 STA $ D 0 11 ; BMM
2320 LDA $D018
2330 ORA #$78 ; CHAR=$20 0 0+$ 4000
2340 STA $D018 ;SCR=$1COO+$4000
2350 LDA $D016
2360 OR A #$10
2370 STA $D016; SET MULTICOLOR
THE ROWER BEHIND COMMODORE 128 GRAPHICS 253
2380 ;
2390 LDA $DD00 ;SELECT BANK 1
2400 AND # $FE
2410 STA SDD00
2420
2430 LDA SREG ;SET SCRRREG
2440 AND #$F8
2450 ORA #$03
2460 STA SREG
2470 ;
2480 LDA #14
2490 STA $D021
2500 ;
2510 PLA
2520 TAY
2530 PLA
2540 TAX
2550 PLA
2560 RTI
2570
2580 ;
2590 BOTTOM LDA $D011 ; SET TEXT
2600 AND # $FF-$ 2 0
2610 STA $D011
2620
2630 LDA $DD0 0 ;CHANGE TOBANK 0
2640 ORA #$01
2650 STA $DD00
2660 •
2670 LDA $D016 ; DISABLE MULTI

2680 AND #$FF-$10
2690 STA $D016
2700
2710 LDA #23 ; MEMORY
2720 STA $D018
2730 ■
2740 LDA #$00
2750 STA $D021
2760 LDA SCROLL
2770 j
2780
2790 DEC SCROLL
2800 LDA SCROLL
2810 STA FLAG 2
2820 CMP #$FF
2830 BNE SLURP
2840 LDA #$07
2850 SLURP STA SCROLL
2860 LDA SREG
2870 AND # $FF-$ 07
2880 ORA SCROLL
2890 STA SREG
2900 BASIRQ JMP (TEMP)
2910 j
2920 •END
For readability, the program is listed as a source file, as though it was entered
through an assembler editor. It is easier to understand as a source file rather than a
listing from the Machine Language Monitor. To enter this program into the Machine
Language Monitor, reference the actual address in place of the variable operand ad
dresses. Most of the actual addresses are listed in the beginning of the program (lines
1010 through 1100). Keep in mind that these are only line numbers for an assembler
254 COMMODORE 128
editor. You will enter the program into a memory location number (address) in the
Machine Language Monitor. In this case, the program is stored in memory starting at
address $0C00. Line 1120 specifies this start address with:
* = $ocoo
Here’s an instruction-by-instruction explanation of the scrolling split screen program.
Line 1010 assigns the variable IVEC to the address $0314, the hardware interrupt
request (IRQ) vector. The interrupt vector is the means by which the Commodore 128
displays split screens and scrolling. By wedging your own routine into the hardware
interrupt vector (in this case scrolling and splitting the screen), it enables you to perform
operations that usually take too long for the microprocessor to perform under an
application program not using interrupts. The interrupt vector is checked for an interrupt
routine every 60th of a second. In this program, the screen is split 60 times per second,
so it appears you have two different screens displayed at the same time. You could not
split the screen without requesting an interrupt; the microprocessor is not able to perform
all the required operations fast enough to keep up with the raster scan of the video
controller. The speed that the screen is continually updated, known as the raster scan,
also occurs at the speed of 60 times per second. For a split screen to occur, you tell the
computer the point on the screen where one type of display ends and the new one (bit
map for example) begins. The way you tell the computer this is by placing the number
of a pixel row, also called a raster row, in the raster compare register located at address
$D012. Line 1030 assigns this address to the variable RASTRO.
You’ll see later in the program that the value placed in the Raster Compare Register
starts the text screen at raster row 201. The raster scan is again interrupted at raster row
50 at the top of the screen to display the bit map screen. This is repeated 60 times every
second, so it appears to the human eye that two different display modes are active at the
same time.
On with the program explanation. Line 1040 defines the BACOL variable for the
background color register zero, located at address $D021. Line 1050 assigns the variable
POINT to location $1802. POINT is used to store the raster row value where the text
screen begins. Line 1060 assigns the variable FLAG to location $FC. FLAG is used later
in the program (lines 2200-2270) to determine where the interrupt occurred, either raster
row 50 or raster row 200.
Lines 1070 and 1080 assign the variables FLAG2 to location $FD and SCROLL to
location $FE respectively. Both FLAG2 and SCROLL store the value of the scrolling
register. SREG is assigned to location $D016, the scrolling register. Only bits 0 through
2 are used as the scrolling bits. The other bits in this address are used for other
purposes. Three scrolling bits are necessary since characters that are scrolled are moved
over seven pixels then shifted to the next character position to the left or right, then
scrolled smoothly again seven more pixels.
Line 1100 assigns the variable TXTPTR to address $FA. This variable marks the
starting address in memory where the scrolled characters are stored.
As was mentioned earlier, line 1120 specifies where the program storage begins in
memory. This is the address where the execution of the program begins in memory. You
will SYS to this address to start the program in BASIC, or GO to this address from the
Machine Language Monitor.
The first sequence of program instructions starting at line 1140 places the contents
of the address (named FRANK) of the beginning of the scrolling text into the memory
locations ($FA and $FB) called TXTPTR and TXTPTR + 1. FRANK is the label in line
2040 which marks the location where the first scrolled character is stored. In this
program the first character is a space; in fact the first forty characters that are scrolled
across the screen are spaces. The forty-first character marks the beginning of the data
‘THIS IS AN EXAMPLE OF SCROLLING’ in line 2060. The scrolled data in lines
2040 through 2130 is stored starting at location $0C98, once this source file is
assembled into object code. In this case, the low byte stored in $FA is $98 and the high
stored in $FB is $0C. The full 16-bit address $0C98 is important, since this is the base
address which you increment as you scroll each letter across the VIC screen. When the
text pointer (TXTPTR) reaches the end of the scrolling text, the base address $0C98 is
again stored in TXTPTR.
The second sequence of instructions starting at line 1190 sets the high value of the
two scrolling variables SCROLL ($FE) and FLAG2 ($FD) to 7. These are used and will
be explained later in the program.
Sequence three starting at line 1230 sets the data direction register to output.
The next module of instructions in lines 1280 through 1300 set the screen size to
38 columns, reducing the screen width by a column on each side. In order to scroll
smoothly, you must set the screen size to 38 columns. Clearing bit 3 of location $D016
sets 38 column size. Setting bit 3 restores the VIC screen to its normal 40 column size.
The extra column on each side of the screen border provides a place for the
scrolled data to scroll smoothly to and from offscreen. This program scrolls left, so each
newly scrolled character is placed in column 39 on the right, before it becomes visible in
column 38. At the same time, the lead character on the left scrolls from column 2 to
offscreen column 1. This occurs in lines 1770 through 2020 and is explained as the
program progresses.
Lines 1330 through 1380 set the text screen color RAM to a white fore
ground. The low er four bits specify the foreground character color in standard
character mode.
The screen memory is stored in video bank 0 where the scrolling text appears at
the bottom fourth of the screen. The bit map and video matrix are stored in video bank
1, the 16K range between $4000 and $7FFF.
The reason only 200 color RAM locations are filled is because only the lower five
rows are visible on the text screen. There is no point clearing the other 800 locations
since they are not visible.
Lines 1410 through 1700 make up the interrupt initialization routine. Line 1430,
labeled LOOP1, sets the interrupt disable bit in the status register. When this bit is set,
interrupts are disabled and none can occur. Only when the interrupt disable bit is cleared
256 COMMODORE 128
(0) can interrupts occur. The last line (1720) of the interrupt initialization routine clears
the interrupt disable and allows interrupts to occur.
Lines 1440 through 1470 store the original contents of the Interrupt Request (IRQ)
vector into temporary storage locations TEMP (low byte $ 1806) and TEMP + 1 (high
byte $1807). This is necessary in order to store the original contents of the IRQ vector
so you can jum p back to this location once the interrupt is serviced as in line 2900.
Lines 1480 through 1510 store the starting location of the interrupt service routine
into the IRQ vector. In this case, MINE is the source file label in line 2170 where the
interrupt service routine begins. In the assembled object file, as in the Machine
Language Monitor, the low byte is $71 and the high byte is $0D, to form the 16 bit
address $0D71. Once the interrupt disable bit is cleared and an interrupt occurs, the
8502 microprocessor finishes executing its current instruction and sets the interrupt
disable status bit, so no other interrupts can occur. The processor then places the
contents of the high byte and low byte of the program counter and the status register on
the stack respectively. Finally, the 8502 fetches the address contained in the IRQ vector
and executes the routine starting at this address, in this case $0D71.
Lines 1530 and 1540 disable the CIA timer in location $DC0E. In addition, line
1550 initializes the variable FLAG to zero. This variable is used later in the program to
figure out where the raster interrupt has occurred, either at the top of the screen (raster
row 49) for bit map mode or near the bottom fourth of the screen (raster row 201) for
standard character mode.
The instructions in lines 1570 through 1590 define the variable POINT ($1802) as
the value of the raster row 49 ($31) where the interrupt occurs to select bit map mode. In
addition, this value is also stored in the Read/Write Raster Register ($D012) for raster
row comparisons later in the program.
The C l 28 VIC screen consists of 200 raster rows, each row one pixel tall, having
320 pixel columns. You know how BASIC addresses the bit map coordinates on a
coordinate plane of 0,0 in the top left corner and 319,199 in the bottom right. However,
the visible raster rows are not labeled in the same way. The visible raster rows start at
50 at the top of the screen and end at 250 at the bottom. These are the same row
numbers as sprites use. Notice there are still 200 rows but that they offset the bit map
coordinate row number by 50. The raster row numbers below 50 and above 250 are off
the visible screen. These offscreen raster rows are referred to as the vertical retrace.
The instructions in lines 1610 and 1620 define the variable POINT + 1 ($1803) as
the value of the raster row 201 ($C9) where the interrupt occurs to select standard
character mode.
Lines 1650 and 1660 enable the raster IRQ Mask Register. Line 1670 sets the
Raster Compare IRQ Flag. By setting bit one in these registers, the raster interrupt
attached to the IRQ line is allowed to occur (once lines 1680 through 1720 are
executed), depending upon whether the physical raster row compares and matches with
either of the values in POINT or POINT + 1 . If either of these match, the interrupt
occurs.
The instructions in lines 1680 through 1700 clear the raster compare high bit (bit
8). This is an extra bit from location $D012 for raster compares.
The instruction in line 1720 clears the interrupt disable status bit, which enables
interrupts to occur. This is the last operation to be performed by the interrupt initializa
tion routine. Now interrupts are ready to occur and be serviced.
Lines 1770 through 1800 check the value of FLAG2. If FLAG2 is positive, the
program branches to the label CHECK in line 1770 and checks the value of FLAG2
again. The value stored in FLAG2 represents the value of the lower three bits in the
horizontal scrolling register at location $D016. The variable SCROLL is also associated
with the variable FLAG2. In the interrupt service routine (in lines 2760 through 2810),
the value in SCROLL is decremented and stored in FLAG2. This value pertains to the
value placed in the actual horizontal scrolling register at $D016. The reason this is
counted is as follows.
The direction of the scrolling is right to left; therefore, you must place the
maximum value (7) in the lower three bits of $D016 and decrement that value by one. If
the program had scrolled left to right, you would initialize the scrolling register to zero
and increment the lower three bits. When the scrolling register value is decremented, the
characters in the screen memory locations which are to be scrolled are moved to the left
by one pixel. Each time the scrolling register is decremented, the characters move
another pixel to the left. When the value (the lower three bits) of the scrolling register
equals zero, you must move the scrolled characters up in screen memory by one location.
This routine is contained in lines 1840 through 1880. After the shift, the lower three bits
of the scrolling register are set back to 7, the characters are again shifted by the VIC
chip 7 pixels to the left and your routine shifts the characters up in memory again by
one. The additional details are covered in the explanation of lines 2760 through 2900.
The instructions in lines 1820 through 1880 shift the text to be scrolled up in
memory by one location for each cycle of the loop. First the X and Y registers are
loaded with the decimal value 39 ($27). The instruction in line 1840 loads the value of
the memory location where the scrolled text begins using indirect Y addressing. The
address is calculated by taking the contents of zero page memory variable TXTPTR
($98) and adding the offset 39 to its contents to arrive at $BF. The effective address
gives the low byte where the scrolled data begins. The first scrolled character is actually
a space. Subsequent data elements are accessed by modifying the value of the Y register.
The store instruction in line 1850 stores the first character of the scrolling text in
screen location 1943 ($0770 + $27), which is the fortieth column of the twenty-third
row. This column is not visible when the screen size is set to 38 columns for horizontal
scrolling. Each newly scrolled character must be placed in this position in order to scroll
smoothly from the offscreen location. Lines 1860 and 1870 decrement the X and Y
registers respectively. Line 1880 branches to the label SHIFT if the Y register is positive
(greater than zero).
The second time through the loop, the low byte of the scrolled character (at
location $98 + $26 = $BE) is stored in screen location 1942 ($0770+ $26), so the
(space) character in $0CBE is stored in screen location $0796. The third time through
the loop, $0CBD is stored in $0795 and so on. This process continues until the X and
Y registers equal zero. So far, only the series of 39 space characters in lines 2040 and
2050 have been shifted across the screen.
258 COMMODORE 128
Once the X and Y registers have been decremented to zero, the TXTPTR is
incremented in line 1890 so that subsequent characters such as “ THIS IS AN EXAM
PLE . . . ” can be scrolled. In the first 39 cycles through the loop (in lines 1840 through
1880) 39 spaces are shifted (scrolled) one character position on the twenty-third character
row on the screen. The next 39 cycles shift 38 spaces and the letter “ T ” in “ THIS” one
character position across the screen. The next 39 cycles, 37 spaces and the letters “ T H ”
in “ THIS” are shifted in memory and scrolled one character position on the screen and
so on. This process occurs until all characters in the data (in lines 2040 through 2130)
are scrolled.
Line 1900 branches to the label MVTIME while TXTPTR (the low byte of the
start of scrolled data) is greater than zero, otherwise TXTPTR + 1 incremented to update
the high byte. Lines 1920 through 1970 check to see if the text pointers are at the end of
the scrolled character data ($0D70 in the assembled program). If the pointers are not at
the end of the character data, the program branches to the label CHECK and the data is
shifted by the VIC chip by seven pixels and the scrolling process repeats again. If the
text pointers are at the end of the scrolled character data, lines 1980 through 2010 set the
text pointers to the beginning of the scrolled data in memory and the process is repeated
continuously as specified by the JMP CHECK instruction.
Lines 2040 through 2130 represent the data to be scrolled by the program. The
data is stored in .BYTE statements as it appears in the Commodore Assembler 64
Development System. In your case, the Machine Language Monitor handles data by
simply storing it in an absolute memory range. In the assembled object code program
the data turns out to be stored in the range $0C98 through $0D70. You will refer to the
data with these addresses and not with a label as in this explanation.
THE INTERRUPT SERVICE ROUTINE

The program instructions in lines 2170 through the end of the program make up the
interrupt service routine. Depending upon the value of the raster comparison, particular
segments of the routine are executed upon the detection of an interrupt. For instance, if
the result of the raster comparison detects an interrupt to occur at raster row 49, lines
2290 through 2560 are executed. This selects bit map mode and performs additional
functions that are explained in a moment. If the raster comparison detects the interrupt
to occur at raster row 201, lines 2590 through 2900 are executed. These instructions
select standard character mode, among other things. Keep in mind that both segments of
the interrupt service routine are executed within a single, complete raster scan of the
screen sixty times per second. Instructions 2170 through 2270 are always executed when
an interrupt occurs.
The instruction in line 2170 clears the raster compare IRQ flag after the interrupt
has occurred. The address of the label MINE, which is loaded into the IRQ vector in
lines 1480 through 1510, tells the 8502 where the interrupt service routine resides upon
the occurrence of an interrupt. In the assembled object code, this is an absolute address
($0D 71).
Lines 2200 through 2270 determine the location (raster row) in which the raster
interrupt has occurred. Line 2200 loads the value of FLAG, which was initialized to
zero, into the accumulator. Line 2210 X O R’s the accumulator with 1, which effectively
places a one in the accumulator for the first pass through this routine. This value is then
stored back into FLAG. In each subsequent occurrence of an interrupt, the value of both
the accumulator and FLAG are toggled between zero and one. The accumulator is then
transferred to the X register in line 2230. Line 2240 loads the value of POINT or POINT
+ 1 depending upon the value in the X register. If the X register equals 0, POINT is
loaded into the accumulator, which specifies the interrupt to occur at raster row 49. If
the X register equals 1, POINT + 1 is loaded into the accumulator, which specifies the
interrupt to occur at raster row 201. Line 2250 stores the accumulator value into the
variable RASTRO. The X register is compared with 1 in line 2260. If the X register
equals 1, the interrupt has occurred at raster row 201 and the program branches to the
instructions in lines 2590. If the value of the X register equals 0, the branch in line 2270
falls through and the instructions in lines 2290 through 2650 are performed.
The instructions in lines 2290 through 2560 perform the operations associated with
bit map mode. Lines 2290 through 2310 select bit map mode. Lines 2320 through 2340
set the video matrix at $1C00 and the bit map at $2000. Both of these start addresses are
offset by the compulsory $4000, since this screen appears in video bank 1 ($4000-$7FFF).
Lines 2350 through 2370 set multi-color mode. Lines 2390 through 2410 select video
bank 1. Lines 2430 through 2460 set the lower two bits of the scrolling register (to the
value 3).
The instructions in 2510 through 2550 restore the original values of the X, Y and
A (accumulator) registers. Line 2560 returns from the interrupt and exits the interrupt
service routine.
Lines 2590 through 2900 perform all the associated text mode operations. Lines
2590 through 2610 select standard character mode. Lines 2630 through 2650 changes
back to video bank 0, the default bank ($0000-$3FFF). Lines 2670 through 2690 disable
multi-color mode, and return to the standard color mode. Lines 2710 and 2720 set the
default screen location 1024 ($0400) and the default start of character memory, with the
decimal value 23 ($17). All numbers which are not preceded by a dollar sign are
assumed to be decimal in this particular assembler editor. Lines 2740 and 2750 set the
background color for the text screen to black.
Lines 2760 through 2890 set the value of the scrolling register, which scrolls the
characters across the screen by 7 pixels, before they are shifted in memory with the
routine in lines 1840 through 1880.
Finally, line 2900 jumps to the default IRQ vector which was saved early in the
program into the variable TEMP. This allows the 8502 to process the normal interrupt
services as though this program’s service routine had not occurred.
Although this program example is long and complex, it contains useful routines
and explanations that have never appeared before in any Commodore text. Study these
routines carefully and add them into your own programs. This section includes a wealth
of information for the novice and experienced software developer alike. Figures 8-28
through 8-32 on the following four pages provide a summary of graphics programming.
260 COMMODORE 128
C H ANG ING VIDEO BANKS M O VING SCREEN RAM
C l 28 POKE 56576, (PEEK (56576) TEXT

BASIC AND 252) OR X
POKE 2604, (PEEK(2604 AND 15)
W HERE X IS THE DECIMAL OR X
VALUE OF BITS 0 AND 1 IN
WHERE X IS THE DECIMAL
TABLE 8-30 ON PAGE 262
VALUE IN TABLE 8-29 ON P. 262
BIT MAP
POKE 2605, (PEEK(2605 AND 15)

OR X
W HERE X IS A VALUE IN
TABLE 8-29 ON P. 262
C l 28 LDA $DD00 TEXT BIT MAP

M ACHINE AND #$F C
LDA $0A2C LDA $0A2D
LANGUAGE ORA # $ X
AND # $ 0 F AND # $ 0 F
STA $DD00 ORA # $ X ORA # $ X
W HERE X IS THE HEX VALUE STA $0A2C STA $0A2D
OF THE BITS IN TABLE 8-30 ON W HERE X IS A HEX EQUIVA
P. 262 LENT OF THE DECIMAL
VALUE IN FIGURE 8-29 ON P.
262
C64 POKE 56576, (PEEK (56576) TEXT OR BIT MAP

BASIC AND 252) OR X
POKE 53272, (PEEK(53272)
W HERE X IS THE DECIMAL AND 15) OR X
VALUE OF BITS 0 and 1 IN
W HERE X IS A DECIMAL
TABLE 8-30 on P. 262
262
C64 LDA $DD00 TEXT OR BIT M AP

MACHINE AND #$F C
LDA $D018
LANGUAGE ORA # $ X
AND # $0F
STA $DD00
ORA # $ X
W HERE X IS THE HEX VALUE STA $D018
OF BITS 0 AND 1 IN
W HERE X IS A HEX
TABLE 8-30 ON P. 262
EQUIVALENT OF THE DECIMAL
262
IN C64 MODE, YOU CAN SET
UP TWO DIFFERENT SCREENS,
ONE FOR TEXT AND THE
OTHER FOR BIT MAP, AS THE
C l 28 KERNAL DOES.
Figure 8-28. Graphics P ro g ram m in g S u m m a ry— P A R T I

M O VIN G CHARACTER MEMORY ACCESSING CHARACTER ROM
TEXT 10 BANK 14: REM SW AP IN CHAR ROM

20 FOR 1 = 0 TO 7
POKE 2604, (PEEK(2604)AND 240) OR Z
30 ? PEEK (I)
40 NEXT
BIT MAP*
50 BANK 15
POKE 2605, (PEEK (2605)AND 240) OR Z
WHERE Z IS A DECIMAL VALUE
IN FIGURE 8-31 ON P. 262
* = ONLY BIT 3 IS SIGNIFICANT
IN BIT MAP MODE
TEXT BIT MAP* LDA #$01

LDA $0A2C LDA $0A2D STA $FF00
AND # $ F 0 AND # $F 0 LDX # $ 0 0
ORA # $ Z ORA # $ Z LDA #$D 000,X
STA $0A2C STA $0A2D LOOP STA $TEM P,X
INX
W HERE Z IS THE HEX EQUIVALENT OF CPX # $ 0 7
A DECIM AL VALUE IN FIGURE 8-31 ON BNE LOOP
P. 262 LDA # $ 0 0
* = ONLY BIT 3 IS SIGNIFICANT IN BMM STA $FF00
TEXT OR BIT MAP* 5 TEM P = 6144

10 POKE 56334,PEEK (56334) AND 254
POKE 53272, (PEEK(53272)AND240) OR Z
20 POKE 1, PEEK (1) AND 251
W HERE Z IS A DECIMAL VALUE IN 30 FOR 1 = 0 TO 7
FIGURE 8-31 ON P. 262 40 POKE (TEM P + I, PEEK (53248 + I)
50 NEXT
* = ONLY BIT 3 IS SIGNIFICANT
60 POKE 1, PEEK (1) OR 4
IN BIT MAP MODE
70 POKE 56334, PEEK (56334) OR 1
80 FOR I = TEMP TO TEM P + 7
90 ? PEEK (I);
100 NEXT
TEXT OR BIT MAP* LDA $DC0E

AND # $ F E
LDA $D018 STA $DC0E
AND #$ F 0 LDA $01
ORA # $ Z AND #$F B
STA $D018 STA $01
LDX # $ 00
WHERE Z IS THE HEX EQUIVALENT LOOP LDA $D000,X
OF A DECIMAL VALUE IN FIGURE STA $TEM P,X
8-31 ON P. 262 INX
CPX # $ 0 7
* = ONLY BIT 3 IS SIGNIFICANT BNE LOOP
IN BIT MAP MODE LDA $01
ORA #$01
STA $01
LDA $DC0E
ORA #$01
STA $DC0E
Figure 8-28. Graphics P ro g ram m in g S u m m a ry— P A R T 2
262 COMMODORE 128
LOCATION*
X BITS DECIMAL HEX
0 OOOOXXXX 0 $0000
16 0001XXXX 1024 $0400 (DEFAULT)
32 0010XXXX 2048 $0800
48 0011XXXX 3072 $0C00
64 0100XXXX 4096 $1000
80 0101XXXX 5120 $1400
96 0110XXXX 6144 $1800
112 0111XXXX 7168 $1C00
128 1000XXXX 8192 $2000
144 1001XXXX 9216 $2400
160 1010XXXX 10240 $2800
176 1011XXXX 11264 $2C00
192 1100XXXX 12288 $3000
208 1101XXXX 13312 $3400
224 1110XXXX 14336 $3800
240 1111XXXX 15360 $3C00
*Remember that the BANK ADDRESS offset of $4000 for each video
bank above zero must be added to the screen memory address.
Figure 8-29. Screen M e m o ry Locations
BANK ADDRESS RANGE VALUE OF BITS 1 & 0 IN $DD00

BINARY DECIM A L
0 $0-$3FFF 1 1 = 3 (DEFAULT)
1 $4000-$7FFF 10 = 2
2 $8000-$BFFF 01 = 1
o
o
o
3 $C 000-$FFFF
II
Figure 8 -3 0 . Video Bank M em o ry Ranges
L O C A T I O N OF C H A R A C T E R M E M O R Y *
VALUE
OF Z BITS DECIMAL HEX
0 XXXXOOOX 0 $0000-$07FF
2 XXXX001X 2048 $0800-$0FFF
4 XXXX010X 4096 $1000-$17FF ROM IMAGE in BANK 0 & 2
(default)*
6 XXXX011X 6144 $1800-$1FFF ROM IMAGE in BANK 0 & 2*
8 XXXX100X 8192 $2000-$27FF
10 XXXX101X 10240 $2800-$2FFF
12 XXXX110X 12288 $3000-$37FF
14 XXXX111X 14336 $3800-$3FFF
Remember to add an offset of $4000 to the start address of character m emory for each
bank above 0; for bank 3 add 3*$4000 = $C000
* = in C64 mode only.
Figure 8 - 3 1. C h arac ter M em o ry Locations

SCREEN DATA COLOR DATA CHARACTER DATA

T EX T B IT M A P TEX T B IT M A P T EX T B IT M A P
C128 BASIC 1024-2023 7168-8167 55296-56295 TAKEN 53248-57343 8192-16383

(DEFAULTS) ($0400-$07E7) ($1C 00- (SD 800- FROM BIT ($D 000- ($2000-$3FFF)
S1FE7) SDBE7) MAP VIDEO SDFFF)
MATRIX
C l 28 1024-2023 7168-8167 55296-56295 *TAKEN T EX T 8192-16383

MACHINE ($0400-$07E7) ($1C00-$1FE7) (SD 800- FROM BIT 53248-57343 ($2000-$3FFF)
LANGUAGE SDBE7) M AP VIDEO (SD 000-
THIS IS THIS IS THIS IS
MATRIX SDFFF)
ALSO PRO ALSO PRO ALSO PRO
GRAMMABLE GRAMMABLE GRAMMABLE
SEE SEE
GRAPHICS GRAPHICS
SUMM ARY SUMM ARY
C64 BASIC 1024-2023 1024-2023 55296-56295 *TAKEN ROM IMAGE NO DEFAULT

(DEFAULTS) ($0400-$07E7) ($0400-$07E7) (SD 800- FROM BIT IS AT MUST
SDBE7) MAP VIDEO 4096-8191* BE PRO
THIS IS
MATRIX ($1000- GRAM M ED
ALSO PRO
1$FFF) SEE
GRAMMABLE
GRAPHICS
SUM M ARY
C64 1024-2023 1024-2023 55296-56295 *TAKEN ROM IMAGE NO DEFAULT

MACHINE ($0400-$07E7) ($0400-$07E7) ($D 800- FROM BIT IS AT
MUST BE
LANGUAGE $DBE7) M AP VIDEO 4096-8191*
THIS IS ALSO PROGRAM- PRO
MATRIX ($1000-$1FFF)
MABLE GRAMM ED
UPPER
SEE
NYBBLE=
GRAPHICS
FOREGROUND
* = actual character SUM M ARY
LOW ER
R O M location =
NYBBLE=
532 4 8 -5 7 3 4 3
BACKGROUND (SDOOO-DFFF)
Figure 8 -3 2 . D e fa u lt Graphics M e m o ry Locations
N O T E : These locations pertain to video bank zero (0) only.

■
0
SPRITES
265
266 COMMODORE 128
SPRITES:
MOVABLE OBJECT BLOCKS
A sprite is a movable bit-mapped object that you can define into a particular shape for
display on the screen. The sprite image can be as large as 24 pixels wide by 21 pixels
tall. Each pixel corresponds to a bit in memory in the sprite storage range; therefore,
each sprite requires 63 bytes of storage. The C l28 has predefined storage locations for
sprite data in the range 3584 ($OEOO) through 4095 ($0FFF).
The C l 28 graphics system has 8 sprites. Each sprite moves on its own indepen
dent plane. A sprite may move in front of or behind objects or other sprites on the
screen, depending on the specified priority. Standard bit-mapped sprites may be any one
of the sixteen available colors. Multi-color sprites may have three colors. The colors that
are assigned to the pixels within the sprite depend on the bit patterns of the image. In
sprite storage memory, the on bits ( 1) enable the sprite pixels to display the color
selected by the sprite color register; the off bits (0) disable the corresponding sprite pixels,
making them transparent and thus allowing the background color to pass through and be
displayed. Sprites also can be expanded to twice the normal size in both vertical and
horizontal directions.
Most of the commercially available graphics software packages for the Commo
dore 128 and C64 rely on sprites. For graphics programming applications, sprites offer
superior animation capabilities. Single sprites are useful for small moving objects.
However, you can adjoin and overlay several sprites to give greater detail to animated
graphic images. For example, suppose you are writing a program that animates a person
running on the screen. You can make the image of the person as a single sprite, but the
effect looks much more realistic if you allocate separate sprites for different parts of the
person’s body. The arms can be one sprite, the body another, and the legs a third. Then,
you can define two additional sprites: one as a second set of legs in a different position,
and the other as a second set of arms in a different position. Position the first set of
arms, the body and the first set of legs on the screen so that they are joined into a full
body. By continually turning on and off the two different sets of arms and legs, the
image appears to be running. This process involves overlaying and adjoining sprites.
The explanation given here is a simplified algorithm, and the actual programming can be
tricky. Sprite programming has been made easy with the new BASIC 7.0 sprite
commands.
The first part of this section explains the new BASIC sprite commands and
illustrates the procedure for overlaying and adjoining sprites. The second part
explains the internal operations of sprites, including storage information, color assign
ments, sprite expansion and addressing the sprite registers in machine language.
SPRITES 267
BASIC 7.0
SPRITE COM M AND SUMMARY
H ere’s a brief description of each BASIC 7.0 sprite command:
C O L L IS IO N : Defines the type of sprite collision on the screen, either sprite to sprite or
sprite to data collision
M O VSPR : Positions or moves sprites from one screen location to another
SPR CO LO R: Defines colors for multi-color sprites
SPRDEF: Enters sprite definition mode to edit sprites
SPRITE: Enables, colors, sets sprite screen priorities, and expands a sprite
SPRSAV: Stores a text string variable into a sprite storage area and vice versa or
copies data from one sprite to another
SSHAPE: Stores the image of a portion of the bit-map screen into a text-string variable
BASIC 7.0
SPRITE CO M M AND FORMATS
COLLISION
Define sprite collision priorities
where:
COLLISION type [^statement]
type Type of collision, as follows:
1 = Sprite-to-sprite collision
2 = Sprite-to-display data collision
3 = Light pen (40 columns only)
statement BASIC line number of a subroutine
E X A M P LE :
COLLISION 1,5000 Detects a sprite-to-sprite collision and program control sent
to subroutine at line 5000.
COLLISION 1 Stops interrupt action which was initiated in above example.
COLLISION 2,1000 Detects sprite-to-data collision and program control directed
to subroutine in line 1000.
MOVSPR
Position or move sprite on the screen (using any of the following four formats):
1. MOVSPR number ,X,Y Place the specified sprite at absolute

sprite coordinate X,Y.
268 COMMODORE 128
2. MOVSPR number, + X, + Y Move sprite relative to its current

position.
3. MOVSPR number, X;Y Move sprite distance x at angle y
relative to its current position.
4. MOVSPR number, angle # speed Move sprite at an angle relative to its
original coordinates, in the clockwise
direction and at the specified speed.
where:
number is sprite’s number (1 through 8)
X ,Y > is coordinate of the sprite location.
A N G LE is the angle (0-360) of motion in the clockwise direction relative to the sprite’s
original coordinate.
SPEED is the speed (0-15) at which the sprite moves.
This statement positions a sprite at a specific location on the screen according to

the SPRITE coordinate plane (not the bit map plane). MOVSPR also initiates sprite
motion at a specified rate. This chapter contains a diagram of the sprite coordinate
plane.
EXA M P LE S:
MOVSPR 1, 150, 150 Position sprite 1 at coordinate 150,150.
MOVSPR 1, + 2 0 , + 3 0 Move sprite 1 to the right 20 (X) coordinates and down

30 (Y) coordinates.
MOVSPR 4, 50; 100 Move sprite 4 by 50 coordinates at a 100 degree

angle.
MOVSPR 5, 45 # 1 5 Move sprite 5 at a 45 degree angle in the clockwise

direction, relative to its original x and y coordinates.
The sprite moves at the fastest rate (15).
NOTE: Once you specify an angle and a speed in the fourth form of the
MOVSPR statement, the sprite continues on its path (even if the sprite
is disabled) after the program stops, until you set the speed to zero (0)
or press RUN/STOP and RESTORE.
SPRCOLOR
Set multi-color 1 and/or multi-color 2 colors for all sprites
SPRCOLOR [smcr-1] [,sm cr-2]
where:
SPRITES 269
sm cr-1 Sets multi-color 1 for all sprites.

sm cr-2 Sets multi-color 2 for all sprites.
Either of these parameters may be any color from 1 through 16.
EXA M PLES:
SPRCOLOR 3,7 Sets sprite multi-color 1 to red and multi-color 2 to blue.
SPRCOLOR 1,2 Sets sprite multi-color 1 to black and multi-color 2 to white.
SPRDEF
Enter the SPRite DEFinition mode to create and edit sprite images.
SPR D EF
The SPRDEF command defines sprites interactively.

Entering the SPRDEF command displays a sprite work area on the screen which is
24 characters wide by 21 characters tall. Each character position in the grid corresponds
to a sprite pixel in the displayed sprite to the right of the work area. Here is a summary
of the SPRite DEFinition mode operations and the keys that perform them:
USER INPUT DESCRIPTION
1 - 8 keys Selects a sprite number at the SPRITE NUMBER? prompt only.

A Turns on and off automatic cursor movement.
CRSR keys Moves cursor.
RETURN key M oves cursor to start of next line.
RETURN key Exits sprite designer mode at the SPRITE
NUMBER? prompt only.
HOME key M oves cursor to top left
corner of sprite work area.
CLR key Erases entire grid.
1-4 keys Selects color source and enables sprite pixels.
CTRL key, 1-8 Selects sprite foreground color (1-8).
Commodore key, 1-8 Selects sprite foreground color (9-16).
STOP key Cancels changes and returns to prompt.
SHIFT RETURN Saves sprite and returns to
SPRITE NUMBER? prompt.
X Expands sprite in X (horizontal) direction.
Y Expands sprite in Y (vertical) direction.
M Multi-color sprite.
C Copies sprite data from one sprite to another.
This SPRite DEFinition area is shown in Figure 9-1.

270 COMMODORE 128
Figure 9 -1 . SPRite D E F inition A re a
SPRITE CREATION PROCEDURE IN

SPRITE D EFIN ITIO N MODE
H ere’s the general procedure to create a sprite in SPRite DEFinition mode:
1. Clear the work area by pressing the shift and CLR/HOME keys at the same
time.
2. If you want a multi-color sprite, press the M key and the cursor ( + ) appears
twice as large as the original one. The double-width cursor appears since
multi-color mode actually turns on two pixels for every one in standard
sprite mode. Multi-color sprites have only half the horizontal resolution of
standard sprites.
3. Select a sprite color. For colors between 1 and 8 , hold down the CONTROL
key and press a key between 1 and 8. To select color codes between 9 and
16, hold down the Commodore (C1) key and press a key between 1 and 8 .
4. Now you are ready to create the shape of your sprite. The numbered keys 1
through 4 fill the sprite and give it shape. For a single-color sprite, use the 2
key to fill a character position within the work area. Press the 1 key to erase
what you have drawn with the 2 key. If you want to fill one character
SPRITES 271
position at a time, press the A key. Now you have to move the cursor
manually with the cursor keys. If you want the cursor to move automatically
to the right while you hold it down, press the A key again. As you fill in a
character position within the work area, you can see the corresponding pixel
in the displayed sprite turn on. The sprite image changes as soon as you edit
the work area.
In multi-color mode, the 2 key fills two character positions in the work
area with the multi-color 1 color, the 3 key fills two character positions with
the multi-color 2 color.
You can turn off (color the pixel in the background color) filled areas
within the work area with the 1 key. In multi-color mode, the 1 key turns
off two character positions at a time.
5. While constructing your sprite, you can move freely in the work area
without turning on or off any pixels using the RETURN, HOME and cursor
keys.
6. At any time, you may expand your sprite in both the vertical and horizontal
directions. To expand vertically, press the Y key. To expand horizontally,
press the X key. To return to the normal size sprite display, press the X or Y
key again.
When a key turns on AND off the same control, it is referred to as
toggling, so the X and Y keys toggle the vertical and horizontal expansion of
the sprite.
7. When you are finished creating your sprite and are happy with the way it
looks, save it in memory by holding down the SHIFT key and pressing the
RETURN key. The Commodore 128 stores the sprites data in the appropri
ate sprite storage area. The displayed sprite in the upper right corner of the
screen is turned off and control is returned to the SPRITE NUMBER
prompt. If you want to create another sprite enter another sprite number and
edit the new sprite just as you did with the first one. If you want to display
the original sprite in the work area again, enter the original sprite number. If
you want to exit SPRite DEFinition mode, simply press RETURN at the
SPRITE NUMBER prompt.
8. You can copy one sprite into another with the O key.
9. If you do not want to SAVE your sprite, press the STOP key. The Commo
dore 128 turns off the displayed sprite and any changes you made are
cancelled. You are returned to the SPRITE NUMBER prompt.
10. To EXIT SPRite DEFinition mode, press the RETURN key while the
SPRITE NUMBER prompt is displayed on the screen without a sprite
number following it.You can exit under either of the following conditions:
■ Immediately after you SAVE your sprite in memory (shift RETURN)

■ Immediately after you press the STOP key
Once you have created a sprite and have exited SPRite DEFinition mode, your
sprite data is stored in the appropriate sprite storage area in the Commodore 128’s
272 COMMODORE 128
memory. Since you are now back in the control of the BASIC language, you have to
turn on your sprite in order to see it on the screen. To turn it on, use the SPRITE
command you learned. For example, you created sprite 1 in SPRDEF mode. To turn it
on in BASIC, color it blue and expand it in both the X and Y directions and enter this
command:
SPRITE 1,1,7,0,1,1,0
Now use the MOVSPR command to move it at a 90-degree angle at a speed of 5,

as follows:
MOVSPR 1, 90 # 5
Now you know all about SPRDEF mode. First, create the sprite, save the sprite
data and exit from SPRDEF mode to BASIC. Next, turn on your sprite with the SPRITE
command. Move it with the MOVSPR command. When you’re finished programming,
SAVE your sprite data in a binary file with the BSAVE command as follows:
BSAVE “ filename” , BO, P3584 TO P4096 (This saves all 8 sprites.)
SPRITE
Turn on and off, color, expand and set screen priorities for a sprite
SPRITE num ber> [,on/off][,fngd][,priority] [,x-exp] [,y-exp] [,mode]

The SPRITE statement controls most of the characteristics of a sprite. The
brackets signify optional parameters. If you omit a parameter, you still must include a
comma in its place.
PARAMETER DESCRIPTION
number Sprite number (1-8)

on/off Turn sprite on (1) or off (0)
foreground Sprite foreground color (1-16)
priority Priority is 0 if sprites appear in front of objects on the screen. Priority
is 1 if sprites appear in back of objects on the screen.
x-exp Horizontal expansion on (1) or off (0)
y-exp Vertical expansion on (1) or off (0)
mode Select standard sprite (0)
or multi-color sprite ( 1 )
Unspecified parameters in subsequent sprite statements take on the characteristics of the

previous SPRITE statement. You may check the characteristics of a SPRITE with the
RSPRITE function.
EXA M PLES:
SPRITE 1,1,3 Turn on sprite number 1 and color it red.
SPRITES 273
SPRITE 2,1,7,1,1,1 Turn on sprite number 2, color it blue, make it pass behind
objects on the screen and expand it in the vertical and
horizontal directions.
SPRITE 6 ,1,1,0,0,1,1 Turn on SPRITE number 6 , color it black. The first 0
tells the computer to display the sprites in front of objects
on the screen. The second 0 and the 1 following it tell the
C l 28 to expand the sprite vertically only. The last 1
specifies multi-color mode. Use the SPRCOLOR com
mand to select the sprite’s multi-colors.
SPRSAV
Store sprite data from a text string variable into a sprite storage area or vice versa.
SPRSAV origin > , destination>
This command copies a sprite image from a string variable to a sprite storage area.
It also copies the data from the sprite storage area into a string variable. Either the origin
or the destination can be a sprite number or a string variable but both cannot be string
variables. If you are copying a string into a sprite, only the first 63 bytes of data are
used. The rest are ignored since a sprite can only hold 63 data bytes.
EXAMPLES:
SPRSAV 1,A$ Copies the bit pattern from sprite 1 to the string variable A$.
SPRSAV B$,2 Copies the data from string variable B$ into sprite 2.
SPRSAV 2,3 Copies the data from sprite 2 to sprite 3.
SSHAPE
Save/retrieve shapes to/from string variables
SSHAPE and GSHAPE are used to save and load rectangular areas of multi-color
or bit-mapped screens to/from BASIC string variables. The command to save an area of
the screen into a string variable is:
SSHAPE string variable, X I, Y l [,X2,Y2]

where:

X I ,Y l Corner coordinate (0,0 through 319,199) (scaled)
X2,Y2 Corner coordinate opposite (X I,Y l) (default is the PC)
Also see the LOCATE command described in Chapter 3 for information on the
pixel cursor.
274 COMMODORE 128
EXA M P LE S:
SSHAPE A$, 10,10 Saves a rectangular area from the coordinates 10,10 to
the location of the pixel cursor, into string variable A$.
SSHAPE B $,20,30,47,51 Saves a rectangular area from top left coordinate
string variable B$.
SSHAPE D$, + 10, + 10 Saves a rectangular area 10 pixels to the right and 10
pixels down from the current position of the pixel cursor.
A D JO IN IN G SPRITES
The following program is an example of adjoining sprites. The program creates an outer
space environment. It draws stars, a planet and a spacecraft similar to Apollo. The spacecraft
is drawn, then stored into two data strings, A$ and B$. The front of the spaceship, the cap
sule, is stored in sprite 1. The back half of the spaceship, the retro rocket, is stored in
sprite 2. The spacecraft flies slowly across the screen twice. Since it is traveling so slowly
and is very far from Earth, it needs to be launched earthward with the retro rockets. After the
second trip across the screen, the retro rockets fire and propel the capsule safely toward Earth.
Here’s the program listing:
5 C O L O R 4 , 1 : C O L O R 0 , 1 : C O L O R 1 , 2 : RE M S E L E C T BLACK BORDER & BKGRND, WHITE FRGRD

10 G R A P H I C 1 , 1 : R E M S E T H I R E S MODE
17 FOR I = 1 T 0 4 0
18 X = I N T ( R N D ( 1 ) * 3 2 0 ) + 1 : R E M DRAW S T A R S
19 Y = I N T ( R N D ( 1 ) * 2 0 0 ) + 1 : R E M DRAW S T A R S
21 DRAW 1 , X , Y : N E X T : R E M DRAW S T A R S
22 BOX 0 , 0 , 5 , 7 0 , 4 0 , , 1 : R E M C L E A R BOX
23 BOX 1 , 1 , 5 , 7 0 , 4 0 : R E M B O X - I N S P A C E S H I P
24 COLOR 1,8:CIRCLE 1,190,90,35,25:PAINT 1,190,95:REM DRAW & PAINT PLANET
25 CIR C LE 1 , 1 9 0 , 9 0 , 6 5 , 1 0 : CIR C LE 1 , 1 9 0 , 9 3 , 6 5 , 1 0 :C IR C L E 1 , 1 9 0 , 9 5 , 6 5 , 1 0 :COLOR 0 , 1
2 6 DRAW 1 , 1 0 , 1 7 TO 1 6 , 1 7 TO 3 2 , 1 0 TO 3 3 , 2 0 TO 3 2 , 3 0 TO 1 6 , 2 3 TO 1 0 , 2 3 TO 1 0 , 1 7
2 8 DRAW 1 , 1 9 , 2 4 TO 2 0 , 2 1 T O 2 7 , 2 5 T O 2 6 , 2 8 : R E M B O T T O M WI N D O W
3 5 DRAW 1 , 2 0 , 1 9 TO 2 0 , 1 7 TO 2 9 , 1 3 TO 3 0 , 1 8 TO 2 8 , 2 3 TO 2 0 , 1 9 : R E M T O P WI N DO W
3 8 P A I N T 1 , 1 3 , 2 0 : RE M P A I N T S P A C E S H I P
4 0 DRAW 1 , 3 4 , 1 0 TO 3 6 , 2 0 TO 3 4 , 3 0 TO 4 5 , 3 0 TO 4 6 , 2 0 TO 4 5 , 1 0 TO 3 4 , 1 0 : R E M SP1
4 2 DRAW 1 , 4 5 , 1 0 TO 5 1 , 1 2 TO 5 7 , 1 0 TO 5 7 , 1 7 TO 5 1 , 1 5 TO 4 6 , 1 7 : R E M ENG1
4 3 DRAW 1 , 4 6 , 2 2 TO 5 1 , 2 4 TO 5 7 , 2 2 TO 5 7 , 2 9 TO 5 1 , 2 7 TO 4 5 , 2 9 : R E M ENG2
4 4 P A I N T 1 , 4 0 , 1 5 : P A I N T 1 , 4 7 , 1 2 : P A I N T 1 , 4 7 , 2 6 : DRAW 0 , 4 5 , 3 0 T O 4 6 , 2 0 T O 4 5 , 1 0
4 5 DRAW 0 , 3 4 , 1 4 T O 4 4 , 1 4 : DRAW 0 , 3 4 , 2 1 T O 4 4 , 2 1 : D R A W 0 , 3 4 , 2 8 T O 4 4 , 2 8
4 7 S S H A P E A $ , 1 0 , 1 0 , 3 3 , 3 2 : REM S A VE S P R I T E I N A$
4 8 S S H A P E B $ , 3 4 , 1 0 , 5 7 , 3 2 : REM S A V E S P R I T E I N B $
5 0 S P R S A V A $ , 1 : REM S P R 1 DATA
5 5 S P R S A V B $ , 2 : RE M S P R 2 DATA
6 0 S P R I T E 1 , 1 , 3 , 0 , 0 , 0 , 0 : REM S E T S P R 1 A T T R I B U T E S
6 5 S P R I T E 2 , 1 , 7 , 0 , 0 , 0 , 0 : REM S E T S P R 2 A T T R I B U T E S
8 2 MOVSPR 1 , 1 5 0 , 1 5 0 : REM O R I G I N A L P O S I T I O N OF SPR1
8 3 MOVSPR 2 , 1 7 2 ,1 5 0 :R E M ORIGINAL PO SITIO N OF S P R 2
8 5 M O V S P R 1 , 2 7 0 # 5 : R E M MOVE S P R 1 A C R O S S S C R E E N
8 7 M O V S P R 2 , 2 7 0 # 5 : R E M MOVE S P R 2 A C R O S S S C R E E N
9 0 FOR 1 = 1 TO 5 9 5 0 : N E X T : R E M DELAY
9 2 M O V S P R 1 , 1 5 0 , 1 5 0 : RE M P O S I T I O N S P R 1 F O R R E T R O R O C K E T L A U N C H
9 3 MOVSPR 2 , 1 7 4 , 1 5 0 : RE M P O S I T I O N S P R 2 FOR R E T RO RO CK ET LAUNCH
9 5 MOVSPR 1 , 2 7 0 # 1 0 : REM S P L I T R O C K E T
9 6 MOVSPR 2 , 90 # 5 : REM S P L I T R O C K E T
9 7 F O R I = l T O 1 2 0 0 : N E X T : REM DELAY
9 8 S P R I T E 2 , 0 : REM TURN O F F R E T R O R O C K E T ( S P R 2 )
9 9 F O R I = l T O 2 0 5 0 0 : N E X T : REM DE LAY
1 0 0 G R A P H I C 0 , 1 : RE M R E T U R N T O T E X T MODE
SPRITES 275
H ere’s an explanation of the program:
Line 5 COLORs the background black and the foreground white.

Line 10 selects standard bit map mode and clears the bit map screen.
Lines 17 through 21 DRAW the stars.
Line 23 BOXes in a display area for the picture of thespacecraft in thetop-left
corner of the screen.
Line 24 DRAWs and PAINTs the planet.
Line 25 DRAWs the CIRCLES around the planet.
Line 26 DRAWs the outline of the capsule portion of the spacecraft.
Line 28 DRAWs the bottom window of the space capsule.
Line 35 DRAWs the top window of the space capsule.
Line 38 PAINTs the space capsule white.
Line 40 DRAWs the outline of the retro rocket portion of the spacecraft.
Line 42 and 43 DRAW the retro rocket engines on the back of the spacecraft.
Line 44 PAINTs the retro rocket engines and DRAWs an outline of the back of theretro
rocket in the background color.
Line 45 DRAWs lines on the retro rocket portion of the spacecraft in the background
color. (At this point, you have displayed only pictures on the screen. You have
not used any sprite statements, so your rocketship is not yet a sprite.)
Line 47 positions the SSHAPE coordinates above the first half (24 by 21 pixels) of the
capsule of the spacecraft and stores it in a data string, A$.
Line 48 positions the SSHAPE coordinates above the second half (24 by 21 pixels) of
the spacecraft and stores it in a data string, B$.
Line 50 transfers the data from A$ into sprite 1.
Line 55 transfers the data from B$ into sprite 2.
Line 60 turns on sprite 1 and colors it red.
Line 65 turns on sprite 2 and colors it blue.
Line 82 positions sprite 1 at coordinate 150,150.
Line 83 positions sprite 2, 23 pixels to the right of the starting coordinate of
sprite 1.
Lines 82 and 83 actually join the two sprites.
Lines 85 and 87 move the joined sprites across the screen.
Line 90 delays the program. This time, delay is necessary for the sprites to complete the
two trips across the screen. If you leave out the delay, the sprites do not have
enough time to move across the screen.
Lines 92 and 93 position the sprites in the center of the screen, and prepare the
spacecraft to fire the retro rockets.
Line 95 propels sprite 1, the space capsule, forward. The number 10 in line 95 specifies
the speed in which the sprite moves. The speed ranges from 0 (stop ) to 15
(fastest).
Line 96 moves the expired retro rocket portion of the spacecraft backward and off the
screen.
276 COMMODORE 128
Line 97 is another time delay so the retro rocket, sprite 2, has time to move off
the screen.
Line 98 turns off sprite 2, once it is off the screen.
Line 99 is another delay so the capsule can continue to move across the screen.
Line 100 returns you to text mode.
SPRITE PROGRAM EXAMPLES
The best way to create sprites is with SPRDEF. The following examples assume you
have created your sprites in SPRDEF mode.
The first example sprite program illustrates the use of the SPRITE and MOVSPR
commands. It positions all eight sprites so they appear to converge on one screen
location, then scatter in all eight directions. Here’s the listing:
10 R E M MOVE S P R I T E E X A M P L E
20 FOR 1 = 1 TO 8
30 MOVSPR 1 , 1 0 0 , 1 0 0
40 NEXT
50 F O R 1 = 1 TO 8
60 SPRITE I , 1 , 1 , 1 , 1 , 1 , 0
70 MOVSPR 1 , 1 * 3 0 # I
80 NEXT
Lines 20 through 40 place all eight sprites at sprite coordinate location 100,100.
At this point, the sprites are not yet enabled, but when they are, all eight are on top of
one another.
Lines 50 and 60 turn on each of the eight sprites in eight different colors. The first
“ I ” is the sprite number parameter. The first “ 1” in line 60 signifies the enabling of
each sprite. The second “ I” specifies the color code for each sprite. The second “ 1”
(the fourth parameter) sets the display priority for all the sprites. A display priority of
one tells the C128 to display the sprites behind objects on the screen. A zero display
priority enables sprites to pass in front of objects on the text or bit-map screen. The fifth
and sixth parameters, both of which are ones ( 1), expand the sprites’ size in both the
vertical and horizontal directions to twice their original size. The final parameter in the
SPRITE statement selects the graphics display mode for the sprites; either standard
bit-map sprites (0) or multi-color bit-map sprites (1). In this example, the sprites are
displayed as standard bit-map sprites.
Line 70 moves the sprites on the screen. The first parameter, I, represents the
sprite number. The second parameter, “ 1*30” , defines the angle at which the sprites
travel on the screen. The pound sign ( # ) notation signifies that the sprites move
according to a particular angle and speed. The final parameter “ I” specifies the speed at
which the sprites travel on the screen. In this example, sprite 1 moves at the slowest rate
of 1, sprite 2 moves at the next highest speed of 2 , while sprite 8 moves the fastest of
the eight sprites at speed 8 . The highest speed a sprite can move is 15.
Finally, line 80 completes the FOR . . . NEXT structure of the loop.
SPRITES 277
Notice that the sprites move continuously even after the program has stopped
RUNning. The reason for this is that sprites are wedged into the interrupt processing of
the C128. To turn off and stop the sprites on the screen, either issue a SPRITE
command that turns them off, or press RUN/STOP and RESTORE.
The second sprite program example provides a simplified adjoining sprite algo
rithm. It moves two adjoined sprites across the screen at a ninety-degree angle,
assuming that your sprites already reside in the sprite storage range between 3584
($0E00) and 4095 ($0FFF). For simplicity, if you don’t have any actual sprite images
stored in the sprite data area, fill the sprite data area with 255 ($FF) from within the
Machine Language Monitor with this command:
F 0E00 OFFF FF
For now, this command turns on all pixels within each sprite. Now you can see
how the adjoining algorithm places and moves sprites 7 and 8 side by side.
H ere’s the listing:
10 REM ADJOINING SPRITE ALGORITHM
20 REM THIS PROGRAM ASSUMES YOUR SPRITES ALREADY EXIST INSPRITE STORAGE
30 1=1 :REM INITIALIZE DISTANCE I
35 SCNCLR
40 MOVSPR 8,50,100:REM SET ORIG POSITION OF SPRITE 8
50 MOVSPR 7,73,100:REM SET ORIG POSITION OF SPRITE 7 TO ADJOIN SPR 8
60 DO
70 SPRITE 8,1,3:REM ENABLE SPR 8
80 SPRITE 7,1,4:REM ENABLE SPR 7
90 MOVSPR 8,1 ;9 0:REM MOVE SPR 8 I UNITS AT A 90 DEGREE ANGLE
100 MOVSPR 7,1 ;90:REM MOVE SPR 8 I UNITS AT A 90 DEGREE ANGLE
110 1=1+1 :REM INCREMENT LOOP
120 LOOP
Line 30 initializes the distance variable I to 1.

Line 40 positions sprite 8 at absolute coordinates 50,100. Since this program
moves two adjoining sprites from the left to right at a ninety-degree angle, sprite 7,
which is attached to sprite 8 must be positioned in such a way that it is touching the right
edge of sprite 8 . Line 50 places sprite 7 on the exact right edge of sprite 8 . Since a sprite
is 24 pixels wide (before expansion), to adjoin two sprites together, place the ajoining
sprite exactly 24 pixels to the right of the top left corner coordinate position of sprite 8 .
The position of a sprite is placed on the sprite coordinate plane according to the upper
leftmost pixel of the sprite. Since the original position of sprite 8 is 50,100, add 24
(inclusive) to the X (horizontal) coordinate to make them join exactly on the respective
edges of both sprites. This is provided your sprites are exactly 24 pixels wide. If you
don’t fill the entire dimensions of a sprite, you may have to adjust the coordinates so
that they meet correctly.
At this point, the sprite coordinates are perfectly adjoined. Line 60 initiates a loop,
so that the distance can be updated to enable the sprites’ movement across the screen.
Lines 70 and 80 enable sprites 8 and 7 and color them red and cyan, respectively.
Lines 90 and 100 move sprite 8 and 7, respectively, at a 90-degree angle
according to the distance specified by the variable I. Line 110 updates the distance of 1
each cycle through the loop. Line 120 circulates the loop until the distance variable I is
equal to 320.
278 COMMODORE 128
The third sprite example provides an algorithm to overlay two sprites and move
them on the screen on a 45-degree angle. Again, this program assumes your sprite data
resides in sprite storage. If your sprite images are not stored there, fill the sprites with
data as you did in the last adjoining example.
Here’s the listing:
10 REM OVERLAY EXAMPLE

20 REM THIS PROGRAM ASSUMES SPRITE DATA RESIDES IN SPRITE STORAGE
30 1=1 :REM INITIALIZE DISTANCE I
35 SCNCLR
40 MOVSPR 8,50,100:REM SET ORIG POSITION OF SPRITE 8
50 MOVSPR 7,50,100:REM SET ORIG POSITION OF SPRITE 7 TO OVERLAY SPR 8
60 DO
70 SPRITE 8,1,3 :REM ENABLE SPR 8
80 MOVSPR 8,I,-45 :REM MOVE SPR 8 I UNITS AT A 45 DEGREE ANGLE
90 SPRITE 8,0,3 :REM TURN OFF SPR8
100 SPRITE 7,1,4 :REM ENABLE SPR 7
110 MOVSPR 7,I;45 :REM MOVE SPR 8 I UNITS AT A 45 DEGREE ANGLE
120 SPRITE 7,0,3 :REM TURN OFF SPR 7
14 0 LOOP
As in the last program, line 30 initializes the distance variable I to 1.

Lines 40 and 50 position sprites 8 and 7, respectively, at coordinate 50,100. At
this point the sprites are not yet enabled, but when they are, sprite 7 will overlay sprite 8
since the lower sprite number has display priority over the higher sprite number.
Line 60 initiates a DO loop to move the sprites along the sprite coordinate plane.
Line 70 enables sprite 8 and colors it red. Line 80 moves sprite 8 a distance of one
coordinate according to the current value of I. Line 90 disables sprite 8 .
Lines 100 through 120 perform the same operations for sprite 7 as lines 70 through
90 did for sprite 8: enable, move a single distance coordinate according to I and disable.
Line 140 repeats the process.
Since this process is repeated so quickly, it appears as though the two sprites
alternate movements. When you create the actual images you will use in your overlay
sprite program, the images between which you alternate will be ones that simulate the
movement of two images and create one animated image.
Create two sprites that appear to form a single animated image. You may have to
perfect the timing of the enabling and disabling of the images to make the animated
image appear more smooth. Nonetheless, you have a basis for animating two objects
into one single moving object.
Although these program examples are written in BASIC, the algorithms are the
same whether you are programming in BASIC or machine language. The next section
discusses sprite operations independent of the BASIC language. Since this section
explained sprites according to BASIC, the next section elaborates on the inner workings
of sprites from a machine level (language) perspective.
SPRITES 279
THE INNER WORKINGS OF SPRITES
You have seen how to create, move, color and expand sprites with the BASIC 7.0 sprite
commands. This section explains how to control sprites outside of the BASIC sprite
commands (except SPRDEF). This tells you which VIC registers are affected and the
specific bits that must be set or cleared to manipulate the sprite features.
Registers of the VIC chip control all aspects of sprites. The enabling of specific
bits in certain VIC registers turns on the features of the eight available sprites. The order
in which you turn on these features is critical to sprite animation. Following is a
summary of the steps necessary to display, color, move and expand sprites. Next to each
step is the VIC chip register or other memory location involved in each element of sprite
programming.
SPR ITE PROGRAM M ING SEQUENCE REGISTERS INVOLVED
1. Create the sprite image Sprite Data Storage: 3584-4095 ($0E00-$0FFF)

This is also programmable. You must change the sprite
pointer values.
2. Point to the sprite data Sprite Data Pointers: 2040-2047 ($07F8-$07FF), or

8184-8191 ($1FF8-$1FFF) when the bit map screen has
been cleared with GRAPHIC 1,1
3. Enable (turn on) the sprite 53269 ($D015) (bits 7 -0 , depending on sprite number)
4. Color the sprite Standard 53287-53294 ($D027-$D02E)

Multi-color 53276 ($D01C), 53285 ($D025), 53286 ($D026)
5. Position the sprite 53248-53264 ($D000-$D010)
6 . Expand the sprite 53271 ($D017) (Y direction), 53277 ($D01D) (X direction)
7. Define sprite display priorities 53275 ($D01B)
8 . Define sprite collision priorities 53278 ($D01E), 53279 ($D01F)
These registers control most sprite characteristics. Once you learn these programming
steps, you will be able to exercise full control over the display and movement of sprites.
CREATING THE SPRITE IMAGE

An easy way to create sprites on the C l 28 is through SPRite DEFinition mode
(SPRDEF). For an explanation of SPRDEF see the SPRDEF entry in the beginning of
this chapter. This section assumes your sprite image is already created, and it resides in
the sprite data storage area. Before leaving SPRDEF, remember to press the SHIFT key
and RETURN at the same time; this allows SPRDEF to store the sprite data in the sprite
data storage area. Press RETURN a second time to exit SPRDEF.
The Commodore 128 has a dedicated portion of memory ranging from decimal
address 3584 ($0E00) through 4095 ($0FFF), where sprite data is stored. This portion of
280 COMMODORE 128
memory takes up 512 bytes. As you know, a sprite is 24 pixels wide by 21 pixels tall. In
standard sprites, each pixel corresponds to one bit in memory. If the bit in a sprite is off
(equal to 0), the corresponding pixel on the screen is transparent, which allows the
background to pass through the sprite. If a bit within a sprite is on (equal to 1), the
corresponding pixel on the screen is turned on in the foreground color as determined by
the sprite color registers. The combination of zeroes and ones produces the image you
see on the screen. Multi-color sprites assign colors differently. See the multi-color sprite
section later in this chapter for details.
Since a sprite is 24 by 21 pixels and each pixel is represented by one bit of storage
in memory, one sprite uses up 63 bytes of memory. See Figure 9-2 to understand the
storage requirements for a sprite’s data.
12345678 12345678 12345678

1
2
3 .........................................................
4 .........................................................
5 .........................................................
6 .........................................................
7 .........................................................
8 .........................................................
9 .........................................................
10 .........................................................
11
12
13 .........................................................
14 .........................................................
15 .........................................................
16 .........................................................
17 .........................................................
18 .........................................................
19 .........................................................
20 ...................................................
21
Each Row = 24 bits = 3 bytes
Figure 9-2. Sprite D ata Requirements
A sprite requires 63 bytes of data. Each sprite block is actually made up of 64

bytes; the extra byte is not used. Since the Commodore 128 has eight sprites and each
one consists of a 64-byte sprite block, the computer needs 512 (8 x 6 4 ) bytes to
represent the data of all eight sprite images.
The area where all eight sprite blocks reside starts at memory location 3584
($0E00) and ends at location 4095 ($0FFF). Figure 9-3 lists the memory address ranges
where each individual sprite stores its data.
SPRITES 281
$0FFF (4095 Decimal)

]— Sprite 8
$ 0 FC 0
]— Sprite 7
$0F80
]— Sprite 6
$0F40
]— Sprite 5
$0F00
]— Sprite 4
$0EC0
]— Sprite 3
$0E80
]— Sprite 2
$0E40
]— Sprite 1
$0E00 (3584 Decimal)
Figure 9 -3 . M em o ry Address Ranges fo r Sprite Storage
Keep in mind that sprites are referred to as 1 through 8 in BASIC, but 0 through 7
in machine language.
SPRITE POINTERS
The VIC chip needs to know where to look for the bit patterns (data) that make up the
sprite image in memory. The sprite pointers are used explicitly for this purpose.
Unlike the Commodore 64, the C128 has automatically filled the sprite data
pointers with values that direct the VIC chip to point to the data stored in the sprite data
range 3584 ($0E00) through 4095 ($0FFF). These sprite data pointers are located at
2040 ($07F8) through 2047 ($07FF) for sprites 0 and 7 respectively. They are also located
in the address range 8184 ($1FF8) through 8191 ($1FFF),once the bitmap screen is
cleared with the GRAPHIC 1,1 command. The default contents of these locations are:
Hexadecimal 38 39 3A 3B 3C 3D 3E 3F
Decimal 56 57 58 59 60 61 62 63
To find the actual location where the sprite data pointers are looking for data in
memory, multiply the contents of the sprite data pointer by 64 (decimal). By multiplying
these values, you’ll see that the pointers look for data in the default sprite storage
locations in Figure 9-3. See Figure 9 -4 for an illustration.
The way the Commodore 128 automatically points to the correct sprite data is conven
ient for programming, since it eliminates a step (provided the original values of the sprite
pointers have not been modified). If you want to store sprite data somewhere else in
memory, however, you’ll have to change the original value of the sprite pointer (from
location 2040 through 2047, or 6184 through 8191) to a value that is equal to:
Start of Sprite Data / 64 = new contents of sprite pointer

282 COMMODORE 128
DATA POINTER START OF

CONTENTS** SPRITE DATA
Sprite 0 Data Pointer = 56 * 64 = 3584 ($0E00)

Sprite 3 Data Pointer = 59 * 64 = 3776 ($0EC0)
Sprite 4 Data Pointer = 60 * 64 = 3840 ($0F00)
Sprite 7 Data Pointer = 63 * 64 = 4032 ($0FC0)
** = This applies to video bank 0 only.
Figure 9-4. Sprite D a ta Locations
The start of sprite data is divided by 64 because the data area is allocated in 64-byte
sections. For example, you want to place your sprite 0 data in the new location 6144
($1800). Divide 6144 by 64 to get 96. Place the value 96 ($60) in address 2040 ($078F).
ENABLING A SPRITE
Once the sprite image has been defined, and the data pointer is pointing to the correct
data, you can turn on the sprite. You do this by placing a value in the Sprite Enable
Register, location 53269 ($D015). The value placed in this register depends on which
sprite(s) you want to turn on. Bits 0 through 7 correspond to sprites 0 through 7. To
enable sprite 0, set bit 0. To enable sprite 1, set bit 1 and so on. The value you
place in the sprite enable register is equal to two raised to the bit position in decimal.
If you are programming in machine language and want to enable more than one
sprite at a time, add the values of two raised to the bit positions together and store the
result in the sprite enable register. For example, to enable sprite 5, raise two to the fifth
power (32 ($20)) and store it as follows:
LDA # $ 2 0
STA $D015
To enable sprites 5 and 7, raise two to the fifth (32 ($20)) and add it to two to the
seventh (128($80)) to obtain the result 160 ($A0):
LDA $A0
STA $D015
An easier way of perceiving the idea is through binary notation in the Machine
Language Monitor as follows:
LDA # % 10100000
STA $D015
To disable the sprite display, clear the bits in the sprite enable register.
SPRITES 283
ASSIGNING COLOR TO SPRITES

Sprites have two kinds of color displays: standard bit-map and multi-color bit-map
sprites. The color assignments to the pixels within the sprites work in a similar way to
standard bit-map and multi-color bit-map modes for the screen.
STANDARD BIT MAP SPRITES

Standard bit-map sprites each have their own color register. The lower four bits of each
sprite color register determine the sprite color as specified by the sixteen C l 28 color
codes. Figure 9-5 shows the standard bit-map sprite color registers.
ADDRESS DESCRIPTION
53287 ($D027) SPRITE 0 COLOR REGISTER

53290 ($D02A) SPRITE 3 COLOR REGISTER
53291 ($D02B) SPRITE 4 COLOR REGISTER
53292 ($D02C) SPRITE 5 COLOR REGISTER
53293 ($D02D) SPRITE 6 COLOR REGISTER
53294 ($D02E) SPRITE 7 COLOR REGISTER
Figure 9-5. Standard Bit Map Sprite Color Registers
Figure 9 -6 lists the color codes that are placed in the standard bit-map sprite color
registers:
0 Black 8 Orange
1 White 9 Brown
2 Red 10 Light Red
3 Cyan 11 Dark Gray
Figure 9 -6 . S p rite C o lo r Codes
In standard bit-map sprites, the data in the sprite block determine how the colors
are assigned to the pixels on the screen within the visible sprite. If the bit in the
sprite storage block is equal to 1. the corresponding pixel on the screen is assigned
the color from the standard sprite color register. If the bit in the sprite data block
284 COMMODORE 128
is equal to zero, those pixels on the sprite are transparent and thebackground data
from the screen passes through the sprite.
M U LTI-C O LO R SPRITES
Multi-color sprites offer a degree of freedom in the use of color, but you trade the
higher resolution of standard sprites for the added color. Multi-color sprites are displayed
in three colors plus a background color. Multi-color bit-map sprites are assigned colors
similar to the way the other multi-color modes work. Before you can assign a sprite
multiple colors, you must first enable the multi-color sprite. The sprite multi-color
register in location 53276 ($D01C) operates in the same manner as the sprite enable
register. Bits 0 through 7 pertain to sprites 0 through 7. To selecta multi-color sprite,
set the bit number that corresponds to the sprite number. This requires that you raise
two to the bit position of the sprite that you want displayed in multi-color. For example,
to select sprite 4 as a multi-color sprite, raise two to the fourth power (16) and place
it in the multi-color sprite register. In machine language, perform the following
instructions:
LDA #$10
STA $D01C
To select more than one multi-color sprite, add the values of two raised to the bit
positions together and store the value in the multi-color sprite register.
The VIC chip provides two multi-color registers (0 and 1), in which to place color
codes. These are the locations of the sprite multi-color registers:
A D D R ESS
Sprite M ulti-Color Register 0 53285 ($D025)

Sprite M ulti-Color Register 1 53286 ($D026)
T h e color codes are those listed in Figure 9-6.
Like multi-color character mode, the pixels in the multi-color sprites are assigned
color according to the bit patterns in the sprite storage block. In this mode, the bits in
the sprite block are grouped in pairs. The bit pair determines how the pixels are assigned
their individual colors, as follows:
B IT P A IR D E S C R IP T IO N
00 TRANSPARENT (SCREEN COLOR)

01 SPRITE M ULTI-COLOR REGISTER # 0 (53285) ($D025)
10 SPRITE COLOR REGISTER
11 SPRITE M ULTI-COLOR REGISTER # 1 (53286) ($D026)
SPRITES 285
If the bit pair is equal to 00, the pixels are transparent and the background from
the screen passes through the sprite. If the bit pattern equals 10 (binary), the color is
taken from the sprite color register (locations 53287-53294) of the sprite being defined.
Otherwise, the other two bit pair possibilities (01 and 11) are taken from sprite
multi-color registers 0 and 1 respectively.
PO SITIO N IN G SPRITES O N THE SCREEN

Each sprite has two position registers to control the sprite’s position on the visible screen:
horizontal (X coordinate) and vertical (Y coordinate) positions. Figure 9-7 gives the
memory locations of the sprite position registers as they appear in the C l28 memory.
LOCATION
D E C IM A L H EX D E S C R IP T IO N
53248 ($D000) SPRITE 0 X POSITION REGISTER

53249 ($D001) SPRITE 0 Y POSITION REGISTER


53258 ($D00A) SPRITE 5 X POSITION REGISTER
53259 ($D00B) SPRITE 5 Y POSITION REGISTER
53260 ($D00C) SPRITE 6 X POSITION REGISTER
53261 ($D00D) SPRITE 6 Y POSITION REGISTER
53262 ($D00E) SPRITE 7 X POSITION REGISTER

53263 ($D00F) SPRITE 7 Y POSITION REGISTER
53264 ($D010) SPRITE X MSB REGISTER
Figure 9 -7 . M em o ry Location of S p rite Position Register
The sprite position registers together plot the sprite on a vertical and horizontal
coordinate. The position of reference for the calculated vertical and horizontal coordi
nate is taken from the upper-left corner pixel within the sprite. Whenever you want to
place the sprite on a particular screen position, calculate the position using the upper left
corner pixel within the sprite. The sprite coordinate plane is not the same as the bit-map
coordinate plane. The bit-map coordinate plane starts in the upper-left corner of the
screen at coordinate 0,0. The lower right corner of the bit map coordinate plane is point
319,199. The sprite coordinate plane starts at point 24,50 in the top-left corner of the
visible screen. The final visible point on the sprite coordinate plane at the bottom-right
com er of the screen is 343,249. Figure 9-8 shows how the sprite coordinate plane
relates to the visible screen.
286 COMMODORE 128
0 ($00) 24 ($18) 296 ($128) 344 ($158)
8 ($08)
29 (S10)-------
50 (132)------- -- 50 ($32)
VISIBLE VIEW ING AREA
NTSC*
40 COLUM NS
208 ($D 0)- 25 ROWS
229 ($E5)
250 ($FA ) — 250 ($FA)

r
I I
488 ($1E8) 24 ($18) 320 ($140) 344 ($158)
‘ N o rth A m e ric a n te le v is io n tra n s m is s io n s ta n d a rd s fo r y o u r h o m e TV.
Figure 9-8. Visible Sprite Coordinates
After seeing the sprite coordinate plane, you may have noticed something unusual.
The vertical coordinate positions have a range of 200. The horizontal coordinate
positions have a range of 320 coordinates. Since the C l28 is an 8-bit computer, the
highest value any register can represent is 255.
How do you position a sprite past the 255th horizontal screen position? The
answer is, you have to borrow a bit from another register in order to represent a value
greater than 255.
An extra bit is already set aside in the Commodore 128 memory in case you want
to move a sprite past the 255th horizontal coordinate. Location 53264 controls sprite
movement past position 255. Each of the 8 bits in 53264 controls a sprite. Bit 0 controls
sprite 0, bit 1 controls sprite 1 and so on. For example, if bit 7 is set, sprite 7 can move
past the 255th horizontal position.
Each time you want a sprite to move across the entire screen, turn on the borrowed
bit in location 53264 when the sprite reaches horizontal position 255. Once the sprite
moves off the right edge of the screen, turn off the borrowed bit so the sprite can move
back onto the left edge of the screen. The following commands allow sprite seven to
move past the 255th horizontal position:
SPRITES 287
LDA $D010
ORA #$80
STA $D010
The number 128 is the resulting decimal value from setting bit 7. You arrive at this
value by raising two to the seventh power. If you want to enable bit 5, raise two to the
fifth power, which, of course, is 32. The general rule is to raise two to the power of the
sprite number that you want to move past the 255th horizontal screen position. Now you
can borrow the extra bit you need to move a sprite all the way across the screen. To
allow the sprite to reappear on the left side of the screen, turn off bit seven again, as
follows:
LDA $D010
AND #$7F
STA $D010
Not all of the horizontal (X) and vertical (Y) positions are visible on the screen.
Only vertical positions 50 through 249 and horizontal positions 24 through 343 are
visible. Location 0,0 is off the screen as is any horizontal location less than 24 and
greater than 343. Any vertical location less than 50 and greater than 249 is also off the
screen. The off-screen locations are set aside so that an animated image can move
smoothly on and off the screen.
EXPA N D IN G THE SIZE OF SPRITES

The VIC chip offers a feature that allows sprites to be expanded in size, in both the
horizontal and vertical directions. When the sprite is expanded, the sprite resolution
does not increase, the pixels within the sprite just cover twice as much area; therefore,
the sprite is twice as large. Here are the locations in memory for vertical and horizontal
sprite expansion:
ADDRESS
Vertical (Y) Sprite Expansion Register 53271 ($D017)

Horizontal (X) Sprite Expansion Register 53277 ($D01D)
These registers operate in the same manner as the sprite enable register.
Bits 0 through 7 pertain to sprites 0 through 7. To expand the sprite size in either
direction, raise two to the bit position and place it in the expansion register(s). For
exam ple, to expand sprite 7 in both directions, perform these machine language
instructions:
LDA # $ 8 0 (% 10000000 = binary notation in the Monitor)

STA $D017
STA $D01D
288 COMMODORE 128
To expand more than one sprite, add two raised to the bit position for the sprite
numbers you want to expand, and store the result in the expansion registers.
To return the sprites to their original size, clear the bits in the expansion
registers.
SPRITE DISPLAY PRIORITIES

In your sprite programs, you have the option of displaying sprites in front of or behind
other sprites or objects on the screen. This is known as defining the sprite’s display
priorities. The VIC chip defines two distinct sprite display priorities:
1. Sprite-to-sprite
2. Sprite-to-data
SPRITE-TO-SPRITE DISPLAY PRIORITIES

Each of the eight sprites available on the Commodore 128 is assigned its own plane in
which it moves independent of other sprites or of the pictures on the screen background.
Visualize the sprite planes as in Figure 9-9.
Screen
Background
Plane
- Sprite plane 7
_ Sprite plane 6
-- Sprite plane 5
~ Sprite plane 4
- Sprite plane 3
- Sprite plane 2
Sprite plane 1
Sprite plane 0
Figure 9-9. Relationship of S p rite Planes
The display priority of each sprite plane depends on the sprite number. The
sprite-to-sprite display priorities are predefined according to the sprite number. This
SPRITES 289
feature is preset by the C l 28 hardware, and is not controlled through a software

register. The lower the sprite number, the higher the priority. The higher priority sprite
passes in front of the lower priority sprite. Sprite 0 has the highest priority; therefore, it
passes in front of any other sprites that may be on the screen should they meet at a
particular screen location. For example, sprites I and 5 are moving toward a common
location on the screen. When the sprites reach the common location, sprite 1 passes in
front of sprite 5, since the lower sprite number has the higher display priority. Keep this
in mind when you want sprites to intersect paths and pass behind or in front of one
another. This is important when overlaying sprites. Assign the sprite you want to pass in
front of other sprites to the lower sprite number.
Portions of sprites (pixels) that are not assigned color (the bits in the sprite storage
block corresponding to those pixels are equal to zero) are transparent. Holes in these
sprites allow background data to pass through the transparent area and create a “ win
dow ” effect. The same window effect occurs when a higher priority sprite with “ holes”
passes in front of a lower priority sprite. Even though one sprite has a higher priority,
if portions of that higher priority sprite are transparent, the lower priority sprite or
the background display data is allowed to pass through the higher priority sprite.
SPRITE-TO-DATA DISPLAY PRIORITIES

The sprite to data (background screen) priority is selected through a software register
within the VIC chip. The Sprite to Background Display Priority Register (location
53275 ($D01B)) specifies whether a sprite passes in front of or behind the objects on the
screen background plane. This feature makes sprites seem more realistic. The default
value of this register is zero; therefore, all sprites pass in front of objects on the screen
background unless you change the bit values in this register. In other words, upon
power-up, all sprites have a higher priority than the screen background.
Bits 0 through 7 pertain to sprites 0 through 7. If a bit in this register is set,
the sprite passes behind the objects on the screen background. The bit number corre
spondence to the sprite number works the same way as the sprite enable register and
many of the other sprite registers. To set these bits, raise two to the power of the bit
position for the sprite number that you want to pass behind the screen background
objects. For instance, to cause sprite 6 to pass behind objects on the screen background,
raise 2 to the sixth power and place the hexadecimal equivalent in location 53275
($D01B) in machine language as follows:
LDA # $ 4 0 (%01000000 = binary rotation in the Monitor)
STA $D01B
To set the sprite-to-data display priorities for more than one sprite, add the values
together for two raised to the respective bit positions and store the result in location
53275 ($D01B).
SPRITE COLLISION PRIORITIES

The VIC chip has a feature that enables you to detect when a collision occurs between
sprites, or between a sprite and screen objects.
290 COMMODORE 128
SPRITE-TO-SPRITE COLLISIONS
A sprite-to-sprite collision occurs when an enabled pixel in the foreground of one sprite
overlaps an enabled pixel from the foreground of another sprite at any point on the sprite
coordinate plane. The collision may also occur at an off-screen coordinate location.
Location 53278 ($D01E) flags whether a sprite-to-sprite collision has occurred. This
register, like most of the sprite registers, has a bit which detects a collision for each
sprite. Bits 0 through 7 pertain to sprites 0 through 7. If a sprite is involved in a
sprite-to-sprite collision, the bit corresponding to the sprite involved in the collision is
set; therefore, at least two bits are always set in a sprite-to-sprite collision. These bits
remain set until they are read at which time the VIC clears the register. You should store
the value of this register in a variable until the collision or conditional code depending
on the collision is fully processed.
Once a sprite-to-sprite collision is detected, the sprite-to-sprite collision Interrupt
Request (IRQ) flag, bit 2 of location 53273 ($D019), is set and an interrupt occurs if
enabled in the IRQ Mask Register at 53274 ($D01A). When this occurs, you can
incorporate an interrupt routine to be activated upon the collision of two sprites.
Therefore, your sprite-to-sprite collision interrupt routine is only executed upon the
condition that two sprites collide. This built-in feature gives you a way to conditionally
wedge an interrupt (IRQ) routine into your application program, depending on the
behavior of the sprites on the screen.
SPRITE-TO-DATA COLLISIONS
A sprite-to-data collision occurs when an enabled pixel in the foreground of a sprite
overlaps a pixel from the foreground of an object on the screen. Location 53279
($D01F) flags whether a sprite-to-data collision has occurred. This register has a bit
which detects a collision for each sprite. Bits 0 through 7 pertain to sprites 0 through 7.
If a sprite is involved in a sprite-to-data collision, the bit corresponding to the sprite
involved in the collision is set. These bits remain set until they are read at which time
the VIC chip clears the entire register. A recommended programming practice is to store
the value of this register in a variable until the collision or conditional code depending
on the collision is fully processed.
Once a sprite-to-data collision is detected, the sprite-to-data collision IRQ flag in
bit 1 of location 53273 ($D019) is set and an interrupt occurs if enabled in the IRQ
Mask Register at 53274 ($D01A). When this occurs, you can incorporate an interrupt
routine to be activated upon the collision of two sprites. Therefore, your sprite-to-data
collision interrupt routine is only executed upon the condition that a sprite has collided
with an object on the screen foreground. Again, this gives you a way to conditionally
wedge an IRQ routine into your application program depending on the behavior of your
animated graphic objects on the screen.
Note that sprite-to-data collisions do not occur with multi-color bit pair 01
(binary). This permits those bits to be interpreted as background display data, without
interfering with sprite-to-data collisions.
10
PROGRAMMING
THE 8O-COLUMN
(8563) CHIP
292 COMMODORE 128
The Commodore 128 computer offers two types of video output: 40-column, composite
video through the VIC chip and 80-column, RGBI through the 8563 chip. The 80-column
display adds an important feature to the Commodore family of home and business
computers: The C128 can be regarded as a business machine. The 8563 chip enables the
C l 28 to display spreadsheets, wordprocessors, database managers and existing CP/M
applications in 80 columns. Now, the latest in the family of inexpensive Commodore
computers runs the Perfect series of business applications, and many other business
applications in C128 mode. In CP/M mode, the C128 runs Wordstar, and many other
popular business applications. In addition, the C l 28 supports all the hardware and
software available for the Commodore 64. The Commodore 128 is truly the complete
personal computer.
THE 8563 VIDEO CHIP

FEATURES
The primary purpose of the 8563 video chip is to display characters on the screen. The
8563 has two sets of characters, each with 256 elements. Unlike the VIC chip, however,
the 8563 can display all 512 characters simultaneously. The VIC chip displays only one
character set at a time.
The 8563 chip supports a limited bit map mode. Bit mapping can be achieved
through your own programs, preferably machine language. The BASIC 7.0 graphics
commands do not support the 80-column screen. Programming the bit-mapped screen in
BASIC is not recommended, since the language is not geared to manipulating single
display bits at a time. Later in this chapter, bit mapping the 80-column screen is
illustrated in machine language.
Another feature of the 8563 is smooth scrolling in the vertical and horizontal
directions. The 8563 chip is equipped with a set of scrolling registers that enable text to
be scrolled up, down, left or right. This is discussed later in the chapter.
PROGRAMMING THE
80-CQLUMN (8563) CHIP
Programming the 8563 video chip is a quite different from programming the VIC chip.
As you know, the registers of the VIC chip are located in the range 53248 ($D000)
through 53296 ($D030) in bank 15. Unlike the VIC chip, the 8563 has only two
memory locations in Commodore 128 I/O memory. $D600 and $D601. This means that
only two memory locations in the Commodore 128 I/O memory pertain to the 8563
video chip. Internally, the 8563 has thirty-seven internal registers, though they are not
addressable in C128 I/O memory. In addition, the 8563 has 16K of RAM of its own that
PROGRAMMING THE 80-CQLUMN (8563) CHIP 293
is independent of the Commodore 128 RAM. You must address locations $D600 and
$D601 as the gateways through which you indirectly address the thirty-seven internal
registers and 16K of 8563 RAM. You cannot directly access any of the thirty-seven
internal registers or the 16K of 8563 RAM.
Location $D600 is the Address Register, and $D601 is the Data Register.
Generally, you place an 8563 register number in the address register ($D600) then either
write to or read from the data register in location $D601. This is a simplified explanation—
the more-detailed information given in the following sections is needed to program the
8563 successfully.
8563 NOTES:
1. You cannot use BASIC PEEK, POKE, or WAIT instructions to access
the 8563, because these commands are implemented using indirect
operations. Any indirect machine instructions (such as LDA ( ),Y or
STA( ),Y) must be avoided because they result in ‘false’ bus states
which are sensed by the 8563 and subsequently acted upon as though
they were valid instructions.
2. You should not, directly or indirectly, access the 8563 during inter
rupts, because there is no way to save and restore the 8563 registers
without disrupting any I/O that might have been in progress at the
time of the interrupt.
DETAILS BEHIND PROGRAMMING

THE 8O-COLUMN CHIP
So far, you have learned that the 8563 chip has:
1. Thirty-seven internal registers.

2. 16K of independent RAM (that is not addressable in the C l 28 memory map).
3. An address register ($D600 and a data register $D601) in C l28 I/O memory.
Figure 10-1 is a summary of the 8563 registers, in the form of a register map:
294 COMMODORE 128
BITS
REG 7 6 5 4 3 2 l 0
0 HT7 HT6 HT5 HT4 HT3 HT2 HT1 HTO H orizontal Total
1 HD7 HD6 HD5 HD4 HD3 HD2 HD1 HDO H orizontal Displayed
2 HP7 HP6 HP5 HP4 HP3 HP2 HP1 HPO Horizontal Sync Position
3 VW 3 VW 2 VW1 VWO HW3 HW2 HW1 HWO Vert/Horiz Sync Width
4 VT7 VT6 VT5 VT4 VT3 VT2 VT1 VTO Vertical Total
5 — — — VA4 VA3 VA2 VA1 VAO Vertical Total Adjust
6 VD7 VD6 VD5 VD4 VD3 VD2 VD1 VDO Vertical Displayed
7 VP7 VP6 VP5 VP4 VP3 VP2 VP1 VPO Vertical Sync Position
8 — — — — — — IM1 IMO Interlace Mode
9 — — — CTV4 CTV3 CTV2 CTV1 CTVO Character Total Vertical
10 — CMl CMO CS4 CS3 CS2 CS1 CSO Cursor Mode, Start Scan
11 — — — CE4 CE3 CE2 CE1 CEO Cursor End Scan Line
12 DS15 DS14 DS13 DS12 DS11 DS10 DS9 DS8 Display Start Address hi
13 DS7 DS6 DS5 DS4 DS3 DS2 DS1 DSO Display Start Address lo
14 CP15 CP14 CP13 CP12 CPU CP10 CP9 CP8 Cursor Position hi
15 CP7 CP6 CP5 CP4 CP3 CP2 CPI CPO Cursor Position lo
16 LPV7 LPV6 LPV5 LPV4 LPV3 LPV2 LPV1 LPVO Light Pen Vertical
17 LPH7 LPH6 LPH5 LPH4 LPH3 LPH2 LPH1 LPHO Light Pen H orizontal
18 UA15 UA14 UA13 UA12 UA11 UA10 UA9 UA8 Update Address hi
19 UA7 UA6 UA5 UA4 UA3 UA2 UA1 UAO Update Address lo
20 AA15 AA14 AA13 AA12 AA11 AA10 AA9 AA8 Attribute Start Adr hi
21 AA7 AA6 AA5 AA4 AA3 AA2 AA1 AAO Attribute Start Adr lo
22 CTH3 CTH2 CTH1 CTHO CDH3 CDH2 CDH1CDHO Character Tot(h), Dsp(V)
23 — — — CDV4 CDV3 CDV2 CDV1CDVO Character Dsp(v)
24 COPY RVS CBRATE VSS4 VSS3 VSS2 VSS1 VSSO Vertical sm ooth scroll
25 TEXT ATR SEMI DBL HSS3 HSS2 HSS1 HSSO Horizontal smooth scroll
26 FG3 FG2 FG1 FGO BG3 BG2 BG1 BGO Foregnd/Bgnd Color
27 A17 AI6 AI5 AI4 AI3 AI2 A ll AIO Address Increment / Row
28 CB15 CBM CB13 RAM — — — — Character Base Address
29 — — — UL4 UL3 UL2 UL1 ULO Underline scan line
30 WC7 WC6 WC5 WC4 WC3 WC2 WC1 wco W ord Count
31 DA7 DA6 DA5 DA4 DA3 DA2 DAI DAO Data
32 BA15 BA14 BA13 BA12 BA11 BA10 BA9 BA8 Block Start Address hi
33 BA7 BA6 BA5 BA4 BA3 BA2 BA1 BAO Block Start Address lo
34 DEB7 DEB6 DEB5 DEB4 DEB3 DEB2 DEB1 DEBO Display Enable Begin
35 DEE7 DEE6 DEE5 DEE4 DEE3 DEE2 DEE1 DEEO Display Enable End
36 — — — — DRR3 DRR2 DRR1DRRO DRAM Refresh rate
37 HSyNCVSyNC Horiz, Vert Sync Polarity
Figure 10 -1 . 8S63 V D C Register Map
The numbers in the left column are the register numbers. When programming the
8563, the registers are referenced by number only, since they have no actual address in
the C l 28 memory map. Within the register map, columns 7 through 0 apply to the bits
within the registers. To the right are the register names, by function. Many of the
registers control more than one operation of the chip, so the names only reference the
primary purpose of the register. These registers are discussed individually at the end of
this chapter. Certain key registers are discussed in the next section. For an explanation
of each register, see the register-by-register description in the back of this chapter.
PROGRAMMING THE 8Q-COLUMN (8563) CHIP 295
The 16K of 8563 RAM is organized by the Kernal in the following default
structure:
TY PE OF M EM ORY 8563 RAM LOCATION
Character Display area (screen) $0000-$07FF

Character Attributes (color) $0800-$0FFF
Character Definitions $2000-$3FFF
The layout can be changed by your application program. Registers 12 and 13 (high
byte/low byte) specify the start of the character display area, which is comparable to
screen RAM in the VIC chip. Registers 20 and 21 (high byte/low byte) specify the
character attribute base address. These registers determine the starting address of display
characteristics of a character such as reverse video, underlining, flashing and color.
Register 28 determines the character set base address, the start of the character defini
tions in 8563 RAM.
As you can see, the 8563 display has the same three components as the
VIC display: screen memory, color memory, and character definitions. All three lie
in the 16K of 8563 RAM and do not show up in the C l28 memory map. These
components of the 8563 video display are discussed in greater detail in the next
section.
The third element of 8563 programming is the address and data registers that
are located in locations $D600 and $D601. These locations are the transitions, the
doorways in and out of the 8563 chip. You cannot directly reference an 8563
register or RAM location. You must indirectly read or write to $D600 and $D601 in
order to communicate between the 8563 chip and the rest of the Commodore 128.
Figure 10-2 gives description of each of the mapped registers and the bits within
them:
$D600 —» address (write): — — R5 R4R3 R2 R1 R0

status (read) : STATUS LP VBLANK — — VERO V E R 1V E R 2
$D601 -> data (r/w) : D7 D6 D5 D4D3 D3 D2 DO
Figure 10-2. Mapped Registers
In location $D600, bit 7 acts as a status bit. When you require a read or write
operation to the 8563 RAM or a register, you must check the value of bit 7 of
$D600. While it is 0, no read or write operations may take place.
296 COMMODORE 128
This is the sequence of steps to perform a read or write operation:
1. Address the registers (by number)or internal RAM location.

2. Check the value of bit 7 until it equals 1 in order for the data to be
valid.
3. Once bit 7 becomes high (1), perform the read or write operation.
Later in this chapter, you’ll see example programs that illustrate the behavior of
these registers.
The rest of the bits in $D600 except bit 6, which involves a light pen are used to
represent a register number for read or write access. Six bits are necessary since the
8563 has 37 registers; therefore, enough bits are needed to represent numbers up to 37.
Location $D601 is strictly a data register that passes values between the 8563
and the 8502 during read and write operations respectively. Each bit in this register
is a data bit. All data that passes between 8563 and the 8502 must pass through this
register.
FUNDAM ENTAL BACKGROUND

A B O U T THE 8563 CHIP
CHARACTERS
As you know, the 8563 chip is primarily designed for displaying characters within an 80
by 25 character matrix. The visible character is an 8 by 8 pixel matrix. The character
can be defined by 16 or 32 bytes and can be displayed less than 8 by 8 pixels.
PROGRAMMING THE 8O-COLUMN (8563) CHIP 297
s b k h b h b h h
a. Visible 8 by 8 pixel character.

b. H o rizo n ta l in terch aracter spacing.
c. V e rtic a l in terch aracter spacing.
Figure 10-3.
The horizontal unit of measure within a character is a pixel. The vertical unit of
measure within a character is called a scan line. A scan line is a single horizontal row of
scanned pixels referred to as one raster. The pixel and the scan line are the smallest units
of measurement by the 8563 chip in the horizontal and the vertical directions, respec
tively. The size of the displayed character image in both pixels and scan lines is
programmable and determined by four registers within the 8563: the Character Total
(R9) (vertical), Character Total (R22) (horizontal), Character Displayed (R23) (vertical)
and Character Displayed (R22) (horizontal) Registers. These registers are described in
the register by register description at the end of this chapter.
FRAMES
A fram e is a two-dimensional array (matrix) of characters, also referred to as simply the
screen. The default frame is the 80-column by 25-line screen in which characters are
displayed. Each character on the frame has a character pointer (a Display RAM byte)
that tells the 8563 which character to display in each character position within the frame.
Think of a frame as a framed picture on a wall containing a two-dimensional array
(matrix) of characters. For example, a frame specified to be 16 by 12 characters is filled
with characters as in Figure 10-4.
The frame also contains the vertical retrace and all other nonvisible portions of the
screen. The frame is programmable in size, similar to a window. The screen frame
determines the size of the viewed image, like the frame surrounding a window.
298 COMMODORE 128
\ /
H AS AD KH AD KS FD K H A
Figure 10-4.
The default frame size is 80 columns by 25 lines. However, the frame size and
characteristics are programmable, so you can change them. The 8563 has several
registers that determine the size and characteristics of the frame. These registers are as
follows:
Horizontal Total (R0)

Horizontal Displayed (R l)
Horizontal Sync Position (R2)
Horizontal Sync Width (R3)
Vertical Total (R4)
Vertical Total Adjust (R5)
Vertical Displayed (R6)
Vertical Sync Position (R7)
Vertical Sync Width (R3)
DRAM Refresh/Scan Line (R36)
These registers are covered in the register-by-register description at the end of the
chapter.
UNDERSTANDING THE 8O-COLUMN

CHARACTER DISPLAY A N D
ATTRIBUTE MEMORY
As you learned earlier in this chapter, the 8563 has three areas of memory that control
the 80-column display: character definitions, character attributes and display screen
memory.
In a previous section, the three main elements of programming the 80-column chip
were introduced: (1) the address and data registers $D600 and $D601 that map into the
C128 RAM memory map, (2) the 16K of independent 8563 RAM, and (3) the 37
internal registers. The 16K of independent 8563 RAM provide the main elements of
video display information: display memory, character attributes and character definitions.
The display memory is comparable to screen memory in the VIC chip. The character
attributes supply color and other information for the characters and parallel the function
of VIC color RAM. The 8563 character definitions correspond to the VIC character
ROM. Though the VIC and 8563 video memory are similar, the operations of the two
are different. This provides a frame of reference to the VIC chip characteristics, so you
can understand the comparison between the operation of both the VIC and 8563 video
processors.
All these video components are stored in the 16K of independent 8563 RAM.
Again, here is the breakdown of the locations where the three sections of video memory
are stored in the 8563 RAM:
TY PE O F M EM ORY 8563 R A M L O C A T IO N
Display Memory $0000-$07FF

Character Attributes $0800-$0FFF
Character Definitions $2000-$3FFF
All of the 8563 video display memory resides in the 16K section of independent
RAM. The 8563 character definitions in RAM require up to 32 bytes per character. If 32
bytes per character are used, the 8563 character definitions use 8K of memory per
character set, or a total of 16K. Otherwise, each character set occupies 4K of indepen
dent 8563 RAM, or a total of 8K for both character sets. The VIC chip requires a total
of 4K of character ROM to represent all the characters, since the VIC chip only requires
8 bytes to represent a character.
The rest of the section describes how the character display area, the character
attributes and the character definitions are addressed and interpreted.
300 COMMODORE 128
UNDERSTANDING
DISPLAY MEMORY
The frame, which was introduced earlier in this chapter, pertains to display memory, the
portion of memory you can see on the screen. Display memory resides in the range
$0000-$07FF in the 16K independent 8563 RAM. Think of the screen as a frame of 25
rows of 80 characters each. A frame may have different proportions, but for the
purposes of this discussion, assume that the frame is 80 by 25.
Each character position (display memory location) within a frame contains a
code that acts as a pointer into the character definitions that tells the 8563 which
character to display in that particular character position within the frame. For the
defined screen of frame size 25 rows by 80 columns, the first 80 pointers define the
character codes for the 80-character positions in the top row of the screen (frame).
Locations 81 through 160 define which characters are displayed in the second row of
the screen (frame), and so on. For the defined screen frame of 80 by 25, 2000
display memory locations are necessary to display the 2000 possible character positions
of the screen.
In text mode, the contents of display RAM are screen display codes like the
ones the VIC chip uses. These are not ASCII CHR$ codes. Screen codes are actually
indexes into the 8563 character RAM definitions, like the screen codes used as indexes
into the VIC character ROM. Since the first character in 8563 character RAM is
the at-sign character, it is assigned screen code 0. The A is the second character,
so it is assigned the screen code 1, and so on. For a complete list of screen codes,
see Appendix D.
ADDRESSING
DISPLAY MEMORY
The contents of registers 12 (high-byte) and 13 (low-byte) define the 16-bit start address
of the first (top-left) character in the frame (display memory). The default start address
of display memory is $0000. The address of the second character position on the
frame, that is, the character in column two, row one, is defined by the contents of
registers 12 and 13 plus 1. Add an index to the contents of registers 12 and 13 to arrive
at the address of subsequent character positions on the frame. For example, to find the
address of the character position in the first column of the second row of an 80 by 25
frame, add an offset of 80 to the contents of registers 12 and 13. To find the address of
the character position in column 1, row 3, add 160 to the contents of registers 12
and 13.
In general, add an offset of 1 for each character position to the right of the
upper-left corner of the frame. To reach the same column in the row below the current
character position, add an offset of 80. Although the screen (frame) appears as a
two-dimensional matrix of characters, the address references in memory are accessed
linearly in sequence. For example, although the character in the top-left corner of the
screen is only 24 character rows away from the character in the lower-left corner of the
screen, in memory, the locations of those characters are really 1920 (24 x 80) memory
locations apart. So a frame, the size of the default screen (80 x 25) requires 2000
memory locations to store the contents of the 2000 character positions on the
frame. The contents of each display RAM location (acting as an indexed pointer into the
character definitions) tells the 8563 which character, out of the 256 available characters
in either character set, to place at any particular position on the 80 by 25 character
frame.
CHARACTER ATTRIBUTES
The character attributes are a set of 8563 RAM locations that assign video enhance
ment qualities to the characters in the frame. Each character position in the frame
is assigned its own attribute byte, so a frame (screen) of 80 columns by 25 lines
requires 2000 attribute bytes. The default location of the attribute bytes for the 80
by 25 frame is stored in the range $0800 through $0FCF in the 16K of independent
8563 RAM.
The attributes (and the corresponding bits) that enhance the video qualities of the
characters in the frame are as follows:
B IT A T T R IB U T E
7 Alternate Character Set

6 Reverse Video Attribute
5 Underline Attribute
4 Blink Attribute
3 Red
2 Green
1 Blue
0 Intensity
The ALTERNATE CHARACTER SET (bit 7), when high (1), enables an
additional 256 characters to be displayed within the frame.
302 COMMODORE 128
The REVERSE VIDEO ATTRIBUTE (bit 6), when high (1), reverses the
foreground and background colors for the corresponding character in display memory.
See also the register-by-register description of the 8563 chip for all the details about
register 24, the Reverse Screen Register.
The UNDERLINE ATTRIBUTE (bit 5), when high (1), enables the UNDER
LINE SCAN LINE (as defined by bits 0 through 4 of register 29) in the foreground
color for that character within the frame. The underline scan line may be any scan line
within the character as determined by the lower four bits of register 29.
When the BLINK ATTRIBUTE (bit 4) is equal to 1, the character associated
with this attribute byte blinks. The blinking is achieved by repeatedly switching back
and forth between the foreground and background colors. The blink rate is determined
by bit 5 of register 24.
Attribute bits 3 through 0 determine the foreground color for the R, G, B and I
respectively. The character within the associated display memory appears in the fore
ground color determined by the R, G, B and I bits. The foreground color of each
character cell can be one of 16 colors. The actual colors on the screen can be more than
red, green and blue. The RGBI signals coming from each input and the combination of
these signals determine the actual screen color. Here are the available colors and
associated color codes. The characters’ background color is determined by bits 0
through 3 of register 26. These bits in register 26 specify the background color for all
the characters on the screen. Note, however, that in machine language these codes are
one less than the BASIC codes.
1 Black 9 Dark Purple

2 White 10 Dark Yellow
3 Dark Red 11 Light Red
4 Light Cyan 12 Dark Cyan
5 Light Purple 13 Medium Gray
6 Dark Green 14 Light Green
7 Dark Blue 15 Light Blue
8 Light Yellow 16 Light Gray
B A S IC C o lo r Codes in 80-Colum n O u tp u t
ADDRESSING CHARACTER ATTRIBUTES

The contents of registers 20 (high-byte) and 21 (low-byte) define the address of the start
of attribute storage. The default start address of attribute storage is $0800. The
contents of these two registers form the 16-bit address that marks the start of attribute
storage.
Address these attributes in the same way as addressing display memory.
To address subsequent character attributes to the right of the character in the HOME
(upper-left) position of the frame, add an offset of 1 to the contents of registers
20 and 21 for each character attribute to the right of the HOME position. To address the
character attribute directly below the HOME position (row 2, column 1), add an offset
of 80 to the contents of registers 20 and 21. To address the character attribute of the
character in the lower-left corner of the 80 by 25 frame, add a decimal offset of 1920
(24 x 80).
DISABLING CHARACTER ATTRIBUTES

Many applications programs for the 8563 chip do not need to assign the character
attributes. The 8563 has a bit that allows attributes to be disabled, saving a considerable
amount of RAM. Bit 6 of register 25 determines whether or not attributes are enabled. If
bit 6 is set (1), attributes are enabled: if the value of bit 6 is 0, attributes are disabled.
When attributes are disabled, the attribute portion of 8563 RAM may be used for other
purposes. If attributes are disabled, only one set of 256 characters is available, and
underlined, blinking and reverse video attributes are not available. In addition, all
characters are displayed in the same foreground color as determined by bit 4 through 7
of register 26. The background color is still determined by the lower four bits (3-0) of
register 26, regardless of whether attributes are enabled or not. The screen border and
the vertical and horizontal intercharacter spaces default to the background color.
Display memory may contain more memory locations and attributes than can be
displayed at a single time in an 80 by 25 frame. This portion of display memory may be
displayed by changing the contents of registers 12 and 13, the Display Start address, and
registers 20 and 21, the Attribute Start address, to reflect a new start of display memory
and attribute storage. The screen (frame) may be scrolled vertically by incrementing or
decrementing the values of the display memory pointer and attribute start addresses.
This provides a way to vertically scroll forward or backward through display memory.
In addition to scrolling vertically through display memory, the 8563 has a feature
that allows scrolling through display memory in the horizontal direction. Both of these
topics are covered later in this chapter.
CHARACTER DEFINITIONS
The character definitions residing in 8563 RAM specify the character patterns you see
on the 80-column screen. The character patterns in memory define the pixel formation
of each member of the character set visually represented on the screen. The character
definitions can be redefined by your application program.
The default memory range for the character definitions is locations $2000 through
$3FFF in 8563 RAM. This range is a total of 8K (8192 bytes). This is enough memory
to represent two character sets of 256 characters each. The attribute bit 7 of each
attribute byte selects either of the character sets for each character position within a
frame.
The 8563 characters are represented by either 16 or 32 bytes, depending on the
value of the lower five bits (4-0) of register 9. If the value of the lower five bits of
304 COMMODORE 128
register 9 is less than 16, each character is defined by 16 bytes in character storage.
Therefore, each character set requires a total of 4096 bytes of 8563 RAM. If the value
of the lower five bits of register 9 is greater than or equal to 16, each character is
defined by 32 bytes of 8563 character RAM. In this case, each character set occupies
8192 bytes (8K) of 8563 character RAM memory and the defined characters are
displayed using more than 16 scan lines each. This occupies the full 16K of 8563 RAM
and leaves no space for attributes. These taller characters are rarely used, however,
because the characters are twice as tall as they are wide and seem out of proportion.
The first byte of each character definition in memory represents the top horizontal
scan line of the character displayed on the screen. Depending on the value of the lower
five bits of register 9, a character will have either 16 or 32 scan lines, each being 8 bits
(pixels) wide.
Although the character definitions in 8563 character RAM are defined by either 16
or 32 8-bit bytes stacked one on top of another, the Display Data Bus is only 8 bits
wide, which limits the actual image of the character to 8 horizontal pixels. The total
width of the character including the horizontal intercharacter spacing may exceed 8
pixels. See the definition of a character in the beginning of the chapter and the
register-by-register description at the end of this chapter for details.
ADDRESSING THE
CHARACTER SET PATTERNS
The default location of the character definition start at address $2000 in 8563 RAM, and
since the starting location is programmable, you may change it. The starting location of
the character set definitions is determined by bits 5 through 7 of register 28. These are
the most significant three bits of the 16-bit address that points to the start of the
character definitions in 8563 RAM. Bits 4 through 0 of register 9 make up the five least
significant bits for the address that points to the start of the 8563 character definitions.
However, if the value of bits 4 through 0 in register 9 is greater than or equal to 16, only
bits 7 through 5 are significant, since the characters are made up of 32 bytes each and
the full 16K of 8563 RAM memory is allocated for the two character sets, 8K for each
set.
READING 8563
(RAM A N D ) REGISTERS
As you know, the 8563 chip has its own 16K of RAM and its own set of registers.
Neither of these appear in the standard Commodore 128 memory map. The only
memory locations that apply to the 8563 chip in Commodore 128 memory are the locations
D600 and $D601. These are the 8563 address and data registers, respectively. In order
to read or write to 8563 RAM or registers, you must pass data through the address and
data registers. This section explains how to read the 8563 registers. The next section
illustrates how to write to the 8563 registers and RAM.
READING AN 8563 REGISTER

Reading an 8563 register requires five distinct programming steps. Although this may be
accomplished in BASIC, it is strongly recommended that you program the 8563 in
machine language. Briefly, here are the five machine language programming steps, the
algorithm, to read an 8563 register. A more detailed explanation and a program to read
an 8563 register follows.
1. Load (immediate) the X register with the register number that you want to
read.
2. Store the contents of the X register (the 8563 register number) into the 8563
address register ($D600).
3. Check the status bit, bit 7 of address register $D600, until it is high(1).
4. If the status bit is low (0), go back and check it again.
5. Otherwise, load the accumulator with the contents of the 8563 data register
$D601. Store the contents of the data register into a C l28 RAM location (a
variable) for future use and print the variable value.
Since many of the 8563 registers are pointers to a16-bit address, aregister pair is
required, one register for the high byte and one for the low byte. In this case, simply
increment the X register and repeat the five-step process to read the register contents of
the low byte. The high-byte register comes first in the register order, therefore the first
read is the high byte of a register pair, increment the X register and repeat the
process for the low byte. When the 8502 uses a register pair to point to a 16-bit
address, the high byte is the low byte plus 1.
Here’s a program that reads the high- (R20) and low- (R21) byte register pair that
acts as a pointer to the start of the attribute storage in the 8563 RAM. The value stored
in locations $FB and $FC return the default start (high/low byte) of the attribute storage.
This pointer can be relocated to point to a section of 8563 RAM in which you define, as
is the case for display RAM and the character patterns.
EX A M P LE :
Reading a register (attribute base address)

306 COMMODORE 128
10 SYS D E C ( " 1 8 0 0 " )

20 A$=HEX$(PEEK( 2 5 1 ))
30 B$=HEX$( PEEK(252) )
40 C$=R IG H T$(A $,2)
50 D$=RIGHT$(B?,2)
60 E$=C$+D$
70 F$="$"
80 G$=F$+E$
90 P R IN T "START ADDRESS = ";G$
RE ADY.
MONITOR
PC SR AC XR 'YR S P
FB000 00 00 00 i0 0 F 8
01800 A2 14 LD X #$14
01802 2 0 OF 18 JSR $ 1 80F
01805 85 FB STA $FB
01807 E8 INX
01808 2 0 OF 18 JSR $ 1 80F
0180B 85 FC STA $FC
0180D 60 RTS
0180E 00 BR K
0180F 8E 00 D6 STX $D600
01812 2C 00 D6 BIT $D600
01815 10 FB BPL $1812
01817 AD 0 1 D6 LDA $D601
0181A 60 RTS
Here are the details behind each instruction.
■ The first instruction ($1800) loads the X register with register number 20 ($14),
the high byte of the attribute start address.
■ The second instruction jumps to the subroutine at $180F.
■ The first instruction of the subroutine stores the contents of the X register,
($14), into the 8563 address register (location $D600 in C128 I/O memory).
■ The second instruction of the subroutine (BIT $D600) tests the UPDATE
READY status bit (7) of the 8563 address register. This is necessary since a
read or write operation does not take place as soon as those instructions are issued.
The Update Ready bit (bit 7) of the 8563 Address register determines when the
read (or write) instruction can be executed. The read operation is pending according
to the value of bit 7 of the address register. While the read operation is pending
(in prelatch), the value of bit 7 is 0. This indicates that the read (or write)
operation data placed in register 31, is not yet valid. Only when the value of bit
7 in the address register is equal to 1, is the data in the data register ($D601) valid.
The behavior of the 8563 chip and the way data is transferred between the 8563
registers and the address and data registers suggests the use of the BIT instruction. The
nature of the 8502 BIT instruction is tailor-made for checking the Update Ready bit
status of location $D600. When the BIT instruction is encountered by the microproces
sor, the value of bit 7 is transferred into the negative flag bit of the 8502 status register.
If the value is equal to 0, the negative condition does not exist. If the value is equal to 1,
a negative condition does exist. Depending upon the value of this bit, you can condition
ally branch to instructions that carry out an operation according to the value of this flag.
T hat’s where the third instruction in the above subroutine ($1815) comes into play.
■ Instruction 3 (BPL $1812) is a conditional instruction that branches depending

upon the value of the negative flag bit in the 8502 status register. In this
program, the BPL (Branch on Result Plus) instruction branches back to
location $1812 and executes the BIT instruction while the negative flag bit is
equal to 0. The BIT instruction is again performed, and the value in bit 7 is
again transferred to the negative flag bit in the 8502 status register. Then the
BPL instruction is encountered in the loop for a second time. If the value of the
status bit is still equal to 0, the branch is performed again. The loop continues
until the negative status bit is equal to one. Only when the status bit is equal to
1 does the loop fall through to the instruction immediately following the branch
instruction.
Prior to the loop falling through to the instruction after the branch, the
Update Ready status bit becomes a 1, it is transferred to the negative flag bit in
the 8502 status register, the branch is skipped, and the data can be read (loaded
with the fourth instruction of the subroutine ($1817) into $D601. Only after the
status bit becomes high (1) is the data in $D601 valid. This is the point where the
data is communicated from the 8563 chip to the Commodore 128 memory. Now
you can read and act upon the contents of the address within register $14 of the
8563 chip.
■ The fourth instruction loads the value stored in the 8563 data register ($D 601)
into the accumulator.
■ The third instruction of the main program stores the value of the data register
into location $FB for future use. Subsequent read operations from ($D601)
destroy the original value, so it is always a good practice to store the contents of
$D601 into a variable or available RAM location.
■ The next instruction, INX, increments the contents of the X register, which
selects the next register number, $15, so the low byte of the 16-bit address
of the start of attribute storage can be read.
■ The rest of the program performs the same sequence of steps above for the
low byte of the start of attribute storage except for the RTS instruction. The
RTS instruction returns program control back to a main program.
The BASIC portion of the program prints the high and low bytes of the 16-bit
address that defines the start of attribute storage.
To find the actual address in 8563 RAM of the start of attribute storage, put the
high byte and low byte together, and the four-digit hexadecimal number defines the start
of attribute storage. Within the monitor, issue this memory command:
M FB
The contents of $FB and $FC (high/low bytes) mark the start of attribute storage.
308 COMMODORE 128
WRITING TO AN
8563 REGISTER
The five-step algorithm to read data from an 8563 register is similar to writing to an
8563 register. In the last section. Step 5 in the read algorithm loaded the accumulator
with the value contained in the 8563 data register ($D601). When writing, perform the
opposite operation. Instead of loading the accumulator as in a read operation, store the
accumulator value into the data register.
H ere’s an algorithm to write to an 8563 register:
1. Load the X register with the register number in which you want to write.
2. Load the accumulator with a value that you are writing to the 8563 register.
3. Store the contents of the X register into the 8563 address register at $D600.
4. Check bit 7, the status bit, of the Address Register until it is high (1).
5. If the status is low (0), loop back and check it again until it is high.
6. When the status bit (7) is high, store the value contained in the accumulator
into the 8563 data register ($D601).
Here’s a sample program that writes to register pair R18 and R19, the Update RAM
location. Register pair R18 and R19 is the gateway into the independent 8563 RAM. For
now w e’ll illustrate how to write to an 8563 register. Later in this chapter, w e’ll take
it a step farther to explain how to read and write to the 16K of independent 8563 RAM.
READY.
MONITOR
PC SR AC XR YR SP
FB000 00 00 00 00 F8
01800 A2 12 LD X #$12
01802 A9 20 LDA #$20
01804 20 IE 18 JSR $ 1 81E
01807 E8 INX
01808 A9 00 LDA #$00
0180A 20 IE 18 JSR $181E
0180D EA NOP
0180E A2 12 LDX #$12
01810 20 2C 18 JSR $ 1 82C
01813 85 FB STA $FB
01815 E8 INX
01816 20 2C 18 JSR $ 182C
01819 85 FC STA $FC
0181B 60 RTS
0181C 00 BRK
0181D 00 BRK
0181E 8E 00 D6 STX $D6 00
01821 2C 00 D6 BIT $D600
01824 10 FB BPL $1821
01826 8D 01 D6 STA $D601
01829 60 RTS
0182A EA NOP
0182B EA NOP
0182C 8E 00 D6 STX $D600
0182F 2C 00 D6 BIT $D600
01832 10 FB BPL $ 1 82F
01834 AD 01 D6 LDA $D601
01837 60 RTS
■ The first instruction (at $1800) loads the X register with the register number
18 ($12).
■ The next instruction loads the accumulator with the high byte of the value you
are placing in register 18 ($20).
■ Instruction 3 calls the subroutine at location $18IE, which is essentially the
same routine as in the read register example in the last section, with one
exception. The instruction in the subroutine immediately before the RTS is a
store instruction to the 8563 data register at $D601. The two instructions before
it form a loop that continually checks the Update Ready status bit pending the
write operation, just as the read program did. The write operation (STA
$D601) is not performed until the status bit 7 in the address register becomes
high (1). Once the status bit becomes equal to a 1, the high byte of the address
in register 18 is written to the data register ($D601). The subroutine returns
control to the main body of the program with the RTS instruction.
■ The program resumes with the fourth instruction INX. This increments the
register number.
■ The fifth instruction (LDA # 00) loads the value of the low byte of the address
to be written to register 19 ($13) into the accumulator.
■ The subroutine at $18 IE is called again, and the same process is repeated.
■ Upon return of the subroutine, the same read routine example from the last
section is performed (the subroutine between $ 182C—$ 1837) to ensure that the
values that were written to registers 18 and 19 are valid. These values are
printed by the BASIC routine. To see those values, use the memory command
within the monitor as follows:
M FB
The first two bytes displayed are:
20 00
This shows that the address in the update location register pair (R18, R19) has
been written to successfully and is valid.
W RITING TO 8563 RAM

Now take the example from the last section one step farther. The last example illustrated
how to write to an 8563 register. It just so happens that the register chosen for the
example is the update location register pair (R18, R19). This is the register pair that
enables you to write to and read from the 16K of independent 8563 RAM. This section
expands on the previous example and shows you how to write to the 16K of 8563 RAM.
The program given below fills display memory in 8563 RAM with characters, in
the process displaying the characters on the screen. The program fills display RAM
located in the default range $0000 through $07CF, with screen code character 102
($66). See Appendix D for a complete list of screen code characters.
H ere’s the listing:
310 COMMODORE 128
READY.
MONITOR
PC SR AC XR YR SP
FB000 00 00 00 00 F8
01800 A2 12 LDX #$12

01802 A9 00 LDA #$00
01804 20 21 18 JSR $1821
01807 E8 INX
01808 20 21 18 JSR $1821
0180B A0 00 LDY #$00
0180D A9 08 LDA #$08
0180F 85 FA STA $FA
01811 A9 66 LDA #$66
01813 20 IF 18 JSR $181F
01816 C8 INY
01817 DO FA BNE $1813
01819 C6 FA DEC $FA
0181B DO F6 BNE $1813
0181D 00 BRK
0181E 00 BR K
0181F A2 IF LDX #$1F
01821 8E 00 D6 STX $D600
01824 2C 00 D6 BIT $D600
01827 10 FB BPL $1824
01829 8D 01 D6 STA $D601
0182C 60 RTS
To execute this program, issue the Go (G) command within the Machine Lan
guage Monitor. In order for this program to operate, I/O memory must be in context in
the currently selected monitor configuration. I/O memory is always in context within
configuration 15 ($F), so issue this command to execute the program:
G F I 800
The first five instructions (memory $1800-$ 1808) and the subroutine stored in
memory locations $1821 through $182C are the same instructions contained in the program
that wrote to registers 18 and 19 in the last section. However, instead of placing the address
$2000 in register pair 18 and 19 as in the last section, the address $0000 is placed
there to mark the default starting location of display memory in 8563 RAM. This is where
the program begins to place screen code 102 in display RAM. The full 2000 bytes (8 pages
of memory minus 48 bytes) of display RAM are filled with this character screen code.
Here’s how the rest of the program works. The instruction stored at location
$180B, LDY # $ 0 0 , initializes the Y register to 0. The next instruction, LDA #$08
loads the accumulator with the value 8. Then, STA $FA, stores the contents of the
accumulator, which initializes this location to 8. Location $FA is used later in the
program as a counter for the number of pages to fill with screen code 102. The
instruction stored at $1811, LDA # $ 6 6 , loads the accumulator with the hexadecimal
value ($66) of decimal screen code 102.
The instruction starting in location $1813, JSR $ 18IF, calls the subroutine located
at address $ 18IF. This subroutine is the crux of the program and actually writes the data
to the display memory (screen). This is essentially the same routine as the subroutine
located between $ 18IE and $1829, the one from the last section. However, there is one
key difference. The first instruction of the subroutine stored starting at $18IF, LDX
# $ 1 F , loads the X register with the register number 31 ($1F). When writing to 8563
RAM, you must address register 31 because it acts as the internal data register of the
8563 chip. All data that will be written to the 8563 RAM must pass through register 31.
The next three instructions:
STX $D600
BIT SD600
BPL $1824
check the status bit (7) of the address register until the value becomes equal to 1
(high). These are the same three instructions used in the last two sections when reading
or writing to a register and in the first subroutine of this program.
Once the status bit of the 8563 address register becomes high, the data residing in
the accumulator (in this case) is written to the external 8563 data register at address
$D601, as in the instruction stored at location $1829. In addition, the data in the
accumulator is written to the address contained in the update register pair (R18, R19).
This occurs since you addressed the internal data register 31 by placing the value 31 in
the 8563 address register at $D600. Upon the status bit becoming high, all data in
register 31 is placed in the current address of 8563 RAM specified by the update register
pair (R18, R19). You placed this address in the update register pair in the very
beginning of the program.
At this point, the 8563 hardware takes over. It automatically increments the
address contained in the update register pair (R18, R19). You as the programmer do not
increment the update register pair’s contents; the 8563 chip does this for you automati
cally. You can say that the update register pair is auto-incrementing. In addition, the
8563 reads the data from the incremented address within 8563 RAM, and places the
contents in the internal 8563 register 31. This also saves programming effort. Remem
ber, both of these tasks are performed by the 8563 hardware, not the programmer.
Since the update register pair automatically increments the 8563 RAM addresses,
all you have to do is continually call the subroutine that writes the data to each
successive display RAM location. That’s what the instructions between addresses $1813
and $181B do in the program. These instructions call the subroutine for eight pages of
display RAM, or a total of 2048 times. The write subroutine writes only one character to
the screen per call, so to fill the entire screen, it must be called 2000 (80 x 25) times.
This sample BASIC program shows how to properly read and write the 8563
registers, and read, modify, write any location in 8563 RAM. The two ML subroutines
are the key- and they can be used from either BASIC as shown here, or from your ML
program.
312 COMMODORE 128
100 V I = D E C ( " 1 8 0 0 " ) : V 0 = D E C ( " 1 8 0 C " ) : BANK15

110 READ A$
115 DO U N T I L A $ = " E N D "
116 POKE VI+I,DEC(A$) (install ML routines)
117 1= 1+1
118 READ A$
119 LOOP
120 DATA 8E,00,D6,2C,00,D6,10,FB,AD,01,D6,60
130 DATA 8E,00,D6,2C,00,D6,10,FB,8D,01,D6,60,END
140 DO
(update location HIGH)
150 PRINT
160 INPUT"ADDRESS (0-16383)";AD (update location LOW)
170 SYS VO, AD/256, 18
180 SYS VO, ADAND255,19
(READ memory data)
190 SYS V I ,,31
200 RREG A
210 PRINT"THE 8563 MEMORY A T ” ;A D ;" IS";A (update location HIGH)
220 INPUT"NEW DATA (0-255) ";A (update location LOW)
230 SYS VO, AD/256, 18
240 SYS VO, ADAND255,19 (WRITE memory data)
250 SYS VO, A ,31
260 LOOP
The following disassembly shows the two ML routines used in the previous BASIC
program and in the Read register, Write register and Write to RAM examples in the
preceding sections . They are
01800 8E 00 D6 STX $D600

01803 2C 00 D6 BIT $D600
01806 10 FB BPL $1803 (routine to READ 8563 register in .X)
01808 AD 01 D6 LDA $D601
0180B 60 RTS
0180C 8E 00 D6 STX $D600

0180F 2C 00 D6 BIT SD600
01812 10 FB BPL $180F (routine to W RITE .A register in .X)
01814 8D 01 D6 STA $D601
01817 60 RTS
BLOCK WRJTE A N D BLOCK COPY

The 8563 chip’s Block Write and Block Copy features are extensions of the write
operation described in the last section. The Block Write feature writes the same data
contained in the accumulator into a specified number of successive 8563 memory
locations. H ere’s the procedure:
1. Place the initial address, where writing in RAM begins, into registers 18
(high byte) and 19 (low byte).
2. Next, load the accumulator with the data you want to place in successive 8563
RAM locations.
3. Load the X register with the number 31 which specifies the data register
within the 8563.
4. Wait for the Update Status Ready bit to become high (1). So far this is the
same procedure as writing to 8563 RAM which was described in the Writing
to 8563 RAM section.
5. To select the Block Write feature, clear (0) bit 7 of register 24, using the
same procedure you use to write to a register.
6. Finally, place the number of successive memory locations you want to write
to 8563 RAM in register 30.
At this point, the 8563 writes the data in register 31 starting from the location deter
mined by registers 18 and 19 through the number of successive memory locations specified
by register 30. The contents of registers 18 and 19 are automatically incremented to access
the next location to be written. Upon completion of the Block Write, registers 18 and 19
contain the last address plus one of the 8563 RAM locations in which it was written.
Note that one write cycle follows the initial write operation to register 31;
therefore, the quantity written to register 30 should be one less than the number of 8563
memory locations in which you want to write.
BLOCK COPY
The Block Copy feature copies (reads and writes) data from one section of 8563 RAM
to another. To copy a block of 8563 RAM:
1. Place the destination address, where copying in RAM begins, into registers
18 (high byte) and 19 (low byte).
2. Wait for the Update Status Ready bit to become high (1).
3. Set (1) bit 7 of register 24, which selects the Block Copy feature. At this
point, the 8502 writes the initial source address for the block copy to registers
32 (high byte) and 33 (low byte).
4. Place the number of successive memory locations to copy into register 30.
The 8563 reads the contents of the first source address and writes the data to the
first destination address. The 8563 RAM addresses are automatically incremented and
copying continues until the number of memory locations specified by register 30 are copied.
READING FROM 8563 RAH

The process of reading from 8563 RAM is the same as reading from a register, except
you read from the data register, register 31.
PROGRAMMING THE 8563 CURSOR

The 80-column cursor is programmable in its image and movement. First, the cursor
image may be programmed in four different ways. Bits 6 and 5 of register 10 control the
following four characteristics of the cursor:
BIT VALUE CH ARACTERISTIC

6 5
0 0 Solid cursor
0 1 No visible cursor
1 0 Slow cursor blink rate (Vi6 the frame rate)
1 1 Fast cursor blink rate C/3 2 the frame rate)
314 COMMODORE 128
Bits 4 through 0 in register 10 determine the bottom scan line in which the cursor
definition begins within the 8 scan line cursor character block. Scan line 0 is at the top
of the cursor character block, while scan line 8 is at the bottom.
The interrupt-driven screen editor has an indirect location which controls the
cursor programming mode. Location $0A2B in C l28 RAM is the shadow register which
controls the different ways to program the 8563 cursor.
Bits 4 through 0 in register 11 specify the top scan line in which the cursor
definition ends within the 8 scan line cursor character block. Bits 4 through 0 in registers
10 and 11 specify the height of the cursor. These registers enable you to program the
cursor as a solid block cursor (by default) or as an underlined cursor.
CURSOR P O S ITIO N IN G
Register 14 (high byte) and 15 (low byte) determine the position of the cursor within the
displayed frame (screen). This register pair indicates the address of the cursor within
8563 display memory. The address of the cursor can be one that is not in the display
area of the current frame. If this address is within the address range of the current frame
(screen), the character position where the cursor is displayed appears in flashing reverse
video, as do most cursors.
The interrupt driven screen editor also has an indirect location that controls the
cursor column and row positioning. Location $0A35 in C128 RAM is the shadow
register for the 8563 cursor column and row selection.
BIT-MAPPING THE
80-CQLUMN DISPLAY
The 8563 chip supports a bit-map mode. The resolution of the 80-column bit-map screen
is 640 by 400 pixels. The 8563 addresses the bytes that make up the bit map across the
width of the screen, instead of in 8 by 8 character cells as with the VIC chip.
Bit-map mode is selected by setting bit 7 of register 25. The Kernal initializes
bit 7 to 0 to select text mode.
The three main components of an 8563 bit-map display, bit-map data, display
memory and color, are similar to the VIC chip in some ways but differ in others. For
example, the VIC has an 8000-byte RAM bit map, in which each bit is assigned an
individual pixel on the visual screen. Since the 8563 bit map is twice the resolution of the
VIC screen, it uses twice the memory: 16,000 bytes (a little less than 16K). The
one-to-one correspondence between bits in memory and pixels on the screen is similar,
but they are addressed differently. The addresses to access those bits are laid out
differently in memory.
Figure 8-25 in chapter 8 illustrates how the VIC bit map is stored in memory. The
VIC screen stores 8 bytes of the bit-map in the 8 bytes that would normally make up a
character in character mode. For instance, the first 8 bytes of the visual bit map are
stored in the first (HOME) character position, which defaults to 8192 ($2000) to 8199
($2007), or the first eight pixels of the first eight raster rows. The character to the
immediate right of the HOME position stores the second eight pixels of the first eight
visual raster rows. The 200 visible VIC screen raster rows are 320 pixels wide. The 200
visible 8563 raster rows are 640 pixels wide. Both the VIC and 8563 screens have
raster rows which are not visible on the screen. This portion is known as the vertical retrace.
The 8563 screen is stored byte by byte across a raster row. The first eight pixels of
the first raster row are stored in the first byte of the 8563 bit-map byte, address $0000 in
8563 RAM. The second eight pixels of the first row are stored in the second byte of the
bit map ($0001 in 8563 RAM). The third eight pixels are stored in location $0002 and
so on. The first raster row is stored in the first eighty memory locations in 8563 RAM,
the second raster row is stored in the second eighty (80-159) 8563 RAM addresses and
so on. See Figure 10-5 below for an illustration.
8563 CHARACTER COLUMNS
RASTER
ROW 1 2 3 4 5 6 7 8 9 10 11 . 80
1 $0000 $0001 $0003 $40004$0005 $0006 $0007 $0009 $000A $000B$000C .. . $004F
2 $0050 $0051 $0052 $0053 $0054 $0055 $0056 $0057 .......................................... $009F
3 S00A 0...........................................................................................................................................
4 $00F0 ...........................................................................................................................................
25 $ 3 E 3 0 ............................................................................................................................. $3E7F
Figure 10 -5 . T h e Relationship B etw een 8563 B it M ap D a ta and Screen Pixels
The collection of the 640 by 200 pixels on the screen form the image in this way: bits
that are set (1) enable screen pixels in the foreground color and clear (0) bits take on the
background color.
TH E BIT-MAP POINTER
The pointer to the start of the 8563 bit map defaults to the start of display memory at
8563 RAM location $0000. The display memory start address is contained in registers
12 (high) byte and 13 (low byte). The contents of this address mark the address of
the first eight pixels of the top raster row on the 8563 screen (frame). Add an index to
access successive pixels in successive raster rows.
Since the full 640-by-200 bit map uses almost all of the 16K available in 8563
RAM, there is no room available in memory for attributes such as color.
COLOR ASSIGNMENTS
The pixel color assignments for the 8563 bit map are based on the value of the bit in
memory which corresponds to the pixel on the screen. If the value of the bit in memory
is equal to zero, the pixel is colored in the background color according to the value of
bits 0 through 3 in 8563 register 26. If the value of the bit in memory corresponding to
the screen pixel is equal to one, the pixel is assigned the foreground color determined by
bits 4 through 7 in register 26.
The full-sized 640-by-200 pixel bit map bit takes up almost the full 16K of 8563 RAM.
You may wonder how you can use the attributes for the corresponding screen locations.
316 COMMODORE 128
You have two choices: either disable the attributes or reduce the size of your bit
map by 2000 bytes, then define and enable the attributes. With a full-size bit map, you
cannot implement the attributes, because there is simply not enough independent 8563 RAM
to do both. If attributes are disabled, the 8563 bit-map screen uses only two colors, as
specified by the background and foreground color codes in register 26.
If you decide to reduce the size of the 8563 bit map and you want to implement
the attributes, set bit 6 of register 25 to enable attributes. Then place the high-byte
and low-byte addresses in registers 20 and 21, respectively, to define the start of
attribute storage. See the Character Attributes section for more information on attributes.
TH E BIT-MAP FRAME SIZE

The size of the bit-map frame defaults to the 640 by 200 raster row dimensions. This
frame (screen) size may be reduced. See the Register-by-Register Description for details
on altering the size of the visual bit-map image.
The next section provides an example of a bit-map program in machine language
containing a program listing and details about the important routines.
8563 BIT MAP PROGRAM

As you know, the 8563 supports bit map mode; however, the BASIC graphics com
mands are not displayed on the 8563 screen. This section contains a program which
transfers the 40-column VIC screen to the 8563 display screen. The screen is still 40
columns wide, though it provides the vehicle with which to understand how the 8563 bit
map screen is addressed. For easier programming, more graphics features and general
functional design use the VIC chip when bit mapping high resolution graphics. The VIC
chip is principally designed for these purposes. The 8563 is primarily designed for the
80-character column display capability.
H ere’s the program listing as it appears in the Machine Language Monitor. Enter
the Monitor and start entering the program at address $1800.
. 01800 20 B3 18 JSR $ 1 8B3

. 01803 A2 19 LDX #$19
. 01805 A9 80 LDA #$80
. 01807 20 77 18 JSR $1877
. 0180A A9 20 LDA #$20
. 0180C 85 FB STA $FB
. 0180E A9 00 LDA #$00
. 01810 85 FA STA $FA
. 01812 A2 12 LDX #$12
. 01814 20 77 18 JSR $1877
. 01817 E8 INX
. 01818 20 77 18 JSR $1877
. 0181B A9 00 LDA #$00
. 0181D 85 Bl STA $Bl
. 0181F A9 19 LDA #$19
. 01821 85 9B STA $ 9B
. 01823 A9 07 LDA #$07
. 01825 85 9C STA $9C
. 01827 A9 27 LDA #$27
PROGRAMMING THE 80-CQLUMN (8563) CH IP 317
01829 85 FE STA $FE

0182B EA NOP
0182C EA NOP
0182D A2 00 LDX #$00
0182F Al FA LDA ($FA,X)
01831 20 75 18 J S R $1875
01834 20 83 18 J S R $1883
01837 EA NOP
01838 EA NOP
01839 EA NOP
0183A C6 FE DEC $FE
0183C DO EF BNE $ 1 82D
0183E A5 Bl LDA $Bl
01840 DO 1C BNE $ 185E
01842 EA NOP
01843 EA NOP
01844 EA NOP
01845 A2 00 LDX #$00
01847 Al FA LDA ( $FA,X )
01849 20 75 18 J S R $1875
0184C 20 8F 18 J S R $ 188F
0184F 20 9B 18 J S R $ 1 89B
01852 C6 9C DEC $ 9C
01854 DO Dl BNE $1827
01856 A9 01 LDA #$01
01858 85 Bl STA $Bl
0185A 4C 27 18 JMP $1827
0185D EA NOP
0185E A9 00 LDA #$00
01860 85 Bl STA $Bl
01862 A2 00 LDX #$00
01864 Al FA LDA ( $FA,X )
01866 20 75 18 J S R $1875
01869 20 8F 18 J S R $ 188F
0186C 20 A9 18 J S R $ 1 8A9
0186F C6 9B DEC $ 9B
01871 DO BO BNE $1823
01873 00 BR K
01874 00 BR K
01875 A2 IF LDX #$1F
01877 8E 00 D6 S T X $D600
0187A 2C 00 D6 B I T $D600
0187D 10 FB BPL $ 1 87A
0187F 8D 01 D6 STA $D601
01882 60 RTS
01883 18 CLC
01884 A5 FA LDA $FA
01886 69 08 ADC #$08
01888 85 FA STA $FA
0188A 90 02 BCC $ 188E
0188C E6 FB INC $FB
0188E 60 RTS
0188F AO 28 LDY #$28
01891 A9 00 LDA #$00
01893 20 75 18 J S R $1875
01896 88 DE Y
01897 DO F8 BNE $1891
01899 60 RTS
0189A EA NO P
0189B 38 SEC
0189C A5 FB LDA $FB
0189E E9 01 SBC #$01
018A0 85 FB STA $FB
018A2 A5 FA LDA $FA
018A4 E9 37 SBC #$37
018A6 85 FA STA $FA
018A8 60 RTS
018A9 E6 FA INC $FA
018AB DO 02 BNE $ 1 8AF
018AD E6 FB INC $FB
318 COMMODORE 128
018AF 60 RTS
018B0 EA NOP
018B1 EA NOP
018B2 EA NOP
018B3 A9 00 LDA # $ 0 0
018B5 85 FC STA $FC
018B7 A9 3F LDA # $ 3 F
018B9 85 FD STA $FD
018BB A2 19 LDX # $ 1 9
018BD A9 80 LDA # $ 8 0
018BF 20 77 18 JSR $1877
018C2 EA NOP
018C3 EA NOP
</>
=#=
o
o
018C4 A9 00 LDA
018C6 A2 12 LDX # $ 1 2
018C8 20 77 18 JSR $1877
018CB E8 INX
018CC EA NOP
018CD EA NOP
018CE 20 77 18 JSR $1877
018D1 A9 00 LDA # $ 0 0
018D3 20 75 18 JSR $1875
018D6 C6 FC DEC $ F C
018D8 DO F7 BNE $ 1 8D1
0 1 8 DA C6 FD DEC $ F D
018DC DO F3 BNE $ 1 8 D l
018DE 60 RTS
The instructions in this explanation are referenced by the memory addresses that
appear in the left column of the machine language monitor output as in the other
machine language program explanations.
The first instruction (J$R $ 18B3) jum ps to a subroutine to clear the bit map (8563)
screen.
The next sequence of instructions ($1803-$ 1809) set bit 7 of 8563 register 25 to
select bit map mode. This bit defaults to 0 for text display. The subroutine starting at
$1877 is the familiar register write routine you have seen earlier in the chapter. Notice
that if the subroutine had started at location $1875, it would have written to 8563 RAM
instead of an 8563 register. This routine is called to do so later in the program.
The trick to the success of this program is addressing both the VIC screen and the
8563 screen correctly. The second sequence of instructions ($ 180A—$ 1816) define the
high byte of the 16-bit address for the start of both the VIC and 8563 screens. First the
X register is loaded with the value 18 ($12) which selects 8563 register 18, the 8563
RAM address high pointer. The accumulator is loaded with the value $20, which is
stored in location $FB, the VIC bit map pointer (high byte). Now the accumulator is
cleared to zero and the write register subroutine places the value zero into the 8563 bit
map (high byte) pointer.
The instructions $1817 through $181A perform the same operation for the low
byte of both the VIC and 8563 bit map pointers.
The instructions between $1820 and $1827 initialize the variables $FC, $FD, $B1
and $9C to zero. The variables $FC and $FD are the low and high byte pointers for the
8563 bit map RAM addresses. The variable $B1 is the column counter for the VIC
screen. The variable $9C is the counter that keeps track of how many addresses to skip
in 8563 RAM. This is explained later in the program.
Here’s where a major difference between the two video chips becomes important—
the way the bit map corresponds to the image on the screen. As you know, the VIC
screen stores eight rows of 8 pixels each to make up one character cell. Each row of the
eight pixels in a character on the screen are stored consecutively in memory. For
example, the HOME character position is stored in default locations $2000 through
$2007 for the bit map. The next character position (as if the display was standard text)
to the right of the HOME Position is stored between $2008 and $2015. So the first
character row is stored between $2000 and $2A00 ($2000 plus (8*8) * 40 = 2560). This
is represented graphically in the preceding section (Figure 10-5).
On the other hand, the 8563 stores a bit map across raster rows on the screen.
Each set of eight pixels of a raster row is stored in a memory location in 8563 RAM
starting at $0000. For example, the first raster row is stored in locations $0000 through
$0050, one location for each group of eight pixels. The 8563 screen is 640 pixels across;
therefore each row needs 80 bytes of storage in 8563 RAM.
Now you see how the two bit maps are addressed differently. To fully understand
how this program gets data from the VIC bit map and places it in the correct spot on the
8563 bit map, this algorithm is provided. The actual instructions to perform these steps
are added later. H ere’s the general scheme you need to follow:
1. Start at the beginning of both the VIC and 8563 bit map locations.
2. Get a VIC bit map byte.
3. Store it in the 8563 bit map.
4. Increment the VIC bit map pointer 8 times (the 8563 pointer increments itself
after a write).
5. Do steps 2, 3 and 4 thirty-nine times.
At this point, you have just transferred the first raster row of the VIC bit map onto
the 8563 bit map.
6. After each raster row is transferred, subtract 311 from the VIC bit map pointer.
This is done 7 times. On the eighth time, the VIC bit map pointer is
incremented by 1 to get the to the start of the next VIC character cell.
7. Perform 40 blank write operations (write zeros to the 40 blank columns) to
get to the beginning of an 8563 raster row. Remember, the 8563 bit map
pointer must be relative to the VIC screen bit map pointer.
If the VIC column counter equals 39, increment the 8563 bit map pointer 41
times to place its relative position at the start of a new character row according to the
VIC bit map.
8. Now repeat the process 25 times, once for each character row.
320 COMMODORE 128
In the program, as discussed, the instructions in $180A through $181A start the
VIC and 8563 bit maps at locations $2000 and $0000 respectively. All these instructions
encompass step 1 of the algorithm.
The instructions that perform step 2 of the algorithm are stored at locations
$182D-$1830.
Step 3 stores the VIC bit map byte at the 8563 bit map position and is performed
by the instruction $1831 and the subroutine stored between $1875 and $1882. This
subroutine writes the value stored in the accumulator into the appropriate address in
8563 RAM as specified by the contents of 8563 registers 18 ($12) and 19 ($13).
The instructions stored at $183A through $183D perform the fourth step of the
algorithm.
The stored in $1883 through $188E update the VIC bit map pointer by 8.
Steps 2, 3 and 4 are performed while the pixel row counter in $9C is less than
seven. If it is less than seven, the next byte of the VIC bit map is loaded and stored into
the 8563 bit map.
Step 6 is contained in the subroutine starting at ($189B). This decrements the VIC
bit map pointer by 311 locations.
Step 7 of the algorithm is contained in the subroutine stored in locations $188F
through $1899.
Finally, step 8 of the algorithm is performed by the instructions $186F through
$1872.
This program allows you to transfer one bit map to the other, but leaves the VIC
bit map as 40 columns. Expand on the algorithm in this section and center the bit map
on the 80-column screen. Write a routine that proportionally spaces the 40-column bit
map and places it on the entire 80-column screen. Also add a routine that clears the
other half of the 80-column bit map. Add these routines to any business application, and
you’ll have a valuable business tool which you can use completely within the 8563 chip
that display your character and bit map data.
SCROLLING THE 8563 SCREEN

The 8563 has a feature that allows the screen to scroll smoothly either vertically (up or
down) or horizontally (left or right). The 8563 scrolling feature can be thought of as a
small virtual screen window which is displayed on a portion of a full size 80-by-25
character screen. The virtual screen is only visible through the window you define at the
specified position of the full-sized screen. Both screens however do exist in the 8563
character RAM storage locations.
VERTICAL SCROLLING
Scrolling in the vertical direction is easier than scrolling horizontally. The general
algorithm to scroll vertically is as follows:
1. Set the default start of the character RAM storage to $2000 in 8563 RAM.
This is done through writing to (or using the default contents of) registers
12 ($0C) and 13 ($0D). Now location $2000 in 8563 RAM marks the start
of the first character in the upper left corner of display memory (the HOME
position).
At this point, the size of the screen is 80-by-25 and the screen RAM
locations are from $2000 through $27CF in 8563 RAM. Therefore, the first
80 bytes store the first 80 characters in the top row of the 8563 screen.
2. Now add 80 ($50) to the low byte of the display RAM Pointer, register 13.
Now the second 80 bytes of the display RAM memory are displayed as the
first character row. The screen is now defined by locations $2050 through
$28IF in 8563 memory. This moves the characters up one character row. The
“ old” first character row has moved off the top of the screen and the “ new ”
character row has moved onto the last character row of the screen from a
previously non-visible range in display area. The new character row should be
defined in 8563 RAM before it is moved onto the screen as the last character
row.
3. Though this moves an entire character row of display RAM up by one row,
the movement is sudden, abrupt and not smooth. The 8563 allows smooth
scrolling to occur vertically, through the lower four bits of register 24 ($18) in
the 8563.
To scroll smoothly one scan line at a time, write the value 1 to register
24. This moves the entire screen up by one scan line. The first previously
visible scan line is now off screen at the top and a new scan line is scrolled
from an off screen location from the bottom onto the last visible scan line of
the screen.
4. To scroll subsequent scan lines, increment the value of the lower four bits of
register 24. By incrementing this value to 2, the entire screen is smoothly scrolled
an additional scan line.
5. Continue to increment the value in register 24 until the value equals the value
stored in the lower four bits of register 9, the total number of vertical scan
lines of a character. Once these values are equal, only the bottom scan line of
the first character row is visible at the top of the screen, and all but the
last scan line of the newly scrolled data is visible at the bottom of the screen.
6. To continue to scroll, reset the value in the lower four bits of register 24 to
zero and add 80 more locations to the contents of the display RAM pointer as
you did in step 2.
If you plan to use attributes, you must increase the value of the attribute
start pointer also to keep up with the display RAM pointer.
322 COMMODORE 128
01800 A2 OC LDX #$oc

01802 A9 20 LDA #$20
01804 85 FB STA $FB
01806 20 60 18 JSR $1860
01809 E8 INX
0180A A9 00 LDA #$00
0180C 85 FA STA $FA
0180E 20 60 18 JSR $1860
01811 EA NO P
01812 A2 18 LDX #$18
01814 A9 01 LDA #$01
01816 A8 TA Y
01817 20 60 18 JSR $1860
01 8 1 A EA NOP
0181B C8 INY
0181C 98 TYA
0181D CO 08 CPY #$08
0181F DO F6 BNE $1817
01821 EA NO P
01822 EA NO P
01823 A9 00 LDA #$00
01825 20 60 18 JSR $1860
01828 A0 00 LDY #$00
0182A E6 FA INC $FA
0182C DO 02 BNE $1830
0182E E6 FB INC $FB
01830 C8 INY
01831 CO 50 CPY #$50
01833 DO F5 BNE $ 1 82A
01835 A2 OC LD X #$oc
01837 A5 FB LDA $FB
01839 20 60 18 JSR $1860
0183C E8 INX
0 1 8 3D A5 FA LDA $FA
0183F 20 60 18 JSR $1860
01842 A9 00 LDA #$00
01844 85 FD STA $FD
01846 A9 00 LDA #$00
01848 85 FC STA $FC
0184A E6 FC INC $FC
0184C DO FC BNE $ 1 84A
0184E E6 FD INC $FD
01850 A5 FD LD A $FD
01852 C9 A0 CMP # $ A0
01854 DO F0 BNE $1846
01856 4C 12 18 JMP $1812
01859 EA NO P
0185A EA NOP
0185B EA NO P
0185C EA NO P
0185D EA NOP
0185E EA NO P
0185F EA NOP
01860 8E 00 D6 STX $D600
01863 2C 00 D6 BIT $D6 00
01866 10 FB BPL $1863
01868 8D 01 D6 STA $D601
0186B 60 RTS
H O R IZ O N T A L SCROLLING
Horizontal scrolling is more complex than vertical scrolling. First, the Display RAM
memory must be set up to match the “ virtual” screen size. If the “ virtual” screen is
132 characters wide, then the Display RAM memory must be organized as if 132
characters were to be displayed, even if only 80 horizontal characters will be displayed
in the 8563 “ w indow .” This means that to display a window of 25 rows of 80

characters of a virtual screen of width 132 characters, the Display RAM memory must
have at least 3300 (132 X 25) memory locations and the same number of attributes.
To display a horizontal “ window” of 80 characters, the 8563 must first read the
first 80 display memory locations and attributes from the first 132-character row, and
then “ skip” the next 52 (132-80) and proceed to read the next 80 pointers and attributes
from the second 132-character row. The amount of the “ skip” must be written to R27.
In this case “ 52” is written to R27 at system initialization.
When horizontal scrolling is not used, a “ zero” written to R27 will cause the
8563 to skip no display memory locations or attributes, allowing the “ virtual” screen in
Display RAM memory to be the same width as the 8563 window.
Now that the 8563 has been set up to view a window 80 characters wide out of a
virtual screen of 132 characters wide, horizontal scrolling may occur by modifying the
address in R12/R13 (DISPLAY START ADDRESS) and R20/R21 (ATTRIBUTE START
ADDRESS). Increasing these addresses by one will scroll the entire screen “ left” one
character. The first character of each character row will be scrolled “ off” the left side
of the screen, all characters on the screen will be moved one space to the left, and one
extra character will be scrolled onto the right side of the screen. Simple manipulation of
the address in R12/R13 and R20/R21 will scroll the window to the left or right.
S M O O TH H O R IZ O N T A L SCROLLING
The 8563 has the ability to scroll smoothly in the horizontal direction. Horizontal
smooth scrolling moves all characters on the screen the desired number of pixels to the
“ left,” up to the width of one character. Before horizontal smooth scrolling can occur,
the 8563 must have something to scroll “ onto” the right side of the screen, so R27 must
be greater than zero. This sets up a virtual screen at least one character greater than the
displayed screen.
The screen is scrolled left by one pixel by changing the contents of R25(3-0) to
“ one” and increasing the contents of R22(3-0) by one. Both registers must be changed
to scroll horizontally.
The number in R25(3-0) and R22(3-0) may be increased, increasing the number
of pixels scrolled, until R25(3-0) equals R22(7^1). This is the maximum amount of
horizontal smooth scrolling allowed. At this point, only one (the right) pixel will exist
for the first character of each character row and the last character will have scrolled onto
the row with only the one (the right) pixel still not visible. To scroll further, the CPU
must change R25(3-0) back to zero and increase R12/R13 (DISPLAY START AD
DRESS) and R20/R21 (ATTRIBUTE START ADDRESS) each by one to scroll hori
zontally by one character. This sequence will have smooth-scrolled left by one character.
324 COMMODORE 128
REGISTER-BY-REGISTER
DESCRIPTION
This section describes each 8563 register in detail by register number. The significant
bits are isolated and explained separately.
The conventions used in the register descriptions are as follows:
Rn (m sb-lsb)
The R stands for “ register.” The letter n represents the register number, ranging
from 0 through 36. The msb stands for “ most significant bit” in that particular
description. Often the most significant bit is assumed to be bit 7; however, in this
discussion, bit 7 is not always the most significant.
The notation lsb stands for “ least significant bit” in the particular register
description. Generally, the least significant bit is assumed to be 0; however, in this
description 0 is not always the least significant.
If a single bit number is enclosed in parentheses, then this is the only bit explained
for that particular function.
Consult this section as a quick reference. See the main body of the chapter for
additional background on programming the 80-column display.
R0 H O R IZ O N T A L T O T A L
The number of characters, minus 1, between successive horizontal sync pulses. This
number includes the displayed part of a character row, the horizontal border and the
blanking interval during horizontal sync pulses.
Rl H O R IZ O N T A L DISPLAYED
The number of characters displayed on each character row. This number sets the width
of the displayed part of the character row.
R2 H O R IZ O N T A L SYNC PO SITIO N
The number of characters from the beginning of the displayed part of a character row to
the start of the horizontal sync pulse. This register controls the horizontal position of the
text on the CRT screen. This number should be greater than the number in Rl
(horizontal displayed).
R3(3-0) H O R IZ O N T A L SYNC W ID TH
The width of the horizontal sync pulse in characters, plus 1.
R3(7-4) VERTICAL SYNC W ID T H

The width of the vertical sync pulse in scan lines. This number sets the duration of the
vertical sync pulse, which may extend past the end of the current frame into the next
frame. In the interlaced sync and video mode, this register should be programmed with a
value equal to twice the number of scan lines desired.
R4 VERTICAL T O T A L
The number of character rows, minus 1, between successive vertical sync pulses. This
number includes the displayed rows, vertical border and the blanking interval during the
vertical sync pulse and determines the vertical sync rate.
R5(4-0) VERTICAL T O T A L ADJUST

The number of scan lines added to the end of the frame for fine adjustment of the
vertical sync rate. In the interlaced sync and video mode, this register should be
programmed with a value equal to twice the number of scan lines desired.
R6 VERTICAL DISPLAYED
The number of character rows displayed in a frame. This number sets the height of the
displayed part of the frame.
R7 VERTICAL SYNC PO SITIO N

The number of character rows, plus 1, from the first displayed character row to the
start of the vertical sync pulse. This number should be greater than the number in R6
(vertical displayed).
R 8(l-0) INTERLACE MODE C O N TR O L

There are three raster-scan display modes: non-interlaced, interlaced sync, and interlaced
sync and video. The non-interlaced mode has been described above. In this mode, each
scan line is refreshed at the vertical sync rate. This mode is selected by setting R 8 (l-0 )
equal to 00 or 10.
In the interlaced sync mode [R 8(l-0) = 01], even and odd fields alternate to
generate frames. The same information is displayed in both odd and even fields but the
vertical sync timing causes the scan lines in the odd fields to be displaced from those in
the even fields by one-half scan line. The spaces between adjacent scan lines are filled,
resulting in a higher-quality character on monitors designed to receive interlaced video.
In the interlaced sync and video mode [R8( 1—0) = 11] even and odd fields
alternate to generate frames. Odd fields display odd scan lines, and even fields display
even scan lines. This doubles the vertical character density on the screen. To display
twice as many character rows, the user must re-initialize the 8563 registers. In this
mode, the vertical sync width [R3(7^4-)] and the vertical total adjust [R5(4—0)] should be
programmed with even values. The duration of each of these events (in scan lines) will
be one half the value programmed into the registers.
R9(4-0) CHARACTER TO TA L, VERTICAL

The number of scan lines, minus 1, in a character. This number includes the displayed
part of a character and the vertical intercharacter spacing below the displayed part. This
register sets the vertical dimension of a character and the vertical dimension of a
character row.
326 COMMODORE 128
R l0(4-0) CURSOR START SCAN LINE

The number of the first (top) scan line of the reverse-video cursor. The number 0 refers
to the first (top) scan line of the character. This register and register 11 (below) define
the cursor height, and can create a variety of effects, from a block cursor to an underline
cursor.
R l0(6-5) CURSOR MODE

The cursor can be programmed to four different modes by programming R10(6-5) to a
00 (solid cursor), 01 (no cursor), 10 (cursor blinking at one sixteenth the frame rate), or
11 (cursor blinking at one thirty-second the frame rate).
Rl 1(4-0) CURSOR END SCAN LINE

The number of the last (bottom) scan line of the reverse-video cursor, plus 1.
RI2, RI3 DISPLAY START ADDRESS (H IG H , LOW)

The address of the first (left) 8 pixels of the top scan line of the frame is defined by R 12
and R13. The most significant 8 bits are in R12, the least significant in R13. These set
the complete 16-bit address for the first 8 pixels. The address of subsequent sets of 8
pixels on that scan line are incremented from the previous address. The bit-mapped pixel
data for a scan line is fetched from adjacent memory locations. A frame of 640
horizontal by 200 vertical pixels will use 16,000 bytes of display 8563 RAM memory
locations (R12, R13) through (R12, R13 + 15999).
The bit map attributes are a set of Display RAM bytes, each of which
corresponds to a character position on the frame. The 8-bit attribute defines additional
characteristics of the character area at that position on the frame. These characteristics
(attributes) are as follows: Bits 3 -0 are foreground R, G, B and I, respectively. Bits 7-4
are background R, G, B and I, respectively. For a frame of i rows of j characters each,
the first j attributes will define the attributes for the first (top) character row, and (i X j)
attributes will need to be defined in display 8563 RAM. For a bit-mapped frame of 25
characters high and 80 characters wide, 2000 attributes will be needed.
RI2, RI3 DISPLAY START ADDRESS (H IG H , LOW)

The address for the pointer of the first (top-left) character of the frame is defined in R12
and R13. The most significant eight bits is in R12, the least significant in R13. These
set the complete 16-bit address for that first pointer. The address of pointers of
subsequent characters are incremented from the previous address, so horizontally adja
cent characters have pointers in adjacent memory locations. A frame of 25 character
rows of 80 characters each will use 2000 pointers of display 8563 RAM memory
locations (R 12/13) through (R12/13 + 1999).
RI4, RI5 CURSOR PO SITIO N

The position of the cursor is set by R14 and R15. These two registers contain the 16-bit
address of the cursor. R14 is the most significant byte, R15 the least significant. If the
address of R14/R15 is the address of a character within the displayed part of the frame,
then that character will be displayed in reverse video for those scan lines determined by
R 10(4—0) and R 1 1(4-0). These registers allow the 8563 to be programmed for a
reverse-video “ block” cursor, or for an “ underline” cursor. The location of the cursor
can be changed by changing the contents of registers R14 and R15.
RI6, RI7 LIG H T PEN VERTICAL, H O R IZ O N T A L

The 8563 supports a light pen function. When the LPEN input pin transitions from a low
to high, the 8563 latches the current vertical and horizontal character counts in the two
light pen registers. The vertical count is latched in R16. LPEN transitions that occur
during the row at the top of the frame will latch a 1 in R 16. Transitions on subsequent
rows will latch 2, 3, etc. The horizontal count is latched in R17. LPEN transitions that
occur during the leftmost character column will latch an 8. Transitions on subsequent
columns will latch 9, 10, etc. The condition of the light pen registers is latched in bit 6 of
the status register. If 0, the light pen registers have not been latched and contain invalid
data. If 1, the registers contain valid latched data. A CPU read of R16 or R17 resets
the Status Register bit.
READ A N D WRITE DISPLAY

8563 RAM MEMORY
RI8, RI9 UPDATE LO C A TIO N
R31 CPU D A TA
The CPU communicates to the 8563 RAM memory via address and data registers in
the 8563. For the CPU to read a memory location, it must place that address in registers
R18 (most significant byte) and R19 (least significant). The 8563 responds by executing
a read of that memory location and placing that data into R 3 1, which the CPU may then
read. During the time that the “ read” is pending and data is not yet valid in R31, the
Update Ready bit of the status register (bit 7) will be 0. Upon completion of the read
cycle, data will be valid in R31 and the bit will be 1. When the CPU reads the data in
R31, the 8563 increments the address in R18/R19 and performs a read of that address.
This allows the CPU to read successive memory locations without repeatedly changing
the addresses in R18/R19.
For the CPU to write data to a memory location, the address must be written to
R18/R19 as in a “ read” described above. Following the automatic read and after the
Update Ready bit is a 1, the CPU should write the desired data to R31. The 8563 then
writes the data to 8563 Memory at the address defined by R18/R19, increments the
address in R18/R19, reads the data from the incremented memory location and places it
in R31. The 8563 then sets the Update Ready bit to a 1. At all times, if the Update
Ready bit is a 0, then CPU access to 8563 memory is pending. Additional access of R18,
R19 or R31 should be avoided until the Update Ready bit is a 1.
R20, R2I ATTRIBUTE START ADDRESS (H IG H , LOW)

The address for the attribute of the first (top-left) character of the frame is defined in
R20 and R21. The most significant 8 bits is in R20, the least significant in R21. These
set the complete 16-bit address for that first attribute. The address of attributes of
328 COMMODORE 128
subsequent characters are incremented from the previous address, so horizontally adjacent
characters have attributes in adjacent memory locations. A frame of 25 character rows of
80 characters each will use 2000 attributes at display 8563 RAM memory locations
(R20/R21) through (R20/R21 + 1999).
R22(3-0) CHARACTER DISPLAYED, H O R IZO N T A L

This number sets the width of the displayed part of the character, and defines the
horizontal intercharacter spacing to the right of the displayed part.
If the defined character is to have no horizontal intercharacter spacing, then
R 22(3-0) should be initialized with a value equal to (the value of R22(7-4) plus 1). This
value may remain constant even with the use of horizontal smooth scrolling.
If the defined character is to have horizontal intercharacter spacing, then R22(3-0)
should be initialized with a value equal to the number of pixels (horizontal) in the displayed
part of a character, minus 1. In this case, the value of R22(3-0) must be changed along
with R25(3-0) to implement horizontal smooth scrolling.
In the pixel double-width mode, the value written to R22(3-0) should be 1 larger
than the numbers described above.
R22(7-4) CHARACTER TO TA L, H O R IZO N T A L

The number of pixels (horizontal) in a character, minus 1. This number includes the
displayed part of a character and the horizontal intercharacter spacing to the right of
the displayed part. This register sets the horizontal dimension of a character.
In the pixel double-width m ode, the value written to R22(7-4) should be the number of
pixels (horizontal) in a character. Do not subtract 1 from the value in this mode.
R23(4-0) CHARACTER DISPLAYED, VERTICAL

The number of scan lines, minus. 1, of the displayed part of a character. This number
sets the height of the displayed part of the character, and defines the vertical intercharacter
spacing below the displayed part.
R23(4-0) can be equal to or less than R9(4-0). If they are equal, then the
displayed part of the character is equal to the total character, and the entire character is
displayed, with no vertical intercharacter spacing.
R24(4-0) VERTICAL SM O O TH SCROLL

The sudden scroll of an entire character row is too abrupt a change for some applica
tions. The 8563 supports a method of vertically scrolling the screen one scan line at a
time, up to one full character row. This is called vertical smooth scroll. This operates
along with vertical scrolling to allow the screen to be scrolled up or down smoothly for
as many character rows as desired.
To scroll the screen one scan line ‘up’ vertically, the CPU should write a 1 to
R24(4—0). This will skip the first scan line of the first character row. The first character
row will then begin with the second scan line, moving it ‘up’ one scan line. All
character rows on the frame will be moved ‘up’ one scan line also. At the bottom of the
Frame, one additional scan line will be displayed of a ‘new’ character row. This will
actually display parts of “ N + 1” character rows instead of the standard N.
The number in R24(4-0) can be increased, increasing the number of scan lines
scrolled, until R24(4—0) equals R9(4-0). This is the maximum amount of smooth scroll
allowed. At this point, only one (the bottom) scan line will exist for the first character
row and the last character row will have scrolled onto the frame with only the last scan
line not visible. To scroll further, the CPU must change R24(4—0) back to 0, and
increase R12/R13 (display start address) and R20/R21 (attribute start address) to scroll
vertically by one character row. This sequence will have smooth scrolled by one
character row.
R24(5) CHARACTER BLINK RATE

When attributes are enabled, the attribute byte for each character contains a bit that
controls the blinking of that character. The rate at which the characters blink is
determined by R24(5). If this bit is a 0, then the character blink rate is one-sixteenth of
the frame rate, if 1, then one thirty-second of the frame rate.
R24(6) REVERSE SCREEN

The complete screen can be reversed (foreground and background colors reversed) by
changing R24(6). With this bit 0, the screen will be displayed with normal foreground/
background colors. A 1 will reverse the foreground and background colors for each
character on the screen. This reversal will re-reverse any character normally reversed by
the cursor and/or the individual character reverse attribute.
R24(7) BLOCK COPY

R30 W ORD C O U N T
To further improve the speed at which the CPU can manipulate 8563 memory, two
additional features have been added to the 8563: Block Write and Block Copy. Block
Write is an extension of the CPU write cycle except that Block Write writes the same
data to more than one successive memory location. This operation is set up by writing
the initial address to R18/R19, waiting for the Update Ready Status bit to be 1, writing
the data to R31 and again waiting for the Update Ready Status bit to be 1. This is the
same as CPU write (above). The CPU must then write a 0 to R24(7) selecting the Block
Write mode, then write to R30 the number of successive memory locations the 8563
should write. Following the write to R30, the 8563 will initiate one or more write cycles
to 8563 memory, writing the data in R31 to successive memory locations. The contents
of R18/R19 will increment with the addresses being written. After the Block Writes are
completed, R18/R19 will contain the address following the last memory location written.
N O TE: One write cycle will follow the initial write of data to R31, so
the quantity written to R30 should be one less than the total number of
memory locations desired to be written.
R25(3-0) SM O O TH H O R IZO N T A L SCROLL

The 8563 has the ability to scroll smoothly in the horizontal direction. Horizontal
smooth scrolling moves all characters on the screen to the ‘left’ the desired number of
330 COMMODORE 128
pixels, up to the width of one character. Before horizontal smooth scrolling can occur,
the 8563 must have something to scroll ‘onto’ the right side of the screen, so R27 must
be greater than 0. This sets up a virtual screen at least one character greater than the
displayed screen.
An unscrolled screen is obtained with the value of R25(3-0) equal to the value of
R22(7-4) (character total, horizontal). The screen is scrolled left by one pixel by
decreasing the contents of R25(3-0) by 1. Additional scrolling occurs with decreasing
values programmed. Maximum scrolling is achieved with a value of 0 in R25(3-0). At
this point, only one (the right) pixel will exist for the first character of each character
row and the last character will have scrolled onto the row with only one (the right) pixel
still not visible. To scroll further, the CPU must change R25(3-0) back to its maximum
value and increase R12/R13 (Display Start Address) and R20/R21 (Attribute Start
Address) each by 1 to scroll horizontally by one character. This sequence will have
smooth scrolled left by one character.
If the character has any nondisplayed pixels {R22(3-0) less than [R22(7-4) plus
one]} then the value of R22(3-0) (character displayed, horizontal) must be changed
along with R25(3-0) for horizontal scrolling. The value of R22(3-0) should be equal to
[(the number of horizontal pixels in the displayed part of a character) plus (the value of
R25(3-0), horizontal smooth scroll)] modulo [(the value of R22(7^1), character total,
horizontal) plus 1], In this case both registers R25(3-0) and R22(3-0) must be changed
to scroll horizontally.
For example, for a character 8 pixels wide with 5 pixels wide displayed:
R22(7-4) R25(3-6) R22(3-0)
Unscrolled 7 7 4
Scroll 1 7 6 3
Scroll 2 7 5 2
Scroll 3 7 4 1
Scroll 4 7 3 0
Scroll 5 7 2 7
Scroll 6 7 1 6
Scroll 7 7 0 5
There are two different versions of the 8563 chip (80-column RGBI) in produc
tion, requiring slightly different register initialization values.
The difference between the 8563-R7A and subsequent revisions (8563-R8 and
8563-R9) is in the horizontal smooth scroll feature, represented by bits 0-3 of register
25. Even if this feature is not utilized, the correct value must be placed into this register
for a normal 80-column display. Software designed specifically for 8563-R7As will
exhibit a problem with the leftmost side of the display when run on a system with R8s or
R9s, and software designed specifically for 8563-R8s and R9s will exhibit a problem
with the rightmost side of the display when run on a system with R7A. To run correctly
on any system, the software must initialize register 25 with the correct data for the
particular 8563 version in the system. The 8563 version can be easily ascertained
through software by reading the 8563 status register, located at $D600 in I/O memory.
Bits 0 -2 contain the version number. Refer to the table below for the actual data to be
used.
8563 rev version ($D600) register 25 value
7A 0 $40
8 1 $47
9 1 $47
This situation does not directly affect software which utilizes the resident C l 28
operating system (Kernal). The Kernal correctly determines which version of the 8563 is
present and initializes it accordingly. Software which relies on the Kernal IOINIT
routine to accomplish this initialization will not have any problems. Software which
performs the initialization itself, or alters the contents of the affected register, may
exhibit the 80-column display anomalies described above, but will otherwise function
normally.
R25(5) PIXEL DOUBLE W ID T H

A change from an 80-column frame to a 40-column frame requires the registers of the
8563 to be initialized differently and requires a different pixel rate. To simplify this
change, a control bit on the 8563 causes the DCLK input signal to be divided by 2. If
R25(4) is a 0, then the pixel width is one DCLK period. If it is a 1, then the pixel width
is two DCLK periods and the internal character clock period is twice the normal period.
The AC timing is the same as normal. The 8563 RAM bus cycles occur at half the
normal rate, with the “ idle” time spent with RAS and CAS “ high.” Note that the
contents of the horizontal initialization registers, and the contents of R22(7^1-) and
R22(3-0) must be initialized with different values in this mode.
R25(5) SEMIGRAPHIC MODE

The displayed part of a character is limited to 8 pixels wide owing to the width of the
Display Data Bus (DD0-DD7). Low-resolution graphics regularly use a “ block-graphics”
character set that may exceed 8 pixels in width. The 8563 allows this in a limited way
by allowing the last displayed pixel of a character to be repeated in the horizontal
intercharacter space, which is normally forced to the background color. If the control bit
R25(5) (semigraphic mode) is a 0, then normal character operation occurs. If it is a 1,
then semigraphics operation occurs, allowing characters to extend into the horizontal
intercharacter space, touching the next character.
R25(6) A TTR IB UTE ENABLE

In some applications, attributes are not necessary. In these cases, the amount of 8563
RAM memory used by the attributes may be excessive, so the 8563 has a control bit to
332 COMMODORE 128
disable attribute usage. If R25(6) is a 0, then attributes will not be fetched by the 8563
and the RAM memory may be used for other purposes. In this case, no character may be
underlined, individually reversed or blinked. Only one character set of 256 characters
may be used. The foreground color of all characters will be the same and will be
determined by R26(7^1). The background color will still be set by R26(3-0). If R25(6)
is a 1, then attributes will be used normally.
R25(7) MODE SELECT (Text/B it Map)

The 8563 is primarily a text display device, but it does support a limited bit-mapped
graphics mode. In the text mode, the displayed frame is an array of characters, each
with a unique pointer to reference the character, and an attribute to define additional
display characteristics. In the bit-mapped mode each pixel is controlled by a unique bit
in 8563 RAM memory. For text operation, bit 7 is initialized to a 0; for bit-mapped
operation, set bit 7. Refer to the Bit-Mapping the 80-column Display section in this
chapter for more information.
R26(3-0) BACKGROUND R, G, B, I
In all cases, the background color is determined by R26(3-0). The borders surrounding
the screen and the vertical and horizontal intercharacter spaces are displayed at the
background color.
R26(7-4) FOREGROUND R, G, B, I
W hen attributes are disabled, the foreground color for all characters is determined by
R26(7-4).
R27 ADDRESS INCREM ENT PER ROW

An additional feature of the 8563 is the ability to scroll horizontally, both left and right
through display 8563 RAM memory. To scroll a window of i horizontal bytes by j
vertical scan lines across a virtual screen of k horizontal by 1 vertical (i < k) there must
be (k x 1) bytes of bit-mapped data in display 8563 RAM memory. The 8563 must
“ skip” (k-i) bytes on every scan line, because only i bytes are displayed of a virtual
line of k bytes in display 8563 RAM memory. The 8563 has the ability to skip a number
of bytes, set by writing to R27. If R27 is 0, then no bytes will be skipped. A nonzero
value in R27 is necessary to allow horizontal smooth-scrolling.
The value in R27 is used to increment the address of the bit-mapped data from one
scan line to the next and to increment the address of the attributes from one character row
to the next. Both the bit-mapped data and the attributes will be incremented by the same
amount, the value of R27.
R28(4) 8563 RAM TYPE (4416/4164)

The data for the displayed text is stored in a dynamic RAM dedicated to the display and
external to the 8563. The CPU accesses this 8563 RAM display memory indirectly
through the 8563 internal registers. The 8563 internally calculates 16-bit RAM ad
dresses, so can address up to 64K bytes of 8563 RAM. The 8563 RAM addresses may
be configured to be compatible with 4164 8563 RAMs (64K by 1) or with 4416 8563
RAMs (16K x 4). Register 28, bit 4 should be initialized to a 0 for operation with
4416s a 1 for operation with 4164s.
4416: Row/Col = A7/A15, A6/A13, A5/A12, A4/A11, A3/A10, A2/A9, A1/A8,

A0/A8
4164: Row/Col = A7/A15, A6/A14, A5/A13, A4/A12, A3/A 11, A2/A10, A1/A9,
A0/A8
R28(7-5) CHARACTER SET START ADDRESS

The starting location of the character set in display 8563 RAM is determined by
R28(7-5). These bits form the most significant 3 bits of the 16-bit address used to
access character data. If R9(4-0) is greater than 15, then only bits 7-6 of R28 will be
used, because each character set will occupy 8192 bytes of 8563 RAM memory.
The character pointers correspond to a displayed character position on the frame.
The 8-bit pointer selects one of the 256 characters in the character set to be displayed at
that position on the frame. For a frame of i rows of j characters each, the first j pointers
will define the characters in the first (top) character row, and (i X j) pointers will need
to be defined in display 8563 RAM. For 25 character rows of 80 characters each, 2000
pointers will be needed.
R29(4-0) UNDERLINE SCAN LINE C O U N T

The number of the scan line used for underline. The underline is defined as a single scan
line. Scan line 0 refers to the scan line at the top of the character.
R32, R33 BLOCK COPY SOURCE ADDRESS

Block Copy is similar to Block Write, except that the data written to 8563 memory is
obtained from other successive memory locations. The Block Copy actually ‘copies’
one part of 8563 memory to another. To set this operation up, the CPU writes the initial
Destination Address to R18/R19, waits for the Update Ready Status bit to be 1 and
writes a 1, to R24(7) selecting the Block Copy mode. The CPU then writes the initial
source address to R32 (most significant byte) and R33 (least significant), then the
number of successive memory locations to be copied should be written to R30. The
8563 will then read the contents of the first source address and write that data to the first
destination address. Additional copies will occur at addresses incremented from the
source and destination addresses, for as many words as are defined in R30.
R34, R35 DISPLAY ENABLE BEGIN, END

During the horizontal and vertical synchronization pulses, the R, G, B and I signals must
be blanked (brought to a low level) to prevent the display of the scanning beam during
retrace times. Two registers allow the user to adjust the beginning (R34) and end (R35)
of the horizontal blanking interval.
Register R34 is programmed with the number of characters from the first dis
played character of a row to the first blanked character in that row. Register R35 is
programmed with the number of characters from the first displayed character of a row to
the last blanked character in that row. Blanking occurs on all scan lines of a frame.
334 COMMODORE 128
R36(3-0) 8563 RAM REFRESH/SCAN LINE

The number of Dynamic RAM refresh cycles every scan line. These refresh cycles occur
on displayed and nondisplayed scan lines, during the displayed part of the screen and
nondisplayed parts (blanked scan lines, vertical borders and vertical sync). Each refresh
address is incremented from the previous refresh address, incrementing through all
65,536 addresses. The least significant byte of the 16-bit Display 8563 RAM memory
address is used as the row address to the 8563 RAM. As the 16-bit refresh address is
incremented, successive rows of the 8563 RAM are refreshed.
il
SOUND AND
MUSIC ON THE
COMMODORE 128
336 COMMODORE 128
IN TR O D U C TIO N
The Commodore 128 has one of the most sophisticated built-in sound synthesizers
available in a personal computer. The synthesizer, called the Sound Interface Device
(SID), is a chip dedicated solely to generating sound and music. The SID chip is capable
of producing three independent voices (sounds) simultaneously. Each of the voices can
be played in one of four types of sounds, called waveforms. The SID chip also has
programmable Attack, Decay, Sustain and Release (ADSR) parameters. These parame
ters define the quality of a sound. In addition, the synthesizer has a filter you can use to
choose certain sounds, eliminate others, or modify the characteristics of a sound or sounds.
All these features add up to a powerful and versatile synthesizer.
To make it easy for you to select and manipulate the many capabilities of the SID
chip, Commodore has developed new and powerful BASIC music statements.
Here are the new sound and music statements available on the Commodore 128:
SOUND
ENVELOPE
VOL
TEMPO
PLAY
FILTER
The beginning of this section explains these sound statements, in an encyclopedia

form at. The second half describes how to program the SID chip in machine
language.
ENVELOPE
Define a musical instrument envelope
ENVELOPE n,[,atk] [,dec] [,sus] [rel][,wf] [,pw]
where:
n Envelope number (0-9)

atk Attack rate (0-15)
dec Decay rate (0-15)
sus Sustain (0-15)
rel Release rate (0-15)
wf Waveform: 0 = triangle
1 = sawtooth
2 = variable pulse (square)
3 = noise
4 = ring modulation
pw Pulse width (0-4095)
SO UND A N D MUSIC O N THE COMMODORE 128 337
A parameter that is not specified will retain its predefined or currently redefined
value. Pulse width applies to the width of the variable pulse waveform (wf = 2) only.
The Commodore 128 has initialized the following ten envelopes:
N A D s R wf pw instrum ent
ENVELOPE 0, 0, 9, 0, o, 2, 1536 piano
ENVELOPE 1, 12, o, 12, 0, 1 accordion
ENVELOPE 2, 0, 0, 15, 0, 0 calliope
ENVELOPE 3, 0, 5, 5, 0, 3 drum
ENVELOPE 4, 9, 4, 4, 0, 0 flute
ENVELOPE 5, 0, 9, 2, 1, 1 guitar
ENVELOPE 6, 0, 9, 0, 0, 2, 512 harpsichord
ENVELOPE 7, 0, 9, 9, 0, 2, 2048 organ
ENVELOPE 8, 8, 9, 4, L 2, 512 trumpet
ENVELOPE 9, 0, 9, 0, 0, 0 xylophone
FILTER
Define sound (SID chip) filter parameters
F IL T E R [fre q ], [,lp ] [,b p ] [,h p ] [,res]
where:
freq Filter cut-off frequency (0-2047)

Ip Low-pass filter on (1), off (0)
bp Band-pass filter on (1), off (0)
hp High-pass filter on (1), off (0)
res Resonance (0-15)
Unspecified parameters result in no change to the current value.

You can use more than one type of filter at a time. For example, both low-pass
and high-pass filters can be used together to produce a notch-(or band-reject) filter
response. For the filter to have an audible effect, at least one type of filter must be
selected and at least one voice must be routed through the filter.
338 COMMODORE 128
EX A M P LE S:
FILTER 1024,0,1,0,2 Set the cutoff frequency at 1024, select the band pass
filter and a resonance level of 2.
FILTER 2000,1,0,1,10 Set the cutoff frequency at 2000, select both the low pass
and high pass filters (to form a notch reject) and set
the resonance level at 10.
PLAY
Defines and plays musical notes and elements
PLAY “ V n,O n,Tn,U n,X n,elem ents”
Vn Voice (n = 1-3)
On Octave (n = 0-6)
TN Tune Envelope Defaults (n = 0-9)
0 = piano
1 = accordion
2 = calliope
3 = drum
4 = flute
5 = guitar
6 = harpsichord
7 = organ
8 = trumpet
9 = xylophone
Un Volume (n = 0-15)
Xn Filter on (n = 1), off (n = 0)
Notes: A ,B ,C ,D ,E ,F ,G
Elements: # Sharp
$ Flat
w Whole note
H Half note
Q Quarter note
I Eighth note
s Sixteenth note
. Dotted
R Rest
M Wait for all voices currently playing to
end current measure
The PLAY statement gives you the power to select voice, octave and tune
envelope (including ten predefined musical instrument envelopes), the volume and the
notes you want to PLAY. All these controls are enclosed in quotes.
All elements except R and M precede the musical notes in a PLAY string.
SO UND AN D MUSIC ON THE COMMODORE 128 339
EXA M P LE S:
PLAY “ V1O4T0U5X0CDEFGAB” Play the notes C ,D ,E ,F,G ,A and

B in voice 1, octave 4, tune enve
lope 0 (piano), at volume 5, with
the filter off.
PLAY “ V 305T6U 7X 1#B $A W .CH D Q EIF” Play the notes B-sharp, A-flat,
a whole dotted-C note, a half
D-note, a quarter E-note and an
eighth F-note.
SO UND
Create sound effects and musical notes
SOUND v,f,d[,dir][,m ][,s][,w ][,p]
where:
v voice (1-3)
f frequency value (0-65535)
d duration (0-32767)
dir step direction (0(up), l(down) or 2(oscillate)) default = 0
m minimum frequency (if sweep is used) (0-65535) default = 0
s step value for sweep (0-32767) default = 0
w waveform (0 = triangle, 1 = sawtooth, 2 = variable, 3 = noise) default= 2
p pulse width (0-4095) default = 2048
The SOUND command is a fast and easy way to create sound effects and
musical tones. The three required parameters v ,f and d select the voice, frequency
and duration of the sound. The duration is in units called jiffies. Sixty jiffies equals
1 second.
The SOUND command can sweep through a series of frequencies which allows
sound effects to pass through a range of notes. Specify the direction of the sweep with
the DIR parameter. Set the minimum frequency of the sweep with M and the step value
of the sweep with S. Select the appropriate waveform with W and specify P as the width
of the variable pulse waveform if selected in W.
EXA M PLES:
SOUND 1,40960,60 Play a SOUND at frequency 40960 in voice 1

for 1 second.
SOUND 2,20000,50,0,2000,100 Output a sound by sweeping through frequen

cies starting at 2000 and incrementing upward
in units of 100. Each frequency is played for
50 jiffies.
340 COMMODORE 128
SOUND 3,5000,90,2.3000,500,1 This example outputs a range of sounds start

ing at a minimum frequency of 3000, through
5000, in increments of 500. The direction of
the sweep is back and forth (oscillating). The
selected waveform is sawtooth and the voice
selected is 3.
EXA M P LE S:
SOUND 1, 49152, 240, 1, 0, 100, 1. 0
F requency________________
Duration _____________ __________
Sweep Direction ____________________
Minimum Sweep Frequency _____________
Step Value for S w e e p __________________
Waveform ____________________________
Pulse Width for Variable Width Waveform
10 DO
20 PRINT"VC FREQ DIR M IN SV WF PW PRINT
30 V = IN T ( RND( 1 ) * 3 ) + l REM V O IC E
40 F=INT(RN D(1)*65535) REM FREQ
50 D=INT(RND(1 )* 3 2 7 6 7 ) REM DURATION
60 DIR=INT(RND(1)*3) REM ST E P D IR
70 M = IN T (R N D ( 1 ) * 6 5 5 3 5 ) REM M IN FREQ
80 S = I N T ( R N D ( 1 ) * 3 2 7 6 7 ) REM ST E P VAL
9 0 W = I N T {R N D { 1 ) * 4 ) REM WAVEFORM
100 P=INT(RND(1)*4095) REM P U L S E W
110 P R I N T V; F ; D I R ; M ; S ; W; P P R I N T : P R I N T
120 S O U N D V , F , D , D I R , M, S , W , P
130 SLEEP 4
140 SOUND V , 0 , 0 , D I R , 0 , 0 , W,
150 LOOP
TEMPO
Define the speed of the song being played
TEMPO n
where n is a relative duration between (0 and 255)
The actual duration for a whole note is determined by using the formula given
below:
whole note duration = 19.22/n seconds
The default value is 8, and note duration increases with n.

SO UND A N D MUSIC ON THE COMMODORE 128 341
EXA M PLES:

TEMPO 1 Defines the TEMPO at the slowest speed.
VOL
Define output level of sound
VOL volume level
This statement sets the volume for SOUND and PLAY statements. VOLUME
level can be set from 0 to 15, where 15 is the maximum volume, and 0 is off. VOL
affects all voices.
EXA M PLES:
VOL 0 Sets volume to its lowest level.

VOL 15 Sets volume for SOUND and PLAY statements to its highest output.
CODING A SONG
FROM SHEET MUSIC
This section provides a sample piece of sheet music and illustrates how to decode notes
from a musical staff and translate them into a form the Commodore 128 can understand.
This exercise is substantially faster and easier if you know how to read music. However,
you don’t have to be a musician to be able to play the tune on your Commodore 128.
For those of you who cannot read music, Figure 11-1 shows how a typical musical staff
is arranged and how the notes on the staff are related to the keys on a piano.
Ill II III 41 III II I

3= m
M iddle
C
Figure 11-1. Musical S taff

342 COMMODORE 128
Figure 11-2 is an excerpt from a composition titled Invention 13 (Inventio 13 in

Italian), by Johann Sebastian Bach. Although this composition was written a few
hundred years ago, it can be played and enjoyed on the most modern of computer
synthesizers, such as the SID chip in the Commodore 128. Here are the opening
measures of Invention 13.
Inventio 13
Figure 11-2. P art o f B ach’s Invention 13
The best way to start coding a song on your Commodore 128 is to break the notes
down into an intermediate code. Write down the upper staff notes on a piece of paper.
Now write down the notes for the lower staff. Precede the note values with a duration
code. For instance, precede an eighth note with an 8, precede a sixteenth note with a 16,
and so on. Next, separate the notes so the notes on the upper staff for one measure are
proportional in time with the notes for one measure on the lower staff.
If the musical composition had a third staff, you would separate it so the duration
would be proportional to the two other upper staffs. Once the notes for all the staffs are
separated into equal durations, a separate and dedicated voice plays each note for a
particular staff. For example, voice 1 plays the upper staff, voice 2 plays the second
staff and voice 3 plays the lowest staff.
L et’s say the upper staff begins with a string of four eighth notes. In addition, say
the lower staff begins with a string of eight sixteenth notes. Since an eighth note is
proportional in time to two sixteenth notes, separate the notes as shown in Figure 11-3.
VI = 8A 8B 8C 8D
V2 = 16D 16E 16F 16G 16A 16B 16C 16D
Figure 11-3. Synchronizing N o tes fo r T w o Voices

SO UND A N D MUSIC ON THE COMMODORE 128 343
The synchronization and timing in a musical composition are critical so you must
make sure the notes in the upper staff for voice 1, for example, are in time agreement
with the notes in the lower staff for voice 2. The first note in the upper staff in Figure
11-3 is an A eighth note. The first two notes for voice 2 are D and E sixteenth notes. In
this case, you must enter the voice 1 eighth note in the PLAY string first, then follow
the voice 2 sixteenth notes immediately after it. To continue the example, the second
note in Figure 11-3 for voice 1 (the upper staff) is a B eighth note. The B eighth note is
equal in time to the two sixteenth notes, F and G, which appear in the bottom staff for
voice 2. In order to coordinate the time, enter the B eighth note in the string for voice 1
and follow it with the two sixteenth notes, F and G, for voice 2.
As a rule, always start with the note with the longer duration. For example, if a
bar starts with a series of two sixteenth notes on the lower staff for voice 2 and the upper
staff starts with an eighth note for voice 1, enter the eighth note in the string first since it
must play for the duration while the two sixteenth notes are being fetched by the
Commodore 128. You must give the computer time to play the longer note first, and
then PLAY the notes of shorter duration, or else the composition will not be synchronized.
Here’s the program that plays Invention 13. Spaces are omitted in the PLAY string
to conserve space. For readability in your programs, add space between elements in the
string. Enter it into your C128 and SAVE it for future use. Now RUN it and sit back
and enjoy the music!
10 R E M I N V E N T I O N 1 3 BA C H
20 TEMPO 6
30 P L A Y " V 1 O 4 T 7 U 8 X 0 " :REM V O I C E 1=0RGAN
40 PLAY"V2O4T0U8X0":REM VOICE 2=PIAN 0
50 REM F I R S T ME AS UR E
60 A$="V201IAV103IEV202QAV103SA04C03BEV202I#GV103SB04DV104ICV202SAEM
70 B$="V104IEV202SA03CV103I#GV202SBEV104IEV202SB03D"
80 REM SECOND MEAS URE
90 C$="V203ICV103SAEV202IAV103SA04CV202I#GV103SBEV202IEV103SB04D"
100 D$="V104ICV202SAEV103IAV202SA03CV104QRV202SBEB03D"
110 RE M RE M T H I R D M E A S U R E
120 E$="V203ICV104SREV202IAV 104SCEV 203ICV103SA 04CV 202IA V102SEG"
130 FS="V103IFV203SD02AV103IAV202SFAV104IDV202SDFV104IFV201SA02C"
140 RE M FOURTH MEASURE
150 G$="V201IBV104SFDV202IDV103SB04DV202IGV103SGBV202IBV103SDF"
160 H$="V103IEV202SGEVl03IGV202SEGV104ICV202SCEV104IEV201SGB"
170 RE M F I F T H MEASURE
180 I$="V201IAV104SECV202ICV103SA04CV103IFV202SDFV104IDV201SB02D"
190 J$="V201IGV103SDBV201IBV103SGBV103IEV202SCEV104ICV201SA02C"
200 RE M S I X T H M E A S U R E
210 K$ = " V 2 0 1 I F V 1 0 4 S C 0 3 A V 2 0 1 I D V 1 0 3 S F A V 1 0 3 I D V 2 0 1 S G 0 2 G V 1 0 3 I B V 2 0 2 S F G "
22 0 M$="V2011AV104SC03AV2021#FV104SCEV201 IBV104SD03BV2021#GVI04SDF"
230 RE M S E V E N T H M E A S U R E
2 40 N$ = " V 2 0 2 I C V 1 0 4 S E C V 2 0 2 1 A V 1 0 4 S E G V 2 0 2 I D V 1 0 4 S F E V 2 0 2 I $ B V 1 0 4 . S D C "
250 0 $ = " V202I#GV103SB04CV202IFV104SDEV202IDV104SFDV201IBV104S#GD"
26 0 RE M E I G H T H M E A S U R E
27 0 P$="V202I#GV104SBDV202IAV104SCAV202IDV104SFDV202IEV103SB04D"
28 0 Q$="V202IFV103S#GBV202I#DV104SC03AV202IEV103SEAV202IEV103SB#G"
290 REM N I N T H M E A S U R E
300 R$="V201HAV103SAECE02QA"
310 PLAY A $ : PLAY B $ : PLAY C $ : P L A Y D $ : PLAY E$
320 PLAY F $ : PLAY G $ : PLAY H S : PLAY I 5 : PLAY J $
330 PLAY K $ : PLAY M $ :P L A Y N $ : PLAY 0 $ : P L A Y P$
340 PLAY Q $ : PLAY R$
344 COMMODORE 128
Here are two sample sounds programs that you can try:
10 REM S O UND L O O P S
20 REM CHANGES S T E P V A L U E, D I R E C T I O N A N D WA V E F O R M
30 S V = 1 0 0 0 : R E M S T E P VALUE
40 DO
50 P R IN T :P R IN T " S V = " ; S V : PRINT
60 : D I R = 0 :REM SWEEP D I R E C T I O N
70 : DO
8 0 : W F = 0 : RE M WA V E F O R M
90 : DO
100 : SOUND 1 , 4 0 0 0 0 , 1 2 0 , D I R , 2 0 0 0 , S V , W F
110 : PRINT "WF=";WF;:WF=WF+1
120 : LOOP U N T I L WF=4
130 : PRINT " D I R = " ; D I R : DIR=DIR+1
140 : LOOP U N T I L D I R = 3
150 SV=SV+2000
1 6 0 LOOP U N T IL SV > 2 0 0 0 0
1 0 REM C l 2 8 - MO DE J O Y S T I C K M U S I C I A N
20 REM P L A YS N O T E S C THROUGH A I N EACH OCTAVE
30 REM TO I N C R E A S E T H E N O T E V A L U E , P U S H J O Y S T I C K F O R WAR D
40 REM TO D E C R E A S E NOTE V A L U E , P U L L BACK
50 REM TO IN C R E A S E OCTAVE VA LU E, PUSH J O Y S T I C K TO THE RIGHT
60 REM TO D E C R E A S E OCTAVE V A L U E , P US H J O Y S T I C K TO T H E LEFT
70 A $ ( 1 ) = " C " : A $ ( 2 ) = " D " : A $ ( 3 ) = " E " : A $ ( 4 ) = " F A $ ( 5 ) = "G " : A $ ( 6 ) = " A " : A $ ( 7 )= "E
8 0 Z = 4 : R E M I N I T I A L I Z E TO FOURTH OCTAVE
9 0 SCNCLR :REM CLEAR SCREEN
1 0 0 I N P U T " E N T E R V O I C E 1 I N S T R U M E N T NUMBER ( 0 - 9 ) " ; N $
1 1 0 I N P U T " E N T E R V O I C E 2 I N S T R U M E N T NUMBER ( 0 - 9 ) " ; 0 $
1 2 0 I N P U T " E N T E R V O I C E 3 I N S T R U M E N T NUMBER ( 0 - 9 ) " ; P $
13 0 P R I N T : P R I N T : P R I N T : P R I N T : P R I N T : P R I N T " INCREASE NOTE”
140 PRIN T:PRINT" ~ " : PRINT
1 5 0 P R I N T " D E C R E A S E OCTAVE < — — > IN C R E A S E OCTAVE"
160 PR IN T: PRIN T" V":PRINT
170 P R IN T : PRINT" DECREASE N O T E " : P R I N T
1 8 0 P R I N T : P R I N T " T O I N C R E A S E OR D E C R E A S E N O T E OR O C T A V E " ;
1 9 0 P R I N T : P R I N T " W I T H O U T P L A Y I N G N O T E S , H O L D DOWN T H E " ;
2 0 0 P R I N T " F I R E BUTTON W HILE P O S I T I O N I N G T H E " ;
2 1 0 P R I N T : P R I N T " J O Y S T I C K I N THE A P P R O P R I A T E D I R E C T I O N " ;
2 2 0 Q $ = " V 1 " + " T " + N $ : R E M V O I C E 1 INSTRUMENT
2 3 0 R $ = " V 2 " + " T " + 0 $ : R E M V O I C E 2 INSTRUMENT
2 4 0 S $ = " V 3 " + " T " + P $ : R E M V O I C E 3 INSTRUMENT
2 5 0 DO
260 DO
270 A = J O Y ( l ) : R E M GET J O Y VALUE
280 I F A = 0 T H E N L O O P : R E M I F Z E R O L O O P AN D G E T A N O T H E R
290 B=A : REM S T O R E T H E V A L U E F O R L A T E R C H E C K I N G
300 Z $ = " O " + S T R $ ( Z ) : REM U P D A T E OC TAV E
310 T $ = Z $ + Q $ + A $ ( I ) : R E M CONSTRUCT V O I C E 1 PLAY STRING
320 U $ = Z S + R $ + A $ ( I ) : R E M CONSTRUCT V O I C E 2 PLAY STRING
330 V $ = Z $ + S $ + A $ ( I ) :REM CONSTRUCT V O I C E 3 PLAY STRING
340
350 I F ( A AND 1 2 8 ) = 1 2 8 T H E N 4 0 0
360 PLAY T $ : R E M PLAY V O I C E 1
370 PLAY U $ : R E M PLAY V O I C E 2
380 PLAY V $ : R E M PLAY V O I C E 3
390
400 IF (A=l OR A = 1 2 9 ) AND ( I < 8 ) T H E N I = I + 1 : R E M I N C R E M E N T N O T E V A L U E
410 I F (A=5 OR A = 1 3 3 ) AN D ( I > 0 ) T H E N 1 = 1 - 1 : REM DE C R E M E N T N O T E V A L U E
420 I F (A=3 OR A = 1 3 1 ) AND ( Z < 6 )T H E N Z = Z + 1 : R E M INCREMENT OCTAVE
430 I F (A=7 OR A - 1 3 5 ) AND ( Z > 0 ) T H E N Z = Z - 1 : R E M DECREMENT OCTAVE
440 A = J O Y ( l ) : R E M GET NEXT J O Y VALUE
450 LOOP WHILE A=B
4 6 0 L O O P : R E M DO I T A G A I N
SOUND A N D MUSIC O N THE COMMODORE 128 345
SOUND AND MUSIC

IN C l28 MODE
BACKGROUND:
THE CHARACTERISTICS OF SOUND
Every sound you hear is actually a sound wave traveling through the air. Like any wave,
a sound (sine) wave can be represented graphically and mathematically (see Figure 11-4).
Figure 11-4. Sine W ave
The sound wave moves (oscillates) at a particular rate (frequency) which deter
mines the overall pitch (the highness or lowness of the sound).
The sound is also made up of harmonics, which are accompanying multiples of the
overall frequency of the sound or note. The combination of these harmonic sound waves
give the note its qualities, called timbre. Figures 11-5 shows the relationship of basic
sound frequencies and harmonics.
Figure I l-S . Frequency and Harm onics
The timbre of a musical tone (i.e., the way a tone sounds) is determined by the
tone’s waveform. The Commodore 128 can generate four types of waveforms: triangle,
sawtooth, variable pulse and noise. See Figure 11-6 for a graphic representation of these
four waveforms.
346 COMMODORE 128
— PULSE W IDTH —
Figure 11-6. Sound W ave fo rm Types

The volume of a sound changes throughout the duration of the note, from when
you first hear it until it is no longer audible. These volume qualities are referred to as
Attack, Decay, Sustain and Release (ADSR). Attack is the rate at which a musical note
reaches its peak volume. Decay is the rate at which a musical note decreases from its
peak volume to its midranged (sustain) level. Sustain is the level at which a musical note
is played at its midrange volume. Release is the rate at which a musical note decreases
from its sustain level to zero volume. The ENVELOPE generator controls the ADSR
parameters of sound. See Figure 11-7 for a graphical representation of ADSR. The
Commodore 128 can change each ADSR parameter to sixteen different levels. This
gives you absolute flexibility over the ENVELOPE generator and the resulting proper
ties of the volume of the sound.
Figure 11-7. A D S R Phases
One of the most powerful Commodore 128 sound statements— the one that
controls the ADSR and waveform— is the ENVELOPE statement. The ENVELOPE
statement sets the different controls in the synthesizer which make each sound unique.
The ENVELOPE gives you the power to manipulate the SID synthesizer.
Here are the definitions of the parameters within the Envelope statement:
Envelope The properties of a musical note specified by the waveform and the attack,
decay, sustain and release settings of the note. For example, the envelope for a
guitar note has a different ADSR and waveform than that for a flute.
W avefo rm The type of sound wave created by the combination of accompanying
musical harmonics of a tone. The accompanying harmonic sound waves are
multiples of, and are based on the overall frequency of the tone. The qualities of
the tone generated by each waveform are recognizably different from one another
and are represented graphically in Figure 11-6.
Pulse W id th The length of time between notes, generated by the pulse waveform.
The Commodore 128 has ten predefined envelopes for ten different musical
instruments. In using the predefined envelopes you do not have to specify the ADSR
parameters, waveform and pulse width settings— this is already done for you. All you
have to do is specify the envelope number. The rest of the parameters are chosen
348 COMMODORE 128
automatically by the Commodore 128. Here are the preselected envelopes for different
types of musical instruments:
ENVELOPE
NUMBER INSTRUMENT ATTACK DECAY SUSTAIN RELEASE W AVEFORM
0 Piano 0 9 0 0 2 1536
1 Accordion 12 0 12 0 1
2 Calliope 0 0 15 0 0
3 Drum 0 5 5 0 3
4 Flute 9 4 4 0 0
5 Guitar 0 9 2 1 1
6 Harpsichord 0 9 0 0 2 512
7 Organ 0 9 9 0 2 2048
8 Trumpet 8 9 4 1 2 512
9 Xylophone 0 9 0 0 0
Figure 11-8. D e fa u lt Param eters fo r E N V E L O P E S ta te m e n t
THE SID FILTER

Once you have selected the ENVELOPE, ADSR, VOLume and TEMPO, use the
FILTER to perfect your synthesized sounds. In your program, the FILTER statement
will precede the PLAY statement. First you should become comfortable with generating
the sound and worry about FILTERing last. Since the SID chip has only one filter, it
applies to all three voices. Your computerized tunes will play without FILTERing, but
to take full advantage of the music synthesizer, use the FILTER statement to increase
the sharpness and quality of the sound.
In the first paragraph of this section, The Characteristics o f Sound, we defined a
sound as a wave traveling (oscillating) through the air at a particular rate. The rate at
which a sound wave oscillates is called the wave’s frequency. Recall that a sound wave
is made up of an overall frequency and accompanying harmonics, which are multiples of
the overall frequency. See Figure 11-5. The accompanying harmonics give the sound its
timbre, the qualities of the sound which are determined by the waveform. The filter
within the SID chip gives you the ability to accent and eliminate the harmonics of a
waveform and change its timbre.
The SID chip filters sounds in three ways: low-pass, band-pass and high-pass
filtering. These filtering methods are additive, meaning you can use more than one filter
at a time. Low-pass filters out frequencies above a certain level you specify, called the
cutoff frequency. The cutoff frequency is the dividing line that marks the boundary of
which frequency level will be played and which will not. In low-pass filtering, the SID
chip plays all frequencies below the cutoff frequency and filters out the frequencies
above it. As the name implies, the low frequencies are allowed to pass through the
filter and the high ones are not. The low-pass filter produces full, solid sounds.
See Figure 11-9.
SOUND AN D MUSIC ON THE COMMODORE 128 349
Figure 11-9. Low-pass F ilter
Conversely, the high-pass filter allows all the frequencies above the cutoff fre
quency to pass through the chip. All the ones below it are filtered out. See Figure 11-10.
The high-pass filter produces tinny, hollow sounds.
Figure 11-10. High-pass F ilte r
The band-pass filter allows a range of frequencies partially above and below the
cutoff frequency to pass through the SID chip. All other frequencies above and below
the band surrounding the cutoff frequency are filtered out. See Figure 11-11.
350 COMMODORE 128
Figure 11-11. Band-pass F ilter
A D VA N C ED FILTERING
Each of the previous FILTERing examples used only one type of filtering at a time. You
can combine the SID chip’s three methods of filtering with each other to achieve
different filtering effects. For example, you can enable the low-pass and high-pass filters
at the same time to form a notch reject filter. A notch reject filter allows the frequencies
below and above the cutoff to pass through the SID chip, while the frequencies close to
the cutoff frequency are filtered out. See Figure 11-12 for a graphic representation of a
notch reject filter.
CUTOFF
FREQUENCY
Figure 11-12. Notch Reject F ilter
You can also add either the low-pass or high-pass filter to the band-pass filter to
obtain interesting effects. By mixing the band-pass filter with the low-pass filter, you
can select the band of frequencies beneath the cutoff frequency and below. The rest are
filtered out.
By mixing the band-pass and the high-pass filters, you can select the band of
frequencies above the cutoff frequency and higher. All the frequencies below the cutoff
are filtered out.
Experiment with the different combinations of filters to see all the different types
of accents you can place on your musical notes and sound effects. The filters are
designed to perfect the sounds created by the other components of the SID chip. Once
you have created the musical notes or sound effects with the SID chip, go back and add
the FILTERing to your programs to make them as crisp and clean as possible.
Now you have all the information you need to write your own musical programs in
Commodore 128 BASIC. Experiment with the different waveforms, ADSR settings,
TEMPOs and FILTERing. Look in a book of sheet music and enter the notes from a
musical scale in sequence within a play string. Accent the notes in the string with the
SID control characters. You can combine your Commodore 128 music synthesizer with
C128 mode graphics to make your own videos or “ movies” complete with sound
tracks.
A U D IO IN P U T TO THE SID CHIP

The SID chip has a little known additional feature that most computers do not offer:
audio input. This allows you to channel your own music, such as an electric guitar, your
voice through a microphone or any other (high impedence, line level) instrument— you
can actually “ ja m ” with the C l 28.
In order to connect properly to the computer, you must devise your own cable or
buy one with the correct connections. Get a cable with an 8-pin DIN connector on the
end you are plugging into the C128. Connect pins 2 (ground), 3 (audio out), and 5 (audio
in) on the 8-pin end.
H ere’s the pinout:
Video Connector—This DIN connector supplies direct audio and composite video
signals. These can be connected to the Commodore monitor or used with separate
components. This is the 40-column output connector.
352 COMMODORE 128
PIN TYPE NOTE
1 LUM/SYNC Luminance/SYNC output

2 GND
3 AUDIO OUT
4 VIDEO OUT Composite signal output
5 AUDIO IN
6 COLOR OUT Chroma signal output
7 NC No connection
8 NC No connection
Run the three connecting wires inside the cable to attach to the audio in, audio out
and ground on the end of your component.
You can connect the C128 to your stereo system or VCR and channel sound
through your speakers. Be careful, however, not to exceed the standard electrical
restrictions. See the SID chip specs in Chapter 16 under the description EXT IN
(Pin 26).
Using the C l28 in these ways opens your eyes to how flexible a personal computer
can be and how it can communicate with a wide range of devices in your home,
including your telephone (with an additional modem), VCR, stereo, TV— and don’t
forget the electric guitar!
PROGRAMMING THE SID CHIP

IN M ACHINE LANGUAGE
This section provides an algorithm and the coded program procedure to play music on
the C128.
First you need to know the key SID locations in the C128 memory map. Here’s an
abbreviated SID memory map, which is exactly the same in both C128 and C64 modes.
Register 0 is located at address 54272 ($D400).
> >>>->->->•>-
W R ITE-O N LY
_l - > _______
> > > - I____________
>
_j —j I____________
—i _ i —j _ i —l—
i I___________ I
z z z z z z z z z z z
Z
o
Z Z Z
o o o o o
Z Z
oz oz zo oz oz oz oz
Q Q a Q
tr <r cc a: cc cc cc cc cc tr cc cc cc cc a: a: a: cc a: cc cc cc cc cc
5 55 55 $ 5 5 5 5 5 5 5 5 5 5 5 5 5 a: cc cc cc
O < ui
UJ f 1 —1
_i Q £
0 2? 0 2?
LO
o 3f H O < ®O o < ;I _
oJ -x E
!= o< (/)
H<
Z H c/j I i -J x
FREQ
c
IL
5 5 o t= s
CL CL O < C/)
o a: 5 5 O
o H
<wD 5 5 o i- o
a. cl o < co
: O O
J5
O
Map
Memory
11-13. SID
Figure
IS S 03 ® < CO O
o >- o
«- — r- o O O O — — o o o o »-
000 00000 000 *-
< j OOOOOOO 0 0 0 0 0 0 0 0 0
0 0 " - N O
N N N N
354 COMMODORE 128
Here’s the general algorithm to output sound from the SID chip in voice 1.
1. Clear the sound chip (For $D400-$D41C)

2. Select ATTACK and DECAY for each voice ($D405)vl
3. Select SUSTAIN and release for each voice ($D406) vl
4. Select a waveform ($D404 for v l)
5. Set the volume ($D 418)
6. Set the frequency of the note with a low byte ($D400) and high byte
($D401) from the Musical Note Table in Figure11-15 at the end of this chapter.
If you want to output a certain frequency level, use this equation:
Fn = F out/.06097
Round off the (Fn) frequency with the integer function and divide the result into the
low byte and high byte.
In decimal:
F hi = Fn/256 is the high byte
and:
Flow = Fn -(2 5 6 * Fhi) is the low byte
To get a feel for the ranges of notes, experiment with the SOUND command in
BASIC.
7. Place the high byte in Frequency Control Register (in $D401) for the
appropriate voice.
8. Place the low byte in the (low byte) frequency control register for the
appropriate voice ($D400).
9. Initiate (GATE) the note or tone. Gating means to start the ATTACK,
DECAY and SUSTAIN cycle of the tone. Set bit 0 of $D404 (V I), $D40B(V2)
and $D412(V3), respectively.
10. Leave bit 0 set for the duration you wish to play the note or tone.
11. Clear the GATE bit.
The durations of the standard musical notes are as follows:
NOTE TYPE DURATION
'/■ 6 128
Vs 256
DOTTED 'A 384
'A 512
*/4 + '/ . 6 640
DOTTED >/4 768
'/2 1024
'/2 + ‘/l6 1152
Vl + Vs 1280
DOTTED Vi 1536
WHOLE 2048
SOUND AN D MUSIC O N THE COMMODORE 128 355
To synchronize three voices, divide each measure into sixteen equal parts. Play
(Gate) the notes with longer duration first, then gate the values of smaller duration, but
of equal proportion, next. For example, one staff has an eighth note. The staff below it
has two sixteenth notes. Play the note of longer duration so they can be gated while the
other faster notes are being fetched.
The following machine language program plays musical notes in one voice. The
musical note data begins at location $1890. The data for each note is stored in the four
byte format starting at $1890 as follows:
Frequency (low byte/high byte)

Duration (low byte/high byte)
For example, the data for the second note starts at location $1894, data for the
third note begins at location $1898 and so on.
For now, sample data has been supplied. Eventually, place your own data beginning
at location $1890. See Figure 11-15 at the end of this chapter for the low- and high-byte
frequency values to place in the SID frequency control registers. Refer to the duration
table on page 354 for the byte values for the note duration.
The NOP instructions are added between instruction sequences for readability.
H ere’s the program listing as it appears in the Machine Language Monitor:
READY.
MONITOR
PC SR AC XR YR S P
; FBOOO 0 0 0 0 0 0 0 0 F8
. 01800 A2 0 0 LDX # $ 0 0
. 01802 8A TXA
. 01803 9 D 0 0 D4 STA $D4 0 0 , X
. 01806 E8 INX
. 01807 E0 19 CPX # $ 1 9
. 01809 DO F 8 BNE $ 1 8 0 3
. 0180B EA NOP
. 0180C EA NOP
. 0180D EA NOP
. 0180E A2 0 5 LD X # $ 0 5
. 01810 A9 F 5 LDA # $ F 5
. 01812 9 D 0 0 D4 STA $D4 00 , X
. 01815 EA NOP
. 01816 EA NO P
. 01817 EA NOP
. 01818 A2 06 LD X # $ 0 6
. 0181A A9 A3 LDA # $ A3
. 0181C 9 D 0 0 D4 STA $D4 0 0 , X
. 0181F EA NO P
. 01820 EA NOP
. 01821 EA NOP
. 01822 EA NOP
. 01823 A2 1 8 LD X # $ 1 8
. 01825 A 9 OF LD A # $ OF
. 01827 9 D 0 0 D4 STA $ D 4 0 0 , X
. 0182A EA NOP
. 0182B EA NOP
. 0182C EA NO P
. 0182D A0 0 0 LDY # $ 0 0
. 0182F A2 0 0 LDX # $ 0 0
. 01831 B9 9 0 1 8 LDA $ 1 8 9 0 , Y
356 COMMODORE 128
01834 F0 3F BEQ $1875

01836 9D 00 D4STA $D4 0 0 , X
01839 E8 INX
0183A C8 INY
0183B B9 9 0 1 8 LDA $ 1890 , Y
0183E 9D 0 0 D4 S T A $D4 0 0 , X
01841 EA NOP
01842 EA NOP
01843 EA NO P
01844 C8 INY
01845 B9 9 0 1 8 LDA $ 1 8 9 0 ,Y
01848 85 FA STA $FA
0184A C8 INY
0184B B9 9 0 1 8 LDA $ 1 8 9 0 ,Y
0184E 85 FB STA $FB
01850 A2 04 LDX #$04
01852 A9 21 LDA #$21
01854 9D 0 0 D4 S T A $D4 00 , X
01857 EA NOP
01858 EA NOP
01859 EA NOP
0185A EA NOP
0185B C6 FA DEC $FA
0185D DO FC BNE $ 1 8 5B
0185F C6 FB DEC $FB
01861 DO F8 BNE $ 1 85B
01863 EA NOP
01864 A2 04 LD X #$04
01866 A9 20 LDA #$20
01868 9D 0 0 D4 S T A $D4 0 0 , X
0186B EA NOP
0186C EA NOP
0186D EA NOP
0186E C8 INY
0186F 4C 2F 18 JMP $ 1 82F
01872 EA NOP
01873 EA NOP
01874 EA NOP
01875 A2 18 LDX #$18
01877 A9 00 LDA #$00
01879 9D 0 0 D4 S T A $D4 00 , X
0187C 00 BRK
>01890 IE 19 80 00 1C 41 80 00
>01898 IE 19 80 00 1C 41 80 00
>018A0 IE 19 80 00 1C 41 00 00
> 0 1 8A8 IE 19 80 00 1C 41 80 00
> 0 1 8B0 IE 19 80 00 IE 19 80 00
> 0 1 8B8 IE 19 80 01 IE 19 80 01
> 0 1 8C0 IE 19 80 04 1C 19 80 04
> 0 1 8C8 IE 19 80 04 1C 41 80 00
> 0 1 8D0 IE 19 80 00 1C 41 00 00
> 0 1 8D8 21 21 00 00 00 00 00 00
> 0 1 8E0 0D 54 08 DO D7 AD 7B 08
> 0 1 8E8 8D 6C 08 AD 7C 08 8D 6 D
Enter the program in memory with the Machine Language Monitor and save it.
Run the program with the following command:
G F I 800
The program plays a series of notes in a single voice (voice 1). Here’s a line by
line description of the instructions in the program.
The instructions stored in locations $1800 through $180A clear the SID registers
($D400-$D418) to zero. This is a recommended programming practice, so you can
assume that all the SID registers are initialized to zero.
The instructions stored in locations $180E through $1814 assign the attack/decay
values in the SID envelope generator. In this program, attack is set at (15:$0F) and the
decay is set at 5. This attack setting makes the sounds seem as if they are far away and
traveling closer, similar to the way a train sounds as it approaches from far away.
The instructions stored in locations $1818 through $18IE select the sustain/release
values for the voice 1 envelope generator. In this case the sustain duration is set on 10
($0A) and the release rate is set to 3. For more information on the ADSR settings, see
the beginning of the chapter.
The next sequence of instructions ($1823-$ 1829) sets the volume to maximum
output (15:$0F). Notice that the last three instruction sequences loaded the X register
with the register number, loaded the accumulator with the appropriate value and stored
the value in the location of the start of the SID plus an offset in the X register. This
programming style is easy to follow and is standardized throughout the program.
The instructions stored in locations $182D through $183E get the low- and
high-byte frequency values for the musical note data, and place them into the low- and
high-byte frequency control registers for voice 1. First the X and Y registers are cleared
to zero. The Y register is used as an index to the musical note data in memory starting at
location $1890. To access successive bytes for frequency and duration, the Y index
register is incremented. The X register is used as an index to access the low- and
high-byte frequency control registers at locations $D400 and $D401. In this program
segment, the X register can only have one of two values; 0 to access location $D400 and
1 to access location $D401.
The first time through the loop, the instruction starting at location $1831 (LDA
$1890,Y) loads the musical data value (low byte frequency) into the accumulator. The
next instruction (BEQ $1875) checks to see if this value is equal to zero. If it is, control
passes to location $1875, the volume is cleared to zero, and the program ends. This
suggests that to end the program, place a zero in the low byte data element. This
mechanism acts as a data terminator, since no additional data is read if the low-byte
value of a musical note equals zero. If the data value does not equal zero, it is stored in
the low byte frequency control register for voice 1.
Both the X and Y registers are incremented, and the next musical data value, this
time at $1891, is loaded into the accumulator with the instruction beginning at location
$ 183B. The next instruction (STA $D400,X) stores the data value in the high byte
frequency control register ($D401) for voice 1.
The instruction at location $1844 increments the Y register. The next load
instruction starting at location $1845 loads the accumulator with the nextdata element in
memory, this time the low byte for the duration of the note. The duration note table is
found on page 357. The low-byte duration is stored in zero page location $FA. The Y
358 COMMODORE 128
register is incremented again and the high byte duration is loaded into the accumulator
and stored in zero page location $FB. These two locations are decremented in the
instructions stored in location $ 185B through $1862.
The instructions stored in locations $1850 through $1857 GATE the note, in other
words play the note. To output audible notes, each voice has a gate bit which initiates the
sound from the envelope generator for a particular voice. Bit 0 of location $D404 is the
gate bit for voice 1. To gate a bit, set bit 0 and also set the bit for the desired wave
form at the same time. In this case, the sawtooth waveform is gated on (3 2 + 1 = $21).
At this point, the sawtooth waveform is gated on and is playing in the voice 1
envelope generator. The selected frequency is output until you turn off the gate bit.
H ere’s where the duration of the note comes into play. The instructions stored in
locations $185B through $1862 decrement the low and high byte duration value stored
in zero page locations $FA and $FB respectively. These instructions act as a time delay
which count through the loops using the duration values from page 354 that are stored in
the musical data region starting at $1890 in memory. When both $FA and $FB are equal
to zero, program control drops through to the instructions stored at $1864 through
$186A. These instructions clear the voice 1 gate bit and stop the envelope generator
from outputting sound.
At this point, the Y register is incremented ($186E) and the program jum ps to
$182F where the X register is cleared to zero. Since the Y register was already
incremented before the JMP instruction in location $186F—71, the low byte frequency
value for the second note is already pointed to by the Y register plus the base address of
$1890. The program then loads that low-byte value and stores it in the location $D400
plus the offset of the X register, in this case 0. That gives the appropriate address $D400
for the voice 1 (low byte) frequency control register. These instructions are executed
repetitively until a low-byte frequency of zero is encountered. When the low-byte
frequency equals zero, control is passed to location $1875, the volume is cleared to
zero, and the program breaks.
This program plays 19 notes, and on the 20th a zero is detected as the low-byte
frequency value (in location $18DC), so volume is set to zero and the program ends.
This data is entered as an example. Place your data there instead according to the note
table in Figure 11-15 at the end of the chapter and the duration table on page 354.
Expand on this example. Add multiple voices using different waveforms. Utilize
the filter to perfect the quality of the musical notes. Use this program as a basis for your
own full-featured musical program.
SYN C H R O N IZA TIO N AND

RING M O D U LATIO N
The 6581 SID chip lets you create more complex harmonic structures through synchro
nization or ring modulation of two voices.
The process of synchronization is basically a logical ANDing of two wave forms.
When either is zero, the output is zero.
Here’s a synchronization algorithm:
SOUND AN D MUSIC O N THE COMMODORE 128 359
1 Clear sound chip registers.

2 Set high frequency voice 1.
3 Set Attack/Decay for voice 1 (A = 13, D = 11).
4 Set high frequency voice 3.
5 Set volume 15.
6 Set start triangle, sync waveform control for voice 1.
7 Timing loop.
8 Clear triangle, sync waveform control for voice 1.
9 W ait, then turn off volume.
The synchronization feature is enabled (turned on) in step 6, where bits 0, 1, and 4
of register SD404 are set. Bit 1 enables the syncing function between voice 1 and voice
3. Bits 0 and 4 have their usual functions of gating voice 1 and setting the triangular
waveform.
Ring modulation (accomplished for voice 1 by setting bit 2 of register $D404 in
step 6 of the algorithm) replaces the triangular output of oscillator 1 with a “ ring
m odulated” combination of oscillators 1 and 3. This produces non-harmonic overtone
structures for use in mimicking bell or gong sounds.
SID REGISTER DESCRIPTION

Following is a register-by-register description of the SID chip in numerical order.
VOICE I
FREQ LO/FREQ HI (Registers 00,01)
Together these registers form a 16-bit number which linearly controls the frequency of
Oscillator 1. The frequency is determined by the following equation:
Fou« = (F„ X FcIk/ 16777216) Hz

Where Fn is the 16-bit number in the Frequency registers and Fc!k is the system clock
applied to the 02 input (pin 6). For a standard 1.0-MHz clock, the frequency is given
by:
Fout = (Fn X 0.06097) Hz

A complete table of values for generating eight octaves of the equally tempered
musical scale with concert A (440 Hz) tuning is provided in Figure 11-15. It should be
noted that the frequency resolution of SID is sufficient for any tuning scale and allows
sweeping from note to note (portamento) with no discernable frequency steps.
PW LO/PW HI (Registers 02,03)

Together these registers form a 12-bit number (bits 4—7 of PW HI are not used) which
linearly controls the Pulse Width (duty cycle) of the Pulse waveform on Oscillator 1.
The pulse width is determined by the following equation:
360 COMMODORE 128
PWout = (PW n/40.95) %

W here PWn is the 12-bit number in the Pulse Width registers.
The pulse width resolution allows the width to be smoothly swept with no discern-
able stepping. Note that the Pulse waveform on Oscillator 1 must be selected in order
for the Pulse Width registers to have any audible effect. A value of 0 or 4095 ($FFF) in
the Pulse Width registers will produce a constant DC output, while a value of 2048
($800) will produce a square wave.
C O N TR O L REGISTER (Register 04)

This register contains eight control bits which select various options on Oscillator 1.
G A T E (B it 0): The GATE bit controls the Envelope Generator for Voice 1. When this
bit is set to a one, the Envelope Generator is Gated (triggered) and the ATTACK/
DECAY/SUSTAIN cycle is initiated. When the bit is reset to a zero, the RE
LEASE cycle begins. The Envelope Generator controls the amplitude of Oscillator
1 appearing at the audio output, therefore, the GATE bit must be set (along with
suitable envelope parameters) for the selected output of Oscillator 1 to be audible.
A detailed discussion of the Envelope Generator can be found in this book in the
chapter on Hardware Overview.
S Y N C (B it I): The SYNC bit, when set to a one, synchronizes the fundamental
frequency of Oscillator 1 with the fundamental frequency of Oscillator 3, produc
ing “ Hard Sync” effects.
Varying the frequency of Oscillator 1 with respect to Oscillator 3 produces a
wide range of complex harmonic structures from Voice 1 at the frequency of
Oscillator 3. In order for sync to occur, Oscillator 3 must be set to some
frequency other than zero but preferably lower than the frequency of Oscillator 1.
No other parameters of Voice 3 have any effect on sync.
R IN G M O D (B it 2): The RING MOD bit, when set to a one, replaces the Triangle
waveform output of Oscillator 1 with a “ Ring Modulated” combination of
Oscillators 1 and 3. Varying the frequency of Oscillator 1 with respect to
Oscillator 3 produces a wide range of non-harmonic overtone structures for
creating bell or gong sounds and for special effects. In order for ring modulation
to be audible, the Triangle waveform of Oscillator 1 must be selected and
Oscillator 3 must be set to some frequency other than zero. No other parameters of
Voice 3 have any effect on ring modulation.
T E S T (B it 3): The TEST bit, when set to a one, resets and locks Oscillator 1 at zero until
the TEST bit is cleared. The Noise waveform output of Oscillator 1 is also reset and
the Pulse waveform output is held at a DC level. Normally this bit is used for testing
purposes, however, it can be used to synchronize Oscillator 1 to external events, allow
ing the generation of highly complex waveforms under real-time software control.
(B it 4): When set to a one, the Triangle waveform output of Oscillator 1 is selected.
The Triangle waveform is low in harmonics and has a mellow, flute-like quality.
(B it 5): When set to a one, the Sawtooth waveform output of Oscillator 1 is selected.
The Sawtooth waveform is rich in even and odd harmonics and has a bright,
brassy quality.
(B it 6): When set to a one, the Pulse waveform output of Oscillator 1 is selected. The
harmonic content of this waveform can be adjusted by the Pulse Width registers,
producing tone qualities ranging from a bright, hollow square wave to a nasal,
reedy pulse. Sweeping the pulse width in real-time produces a dynamic “ phas
ing” effect which adds a sense of motion to the sound. Rapidly jumping between
different pulse widths can produce interesting harmonic sequences.
N O IS E (B it 7): When set to a one, the Noise output waveform of Oscillator 1 is
selected. This output is a random signal which changes at the frequency of
Oscillator 1. The sound quality can be varied from a low rumbling to hissing
white noise via the Oscillator 1 Frequency registers. Noise is useful in creating
explosions, gunshots, jet engines, wind, surf and other unpitched sounds, as well
as snare drums and cymbals. Sweeping the oscillator frequency with Noise
selected produces a dramatic rushing effect.
One of the output waveforms must be selected for Oscillator 1 to be audible,
however, it is NOT necessary to de-select waveforms to silence the output of
Voice 1. The amplitude of Voice 1 at the final output is a function of the Envelope
Generator only.
NOTE: The oscillator output waveforms are NOT additive. If more than
one output waveform is selected simultaneously, the result will be a
logical ANDing of the waveforms. Although this technique can be used
to generate additional waveforms beyond the four listed above, it must be
used with care. If any other waveform is selected while Noise is on, the
Noise output can “ lock u p .” If this occurs, the Noise output will remain
silent until reset by the TEST bit or by bringing RES (pin 5) low.
A TTA CK /D E C A Y (Register 05)

Bits 4 -7 of this register (ATK0-ATK3) select 1 of 16 ATTACK rates for the Voice 1
Envelope Generator. The ATTACK rate determines how rapidly the output of Voice 1
rises from zero to peak amplitude when the Envelope Generator is gated. The 16
ATTACK rates are listed in Figure 11-14.
Bits 0-3 (DCY0-DCY3) select 1 of 16 DECAY rates for the Envelope Generator.
The DECAY cycle follows the ATTACK cycle and the DECAY rate determines how
rapidly the output falls from the peak amplitude to the selected SUSTAIN level. The 16
DECAY rates are listed in Figure 11-14.
SUSTAIN/RELEASE (Register 06)

Bits 4—7 of this register (STN0-STN3) select 1 of 16 SUSTAIN levels for the Envelope
Generator. The SUSTAIN cycle follows the DECAY cycle and the output of Voice 1
will remain at the selected SUSTAIN amplitude as long as the gate bit remains set. The
SUSTAIN levels range from zero to peak amplitude in sixteen linear steps, with a
SUSTAIN value of 0 selecting zero amplitude and a SUSTAIN value of 15 ($F)
selecting the peak amplitude. A SUSTAIN value of 8 would cause Voice 1 to SUSTAIN
at an amplitude one-half the peak amplitude reached by the ATTACK cycle.
362 COMMODORE 128
Bits 0 -3 (RLS0-RLS3) select 1 of 16 RELEASE rates for the Envelope Genera

tor. The RELEASE cycle follows the SUSTAIN cycle when the Gate bit is reset to zero.
At this time, the output of Voice 1 will fall from the SUSTAIN amplitude to zero
amplitude at the selected RELEASE rate. The 16 RELEASE rates are identical to the
DECAY rates as listed in Figure 11-14.
N O TE: The cycling of the Envelope Generator can be altered at any

point via the Gate bit. The Envelope Generator can be gated and re
leased without restriction. For example, if the gate bit is reset before the
envelope has finished the ATTACK cycle, the RELEASE cycle will
immediately begin, starting from whatever amplitude had been reached.
If the envelope is then gated again (before the RELEASE cycle has
reached zero amplitude), another ATTACK cycle will begin, starting
from whatever amplitude had been reached. This technique can be used
to generate complex amplitude envelopes via real-time software control.
VALUE ATTACK RATE D E C A Y / REL EASE RATE
DEC (HEX) (TIM E/CYCLE) TIM E/CYCLE)
0 (0 ) 2 ms 6 ms
1 (1) 8 ms 24 ms
2 (2 ) 16 ms 48 ms
3 (3) 24 ms 72 ms
4 (4) 38 ms 114 ms
5 (5) 56 ms 168 ms
6 (6 ) 68 ms 204 ms
7 (7) 80 ms 240 ms
8 (8 ) 100 ms 300 ms
9 (9) 250 ms 750 ms
10 (A) 500 ms 1.5 s
11 (B) 800 ms 2.4 s
12 (C) 1 s 3 s
13 (D) 3 s 9 s
14 (E) 5 s 15 s
15 (F) 8 s 24 s
N O TE: Envelope rates are based on a 1.0-MHz 02 clock. For other 02

frequencies, multiply the given rate by 1 MHz/02. The rates refer to the
amount of time per cycle. For example, given an ATTACK value of 2,
the ATTACK cycle would take 16 ms to rise from zero to peak ampli
tude. The DECAY/RELEASE rates refer to the amount of time these
cycles would take to fall from peak amplitude to zero.
Figure 11-14. Envelope Rates

VO ICE 2
Registers $07-$0D control Voice 2 and are functionally identical to registers $00-$06
with these exceptions:
1. When selected, SYNC synchronizes Oscillator 2 with Oscillator 1.

2. When selected, RING MOD replaces the Triangle output of Oscillator 2 with
the ring modulated combination of Oscillators 2 and 1.
VO ICE 3
Registers $0E-$14 control Voice 3 and are functionally identical to registers $00-$06
with these exceptions:
1. When selected, SYNC synchronizes Oscillator 3 with Oscillator 2.

2. When selected, RING MOD replaces the Triangle output of Oscillator 3 with
the ring modulated combination of Oscillators 3 and 2.
Typical operation of a voice consists of selecting the desired parameters: fre

quency, waveform, effects (SYNC, RING MOD) and envelope rates, then gating the
voice whenever the sound is desired. The sound can be sustained for any length of time
and terminated by clearing the gate bit. Each voice can be used separately, with
independent parameters and gating, or in unison to create a single, powerful voice.
W hen used in unison, a slight detuning of each oscillator or tuning to musical intervals
creates a rich, animated sound.
FILTER
FC LO/FC HI (REGISTERS $15,$16)
Together these registers form an 11-bit number (bits 3-7 of FC LO are not used) which
literally controls the Cutoff (or Center) Frequency of the programmable Filter. The
approximate Cutoff Frequency ranges from 30 Hz to 12 KHz.
RES/FILT (REGISTER $17)

Bits 4 -7 of this register (RES0-RES3) control the resonance of the filter. Resonance is a
peaking effect which emphasizes frequency components at the Cutoff Frequency of the
Filter, causing a sharper sound. There are sixteen resonance settings ranging linearly
from no resonance (0) to maximum resonance (15 or $F). Bits 0-3 determine which
signals will be routed through the Filter:
When set to a zero, Voice 1 appears directly at the audio output and the
F IL T I ( B it 0):
Filter has no effect on it. When set to a one, Voice 1 will be processed through
the Filter and the harmonic content of Voice 1 will be altered according to the
selected Filter parameters.
364 COMMODORE 128
F IL T 2 (B it I): Same as bit 0 for Voice 2.

F IL T 3 (B it 2): Same as bit 0 for Voice 3.
F IL T E X (B it 3): Same as bit 0 for External audio input (pin 26).
M O D E/VO L (REGISTER $18)

Bits 4 -7 of this register select various Filter mode and output options:
LP (B it 4): When set to a one, the Low-Pass output of the Filter is selected and sent to
the audio output. For a given Filter input signal, all frequency components below
the Filter Cutoff Frequency are passed unaltered, while all frequency components
above the Cutoff are attenuated at a rate of 12 dB/Octave. The Low-Pass mode
produces full-bodied sounds.
BP (B it 5): Same as bit 4 for the Bandpass output. All frequency components above
and below the Cutoff are attenuated at a rate of 6 dB/Octave. The Bandpass mode
produces thin, open sounds.
H P (B it 6): Same as bit 4 for the High-Pass output. All frequency components above
the Cutoff are passed unaltered, while all frequency components below the Cutoff
are attenuated at a rate of 12 dB/Octave. The High-Pass mode produces tinny,
buzzy sounds.
3 O FF (B it 7): When set to a one, the output of Voice 3 is disconnected from the direct
audio path. Setting Voice 3 to bypass the Filter (FILT 3 = 0) and setting 3 OFF
to a one prevents Voice 3 from reaching the audio output. This allows Voice 3 to
be used for modulation purposes without any undesirable output.
N O TE: The Filter output modes ARE additive and multiple Filter modes
may be selected simultaneously. For example, both LP and HP modes
can be selected to produce a Notch (or Band Reject) Filter response. In
order for the Filter to have any audible effect, at least one Filter output
must be selected and at least one Voice must be routed through the Filter.
The Filter is, perhaps, the most important element in SID as it allows the
generation of complex tone colors via subtractive synthesis (the Filter is
used to eliminate specific frequency components from a harmonically rich
input signal). The best results are achieved by varying the Cutoff Fre
quency in real-time.
Bits 0 -3 (VOLO-VOL3) select 1 of 16 overall Volume levels for the final composite
audio output. The output volume levels range from no output (0) to maximum
volume (15 or $F) in sixteen linear steps. This control can be used as a static
volume control for balancing levels in multi-chip systems or for creating dynamic
volume effects, such as Tremolo. Some Volume level other than zero must be
selected in order for SID to produce any sound.
SOUND A N D MUSIC ON THE COMMODORE 128 365
MISCELLANEOUS
P O T X (REGISTER $19)
This register allows the microprocessor to read the position of the potentiometer tied to
POTX (pin 24), with values ranging from 0 at minimum resistance, to 255 ($FF) at
maximum resistance. The value is always valid and is updated every 512 02 clock
cycles. See the Pin Description section for information on pot and capacitor values.
PO TY (REGISTER $ IA )
Same as POTX for the pot tied to POTY (pin 23).
OSC 3/RANDOM (REGISTER $ IB)

This register allows the microprocessor to read the upper 8 output bits of Oscillator 3.
The character of the numbers generated is directly related to the waveform selected. If
the Sawtooth waveform of Oscillator 3 is selected, this register will present a series of
numbers incrementing from 0 to 255 ($FF) at a rate determined by the frequency of
Oscillator 3. If the Triangle waveform is selected, the output will increment from 0 up
to 255, then decrement down to 0. If the Pulse waveform is selected, the output will
jum p between 0 and 255. Selecting the Noise waveform will produce a series of random
numbers, therefore, this register can be used as a random number generator for games.
There are numerous timing and sequencing applications for the OSC 3 register; how
ever, the chief function is probably that of a modulation generator. The numbers
generated by this register can be added, via software, to the Oscillator or Filter
Frequency registers or the Pulse Width registers in real-time. Many dynamic effects can
be generated in this manner. Siren-like sounds can be created by adding the OSC 3
Sawtooth output to the frequency control of another oscillator. Synthesizer “ Sample
and H old” effects can be produced by adding the OSC 3 Noise output to the Filter
Frequency control registers. Vibrato can be produced by setting Oscillator 3 to a
frequency around 7 Hz and adding the OSC 3 Triangle output (with proper scaling) to
the Frequency control of another oscillator. An unlimited range of effects are available
by altering the frequency of Oscillator 3 and scaling the OSC 3 output. Normally, when
Oscillator 3 is used for modulation, the audio output of Voice 3 should be eliminated (3
OFF = 1).
ENV 3 (Register $ IC )
Same as OSC 3, but this register allows the microprocessor to read the output of the
Voice 3 Envelope Generator. This output can be added to the Filter Frequency to
produce harmonic envelopes, WAH-WAH, and similar effects. “ Phaser” sounds can be
created by adding this output to the frequency control registers of an oscillator. The
Voice 3 Envelope Generator must be Gated in order to produce any output from this
register. The OSC 3 register, however, always reflects the changing output of the
oscillator and is not affected in any way by the Envelope Generator.
366 COMMODORE 128
EQUAL-TEMPERED
MUSICAL SCALE VALUES
The table in Figure 11-15 lists the numerical values which must be stored in the SID
Oscillator frequency control registers to produce the notes of the equal-tempered musical
scale. The equal-tempered scale consists of an octave containing twelve semitones
(notes): C ,D ,E ,F,G ,A ,B and C # ,D # ,F # ,G # ,A # . The frequency of each semitone is
exactly the 12th root of 2 ( ^ 2 ) times the frequency of the previous semitone. The
table is based on a 02 clock of 1.02 MHz. Refer to the equation given in the Register
Description for use of other master clock frequencies. The scale selected is concert
pitch, in which A -4 = 440 Hz. Transpositions of this scale and scales other than the
equal-tempered scale are also possible.
Although the table in Figure 11-15 provides a simple and quick method for
generating the equal-tempered scale, it is very memory inefficient as it requires 192
bytes for the table alone. Memory efficiency can be improved by determining the note
value algorithmically. Using the fact that each note in an octave is exactly half the
frequency of that note in the next octave, the note look-up table can be reduced from
ninety-six entries to twelve entries, as there are twelve notes per octave. If the twelve
entries (24 bytes) consist of the 16-bit values for the eighth octave (C-7 through B-7),
then notes in lower octaves can be derived by choosing the appropriate note in the eighth
octave and dividing the 16-bit value by two for each octave of difference. As division by
two is nothing more than a right-shift of the value, the calculation can easily be
accomplished by a simple software routine. Although note B-7 is beyond the range of
the oscillators, this value should still be included in the table for calculation purposes (the
MSB of B-7 would require a special software case, such as generating this bit in the
CARRY before shifting). Each note must be specified in a form which indicates which
of the twelve semitones is desired, and which of the eight octaves the semitone is in.
Since 4 bits are necessary to select one of twelve semitones and 3 bits are necessary to
select one of eight octaves, the information can fit in one byte, with the lower nybble
selecting the semitone (by addressing the look-up table) and the upper nybble being
used by the division routine to determine how many times the table value must be
right-shifted.
MUSIC NOTE VALUES

Figure 11-15 contains a complete list of N o te# , actual note, and the values to be stored
into the HI FREQ and LOW FREQ registers of the sound chip to produce the indicated
note.
Figure 11-15. Music N o te Values
MUSICAL NOTE OSCILLATOR FREQ
NOTE OCTAVE DECIM AL HI LOW
0 C -0 268 1 12
1 C # -0 284 1 28
2 D -0 301 1 45
3 D # -0 318 1 62
4 E -0 337 1 81
5 F -0 358 1 102
6 F # -0 379 1 123
7 G -0 401 1 145
8 G # -0 425 1 169
9 A -0 451 1 195
10 A # -0 477 1 221
11 B- 0 506 1 250
16 C -l 536 2 24
17 C # -l 568 2 56
18 D -l 602 2 90
19 D # -l 637 2 125
20 E -l 675 2 163
21 F -l 716 2 204
22 F # -l 758 2 246
23 G—1 803 3 35
24 G # -l 851 3 83
25 A -l 902 3 134
26 A # -l 955 3 187
27 B -l 1012 3 244
32 C -2 1072 4 48
33 C# - 2 1136 4 112
34 D -2 1204 4 180
35 D# - 2 1275 4 251
36 E -2 1351 5 71
37 F -2 1432 5 152
38 F # -2 1517 5 237
39 G -2 1607 6 71
40 G # -2 1703 6 167
41 A -2 1804 7 12
42 A# -2 1911 7 119
43 B -2 2025 7 233
48 € -3 2145 8 97
49 C # -3 2273 8 225
50 D -3 2408 9 104
51 D # -3 2551 9 247
52 E -3 2703 10 143
53 F -3 2864 11 48
54 F # -3 3034 11 218
55 G -3 3215 12 143
56 G # -3 3406 13 78
368 COMMODORE 128
Figure 11—15. Music N o te Values (continued)
NOTE OCTAVE DECIMAL HI LOW
57 A -3 3608 14 24
58 A# -3 3823 14 239
59 B -3 4050 15 210
64 C -4 4291 16 195
65 C # -4 4547 17 195
66 D -4 4817 18 209
67 D # -4 5103 19 239
68 E -4 5407 21 31
69 1 -4 5728 22 96
70 F # -4 6069 23 181
71 G^4 6430 25 30
72 G # -4 6812 26 156
73 A -4 7217 28 49
74 A# -4 7647 29 223
75 B -4 8101 31 165
80 C -5 8583 33 135
81 C # -5 9094 35 134
82 C -0 9634 37 162
83 C # -0 10207 39 223
84 D -0 10814 42 62
85 F -5 11457 44 193
86 F # -5 12139 47 107
87 G -5 12860 50 60
88 G # -5 13625 53 57
89 A -5 14435 56 99
90 A# -5 15294 59 190
91 B -5 16203 63 75
96 C- 6 17167 67 15
97 C# - 6 18188 71 12
98 D- 6 19269 75 69
99 l) # - 6 20415 79 191
100 E- 6 21629 84 125
101 F- 6 22915 89 131
102 F# - 6 24278 94 214
103 G- 6 25721 100 121
104 G# - 6 27251 106 115

105 A- 6 28871 112 199
106 A# - 6 30588 119 124
107 B- 6 32407 126 151
112 C -7 34334 134 30
113 C # -7 36376 142 24
114 D -7 38539 150 139
115 D # -7 40830 159 126
116 E -7 43258 168 250
NOTE OCTAVE DECIM AL HI LOW
117 F -7 45830 179 6

118 F # -7 48556 189 172
119 G -7 51443 200 243
120 G # -7 54502 212 230
121 A -7 57743 225 143
122 A# -7 61176 238 248
123 B -7 64814 253 46
Figure 11-15. Music N o te Values (continued)
FILTER SETTINGS
LOCATION CONTENTS
54293 Low cutoff frequency (0-7)
54294 High cutoff frequency (0-255)
54295 Resonance (bits 4-7)

Filter voice 3 (bit 2)
54296 High pass (bit 6 )
Bandpass (bit 5)
Low pass (bit 4)
Volume (bits 0-3)
Figure 11-16. F ilte r Settings

P
n
n
12____
INPUT/OUTPUT
GUIDE
372 COMMODORE 128
IN TR O D U C TIO N
Besides being a calculator and a data manipulator, the Commodore 128 is a communica
tor. Through the various connecting ports, it can transmit and receive information to
many different kinds of peripheral equipment. Because of the variety of devices, each
must have its unique access codes and specific instructions (software). This chapter
includes the details of each connection and the associated technique of communicating
with various devices.
IN P U T /O U T P U T PORTS
M ost of the connections are considered as input and output because information is
transmitted in both directions. These connections include the Commodore Serial Port for
the disk drive(s) and printer(s), the User Port for external communication with modems,
an expansion port for additional memory, a Datassette connector for a tape storage
device, two monitor ports for both the composite video and the RGBI monitors, a TV
connector, and game ports for various control devices such as joysticks, paddles and
graphics tablets. The port pinout details begin on page 394.
DISK DRIVE
The disk drive is the most popular information storage device today. It stores programs
and data as magnetic signals on a floppy disk of flexible plastic coated with iron oxide.
Information (characters, letters and numbers) represented by a series of binary digits is
magnetically impressed upon the disk in one of thirty-five concentric ring tracks. Each
ring is divided into 17 to 21 sectors each. Each ring can store about 5000 bytes per
track. A 1541 disk drive has a capacity of about 170K. The 1571 disk drive is
double-sided, so a disk formatted on the 1571 holds twice as much data, about 340K.
FORM ATTING A DISK

Blank disks from the factory are not assigned tracks or sectors. A disk must be
formatted by your computer/disk drive according to the disk operating system. You
cannot store anything on a brand-new disk until this process, called Formatting or
newing, is completed. The Commodore disk drive has built-in microprocessors that
properly organize a blank disk when instructed to do so by the computer. The eighteenth
track is used as a directory for the disk drive. All the addresses for all the blocks (over
600) are stored on track 18 as the Block Allocation Map, or more simply the directory.
Prior to formatting a disk, decide on an identification name and a two-digit code.
Select a name of up to 16 characters that will help identify your disk. For example, the
disk name could be your name, or a topic such as games. The two-digit code can be almost
any letter-number combination. This code immediately alerts the disk drive whenever
IN P U T /O U TP U T GUIDE 373
you change disks— so, for your own protection, try to assign each disk a unique code.
The format command for C l28 mode is as follows:
HEADER “ disk name” , lxy,DO,ON U8
For C64 or C l28 modes, use:
OPEN 15,8,15, “ NO:diskname, xy”
Note that xy is any two-digit code you select, such as 01, 45, W3, 40, or WD.
The 0 after the D or N represents the default drive number for a single disk drive or the
first drive of a dual drive unit where both share a common device number. The 8 in each
statement represents the device number. A second drive would be set to number 9, a
third drive to 10, etc. With a single drive, the HEADER command can simply be:
HEADER“ disk nam e” ,Ixy
The formatting process typically takes about 80 seconds on the 1541 disk drive,
and considerably less on the 1571 drive. Once this operation is completed, this disk can
be used for storing programs and data. Do not format it again— otherwise everything
saved on it will be erased. To simply “ w ipe” an already formatted disk, use the above
instructions but do not include the comma followed by an identification number. This
process will maintain the original format, use the new name and wipe the directory
clean— a process considerably faster than total formatting.
The OPEN command establishes a communication link between the computer and
a specific device number. Disk drives are factory set as device number 8 and can be
converted to device 9, 10, 11 or 12. See your disk drive manual for details.
To verify you have properly formatted the disk, see the section below on reading
the directory.
SAVING PROGRAMS
The saving process for storing programs on a disk does not remove the program from
the computer. It simply “ copies” the program or file from the computer onto the disk in
the disk drive. Nothing is lost in the process. BASIC programs are saved to a formatted
disk with the following commands:
C128 or C64 SAVE“ program name” ,8
C128 only DSAVE“ program nam e”
DSAVE“ program nam e” ,DO,U9
(if the drive is device number 9).

In C l 28 mode, machine language programs or binary files can be saved through
these BASIC commands:
BSAVE“ filenam e” ,Bb,Ps TO Pe

or BSAVE1‘filename” ,D 0,U 9,Bb,Ps TO Pe
(if the drive is device number 9).
374 COMMODORE 128
m
In these formats:
“ b ” is the memory bank number (0 through 15)
“ s” is the starting address
“ e ” is the ending address
E X A M P LE :
BSAVE“ filename” ,B2,P3584 TO P4096
This saves the binary file named “ filename” from memory bank 2, in the range of
memory between 3584 to 4096 decimal.
The following command saves a BASIC program in either C128 or C64 modes:
SAVE “ filename” ,8
where 8 represents the device number.
REPLACING A FILE OR PROGRAM

When re-saving a program under the same filename, you must use the at sign (@ ) in the
SAVE command format, as follows:
In BASIC 7.0: DS A VE“ © program name”

In BASIC 2.0: SAVE“ @0:program name” ,8
The @ symbol tells the computer to store this program in place of the one of the
same name. Without this symbol, the disk drive will not allow the new program to be
stored. Note that during the replace process, the new program is saved first, then the old
one is automatically deleted and the directory updated. Use V ER IFY to make sure you
are saving correctly.
VERIFYING A PROGRAM OR FILE

You may want to verify that a program was accurately saved on the disk. The VERIFY
command compares the program in memory with the program on the disk. If the
programs are identical byte for byte, an OK message is displayed.
In BASIC 7.0, use the following:
D VERIFY” filename”
or
D VERIFY ‘‘filename ” , DO, U9
The first command above verifies a program using a single disk drive assigned as Unit 8,
and the second command above verifies a program using a single disk drive assigned as
unit 9.
In 2.0 BASIC, use:
VERIFY ‘ ‘fi lename ” , 8
or VERIFY “ filename” ,8,1
(to verify the program from the memory location from which it was SAVEd, primarily
machine language programs).
After your program has been saved and verified, it is stored on your disk for future
use.
TH E DIRECTORY
Once a disk is formatted, it has a directory, or catalog listing of the files stored on the
disk. Prior to loading a program, you may want to view the directory to find out exactly
what is stored on the disk. When you call for the directory, the screen displays the disk
name and identification number followed by a list of programs or files presently stored.
O f course, a newly formatted disk will display only its name and number. To see the
directory, use the commands listed below.
In BASIC 7.0, use the following:
DIRECTORY
or CATALOG
or CATALOG D0,U9
or DIRECTORY D0,U9 (for a dual disk drive, device number 9)
The program in memory is not affected by this command.

In BASIC 2.0, use the following command:
LOAD“ $0” ,8:LIST
This command erases the resident program. The 0 is optional with a single-drive
unit.
To print a hard copy of a directory, you must use the BASIC 2.0 method. You
must first open a channel to the printer. This is covered below in the section Output to a
Printer.
RETRIEVING PROGRAMS OR
FILES FROM DISKS
Once you have files or programs stored on a disk, you retrieve them with an appropriate
version of the LOAD command.
In BASIC 7.0, use the following command:
DLOAD “ filename”
or DLOAD“ filename” ,DO,U9

(for disk drive unit 9, drive 0)
To retrieve a program or binary file saved by the BSAVE command or by the

machine language monitor, use the following BASIC 7.0 command:
BLO AD ‘‘filename ” , B b, Ps
376 COMMODORE 128
or BLOAD“ filename” ,DO,U9,Bb,Ps

(for disk drive unit 9, drive 0)
where b is the bank number and s is the starting address.
In BASIC 2.0, use the following command formats:
To load a BASIC program: LOAD“ filename” ,8

To load a machine language program: LOAD“ filename” ,8 ,l
The relocate flag is either a 1 or a 0, which is the default value. The 1 tells the
computer to LOAD from the point where it was SAVEd.
CREATING A N D STORING FILES
A file is a series of related data: characters, numbers and letters. Anything stored on the
disk is considered a file. A program is a specific type of file that can be executed or run.
Data files are files that contain raw data used by a program. These files are classified as
either sequential, relative, user or random. The directory lists all the files and indicates
the type of file: PRG for program, SEQ for sequential file, REL for relative file and
USR for user files.
A program is a file that contains operating instructions for the computer. A
sequential file is a series of data stored in sequence and can be read only in sequential
order. A relative file is a file stored in sequential order and can be accessed directly at
any location using a pointer.
The creation of a sequential file is initiated with an OPEN statement. A file is
OPENed with the following information:
■ File number: 0 through 255; 15 is reserved for special commands and error
checking.
■ Device number: default is 8 for a disk drive; 4 for a printer (4—30 are available
for serial bus devices).
■ Channel number: 0 through 255; the line of communication established between
devices (also known as secondary address)
Additional (optional) information may also be specified with quotes:
■ Drive number followed by a colon: default is 0 for asingle drive;1 may be

specified on a dual drive.
■ Filename: the name of the file, up to 16 characters.
■ Filetype: SEQ = sequential, PRG = program, REL = relative, USR = user
■ Operation: (R = read, W = write)
The following command:

OPEN 111, 8, 12,“ 0:NAME, S ,W ”
opens file number 111 on device number 8, using channel 12. The file is opened to
default drive 0, with the filename “ N A M E,” as a sequential file. The W specifies that a
write operation will take place. To store data, it is “ written” on the disk. Data is
retrieved by “ reading” the disk.
Enter, save and run the following programs as sample routines to create, save and
read a sequential file. See your disk drive manual for more information.
10 REM STORE DATA

11 REM NOTE PROGRAM STOPS WHEN YOU RESPOND WITH “ END” .
20 OPEN 5 ,8 ,5 ,“ EX A M PLE,S,W ”
25 DO
30 INPUT “ NAME OF A FRIEND” ;N$
40 PR IN T #5,N $
50 IF N $ = “ END’T H E N CLOSE 5:END
60 LOOP
N O TE: In C128 mode, the Commodore 128 has a set of disk input/
output commands. These include the DLOAD and DSAVE commands
already described, as well as the DOPEN, DCLOSE and other com
mands. See the BASIC 7.0 Encyclopedia in Chapter 2 or your disk drive
manual for details on any of these commands.
Line 20 tells the disk drive to open a sequential file for writing data. Line 30
obtains the information from the keyboard by requesting it on the screen. The PR IN T #
statement in line 40 specifies the file number where the information is to be written.
Line 40 is the statement that actually writes the data to the channel and creates the
contents of the file. Line 50 also puts a recognizable end-of-file marker—the word
“ end” at the end of the file. Line 50 closes the file after all data has been written.
H ere’s a program that reads the data from the sequential file you just created:
10 REM RETRIEVE FILES
20 OPEN 6 ,8 ,6 ,“ EXAM PLE,S.R”
25 DO
30 IN PU T#6,L$
40 IF L$ = “ END’T H E N CLOSE6:END
50 PRINT L$
60 LOOP
Line 20 opens a channel to the sequential file called EXAMPLE for reading. Once
the end item is found, the file is properly closed. As each item is retrieved using the
IN PU T # statement, it is printed to the screen with a simple PRINT statement. Any
sequential file in which all writing has been completed and which is already closed can
be opened only as a read file; it cannot be written to unless an APPEND operation is
specified or the @ prefix is used.
Files can be expanded in size with the BASIC 7.0 APPEND command as follows:
378 COMMODORE 128
A P P E N D #7,“ EXAM PLE”

This command opens the file called “ EXAM PLE.” Any data sent to it in a PRIN T#7
statement will be appended. Any file number may be specified as long as the PRINT file
number matches the APPEND file number.
Once all the data is sent, the channel is closed with
D CLO SE#7
In BASIC 2.0, use this command:
O PEN 7,8,7, “ 0:EXAM PLE,A”
The letter A allows subsequent P R IN T # statements to enter data at the end of the
file “ Exam ple.” This BASIC 2.0 technique also requires a CLOSE statement: CLOSE7.
OTHER FILES
C128 BASIC includes several direct disk drive housekeeping commands including
SCRATCH, RENAME, CONCAT, COPY and COLLECT as well as a relative file
command RECORD. Note that these operations are also available in the C64 mode if
combined with an OPEN statement on channel 15. See your C l 28 System Guide
Encyclopedia or disk drive manual for details.
In addition, see your disk drive manual for details on relative files, random files
and direct program access to the disk drive microprocessor via channel 15.
C H A N G IN G THE DISK DRIVE

DEVICE NUMBER
The disk drive device number is usually 8, but it can be changed to be device 9, 10, or
11 by either a hardware or software change. The following command changes your drive
to unit 9 through software until power is turned off:
OPEN 15,8,15 :PR1NT# 15, ‘ ‘M -W ’ ’CHR$( 119)CHR$(0)CHR$(2)
CH R$(41)CHR$(73):CLOSE 15
The 1571 has a device number switch on the back.
O U TP U T TO A PRINTER
Your Commodore 128 computer transmits data over the serial port to many different
styles of printers. The most popular Commodore printers today are the daisy wheel and
dot matrix types. The daisy wheel system uses metal or plastic molded letters similar to
those on a typewriter and produces the finest impressions. The fingers that hold these
characters are mounted on a center post, and the wheel resembles a daisy flower. The
printer head spins the wheel until the correct letter is in position behind a hammer,
which presses it against a ribbon onto the paper.
The more versatile system is the dot matrix. Using a column of metal pins, a
series of dots is impressed on paper through an inked ribbon as the printer head moves
across the page. The pattern of dots is based upon the binary code of each character.
The infinite number of potential dot patterns makes it possible to print any style
character imaginable from italics to Japanese depending only upon the printer’s internal
software (firmware).
Although many printer manufacturers claim their system is compatible with the
Commodore 128 computer, if Commodore does not authorize the use, a special printer
interface may be required. Printers cited as Centronics parallel or Serial do not directly
match any of the Commodore computer ports. Commodore does not support any third-
party printers or printer interfaces. The Commodore Serial Port is unique and not the
same as the RS232 Serial Port, which many printer manufacturers specify.
The Commodore computer can display 512 characters. The second set of 256
includes both upper and lower case letters; the first includes upper case only with
graphics. This is possible by using identical codes accessed separately. Consequently,
only Commodore dot matrix printers are preprogrammed to print all the available
Commodore characters. Owing to these differences, Commodore’s ASCII codes are not
identical to the true ASCII standard. Therefore, special modifications are required when
interfacing with non-Commodore equipment.
In order to operate a printer, a line of communication with the computer must be
opened using the OPEN statement. The printer is usually device number 4, though often
switch selectable to 5. The command:
OPEN 3,4
establishes a link to the printer as device number 4. The first number can be any
number between 1 and 255. Numbers beyond 128 force a line feed after a carriage return.
PRINTING A PROGRAM LISTING

The following commands will produce a printed listing:
OPEN3,4:CMD3:LIST
The CMD command redirects all screen output to the printer. Any subsequent
output is sent to the printer until turned off with:
PRINT#3:CLOSE3
These statements are necessary to clear the channel and ensure that all commands
have been properly sent to the printer.
N O TE: To print a listing with double spacing, a line feed can be forced
using a file number greater than 127 such as:
OPEN 131,4:CMD 131 :LIST:PRINT# 131 :CLOSE 131

380 COMMODORE 128
PRINTING A TYPED LINE

A PR IN T # command to an opened print file number outputs the associated data to the
printer. The following commands output whatever is typed:
OPEN3 ,4:PR IN T#3, ‘‘TYPE ANYTHING’ ’

After R E T U R N is depressed, the line is sent to the printer and anything
under quotes is printed on paper.
PRINTING TH RO UG H A PROGRAM
Instructions to a printer require a P R IN T # statement and can be mixed with the
PRINT statement for video display. The following program prints five lines of text for
each PR IN T # statement while using the screen display to obtain the input:
10 OPEN 3,4,7
20 DO UNTIL L = 5
25 IF L = 0THEN30 :ELSE INPUTA$:GOT040
30 INPUT“ PLEASE TYPE” ;A$
40 PR IN T#3,A $
50 L= L+ 1
60 LOOP
70 CLOSE3
This is the beginning of a simple word processor and could be expanded into a
complex system through printer control.
CONTRO LLING YO UR PRINTER

Each printer capability (such as lower case characters, line feed, double width charac
ters, reverse characters, dot graphics and special print fonts) is controlled by sending the
proper control code from the computer. Depending upon the printer being used, there
are two ways of sending code— either via the CHR$ function or the secondary address.
Standard codes are sent via the BASIC CHR$(X) function. ASCII codes under 27
are generally reserved for printer codes (such as 14 for the double width feature). Codes
over 27 represent specific keyboard characters. Printer control codes over 27 (e.g., 145)
are preceded with the ESCape code 27. Some systems will accept CHR$(27) followed
by either CHR$(X) or the equivalent character within quotes. Thus, boldface type
represented by Code 33 could be turned on with:
P R IN T #3, CHR$(27)CHR$(33)
or P R IN T #3, CHR$(27)“ !”
and turned off with:
P R IN T # 3, CHR$ (27)CHR$(34)
assuming channel 3 was previously opened.
Many word processors can accommodate special printer controls with at least one
of these methods. Refer to your printer and word processing manuals for the technique
o f sending control characters to your printer.
The alternate method of sending control codes to a printer is via the secondary
address in the OPEN statement. An OPEN statement without a secondary address
defaults to 0. Hence OPEN3,4 is the same as OPEN3,4,0. Generally a secondary
address of 7 will cause the printer to print with upper case and lower case characters:
OPEN3,4,7
Other secondary addresses are often available and will perform various duties in
accordance with your printer’s design. See your manual for specifics. Most word
processors allow the use of secondary addresses as embedded codes.
The following program will open a channel to a sequential file on disk, read data
and output to a printer:
10 OPEN 3 ,8 ,3 ,“ EX A M PLE,S,R”
20 OPEN 4,4,0
25 DO
30 IN PU T#3,L$
40 IF L$ = “ END’ 'THEN80ELSE50
50 PR IN T #4,L $
60 PRINT L$
70 LOOP
80 CLOSE4:DCLOSE3:END
O U TP U T TO A MODEM
A modem is a device used to transmit and receive data across telephone lines.
Generally, a computer uses an RS232 port to communicate with the external
world. Commodore’s RS232C Port (also known as the User Port) is a variation of the
RS232 standard port. The RS232 has 0 to 5 TTL voltage levels rather than the standard
plus/minus 12 volts.
The OPEN statement for the modem (device 2) sets up the parameters to match the
speed (baud rate) and format (number of data and stop bits) of the host computer by
sending a Control Register code. A second required code, the Command Register code,
defines the parity, duplex and handshake. For example:
OPEN3,2,C),CHR$(6)CHR$(l 12)
opens a channel to the User Port (device # 2 ) and establishes a 300-baud rate, 8-bit
word, single stop bit (Control Register (6)), even parity and half duplex (Command
Register (112)). See Figure 12-1, the Control Register Map, and Figure 12-2, the
Command Register Map.
382 COMMODORE 128
THE RS-232 CHANNEL

The RS-232 capabilities can be accessed from either BASIC or machine language.
The BASIC commands IN P U T # and G E T # are used to fetch data from the receiving
buffer while P R IN T # and CM D place data into the transmitting buffer.
The RS-232 Kernal routines run under the control of the 6526 CIA # 2 device
timers and interrupts. The 6526 chip generates NMI (Non-Maskable Interrupt) requests
for RS-232 processing. This allows background RS-232 processing to take place during
BASIC and machine language programs. There are built-in hold-offs in the Kernal,
cassette and serial bus routines to prevent the disruption of data storage or transmission
by the NMIs that are generated by the RS-232 routines. During cassette or serial bus
activities, data can not be received from RS-232 devices. But because these hold-offs
are only local, (assuming you’re careful about your programming) no interference
should result.
There are two buffers in the Commodore 128 memory to help prevent the
loss of data when transmitting or receiving RS-232 information.
The Commodore 128 RS-232 I/O buffers consist of two first-in/first-out (FIFO)
buffers, each 256 bytes long. In C128 mode, the RS-232 I/O buffers are permanently
allocated. The input buffer is located between 3072 and 3327 ($0C00-$0CFF). The
output buffer is located between 3328 and 3583 ($0D00-$0DFF). You cannot harm
program storage by opening and closing the RS-232 channel as can happen with the
Commodore 64. By permanently allocating the buffers in a fixed location, no “ OUT OF
M EM ORY” error can occur, and your program in memory will not get “ stepped o n .”
OPENING THE RS-232 CHAN N EL

Only one RS-232 channel should be open at any time; a second OPEN statement will
cause the buffer pointers to be reset. Any characters in either the transmit buffer or the
receive buffer will be lost.
Up to 4 characters can be sent in the filename field. The first two are the control
and command register characters; the other two are used for user-defined, non-standard
baud rates. Baud rate, parity and other options can be selected through this feature.
No error checking is done on the control word to detect a nonimplemented baud
rate. Any illegal control word will cause the system output to operate at a very slow rate
(below 50 baud).
BASIC S Y N T A X
OPEN lfn,2,0” <control re g .> <com m and re g .> C opt baud low > < o p t baud h ig h > ”
The logical file number (lfn) then can be any number from 1 through 255. But be
aware of the fact that if you choose a logical file number that is greater than 127, then a
line feed will follow all carriage returns.
TH E C O N TR O L REGISTER
The C ontrol R egister is a single-byte required to specify the baud rates and other RS-232
options. Select the parameters you want from the Control Register Map (Figure 12-1)
by placing the proper bit value in binary form. A selection, for example, from left to
right of two stop bits, 7-bit word and 1200-baud rate yields:
1010 1000 (in binary), which is decimal 168 ($A8) and results in
OPEN 3,2,0,CHR$(168)
defining these parameters.
128 64 32 16 8 42 0
@@0 0 || [| 0 0
BAUD RATE
0 0 0 0 USER RATE
STOP BITS — 0 0 0 1 50 BAUD

0-1 STOP BIT 0 0 75
0 1
1-2 STOP BITS
0 0 1 1 110
0 1 0 0 134.5
0 1 0 1 150
0 1 1 0 300
W ORD L E N G T H '
0 1 1 1 600
BIT DATA
6 5 W ORD LENG TH 1 0 0 0 1200
0 0 8 BITS 1 0 0 1 1800 [N l]
0 1 7 BITS 1 0 1 0 2400 [N l]
1 0 6 BITS 1 0 1 1 3600 [N l]
1 1 5 BITS 1 1 0 0 4800 [N l]
1 1 0 1 7200 [N l]
UNUSED
1 1 1 0 9600 [N l]
1 1 1 1 19200 [N l]
Figure 12-1. C o n tro l Register M ap
TH E C O M M A N D REGISTER
The C o m m a n d R egister is a single byte that defines parity, duplex and handshaking.
Select the parameters from the Command Register Map in Figure 12-2 by creating the
binary code and converting it to a decimal value. Thus:
384 COMMODORE 128
OPEN3,4,O.CHR$(168)CHR$(32)
specifies odd parity, full duplex. A baud rate other than those listed in the Control
Register Map is also selectable by choosing 0000 for bits 0 through 3 in the Control
Register and including this rate in the OPEN statement as two bytes following the
Command Register.
The “ opt baud low ” and high byte values are calculated from the desired
baud rate as a function of the system frequency (1.022727E6 for the USA NTSC
system or 0.985248E6 PAL European system). Divide the frequency by the desired
rate. Divide this in half and subtract 100. This represents a two-byte decimal number.
The high value is the number of times it can be divided by 256. The remainder is the
low byte.
PARITY O P T IO N S - L -H A N D S H A K E
BIT BIT BIT
7 6 5 O P E R A TIO N S 0-3 LIN E
PARITY D IS A B LE D , N O N E 1-X LIN E
- - 0
G EN ER A TE D /R EC E IV E D
ODD PARITY
0 0 1
R EC EIVER /TR AN SM ITTER
EVEN PARITY
0 1 1
R EC EIVER /TR AN SM ITTER
M A R K T R A N S M IT TE D
1 0 1
PARITY C H E C K D IS AB LE D
1 SPACE T R A N S M IT TE D
1 1
PARITY C H E C K D IS AB LE D
D U P L E X -----------------
0-F U LL DUPLEX
1-H A LF DUPLEX
U NUSED
UNUSED
UNUSED
Figure 12 -2 . C o m m an d Register Map

E X A M P LE :
1000 Baud: Divide the system frequency by the baud rate:

1022727/1000= 1022.727
Divide by 2 and subtract 100: 1022.727/2 - 100 = 411.3635 (or rounded up to 412)
Convert this into two byte values: Division by 256 results in 1 with a remainder of
156. The following statement therefore defines a 1000-baud, 7-bit, single stop bit, odd
parity, full duplex:
OPEN3,2,0,CHR$(32)CHR$(32)CHR$(156)CHR$(1)
where the Control Register value in binary is 00100000 and the Command Register in
binary is 00100000.
NO TE: In a C64 BASIC program, the RS-232 OPEN command should be

performed before creating any variables or arrays because an automatic
CLR is performed when an RS-232 channel is OPENed owing to the
allocation of 512 bytes at the top of memory. This is not the case in C l 28
mode, since the RS-232 I/O buffers are permanently allocated.
G ETTING DA TA FROM THE

RS-232 C H AN N EL
The Commodore 128 receiver buffer ($OCOO-$OCFF) will hold up to 255 characters before
overflowing. This is indicated in the RS-232 status word ST. If an overflow occurs, sub
sequent characters are lost. Keep the buffer as clear as possible. To receive data at high
speeds, use machine language routines or use the FAST command to process at 2MHz
with the 80-column screen (in C l 28 mode only).
Use G E T # with the file number corresponding to the OPEN statement. Machine
code Kernal entries include BASIN and GETIN.
SENDING DA TA TO THE
RS-232 C H A N N EL
When sending data, the output buffer ($0D00-$0DFF) can hold 255 characters before a
full buffer hold-off occurs. The system will wait in the CHROUT routine until trans
mission is allowed or the R U N / S T O P and R E ST OR E keys are used to
recover the system.
Use CMD or PR IN T# with the proper file number to send data. Acceptable
Kernal entries include BSOUT (CHROUT).
CLOSING THE RS-232 CHANNEL

Use CLOSE followed by the file number to close the RS-232 channel. Closing the file discards
all data in the buffers at the time of execution, stops all transmitting and receiving and sets
the RTS and transmitted data lines high. In C64 mode (only) CLOSE removes both buffers.
386 COMMODORE 128
N O TE: Care should be taken to ensure all data is transmitted before

closing the channel. A way to check this from BASIC is:
10 DO
20 LOOP WHILE PEEK (2575) and 3
30 CLOSE lfn
Location 2575 is the RS-232 “ ENABL” location.
Table 12-1 defines the User Port RS-232 channel lines.
(6526 DEVICE # 2 Loc. $DDOO-$DDOF)
PIN 6526 IN/

ID ID DESCRIPTION EIA ABV OUT MODES
c PBO Received Data (BB) Sin IN 1 2

D PB1 Request to Send (CA) RTS OUT 1*2
E PB2 Data Terminal Ready (CD) DTR OUT 1*2
F PB3 Ring Indicator (CE) Rl IN 3
H PB4 Received Line Signal (CF) DCD IN 2
J PB5 Unassigned ( ) XXX IN 3

K PB 6 Clear to Send (CB) CTS IN 2
L PB7 Data Set Ready (CC) DSR IN 2
B FLAG2 Received Data (BB) Sin IN 1 2

M PA2 Transmitted Data (BA) Sout OUT 1 2
A GND Protective Ground (AA) GND 1 2

N GND Signal Ground (AB) GND 1 2 3
MODES:
1. 3-Line Interface (Sjn,S0Ut,GND).
2. X-Line Interface.
3. User Available Only (Unused/unimplemented in code).
* These lines are held high during 3-Line mode.
T a b le 12-1. User P o rt Lines
[7 ] [6 ] [5 ] [4 ] [3 ] [2 ] [1 ] [ 0 ] ( M a c h i n e L a n g . — RSSTAT)
: : : : : : : : _ P A R I T Y ERROR BIT
: : : : : : :______ F R A M I N G ERROR BIT
: : : : : :___________ RECEIVER BUFFER O V E R R U N BIT
: : :________________RECEIVER BUFFER — EMPTY
(USE TO TEST AFTER A G E T # )
: .-____________________ CTS S IG N A L M I S S I N G BIT
: : :________________________ UN USED BIT
:_____________________________ DSR S IG N A L M I S S I N G BIT
:__________________________________ BREAK DETECTED BIT
Figure 12-3. RS-232 Status Register

N O TES to Figure 12-3:

The RS-232 Status Register can be read with these statements.
If the STATUS = 0 , then no error has been detected.
S = PEEK (2580): PRINT S : POKE 2580,0
If the RS-232 status is read by BASIC with the above statements or by
using the KERNAL READST routine, the RS-232 status word is cleared
when you exit. If multiple uses of the STATUS word are necessary, the
S variable above should be assigned to another variable. For example:
SR = S: REM ASSIGNS S TO SR
The RS-232 status is read (and cleared) only when the RS-232 channel
was the last external I/O used.
SAMPLE BASIC PROGRAM

USING RS-232 CHAN N EL
This simple terminal program works in both C128 and C64 modes:
10 REM SIMPLE DUMB TERMINAL EMULATOR

20 REM THIS PROGRAM REQUIRES A MODEM
30 OPEN5,2,3,CHR$(6)+CHR$(0)
40 DIMF%(255),T%(255)
50 FORJ=3 2T06 4 :T% (J )=J :NEXT
60 T%(13)=13:T%(20)=8:RV=18:CT=0
70 FORJ=6 5T090:K=J+3 2:T%(J)=K:NEXT
80 FORJ=91T095:T%(J)=J:NEXT
90 F O R J = l 9 3 T 0 2 1 8 : K = J - 1 2 8 : T % ( J ) = K : NEXT
100 T%(146)=16:T%(133)=16
110 T%(137)=3:T%(134)=17:T%(138)=19
120 FORJ=0TO255
130 K = T % (J )
140 IFKOOTHENF%(K)=J:F% (K+128 )= J
150 NEXT
160 PRINT" "CHRS(147)
170 GET#5,A$
180 IFA$=""ORST<>0THEN220
190 PRINT" "C H R $ ( 1 5 7 ) ; C H R $ ( F%( A S C (A$ ) ) ) ;
200 IFF%(A S C {A $ ))=34THENPOKE212,0
210 GOT0170
220 PRINTCHR?(R V )" "CHR$(157);C HR$(146);:GETA$
230 IFA$<>""THENPRINT# 5,CHR$(T % (A S C (A $ ) ) ) ;
240 CT=CT+1
250 IFCT=8THENCT=0:RV=164-RV
260 GOT0170
ZERO-PAGE MEMORY LO CATIONS A N D

USAGE FOR RS-232 SYSTEM INTERFACE
The zero-page RS-232 storage locations for both C128 and C64 modes are as follows:
$00A 7-IN B IT — Receiver input bit temp storage.
$00A 8-B IT C I— Receiver bit count in.
$OOA9-RINONE— Receiver flag start bit check.
388 COMMODORE 128
$00AA-RIDATA— Receiver byte buffer/assembly location.

$00AB-RIPRTY— Receiver parity bit storage.
$00B4-BITTS — Transmitter bit count out.
$00B5-NXTBIT — Transmitter next bit to be sent.
$OOB6-RODATA— Transmitter byte buffer/disassembly location.
All the above zero-page locations are used locally and are given only as a guide to
understand the associated routines. These cannot be used directly by the BASIC or
Kernal-level programmer to do RS-232-type programming functions. The system RS-232
routines must be used.
NONZERO-PAGE MEMORY LOCATIO NS A N D

USAGE FOR RS-232 SYSTEM INTERFACE
The general (nonzero-page) RS-232 storage locations for C l28 mode are as follows:
(The addresses in parentheses are for C64 mode.)
$OA10($0293)-M51CTR— Pseudo 6551 Control Register (see Figure 12-1).

$0A11($0294)~M51CDR— Pseudo 6551 Command Register (see Figure 12-2).
$0A12($0295)-M 51AJB— Two bytes following the Control and Command Registers in
the file name field. These locations contain the user-defined baud rate for the start
of the bit test during the interface activity, which, in turn, is used to calculate baud rate.
$0A14($0297)— RSSTAT— The RS-232 Status Register (see Figure 12-3).
$0A15($0298)-BITNUM — The number of bits to be sent/received.
$OA16($0299)-BAUDC)F— Two bytes that are equal to the time of one bit cell (based on
system clock/baud rate).
$0A18($029B)-RIDBE— The byte index to the end of the receiver FIFO buffer.
$0A19($029C)-RIDBS — The byte index to the start of the receiver FIFO buffer.
$OA1A($()29D)-RODBS— The byte index to the start of the transmitter FIFO buffer.
$OA1B($()29E)-RODBE— The byte index to the end of the transmitter FIFO buffer.
$0A0F($02A1)-ENABL— Holds current active interrupts in the CIA # 2 ICR. When bit
4 is turned on, the system is waiting for the Receiver Edge. When bit 1 is turned
on, then the system is receiving data. When bit 0 is turned on, then the system is
transmitting data.
O U TP U T TO A SCREEN
The most common form of computer output (called soft copy) is a display tube that
displays luminescent dots on a screen controlled by electrical signals. Although the
television is a specialized monitor and limited by the dot density prescribed by FCC
standards for proper reception, it is available in most homes and becomes a ready-to-use
output device for the home computer. The C128 computer sends radio frequency signals
through the TV port to the television’s antenna and tuner.
Monitors for computers produce better-quality displays since they are direct-wired
and involve no tuners. The C l28 includes a port to accept the traditional monitors as
well as Commodore’s 40-column 1702 color monitor. The C l28 80-column color
capability is enhanced with the new 40/80-column color monitor Model 1902. In
this RGB I monitor, signals separately control each of the three color guns (red, green,
blue) for maximum clarity and detail.
The screen can be accessed as an input or output device through the OPEN
command as device 3.
OPEN 1,3
The screen is the default output device on the C l28. Typing on the keyboard (device 0)
results in a screen character display output whenever the monitor is powered up. No
special commands are normally required. However, once control is sent to another
device, reversion to the screen may be desired.
Control of the screen display is available primarily with the PRINT command;
associated punctuation including the quote and the semicolon and common, BASIC
functions including the TAB(X), SPC(X), and CHR$(X) as well as the POKE com
mand. In addition, there are color controls, sprite animation and graphics keys.
If using the dual 1902 monitor or two separate 40- and 80-column monitors, it is
possible to swap control between the two screens via the 4 0 / 8 0 switch or pressing
the ESC and X keys. If one is used for text and the other for graphics, use the
GRAPHIC command.
The following command will also swap the display output device:
PRINT CHR$(27)“ X ”
In addition to a screen display, sound commands can be used through a Commodore
monitor or the television speakers. See Chapter 11, Sound and Music.
O U TP U T TO THE DATASSETTE
Commodore programs and files can also be stored on tape cassette using a special
recorder called the Datassette. Unlike most tape recorders, which store analog sinusoidal
signals, the Datassette stores digital pulse signals that are more distinct, resulting in
a reliable storage format. Compared to a disk storage system, tape is somewhat slower
and limited to program and sequential files. Access to the files on tape is sequential.
Just as DLOAD automatically loads a program from disk, LOAD loads a
program from tape. Similarly SAVE and VERIFY automatically default to tape. A
LOAD for machine code requiring a secondary address would be:
LOAD1‘NAM E’ ’,1,1
Access to tape for sequential files is made with an OPEN statement to device 1:
OPEN 1,1
A secondary address of 0 (default value) assumes the file is opened for reading
(using IN PU T # or G E T #). A secondary address for writing (PRIN T# statements)
requires a 1 or 2, where 2 includes an end-of-tape marker.
390 COMMODORE 128
OPEN 1,1,2 ‘ ‘FILENAM E”
The end-of-tape marker prevents searching for files beyond the data file you are looking
for. It is written when the CLOSE command is given.
N O TE: Do not connect the Datassette and the 1571 disk drive to the C l 28
at the same time. If they are used at the same time, the data transfers may
not be reliable.
IN PU T FROM THE
CONTROLLER PORTS
The C l 28 has two controller ports (numbered 1 and 2) that allow the use of control
devices such as a mouse, joystick, paddle or light pen. Each game port will accept either
one joystick, one mouse, or one paddle pair. A light pen can be plugged only into Port 1.
JOYSTICKS
A digital joystick has five distinct switches. Four of the switches are used for direction
and one is used for the fire button. The joystick switches are arranged as shown in
Figure 12-A.
(Top)
FIRE
(Switch 4)
UP
(Switch 0)
LEFT RIGHT
(Switch 2) (Switch 3)
DOW N
(Switch 1)
Figure 12-4. Joystick Switch A rra n g e m e n t
These switches correspond to the lower 5 bits of the data in location 56320 or
56321. Normally the bit is set to a 1 if a direction is not chosen or the fire button is not
pressed. When the fire button is pressed, the bit (bit 4 in this case) changes to a 0. To
read the joystick from C64 BASIC, the following subroutine should be used:
IN P U T /O U TP U T G UIDE 391
10 REM J O Y S T I C K RE AD R O U T I N E F O R C 1 2 8 OR C 6 4 MODES
20 JV=PEEK(56321)
30 I F ( J V AND 1 5 ) < 1 5 T H E N P R I N T 1 5 - ( J V A N D 1 5 )
40 F I R E = J V AND 1 6
50 I F F I R E 0 1 6 THEN P R I N T " F I R E "
6 0 GOTO 2 0
NOTE: For the first joystick, set JV = PEEK (56320).
The values for JV correspond to the joystick directions shown in Table 12-2.
JV E Q U A L T O D IR E C T IO N
0 NONE
1 UP
2 DOWN
3 —
4 LEFT
5 UP & LEFT
6 DOWN & LEFT
7 —
8 RIGHT
9 UP & RIGHT
10 DOWN & RIGHT
T a b le 12 -2 . Joystick D irection Values
A machine language routine that accomplishes the same task is given in the
section on shift and rotate instructions in Chapter 5, Machine Language on the
Commodore 128.
BASIC 7.0 includes special joystick functions. To determine the position of joystick
N, PRINT JOY(N) returns the position of the joystick. A number between 0 and 8
inclusive specifies the relative position where 0 is stationary and 1 through 8 represent
clockwise positions. Values between 128 and 136 signify the fire button is pressed as
well. Thus if JOY(2) returns 131, it means joystick number 2 fires to the right (position 3).
1 0 REM C l 2 8 - M O D E S P R I T E - M O V E J O Y S T I C K R O U T I N E
2 0 S P R I T E 1 , 1 , 7 : REM E N A B L E S P R I T E 1
3 0 M O V S P R 1 , 1 6 0 , 1 5 0 : REM P L A C E S P R I T E 1 ON C E N T E R O F T H E S C R E E N
4 0 DO
50 : DO
60 : A = J O Y ( l ) : R E M GET J O Y S T I C K D I R E C T I O N VALUE
70 : I F A = 0 THEN L O OP :R EM LOOP I F ZERO - T H I S KEEPS S P R I T E STATIONARY
80 : B = A : R E M S E T B EQUAL TO A
90 : Z = ( A - l ) * 4 5 :REM CALCULATE ANGLE
100 : M O V S P R 1 , 5 ; Z : RE M MOVE S P R I T E 5 U N I T S A T T H E C A L C U L A T E D A N G L E
110 : A = J O Y ( l ) : R E M CHECK J O Y S T I C K VALUE AG AIN
1 2 0 : L O O P W H I L E A = B : R E M MOVE I N T H E S A ME D I R E C T I O N W H I L E NEW V A L = O L D J O Y V A L U E
1 3 0 L O O P : RE M O T H E R W I S E C H E C K J O Y V A L U E A G A I N AND MOVE I N D I F F E R E N T D I R E C T I O N
392 COMMODORE 128
Using Boolean logic and the masking technique, a comparison can be made to
determine whether the fire button was depressed:
IF (JOY(N)AND 128)= 128 THEN PRINT“ FIRE”
TH E 1350 MOUSE
The Commodore 128 makes use of an additional new peripheral device known as a
mouse. The Commodore mouse is a user interface that allows you to control the
operations of a program, such as selecting menu or icon options, without the need to
type on the keyboard. It attaches to the controller ports like a joystick, paddle or light
pen. Programs that support a mouse usually have a pointer indicator to let you know
where the mouse is pointing on the screen. To move the pointer on the screen, place the
mouse on a flat surface and move it in the direction you want to move the pointer. Lift
the mouse off the surface and the screen indicator remains stationary. The left hand
button responds like a joystick fire button. The right hand button may have additional
functions as specified by a particular software package.
As far as BASIC is concerned, the mouse is supported by the JOY function, the
same way a joystick is controlled. The previous joystick program example, which
moves a sprite on the screen, also works with the mouse. The Joystick Musician
program in Chapter 11 also works with the mouse. Any BASIC program using the JOY
joystick function should operate with the Commodore mouse.
The mouse also works in C64 mode, the same way a joystick routine would in
C64 mode. The previous C64 mode joystick example program supports the mouse in
C64 mode.
PADDLES
A paddle is connected to both CIA-1 and the SID chip (MOS 6581 Sound Interface Device)
through a controller port. The paddle value is read via the SID registers 54297 ($D419) and
54298 ($D41A). Paddles are not reliable when read from BASIC alone! The best way to
use paddles, from BASIC or machine code, is to use a machine language routine (SYS
to it from BASIC then PEEK the memory locations used by the subroutine).
The following BASIC program moves a sprite with paddles:
1 0 REM C 1 2 8 - M O D E S P R I T E - M O V E P A D D L E R O U T I N E
2 0 REM P L A C E DUAL P A D D L E S I N C O N T R O L L E R P O R T A
3 0 R E M P A D D L E 1 C O N T R O L S H O R I Z O N T A L S P R I T E MO V E ME N T
4 0 REM P A D D L E 2 C O N T R O L S V E R T I C A L MOVEMENT
50 SPR IT E 1 , 1 , 7
6 0 DO
7 0 A = P O T ( 1 ) : REM G E T P O S I T I O N ( X ) O P P A D D L E 1
8 0 B = P O T ( 2 ) : RE M G E T P O S I T I O N ( Y ) O F P A D D L E 2
9 0 P R I N T A ; B : RE M P R I N T V A L U E S
1 0 0 M O V S P R 1 , A , B : R E M MOVE S P R I T E 1 A C C O R D I N G T O P A D D L E V A L U E S
1 1 0 I F ( A > 2 5 5 ) OR ( B> 2 5 5 ) TH EN B E G I N : R E M CHECK FOR F I R E BUTTON
120 : SOUND 2 , 6 4 0 0 0 , 6 0 , 0 , 1 0 0 0 , 1 6 0 0 0 , 3 : RE M F I R E
130 : CO LOR 0 , 3 : REM CHANGE S C R E E N COLOR
140 : SLEEP 1 : REM DELAY
150 : CO L O R 0 , 1 : REM C H A N G E S C R E E N C O L O R BACK
160 B E N D : REM END T H E B E G I N / B E N D S T R U C T U R E
1 7 0 L O O P : R E M DO I T A G A I N
LIG H T PEN
The light pen input latches the current screen position into a pair of registers (LPX,
LPY) on a low-going edge. The X position register 53267 ($D013) will contain the 8 MSB
of the X position at the time of transition. Since the X position is defined by a 512-state
counter (9 bits), resolution to two horizontal dots is provided. Similarly, the Y position
is latched in its register 53268 ($D014), but here 8 bits provide single raster resolution
within the visible display. The light pen latch may be triggered only once per frame, and
subsequent triggers within the same frame will have no effect. Therefore, you must take
several samples before turning the pen to the screen (three or more samples average),
depending upon the characteristics of your light pen. The following program is a light
pen draw routine in C l 28 mode:
1 0 RE M C l 2 8 MO D E L I G H T P E N DRAW R O U T I N E
2 0 REM P L A C E L I G H T P E N I N C O N T R O L L E R P O R T 1
3 0 C O L O R 0 , 2 : C O L O R 4 , 2 : C O L O R 1 , 1 : REM S E L E C T S C R E E N COLORS
4 0 G R A P H I C 1 , 1 : R E M S E L E C T B I T MAP MODE
5 0 DO
60 : X=PEN(0) : RE M G E T H O R I Z O N T A L COORDINATE
70 : Y=PEN(1) : REM G E T V E R T I C A L C O O R D I N A T E
80 : DRAW 1 , X , Y : R E M P L O T P O I N T
90 L O O P : R E M DO I T A G A I N
100 :
110 REM I F YOU L I K E T H E D R A W I N G ,
120 REM B R E A K T H E P R O G R A M AND
130 REM SAVE I T AS A BINARY F I L E
140 REM FROM $ 2 0 0 0 - $ 3 F F F .
O U TP U T CONTROL TO
ALL DEVICES
Each peripheral device is identified by a device number, as listed in Table 12-3.
DEVICE DEVICE NUMBER SECONDARY ADDRESSES
Keyboard 0
Datasette 1 0 ,1 ,2
Modem (RS232) 2 0
Screen 3 0 ,1
Printer 4 or 5 0 ,7 ,etc
Disk Drive 8,9,10,11 2 -14,15
T a b le 12-3. Device N u m b ers
Serial devices can be numbered 4 through 30, but they usually conform to the
above conventions. All devices can be accessed in a single OPEN statement by using
an integer variable:
394 COMMODORE 128
10 INPUT "SELECT DEVICENUMBER";D*

20 INPUT " S E L E C T S E C O N D A R Y A D D R E S S " ; S%
30 I N P U T "NAME OF F I L E OR C O NTR OL R E G I S T E R " ; A $
40 OPEN 9 , D % , S % , A $
50 GOSUB 80
60 CLOSE 9
70 EN D
80 REM P L A C E S U B R O U T I N E HE R E
90 RETURN
THE IN P U T/O U TP U T
PINOUTS
THE SERIAL PORT

The Serial Port is a daisy chain arrangement designed to let the Commodore 128 computer
communicate with more than one device, such as Commodore disk drives and printers.
Up to five devices can be connected at one time. The serial bus can transmit control
signals from the computer, send data onto the bus, and receive data from the bus. Data
or control signals are routed to the proper device by opening a specific bus address
ranging from 4 to 31. Typically, addresses 4 and 5 are reserved for printers, and
addresses 8 through 11 are for the disk drive.
Six lines are used in serial bus operation— three input and three output. The three
input lines bring data, control and timing signals into the computer. The three output
lines send data, control and timing signals from the computer to the external devices.
See Hardware Specifications (Chapter 16) for details on the C l28 serial bus protocol.
SERIAL I/O
PIN TYPE
1 SERIAL SRQIN
2 GND
3 SERIAL ATN IN/OUT
4 SERIAL CLK IN/OUT
5 SERIAL DATA IN/OUT
6 RESET
Figure 12-5. Serial P o rt Pinouts
USER PORT (RS-232 CHANNEL)

The User Port (RS-232 Channel) connects the Commodore 128 to the outside
world, such as another Commodore computer. It is directly connected to the 6526 CIA
# 1 chip. Figure 12-6 shows the User Port pinouts. Table 12-4 defines the pinouts.
IN P U T /O U TP U T G UIDE 395
1 2 3 4 5 6 7 8 9 10 11 12
A B C D E F H J K L M N
Figure 12 -6 . U ser P o rt Pinouts
PIN
T O P SIDE DESCRIPTION NOTES
1 GROUND
2 + 5V (100 mA M AX.)
3 RESET By grounding this pin, the Commodore 128 will do
a cold start, resetting completely. The pointers to
a BASIC program will be reset, but memory will
not be cleared. This is also a RESET output for
the external devices.
4 CNT1 Serial port counter from CIA-1 (see CIA specs).
5 SP1 Serial port from CIA-1 (see 6526 CIA specs).
6 CNT2 Serial port counter from CIA-2 (see CIA specs).
7 SP2 Serial port from CIA-1 (see 6526 CIA specs).
8 PC2 Handshaking line from CIA-2 (see CIA specs).
9 SERIAL This pin is connected to the ATN line o f the serial
ATN IN bus.
10 9 VAC + phase Connected directly to the Commodore 128 trans
11 9 V A C -phase former (100 mA M ax.).
12 GROUND
BOTTOM SIDE
A GROUND The Commodore 128 gives you control over Port
B FLAG2 B on CIA chip 1. Eight lines for input or output
C PBO are available, as well as two lines for handshaking
D PB1 with an outside device. The 1/0 lines for Port B
E PB2 are controlled by two locations. One is the Port
F PB3 itself, and is located at 56577 ($DD01 HEX).
H PB4 Naturally you PEEK it to read an INPUT, or
J PB5 POKE it to set an OUTPUT. Each of the eight
K PB 6 I/O lines can be set up as either an INPUT or an
L PB7 OUTPUT by setting the Data Direction Register
M PA2 properly.
N GROUND
T a b le 12 -4 . U ser P o rt Pin Descriptions
The D a ta D ire c tio n R eg iste r has its location at 56579 ($DD03 hex). Each of the
eight lines in the port has a bit in the 8-bit Data Direction Register (DDR) that controls
whether that line will be an input or an output. If a bit in the DDR is a 1, the
396 COMMODORE 128
corresponding line of the port will be an output. If a bit in the DDR is a 0, the
corresponding line of the port will be an input. For example, if bit 3 of the DDR is set to 1,
then line 3 of the port will be an output. If the DDR is set like this:
BIT # : 7 6 5 4 3 2 10
VALUE: 0 0 11 1 0 0 0
lines 5, 4 and 3will be outputs since those bits are l ’s. The rest of the lines will be
inputs, since those lines are 0 ’s.
To PEEK or POKE the User Port, it is necessary to use both the DDR and the port
itself.
Remember that the PEEK and POKE statements need a number from 0 to 255.
The numbers given in the example must be translated into decimal before they can be
used. The value would be:
25 + 24 + 23 = 32 + 16 + 8 = 56
Notice that the bit number for the DDR is the same number that is equal to 2 raised to a
power to turn the bit value on.
(16 = 2 | 4 = 2 x 2 x 2 x 2 , 8 = 2 | 3 = 2 x 2 x 2 )
The two other lines, flagl and PA2, are different from the rest of the User Port.
These two lines are mainly for handshaking, and are programmed differently from
port B.
Handshaking is needed when two devices communicate. Since one device may run
at a different speed than another device, it is necessary to give each device some way of
knowing what the other device is doing. Even when the devices are operating at the
same speed, handshaking is necessary to communicate when data is to be sent, and if it
has been received. The flagl line has special characteristics that make it well suited for
handshaking.
Flagl is a negative-edge-sensitive input that can be used as a general-purpose
interrupt input. Any negative transition on the flag line will set the flag interrupt
bit. If the flag interrupt is enabled, this will cause an Interrupt Request. If the
flag bit is not enabled, it can be polled from the Interrupt Register under program control.
PA2 is bit 2 of port A of the CIA. It is controlled like any other bit in the port.
The port is located at 56576 ($DD00). The Data Direction Register is located at 56578
($DD02).
For more information on the 6526, see the hardware chapter.
THE COMPOSITE
VIDEO CO N N EC TO R
This DIN connector supplies direct audio and composite video signals. These can
be connected to the Commodore monitor or used with separate components. This is the
40-column output connector. Figure 12-7 shows the pinouts for the composite video
connector. Table 12-5 describes the pinouts.
Figure 12-7. C om posite Video C o n necto r Pinouts
PIN TYPE NOTE

2 GND
3 AUDIO OUT
5 AUDIO IN
7 NC No connection
8 NC No connection
T ab le 12-5. C om posite Video C o n necto r Pin Descriptions
THE RGBI
VIDEO C O N N EC TO R
The RGBI video connector is a 9-pin connector that supplies an RGBI (Red/Green/Blue/
Intensity) signal. This is the 80-column output. Figure 12-8 shows the RGBI pinouts.
Table 12-6 defines the RGBI pinouts.
Figure 12-8. RGBI C o n n e cto r P inouts

398 COMMODORE 128
PIN SIGNAL
1 Ground
2 Ground
3 Red
4 Green
5 Blue
6 Intensity
7 Monochrome
8 Horizontal Sync
9 Vertical Sync
T a b le 12-6. RGBI C o n necto r Pin Descriptions
THE CASSETTE C O NNECTO R

A 1530 Datassette recorder can be attached to the cassette port to store programs and
information. Figure 12-9 shows the cassette port pinouts. Table 12-7 describes the
pinouts.
1 2 3 4 5 6
A B C D E F
Figure 12 -9 . Cassette P o rt Pinouts
PIN TY PE
A -l GND
B-2 + 5V
C-3 CASSETTE MOTOR
D-4 CASSETTE READ
E-5 CASSETTE WRITE
F-6 CASSETTE SENSE
T a b le 12 -7 . Cassette P o rt Pin Descriptions
TH E CONTROLLER PORTS
There are two controller ports, numbered 1 and 2. Each controller port can accept a
joystick, mouse or game controller paddle. A light pen can be plugged only into Port 1, the
port closest to the front of the computer. Use the ports as instructed with the software.
Figure 12-10 shows the Controller Port pinouts. Table 12-8 describes the pinouts.
CONTRO L PORT I C O N T R O L PORT 2
PIN TYPE NOTE PIN TYPE NOTE
1 JOYAO 1 JOYBO
2 JOYA1 2 JOYB1
3 JOYA2 3 JOYB2
4 JOYA3 4 JOYB3
5 POT AY 5 POT BY
6 BUTTON A/LP 6 BUTTON B
7 + 5V M AX. 50mA 7 + 5V M AX. 50mA
8 GND 8 GND
9 POT AX 9 POT BX
T a b le 12-8. C o n tro lle r P o rt Pin Descriptions
THE EXPANSION PORT

Vhe expansion port connector is a 44-pin female edge connector accessing the comput
er’s address and data buses. It accepts Commodore’s preprogrammed game and utility
software cartridges. As a parallel port, it can accept Commodore’s IEEE peripherals
with an IEEE interface for controlling instrumentation and other devices. RAM expan
sion modules will also connect to this port and accept the BASIC commands of FETCH,
STASH and SWAP. x
Figure 12-11 shows the expansion port pinouts. Table 12-9 describes the pinouts.
2221 20 191B 17 1615 14 13 12 11 10 9 8 7 6 5 4 3 2 1
Z V X W V U T S R P N M L K J H F E D C B A
(view of port from the back of the C128)
Figure I 2 - I I . Expansion P o rt Pinouts

400 COMMODORE 128
PIN TYPE PIN TYPE
1 GND A GND
2 + 5V B ROMH
3 + 5V C RESET
4 IRQ D NMI
5 R/W E S 02
6 Dot Clock F A15
7 I/O 1 H A14
8 GAME J A13
9 EXROM K A12
10 I/O 2 L A ll
11 ROML M A10
12 BA N A9
13 DMA P A8
14 D7 R A7
15 D6 S A6
16 D5 T A5
17 D4 U A4
18 D3 V A3
19 D2 W A2
20 Dl X A1
21 DO Y A0
22 GND Z GND
T a b le 12-9. Expansion P o rt Pin Descriptions

13
THE
COMMODORE 128
OPERATING
SYSTEM
402 COMMODORE 128
The Commodore 128 operating system controls, directly or indirectly, all functions of
your computer. The Commodore 128 operating system is housed in a ROM chip called
the K ern al, which contains about 16K of machine language instructions. These instruc
tions make up the routines that control all the machine’s functions— even the ones you
take for granted. For instance, the Kernal controls all input and output functions,
including receiving the characters from the keyboard when you type, sending text to a
printer, and displaying graphics and text on the screen. Every task performed by the
computer other than application program activities is controlled by the Kernal. The
Kernal even manipulates and executes the application programs you load or type into
your com puter’s memory.
TAKING FULL ADVANTAGE OF THE

COMMODORE 128 OPERATING SYSTEM
The 16K of operating system instructions contained in the Kernal are available for use
with your own programs. Instead of “ reinventing the wheel” and duplicating code, you
can call (that is, use) these Kernal routines in your own programs. You do this through
the Kernal Jump Table, which consists of a series of ROM entry points in which you
can call machine language routines already available in the Commodore 128 Kernal. By
calling these routines, which handle the most fundamental functions of the computer,
you avoid duplicating code. This helps you utilize your computer to its fullest potential.
The Kernal Jump Table also facilitates compatibility. If the Commodore 128
operating system is modified or upgraded, which happens frequently in the microcom
puter industry, the entry points in the jum p table are revised to reflect address changes
o f Kernal subroutines. The key to keeping applications programs compatible from one
version of the operating system to another is to enter the operating system through the
Kernal Jump Table. Instead of jumping directly to a subroutine (JSR), use the Kernal
Jump Table as the entry point, since it contains the correct address vector to the
specified routine, regardless of the version of the Kernal being used. If you always enter
the operating system routines from the Kernal Jump Table, the address of the desired
routine will always be reached. On the other hand, if you bypass the Kernal Jump Table
and jum p directly to the address where you think the routine resides, you may cause an
error, because the starting point of the desired routine may have changed from one
version of the operating system to another.
THE COMMODORE 128 OPERATING SYSTEM 403
HOW TO USE (CALL)

THE KERNAL ROUTINES IN
YOUR OWN PROGRAMS
Most of the system subroutines require specific parameters, or values, which must be
loaded into the accumulator, X or Y index registers. Some Kernal subroutines require
additional preparatory routines to be called before invoking the specific target routine.
Each Kernal subroutine terminates with an RTS instruction that tells the microprocessor
to return from the subroutine. Many of the Kernal routines return values that are placed
in the accumulator, X or Y registers. Some even return error codes in the status register,
which can be acted on in your applications program.
Here is the procedure for calling a Kernal subroutine:
1. Place the necessary preparatory values in the required registers— either the A,
X or Y registers.
2. Enable the appropriate system configuration. For example, if the routine
requires the Kernal to be present, invoke either configuration 12, 13, 14 or 15,
since these make the Kernal available. Bank 15 is the default configuration.
3. Call the subroutine with the JSR instruction, using the address of the jump
vector as shown in the Kernal Jump Table in Figure 13-1.
For example, the starting address of the routine ACPTR is stored, starting
in location $FFA5. Actually, the operation code (opcode) for the JMP instruc
tion is stored at location $FFA5 and the address of the entry point of the
routine is stored at locations $FFA6 and $FFA7. What really happens is that
the JSR instruction in your application program transfers control to the jump
table address ($FFA5, for example); then the JMP instruction at $FFA5 jumps
to the subroutine at the address specified in locations $FFA6 and $FFA7. In
other words, when you issue a JSR instruction to the Kernal Jump Table, you
actually perform two jumps: one (JSR) to the jump table, and then a second
jum p (JMP) to the actual starting address of the routine.
The routine terminates with an RTS instruction that is already part of the
Kernal subroutine. Your application program resumes with the instruction
immediately following the JSR instruction.
4. Upon return from the subroutine, check the status register for any error conditions. If
an error condition is present, take precautions in your application
program to handle the error and act on the value of the status register.
Figure 13-1 is the C l 28 system vector and jump table that includes the
name, address and function description of the operating system Kernal routines.
404 COMMODORE 128
C l 28 S Y S T E M V E C T O R S
1. $FFFB SYSTEM ;operating system vector (RAM I)

2 . $FFFA NMI p rocessor NMI vector
3. $FFFC RESET ;processor RESET vector

4. $FFFE IRQ p rocessor IRQ/BRK vector
C B M S T A N D A R D K E R N A L JU M P T A B L E C A LL S
1. $FF81 JM P CINT ;init screen editor and devices

2. $FF84 JM P IOINIT ;init I/O devices
3. $FF87 JM P RAMTAS ;init RAM and buffers
4. $FF 8 A JM P RESTOR ;init indirect vectors (system)
5. $FF 8 D JM P VECTOR ;init indirect vectors (user)
6. $FF90 JM P SETMSG ;kernal messages on/off
7. $FF93 JM P SECND ;serial: send SA after LISTN
8. $FF96 JM P TKSA ;serial: send SA after TALK
9. $FF99 JM P M EM TOP ;set/read top of system RAM
10. $FF9C JM P M EMBOT ;set/read bottom o f system RAM
11. $FF9F JM P KEY ;scan keyboard
12. SFFA2 JM P SETTMO preserved)
13. $FFA5 JM P ACPTR ;serial: byte input
14. $FFA 8 JM P CIOUT ;serial: byte output
15. $FFAB JM P UNTLK ;serial: send untalk
16. $FFAE JM P UNLSN ;serial: send unlisten
17. $FFB1 JM P LISTN ;serial: send listen
18. $FFB4 JM P TALK ;serial: send talk
19. $FFB7 JMP READSS ;read I/O status byte
20. $FFBA JM P SETLFS ;set channel LA, FA, SA
21. $FFBD JM P SETNAM ;set filename pointers
22. $FFC0 JM P (IOPEN) ;open logical file

23. $FFC3 JM P (ICLOSE) ;close logical file
24. $FFC 6 JM P (ICHKIN) ;set input channel
25. $FFC9 JM P (ICKOUT) ;set output channel

26. $FFCC JM P (ICLRCH) ;restore default channels
27. $FFCF JM P (IBASIN) ;input from channel
28. $FFD2 JM P (IBSOUT) ;output to channel

29. $FFD5 JM P LOAD ;load from file
30. $FFD 8 JM P SAVE ;save to file
31. $FFDB JMP SETTIM ;set internal clock

32. $FFDE JM P RDTIM ;read internal clock
33. $FFE1 JMP (ISTOP) ;scan STOP key
34. $FFE4 JM P (IGETIN) ;read buffered data

35. $FFE7 JM P (ICLALL) ;close all files and channels
36. $FFEA JM P UDTIM in crem en t internal clock
37. $FFED JM P SCRORG ;get current screen window size
38. $FFF0 JM P PLOT ;read/set cursor position
39. $FFF3 JM P IOBASE ;read base address o f I/O block
N E W C l 28 K E R N A L JU M P T A B L E C A LLS
1. $FF47 JM P SPIN SPOUT ;setup fast serial ports for I/O
2. $FF4A JM P CLOSE ALL ;close all files on a device

3. $FF4D JM P C64MODE ;reconfigure system as a C64
4. $FF50 JM P DMA CALL ;send command to DMA device
5. $FF53 JM P BOOT CALL ;boot load program from disk

6. $FF56 JM P PHOENIX ;init function cartridges
7. $FF59 JM P LKUPLA ;search tables for given LA
8. $FF5C JM P LKUPSA ;search tables for given SA

9. $FF5F JM P SW APPER ;switch between 40 and 80 columns
10. $FF62 JM P DLCHR ;init 80-col character RAM
11. $FF65 JM P PFKEY ;program a function key

12. $FF 6 8 JM P SETBNK ;set bank for I/O operations
13. $FF 6 B JM P GETCFG ;lookup MMU data for given bank
14. $FF 6 E JM P JSRFAR ;gosub in another bank

15. $FF71 JM P JMPFAR ;goto another bank
16. $FF74 JM P INDFET ;LDA (fetvec),Y from any bank
17. $FF77 JM P INDSTA ;STA (stavec),Y to any bank

18. $FF7A JM P INDCMP ;CMP (cm pvec),Y to any bank
19. $FF7D JM P PRIMM ;print immediate utility
Figure 13-1. U ser-C allab le Kernal Routines
Figure 13-2 lists the conventions used in the description of each Kernal subrou
tine. The figure is followed by descriptions of the C l28 system vectors and Kernal
subroutines. Included in each description are the subroutine name, call address, prepara
tory routines needed (if any), the registers affected, the error codes associated with each
routine, how to use them and an example of each kernal subroutine call.
406 COMMODORE 128
USER CALLABLE
KERNAL ROUTINE C O N VE N TIO N S
C all Address: This is the call address of the Kemal routine, given in hexadecimal.
Function N am e: Name of the Kemal routine.
Register: Registers, memory and flags listed under this heading are used to pass
parameters to and from the Kemal routines.
P rep arato ry Routines: Certain Kemal routines require that data be set up by prepara
tory routines before the target routine can be called. The necessary routines are
listed here.
E rro r Returns: A return from a Kemal routine with the carry set indicates an error
was encountered in processing. The accumulator will contain the number of the
error.
E rro r Codes: Below is a list of error messages that can occur when using the
Kemal routines. If an error occurs during a Kemal routine, the carry bit of the
accumulator is set, and the number of the error message is returned in the
accumulator.
N O TE: Some Kemal I/O routines do not use these codes for error
messages. Instead, errors are identified using the Kemal READST routine.
NUMBER MEANING
0 Routine terminated by the STOP key

1 Too many open files
2 File already open
3 File not open
4 File not found
5 Device not present
6 File is not an input file
7 File is not an output file
8 File name is missing
9 Illegal device number
41 File read error
Registers A ffected: All registers, memory and flags used by the Kemal routine are
listed here.
Examples: An example of each Kemal routine is listed.
Description: A short explanation of the function of the Kernal routine is given here.
Figure 13-2. U ser-C allable K ernal R outine C onventions

C l28 SYSTEM VECTORS
The vectors listed below, with the exception of the SYSTEM vector, are located not
only in system ROM but in each RAM bank. The beginning and end of the system
interrupt handlers are found in the top page ($FFxx) of all memory configurations as
well. The reason is simple: An interrupt can occur anytime, from any memory configu
ration. The registers and memory configuration must be preserved prior to bringing the
operating system into context to process the interrupt. They must similarly be restored
before control is finally returned to the interrupted code. Note that the system vectors
are indirect jum ps, and are not usually called by the user since they terminate with an
RTI instruction, not an RTS instruction. In other words, they process interrupted events,
not subroutine calls.
1. $FFF8 SYSTEM ;operating system vector (RAMI)
The SYSTEM vector and accompanying KEY string provide applications software
with a means of regaining system control after a hardware reset. With this vector,
software may be protected from an otherwise unrecoverable situation. The KEY
string provides the distinction between a “ warm” reset and a “ cold” power-up. If
the system has just powered up, the KEY string is missing and thus the SYSTEM
vector is invalid; the CBM KEY is installed and the SYSTEM vector is set to C128
mode. If the system was reset (i.e., KEY was found), the SYSTEM vector is
moved to common RAM at $02 and an indirect JMP is performed. In most cases,
the user need only call IOINIT beore resuming control. The layout in RAM-1 is:
$FFF5 ‘C ’ ($43)
$FFF6 ‘B ’ ($42)
$FFF7 ‘M ’ ($4D)
$FFF8 SYSTEM vector low
$FFF9 SYSTEM vector high
For example, suppose a programmer set out to do some heavy-duty “ hacking”

using the built-in Monitor. Fully realizing the likelihood of losing control in a crash,
the programmer could redirect the SYSTEM vector to the Monitor itself and thus
regain control with minimal RAM loss, simply by pressing the RESET button. To
accomplish this, the programmer would enter:
a. >1FFF8 00 13 :aim SYSTEM to $1300

b. A 1300 JSR $FF84 :call IOINIT
JMP $B000 :call Monitor
2. $FFFA NMI ;processor NMI vector
The Non-Maskable Interrupt (NMI) vector is activated whenever the processor

NMI pin detects a negative edge. There are two possible sources of N M I’s under
normal conditions: the REST ORE key and RS-232 I/O. In the event of an
NMI, the operating system disables IRQ’s, saves the registers and current memory
408 COMMODORE 128
configuration on the stack, brings the system configuration (ROM ’s, I/O, RAMO)
into context, and executes an indirect jump through the RAM vector located at
$318. The system NMI handler clears the ICR register of CIA-2, from which it
determines the source of the interrupt. If it is from timer A, control passes to the
RS-232 transceiver. If not, the R E S T O R E key is assumed, and for safety,
the CBM convention of requiring the S T O P key to be simultaneously de
pressed is checked. If the STOP key is depressed, all system indirect vectors are
restored, IOINIT and CINT are called, and the SYSTEM—VECTOR (do not
confuse with the SYSTEM vector!) is taken. Control is returned upon restoration of
the registers and memory configuration. NM I’s may be disabled by causing an initial
NMI from timer A, but never reading the ICR to clear it, thus keeping the NMI
signal grounded.
Application software can intercept an NMI event by modifying either of the
two RAM vectors mentioned above. The NMI indirect vector at $318 in common
RAM will pass control whenever an NMI occurs. The SYSTEM_VECTOR, located
at $A00 in RAMO, will pass control after STOP/RESTORE is detected and handled.
For example, suppose a situation similar to the one illustrated previously for
the SYSTEM vector occurs. To return control to the Monitor whenever STOP/
RESTORE is detected, enter:
a. > A 00 00 BO :aim SYSTEM_VECTOR to Monitor
Similarly, to perform an action anytime an NMI occurs (e.g., RESTORE

alone), use the NMI indirect. For example, increment the VIC border color when
ever you press R E S T O R E (or cause any other NMI). Enter:
a. > 318 00 13 :aim indirect to $1300

b. A 1300 INC $D020 :change border color
JMP $FF33 :retum from interrupt
3. $FFFC RESET ;processor RESET vector
The processor RESET vector is activated whenever the system RESET signal is
low. It is low at power-up, and is pulled low by pressing the RESET button. This
signal effects not only the processor but most of the I/O devices found in the
system. In fact, RESET is the one processor control signal that is shared between
the two processors (8502 and Z80) of the C128. The Z80 gains initial (default)
control of the system while the 8502 is held in a waiting state. When the 8502
finally starts after a reset, the Kemal initialization routine START always receives
control and immediately performs the following actions:
1. Brings the system map into context.
2. Disables IRQ’s.
3. Resets the processor stack pointer.
4. Clears decimal mode.
5. Initializes the MMU.
6. INSTALLS the Kernal RAM code.
Up to this point there is no provision for user code. The next two routines in the
initialization path actually look for installed user code:
7. SECURE: Check and initialize the SYSTEM vector.

8. POLL: Check for a ROM cartridge.
POLL first scans for any installed C64 cartridges. They are recognized by
either the GAME or EXROM signal being pulled low. If so, G 064 is executed
(see the Kernal jump entry for details). Polling for C64 cartridges is actually
redundant at this point since the Z80 processor, which powers up initially, did this
already. POLL then scans for installed C l 28 cartridges and function ROM ’s. They
are recognized by the existence of the O key in any of the four function ROM
slots (two internal, two external) and are polled in the order external low (16 or
32KB), external high (16KB), internal low (16 or 32KB), internal high (16KB).
The entire format is:
$x000 —» cold start entry

$x003 - » warm start entry (unused)
$x006 ID. ($01-$FF)
$x007 -> “ CBM ” key string
where x = $8— or $C— .
The ID of any C l 28 cartridge found is entered into the Physical Address Table
(PAT) located at SAC1-SAC4. ID ’s must be non-zero. Cartridges may recognize
each other by examining the PAT for particular ID ’s. An ID of 1 indicates an
auto-start cartridge, and its cold start entry will be called immediately. All others
will be called later (see PHOENIX jum p), as will any auto-starters that RTS is to
POLL. A cartridge can determine where it is installed by examining CURBNK,
located at $ACO. Because it is possible for a cartridge to be called with interrupts
enabled, the following diversion from the above format is recommended (the warm
start entry is never called by the system):
$x000 SEI
$x001 JMP STARTUP
$x004 NOP
$x005 NOP
The balance of the C l28 initialization is:
9. IOINIT: Initialize I/O devices.
10. Check for STOP AND O keys.
11. RAMTAS: Initialize system RAM.
12. RESTOR: Initialize system indirects.
13. CINT: Initialize video displays.
14. Enable IR Q ’s (except foreign systems).
15. Dispatch.
IOINIT is perhaps the major function of the Reset handler. It initializes both C IA ’s
(timers, keyboard, serial port, user port, cassette) and the 8502 port (keyboard,
410 COMMODORE 128
cassette, VIC bank). It distinguishes a PAL system from an NTSC one and sets
PALCNT ($A03) if PAL. The VIC, SID and 8563 devices are initialized, including
the downloading of character definitions to 8563 display RAM (if necessary). The
system 60Hz IRQ source (the VIC raster) is started. IOINIT is callable by the user
via the jum p table.
During initialization, the user may press certain keys as a means of selecting
an operating mode. One key checked is the Commodore key O , indicating C64
mode is desired. While this key was scanned much earlier by the Z80 to speed
the switchover to C64 mode, there is a redundant check for it here. The only other
key scanned at this time is the S T O P key, which signals a request by the user
to power up into the Monitor utility. Note that control does not pass from the
initialization process at this point; the Kemal needs to know if RAMTAS should be
skipped. Only if the S T O P key is depressed and this was a “ w arm ” reset
(vs. “ cold” power-up) can RAMTAS be skipped.
RAMTAS clears all page-zero RAM, allocates the cassette and RS-232
buffers, sets pointers to the top and bottom of system RAM (RAM-0), and
installs the SYSTEM—VECTOR (discussed earlier under NMI) that points to BA
SIC cold start. Lastly it sets a flag, DEJAVU ($A02), to indicate to other routines
that system RAM has been initialized. This is the difference between a “ cold”
and a “ w arm ” system. If DEJAVU contains $A5, the system is “ warm ” and
SYSTEM _VECTOR is valid. Many programmers debugging code need to recover
from a system hang or crash via R E S E T but do not want RAM cleared. This
is why the S T O P key is scanned, RAMTAS is skipped, and the Monitor
(rather than BASIC) is selected. RAMTAS is callable by the user via the jump
table.
RESTOR initializes the Kemal indirect vectors. This must be done before
many system routines will function. Applications that complement the operating
system via “ wedges” must install them after they are initialized. RESTOR is user
callable from the jump table (see also the VECTOR call).
C IN T is the Editor’s initialization routine. Both 40- and 80-column display
modes are prepared, editor indirect vectors installed, programmable key definitions
assigned, and the 40/80 key scnaned for default display determination. CINT is also
a jum p table entry.
Finally, the IR Q ’s are enabled and control is passed to either BASIC
initialization, G 0 6 4 code, or the ML Monitor. BASIC will call the Kernal
PHOENIX routine upon the conclusion of its initialization, which will call any
installed C l 28 cartridges (any ID) and attempt to auto-boot an application from
disk.
An initialization status byte, INIT_STATUS ($A04), marks the progress of
the initialization process. It is cleared automatically at the beginning of the Reset
code, and as specific stages are completed, a particular bit is set. The layout is:
B7 —> 8563 characters installed

B6 —* CINT performed
BO —» BASIC initialized
Any IOINIT calls, including Reset, will not result in 8563 character RAM initial
ization if B7 is set. Similarly, CINT will not initialize the keyboard matrix lookup
tables and the programmable key definitions if B6 is set. This is how N M I’s, for
example, can call IOINIT and CINT without destroying users’ setups. Finally,
BASIC initialization must be complete before BO is set. This determines whether
the IRQ handler, for example, should call the BASIC IRQ routines. Note that the
following sequence of events should be performed for a BASIC programmer to
recover from a crash via R E S E T
1. Hold down S T O P key to enter Monitor.

2. Press and release RESET button.
3. Release S T O P key.
4. Enter: > A 0 4 C l :re-enable BASIC IRQ.
5. Enter: X :exit Monitor to BASIC.
This sequence is necessary because INIT_STATUS was cleared by the reset, and
the BASIC initialization was skipped, leaving BO reset. If BO had not been set,
BASIC IRQ routines such as SOUND, PLAY, and SPRITE handlers would not
have functioned, usually resulting in an apparent “ hang.”
4. $FFFE IRQ ;processor IRQ/BRK vector
This hardware vector is taken whenever the IRQ pin of the processor is pulled low,
or the processor executes a BRK instruction. For proper operation an IRQ must
occur sixty times every second [NTSC (60Hz) presents no problems, but adjust
ments have to be made for PAL (50Hz) systems]. The usual source of IRQ ’s in the
C128 is the VIC raster, which is unleashed during system initialization by IOINIT.
In the event of an IRQ or BRK, the operating system saves the registers and current
memory configuration on the stack and brings the system bank (ROM ’s, I/O,
RAMO) into context. The processor status at the time of interruption is then read
from the system stack to determine if the interrupt was an IRQ or a BRK. The C128
uses the following code to accomplish this:
TSX ;get stack pointer

LDA $105,X ;retrieve processor status
AND # $ 1 0 ;examine BRK flag
If the BRK flag is set, control passes to the ML Monitor through the BRK
indirect vector at $316, which usually points to the Monitor BREAK entry. Here
the program counter (PC), processor status, registers, memory configuration and
stack pointer are retrieved and displayed.
If the BRK flag is 0, an IRQ is assumed and control passes through the IRQ
indirect vector at $314, normally to the system IRQ handler. The following processes
are then performed in the order shown:
412 COMMODORE 128
1. IR Q ’s disabled.
2. Editor: split screen handler.
3. Editor: clear VIC raster IRQ.
4. IRQ ’s enabled.
5. Editor: keyboard scan.
6. Editor: VIC cursor blink.
7. Kernal: “ jiffie” clock.
8. Kernal: cassette switches.
9. Kernal: clear CIA-1 ICR.
10. BASIC: sprites, sounds, etc.
11. Return from interrupt.
As indicated in the preceding description, the C l 28 operating system uses the

IRQ heavily. In particular, the Editor split screen handler has rather strict require
ments that programmers must recognize and accommodate. An all-text screen or a
fully bit mapped screen presents no particular problem, but a split text and bit map
screen requires twice the IRQ frequency (two every sixtieth of a second). Thus, the
Editor (and consequently, IRQ-dependent applications) must distinguish between
the “ m ain” 60Hz IRQ and the “ middle” IRQ. Only during the main IRQ will all
the actions listed above be performed; the middle IRQ is only used by the Editor to
split the screen. The Editor IRQ routine sets the carry flag to designate a main IRQ.
Moreover, there is no margin in the timing requirements of a split screen. Program
mers should note the way the Editor uses the VIC during IRQ’s and avoid direct VIC
I/O if the Editor screen operations are enabled. There is a flag byte called G RA PH M ,
located at $D8, which can be set to $FF to disable the Editor’s use of the VIC.
The Editor is also responsible for scanning the C128 keyboard. Programmers
should note that SCNKEY has two indirect jum ps, K EY LO G and K E Y PU T, it
takes during its execution. The keyboard is controlled via CIA-1 PRA, PRB, VIC
register # 4 7 (extended key matrix), and the 8502 port (bit 6 = CAPS LOCK). The
SCNKEY routine is callable from the jump table.
The balance of the IRQ routines up to calling BASIC are self-explanatory.
The Kernal software clock is maintained by UDTIM, which is in the jum p table.
The IRQ processing makes one last call to BASIC_IRQ ($4006), but only if
INIT_STATUS bit 0 is set indicating BASIC is ready to handle IRQs (this was
discussed earlier in the RESET section). BASIC_IRQ is a heavy user of the VIC
and SID, and the same precautions should be taken regarding direct VIC and SID
I/O as with the Editor. BASIC_IRQ also utilizes a hold-off byte called IRQ_W RAP
_FLAG. It is normally used by the system to block IRQ-ed IRQ calls, but can be
set by the user with the effect of disabling the BASIC—IRQ handler. Alternatively
you could clear bit 0 of the INIT_STATUS byte as mentioned above and achieve
the same result.
The use of the IRQ indirect ($314) by an application usually requires more
care than most other wedges for several reasons. The likelihood of an IRQ occur
ring while the wedge is being installed is greater, there exists the possibility that the
user or some other software has already wedged the vector, and usually it is
desirable for the system IRQ functions to continue normally (e.g., keyscan) as
opposed to replacing them totally with our own (as we did with the NMI examples).
The following examples accomplish these objectives as well as masking out all but
the main IRQ. First we must install our IRQ handler. This example converts the
4 0 / 8 0 key into a SLOW/FAST key:
A 1302 BIT $D 011 ;test VIC reg 17

BPL $1324 ;branch if wrong IRQ
LDA $D505 preparation
ORA #$80 ;b7 of MMU MCR
STA $D505 ;set for input
LDA $D 011 ;assume FAST setup
AND #$6F ;blank VIC (RC8 = 0)
LDX #$01 ;2MHz
BIT $D505 ;test 4 0 / 8 0 key
BPL $ 13 IE ;branch if down (FAST)
ORA #$ 1 0 ;unblank VIC
DEX ;lM Hz
131E STX $D030 ;set speed
STA $D 011 ;set blank bit
1324 JMP ($1300) ;continue system IRQ
Now we need a small routine to actually wedge our code into the system IRQ. The
following code saves the current IRQ vector for our handler above to exit through
and substitutes a pointer to our code:
A 1400 SEI ;prevent interruptions

LDA $314 ;get current IRQ lsb
STA $1300 ;and save it
LDA $315 ;get current IRQ msb
STA $1301 ;and save it
LDA #$02 ;get our IRQ lsb
STA $314 ;and substitute it
LDA #$13 ;get our IRQ msb
STA $315 ;and substitute it
CLI ;re-enable IRQ processing
RTS
Enable the wedge by typing J 1400, that’s all there is to it. Depressing the
Locking 4 0 / 8 0 key now puts you in FAST mode; releasing it SLOWS you
down, and the keyscan, etc., continues to function. Note, however, that on this
414 COMMODORE 128
split screen our code throws off the timing, making for an unattractive display.
There are really only three things to watch out for when toying with the C l 28
system IRQ: First, be sure to keep the raster compare value on-screen to keep the
IRQ ’s happening (the best way is to keep RC8 zero as in example above); second,
never attempt to access the 8563 during an IRQ if there is any chance that it is in
use; finally, be sure the source of the IRQ is being cleared.
CBM STANDARD KERNAL CALLS
The following system calls make up the set of standard CBM system calls for the
Commodore 128 Personal Computer. Several of the calls, however, function somewhat
differently or may require slightly different setups than C64 mode. This was necessary
to accommodate specific features of the system, notably the 40/80 column windowing
Editor and banked memory facilities. As with all Kemal calls, the system configuration
(high ROM, RAM-0 and I/O) must be in context at the time of the call.
1. $FF81 CINT initialize screen editor and devices
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
,X used
.Y used
Memory: init Editor RAM
init Editor I/O
Flags: none
E X A M P LE :
SEI
JSR $FF81 initialize screen editor
CLI
C IN T is the Editor’s initialization routine. Both 40- and 80-column display

modes are prepared, editor indirect vectors installed, programmable key definitions
assigned, and the 4 0 / 8 0 key scanned for default display determination. CINT
sets the VIC bank and VIC nybble bank, enables the character ROM, resets SID
volume, places both 40- and 80-column screens and clears them. The only thing it
does not do that pertains to the Editor is I/O initialization, which is needed for IRQ’s
(keyscan, VIC cursor blink, split screen modes), key lines, screen background
colors, etc. (see IOINIT). Because CINT updates Editor indirect vectors that are
used during IRQ processing, you should disable IRQ ’s prior to calling it. CINT
utilizes the status byte INIT_STATUS ($A04) as follows:
$A04 bit 6 = 0 —> Full initialization.

(set INIT_STATUS bit 6)
1 —> Partial initialization.
(not key matrix pointers)
(not program key definitions)
CINT is also an Editor jum p table entry ($C000).
2. $FF84 IOINIT ;init I/O devices
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: initialize I/O
Flags: none
E X A M P LE :
SEI
JSR $FF84 initialize system I/O
CLI
IO IN IT is perhaps the major function of the Reset handler. It initializes both

C IA ’s (timers, keyboard, serial port, user port, cassette) and the 8502 port (key
board, cassette, VIC bank). It distinguishes a PAL system from an NTSC one and
sets PALCNT ($A03) if PAL. The VIC, SID and 8563 devices are initialized,
including the downloading of character definitions to 8563 display RAM (if neces
sary). The system 60Hz IRQ source, the VIC raster, is started (pending IRQs are
cleared). IOINIT utilizes the status byte INIT_STATUS ($A04) as follows:
$A04 bit 7 = 0 —>- Full initialization.
(set INIT—STATUS bit 7)
= 1 —» Partial initialization.
(not 8563 character definitions)
416 COMMODORE 128
You should be sure IRQ’s are disabled before calling IOINIT to avoid interrupts
while the various I/O devices are being initialized.
3. $FF87 RAMTAS ;init RAM and buffers
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: initializes RAM
Flags: none
E X A M P LE :
JSR $FF87 initialize system RAM
RAMTAS clears all page-zero RAM, allocates the cassette and RS-232
buffers, sets pointers to the top and bottom of system RAM (RAM 0) and points
the SY STEM .V ECTO R ($A00) to BASIC cold start ($4000). Finally, it sets a
flag, DEJAVU ($A02), to indicate to other routines that system RAM has been
initialized and that the SYSTEM—VECTOR is valid. It should be noted that the
C128 RAMTAS routine does not in any way test RAM.
4. $FF8A RESTOR ;init Kemal indirects
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: Kemal indirects restored
Flags: none
E X A M P LE :
SEI
JSR $FF8A ;restore Kernal indirects
CLI
RESTOR restores the default values of all the Kemal indirect vectors from
the Kemal ROM list. It does not affect any other vectors, such as those used by the
Editor (see CINT) and BASIC. Because it is possible for an interrupt (IRQ or NMI)
to occur during the updating of the interrupt indirect vectors, you should disable
interrupts prior to calling RESTOR. See also the VECTOR call.
5. $FF8D VECTOR ;init or copy indirects
PREPARATION:
Registers: .X = adr (low) of user list
.Y = adr (high) of user list
Memory: system map
Flags: .C = 0 —* load Kemal vectors
,C = 1 — * copy Kemal vectors
Calls: none
RESULTS:
Registers: .A used
.Y used
Memory: as per call
Flags: none
E X A M P LE :
LDX #save_lo
LDY #save_hi
SEC
JSR $FF87 ;copy indirects to “ save”
VECTOR reads or writes the Kernal RAM indirect vectors. Calling VEC
TOR with the carry status set stores the current contents of the indirect vectors to
the RAM address passed in the .X and .Y registers (to the current RAM bank).
Calling VECTOR with the carry status clear updates the Kernal indirect vectors
from the user list passed in the .X and .Y registers (from the current RAM bank).
Interrupts (IRQ and NMI) should be disabled when updating the indirects. See also
the RESTOR call.
418 COMMODORE 128
6. $FF90 SETMSG ;Kemal messages on/off
PREPARATION:
Registers: .A = message control
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: MSGFLG ($9D)updated
Flags: none
E X A M P LE :
LDA # 0
JSR $FF90 ;turn OFF allKernal messages
SETMSG updates the Kernal message flag byte MSGFLG ($9D) that deter
mines whether system error and/or control messages will be displayed.BASIC
normally disables error messages always and disables control messages in Run
mode. Note that the Kernal error messages are not the verbose ones printed by
BASIC, but simply the I/O ERROR # message that you see when in the Monitor,
for example. Examples of Kernal control messages are LOADING, FOUND, and
PRESS PLAY ON TAPE. The MSGFLG control bits are:
MSGFLG bit 7 = 1 —* enable CONTROL messages
bit 6 = 1 —> enable ERROR messages
7. $FF93 SECND iserial: send SA after LISTN
PREPARATION:
Registers: .A = SA (secondary address)
Memory: system map
Flags: none
Calls: LISTN
RESULTS:
Registers: .A used
Memory: STATUS ($90)
Flags: none
E X A M P LE :
LDA # 8
JSR $FF81 ;LISTN device 8
LDA # 1 5
JSR $FF93 ;pass it SA # 1 5
SECND is a low-level serial routine used to send a secondary address (SA) to

a LISTNing device (see LISTN Kemal call). An SA is usually used to provide setup
information to a device before the actual data I/O operation begins. Attention is
released after a call to SECND. SECND is not used to send an SA to a TALKing
device (see TKSA). (Most applications should use the higher level I/O routines: see
OPEN and CKOUT).
8. $FF96 TKSA ;serial: send SA after TALK
PREPARATION:
Registers: .A = SA (secondary address)
Memory: system map
Flags: none
Calls: TALK
RESULTS:
Registers: .A used
Flags: none
E X A M P LE :
LDA # 8
JSR $FFB4 ;TALK device 8
LDA # 1 5
JSR $FF93 ;pass it SA # 1 5
TKSA is a low-level serial routine used to send a secondary address (SA) to a

device commanded to TALK (see TALK Kemal call). An SA is usually used to
provide setup information to a device before the actual data I/O operation
begins. (Most applications should use the higher-level I/O routines; see OPEN and
CHKIN).
9. $FF99 MEMTOP ;set/read top of system RAM
PREPARATION:
Registers: .X = lsb of MEMSIZ
.Y = msb of MEMSIZ
Memory: system map
Flags: C = 0 —* set top of memory
.C = 1 —» read top of memory
Calls: none
420 COMMODORE 128
RESULTS:
Registers: .X = lsb of MEMSIZ
.Y = msb of MEMSIZ
Memory: MEMSIZ ($A07)
Flags: none
E X A M P LE :
SEC
JSR $FF99 ;get top of user RAMO
DEY
CLC
JSR $FF99 ;lower it 1 block
MEMTOP is used to read or set the top of system RAM, pointed to by

MEMSIZ ($A07). This call is included in the C l 28 for completeness, but neither the
Kernal nor BASIC utilizes MEMTOP since it has little meaning in the banked
memory environment of the C l 28 (even the RS-232 buffers are permanently
allocated). Nonetheless, set the carry status to load MEMSIZ into .X and .Y, and
clear it to update the pointer from .X and .Y. Note that MEMSIZ references only
system RAM (RAMO). The Kernal initially sets MEMSIZ to $FF00 (MMU and
Kernal RAM code start here).
10. $FF9C MEMBOT ;set/read bottom of system RAM
PREPARATION:
Registers: .X = lsb of MEMSTR
.Y = msb of MEMSTR
Memory: system map
Flags: .C = 0 —* set bot of memory
.C = 1 —» read bot of memory
Calls: none
RESULTS:
Registers: .X = lsb of MEMSTR
.Y = msb of MEMSTR
Memory: MEMSTR ($A05)
Flags: none
E X A M P LE :
SEC
JSR $FF9C ;get bottom of user RAMO
INY
CLC
JSR $FF9C ;raise it 1 block
MEMBOT is used to read or set the bottom of system RAM, pointed to by

MEMSTR ($A05). This call is included in the C128 for completeness, but neither
the Kemal nor BASIC utilizes MEMBOT since it has little meaning in the banked
memory environment of the C128. Nonetheless, set the carry status to load MEMSTR
into .X and .Y, and clear it to update the pointer from .X and .Y. Note that
MEMSTR references only system RAM (RAMO). The Kemal initially sets MEMSTR
to $1000 (BASIC text starts here).
11. $FF9F KEY ;scan keyboard
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: keyboard buffer
keyboard flags
Flags: none
E X A M P LE :
JSR $FF9F ;scan the keyboard
KEY is an Editor routine that scans the entire C l28 keyboard (except the
40/ 80 key). It distinguishes between ASCII keys, control keys, and program
mable keys, setting keyboard status bytes and managing the keyboard buffer. After
decoding the key, KEY will manage such features as toggling cases, pauses or
delays, and key repeats. It is normally called by the operating system during the
60Hz IRQ processing. Upon conclusion, KEY leaves the keyboard hardware driv
ing the key line on which the S T O P key is located.
There are two indirect RAM jum ps encountered during a keyscan: KEYVEC
($33A) and KEYCHK ($33C). KEYVEC (alias KEYLOG) is taken whenever a key
depression is discovered, before the key in .A has been decoded. KEYCHK is taken
after the key has been decoded, just before putting it into the key buffer. KEYCHK
carries the ASCII character in .A, the keycode in .Y, and the shiftkey status in .X.
The keyboard decode matrices are addressed via indirect RAM vectors as
well, located at DECODE ($33E). The following table describes them:
$33E Mode 1 normal keys
$340 Mode 2 SHIFT keys
$342 Mode 3 O keys
$344 Mode 4 C O N T R O L keys
$346 Mode 5 CAPS LOCK keys
$348 Mode 6 ALT keys
422 COMMODORE 128
The following list briefly describes some of the more vital variables utilized
or maintained by KEY:
ROWS $DC01 I/O port outputting keys

COLM $DC00 I/O port driving C64 keys
VIC # 4 7 $D02F I/O port driving new keys
8502 P6 $0001 I/O port sensing CAPS key
NDX $D0 -> keyboard buffer index
KEYD $34A -* keyboard buffer
XMAX $A20 — > keyboard buffer size
SHFLAG $D3 shift key status
RPTFLG $A22 repeat key enables
LOCKS $F7 pause, mode disables
KEY is also found in the Editor jump table as SCNKEY at $C012
12. $FFA2 SETTMO preserved)
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: TIMOUT ($A0E)
Flags: none
E X A M P LE :
LDA #value
JSR $FFA2 ;update TIMOUT byte
SETTMO is not used in the C l 28 but is included for compatibility and

completeness. It is used in the C64 by the IEEE communication cartridge to disable
I/O timeouts.
13. $FFA5 ACPTR ;serial:byte input
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: TALK
TKSA (if necessary)
RESULTS:
Registers: .A = data byte
Flags: none
E X A M P LE :
JSR $FFA5 ;input a byte from serial bus

STA data
A C PT R is a low-level serial I/O utility to accept a single byte from the

current serial bus TALKer using full handshaking. To prepare for this routine,
a device must first have been established as a TALKer (see TALK) and passed
a secondary address if necessary (see TKSA). The byte is returned in .A.
(Most applications should use the higher-level I/O routines; see BASIN and
GETIN).
14. $FFA8 CIOUT ;serial: byte output
PREPARATION:
Registers: .A = data byte
Memory: system map
Flags: none
Calls: LISTN
SECND (if necessary)
RESULTS:
Registers: .A used
Flags: none
E X A M P LE :
LDA data
JSR $FFA8 ;send a byte via serial bus
C IO U T is a low-level serial I/O utility to transmit a single byte to the current

serial bus LISTNer using full handshaking. To prepare for this routine, a device
must first have been established as a LISTNer (see LISTN) and passed a secondary
address if necessary (see SECND). The byte is passed in .A. Serial output data is
buffered by one character, with the last character being transmitted with EOI after a
call to UNLSN. (Most applications should use the higher level I/O routines; see
BSOUT.)
424 COMMODORE 128
15. $FFAB UNTLK ;serial: send untalk
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
Flags: none
E X A M P LE :
JSR $FFAB ;UNTALK serial device
U N TLK is a low-level Kemal serial bus routine that sends an UNTALK

command to all serial bus devices. It commands all TALKing devices to stop
sending data. (Most applications should use the higher-level I/O routines; see
CLRCH.)
16. $FFAE UNLSN ;serial: send unlisten
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
Flags: none
E X A M P LE :
JSR $FFAE ;UNLISTEN serial device
UNLSN is a low-level Kemal serial bus routine that sends an UNLISTEN

command to all serial bus devices. It commands all LISTENing devices to stop
reading data. (Most applications should use the higher-level I/O routines; see
ICLRCH.)
THE COMMODORE 128 OPERATING SYSTEM 42S
17. $FFB1 LISTN ;serial: send listen command
PREPARATION:
Registers: .A = device (0-31)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
Flags: none
E X A M P LE :
JSR $FFB1 ;command device to LISTEN
LISTN is a low-level Kernal serial bus routine that sends a LISTEN

command to the serial bus device in .A. It commands the device to start read
ing data. (M ost applications should use the higher-level I/O routines; see
ICKOUT.)
18. $FFB4 TALK ;serial: send talk
PREPARATION:
Registers: .A = device (0-31)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
Flags: none
E X A M P LE :
JSR $FFB4 ;command device to TALK
TA LK is a low-level Kemal serial bus routine that sends a TALK command

to the serial bus device in .A. It commands the device to start sending data. (Most
applications should use the higher-level I/O routines; see ICHKIN.)
426 COMMODORE 128
19. $FFB7 READSS ;read I/O status byte
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A = STATUS ($90 or $A14)
Memory: STATUS cleared if RS-232
Flags: none
E X A M P LE :
JSR $FFB7 ;STATUS for last I/O
READSS (alias READST) returns the status byte associated with the last I/O
operation (serial, cassette or RS-232) performed. Serial and cassette tape operations
update STATUS ($90) and RS-232 I/O updates RSSTAT ($A14). Note that to
simulate an ACIA, RSSTAT is cleared after it is read via READSS. The last I/O
operation is determined by the contents of FA ($BA); thus applications that drive
I/O devices using the lower-level Kernal calls should not use READSS.
20. $FFBA SETLFS ;set channel LA, FA, SA
PREPARATION:
Registers: .A = LA (logical # )
.X = FA (device # )
.Y = SA (secondary adr)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: LA, FA, SA updated
Flags: none
E X A M P LE :
See OPEN
SETLFS sets the logical file number (LA, $B8), device number (FA, $BA)
and secondary address (SA, $B9) for the higher-level Kernal I/O routines. The LA
must be unique among OPENed files and is used to identify specific files for I/O
operations. The device number range is 0 to 31 and is used to target I/O. The SA is
a command to be sent to the indicated device, usually to place it in a particular
mode. If the SA is not needed, the .Y register should pass $FF. SETLFS is often
used along with SETNAM and SETBNK calls prior to OPENs. See the Kemal
OPEN, LOAD and SAVE calls for examples.
21. $FFBD SETNAM ;set filename pointers
PREPARATION:
Registers: .A = string length
.X = string adr low
.Y = string adr high
Memory: system map
Flags: none
Calls: SETBNK
RESULTS:
Registers: none
Memory: FNLEN, FNADR updated
Flags: none
E X A M P LE :
See OPEN
SETNAM sets up the filename or command string for higher-level Kemal I/O
calls such as OPEN, LOAD and SAVE. The string (filename or command) length is
passed in .A and updates FNLEN ($B7). The address of the string is passed in .X
(low) and .Y (high). See the companion call, SETBNK, which specifies in which
RAM bank the string is found. If there is no string, SETNAM should still be called
and a null ($00) length specified (the address does not matter). SETNAM is often
used along with SETBNK and SETLFS calls prior to OPENs. See the Kernal
22. $FFC0 OPEN ;open logical file
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: SETLFS, SETNAM, SETBNK
428 COMMODORE 128
RESULTS:
Registers: .A = error code (if any)
.X used
.Y used
Memory: setup for I/O
STATUS, RSSTAT updated
Flags: .C — 1 —> error
E X A M P LE : OPEN 1,8,15,“ 10”

LDA #length fnlen
LDX # < filen am e fnadr (command)
LDY # > filen am e
JSR $FFBD SETNAM
LDX # 0 fnbank (RamO)

JSR $FF68 SETBNK
LDA #1 la
LDX #8 fa
LDY #15 sa
JSR $FFBA SETLFS
JSR $FFC0 OPEN
BCS error
filename .BYTE ‘10’
length = 2
OPEN prepares a logical file for I/O operations. It creates a unique entry in
the Kemal logical file tables LAT ($362), FAT ($36C) and SAT ($376) using its
index LDTND ($98) and data supplied by the user via SETLFS. There can be up to
ten logical files OPENed simultaneously. OPEN performs device-specific opening
tasks for serial, cassette and RS-232 devices, including clearing the previous status and
transmitting any given filename or command string supplied by the user via SETNAM
and SETBNK. The I/O status is updated appropriately and can be read via READSS.
The path to OPEN is through an indirect RAM vector at $31 A. Applications
may therefore provide their own OPEN procedures or supplement the system’s by
redirecting this vector to their own routine.
23. $FFC3 CLOSE ;close logical file
PREPARATION:
Registers: .A = LA (logical # )
Memory: system map
Flags: .C (see text below)
Calls: none
RESULTS:
.X used
.Y used
Memory: logical tables updated
Flags: .C = 1 —* error
E X A M P LE :
LDA #1 :1a
JSR $FFC3 ;CLOSE
BCS error ;(tape files only)
CLOSE removes the logical file (LA) passed in .A from the logical file tables
and performs device-specific closing tasks. Keyboard, screen and any unOPENed
files pass through. Cassette files opened for output are closed by writing the last
buffer and (optionally) an EOT mark. RS-232 I/O devices are reset, losing any
buffered data. Serial files are closed by transmitting a CLOSE command (if an SA
was given when it was opened), sending any buffered character, and UNLSTNing
the bus.
There is a special provision incorporated into the CLOSE routine of systems
featuring a BASIC DOS command. If the following conditions are all true, a full
CLOSE is not performed; the table entry is removed but a CLOSE command is not
transmitted to the device. This allows the disk command channel to be properly
OPENed and CLOSEd without the disk operating system closing all files on its
end:
.C — 1 —> indicates special CLOSE

FA > = 8 —>device is a disk
SA = 15 —* command channel
The path to CLOSE is through an indirect RAM vector at $31C. Applications

may therefore provide their own CLOSE procedures or supplement the system’s by
24. $FFC6 CHKIN ;set input channel
PREPARATION:
Registers: .X = LA (logical # )
Memory: system map
Flags: none
Calls: OPEN
430 COMMODORE 128
RESULTS:
.X used
.Y used
Memory: LA, FA, SA, DFLTN
Flags: .C = 1 —» error
E X A M P LE :
LDX # 1 ;la
JSR $FFC6 ;CHKIN
BCS error
CHKIN establishes an input channel to the device associated with the logical
address (LA) passed in .X, in preparation for a call to BASIN or GETIN. The
Kemal variable DFLTN ($99) is updated to indicate the current input device and the
variables LA, FA and SA are updated with the file’s parameters from its entry in
the logical file tables (put there by OPEN). CHKIN performs certain device specific
tasks: screen and keyboard channels pass through, cassette files are confirmed for
input, and serial channels are sent a TALK command and the SA transmitted (if
necessary). Call CLRCH to restore normal I/O channels.
CHKIN is required for all input except the keyboard. If keyboard input is
desired and no other input channel is established, you do not need to call CHKIN or
OPEN. The keyboard is the default input device for BASIN and GETIN.
The path to CHKIN is through an indirect RAM vector at $ 3 IE. Applications
may therefore provide their own CHKIN procedures or supplement the system ’s by
25. $FFC9 CKOUT ;set output channel
PREPARATION:
Registers: .X = LA (logical # )
Memory: system map
Flags: none
Calls: OPEN
RESULTS:
.X used
.Y used
Memory: LA, FA, SA, DFLTO
Flags: .C = 1 —* error
E X A M P LE :
LDX # 1 ;la
JSR $FFC9 ;CKOUT
BCS error
CKOUT establishes an output channel to the device associated with the

logical address (LA) passed in .X, in preparation for a call to BSOUT. The Kernal
variable DFLTO ($9A) is updated to indicate the current output device and the
variables LA, FA and SA are updated with the file’s parameters from its entry in
the logical file tables (put there by OPEN). CKOUT performs certain device
specific tasks: keyboard channels are illegal, screen channels pass through, cassette
files are confirmed for output, and serial channels are sent a LISTN command and
the SA transmitted (if necessary). Call CLRCH to restore normal I/O channels.
CKOUT is required for all output except the screen. If screen output is
desired and no other output channel is established, you do not need to call CKOUT
or OPEN. The screen is the default output device for BSOUT.
The path to CKOUT is through an indirect RAM vector at $320. Applications
may therefore provide their own CKOUT procedures or supplement the system’s by
26. $FFCC CLRCH ;restore default channels
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
Memory: DFLTI, DFLTO updated
Flags: none
E X A M P LE :
JSR $FFCC ;restore default I/O
CLRCH (alias CLRCHN) is used to clear all open channels and restore the
system default I/O channels after other channels have been established via CHKIN
and/or CHKOUT. The keyboard is the default input device and the screen is the
default output device. If the input channel was to a serial device, CLRCH first
UNTLKs it. If the output channel was to a serial device, it is UNLSNed first.
432 COMMODORE 128
The path to CLRCH is through an indirect RAM vector at $322. Applications

may therefore provide their own CLRCH procedures or supplement the system’s by
27. $FFCF BASIN ;input from channel
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: CHKIN (if necessary)
RESULTS:
Registers: .A = character (or error code)
Memory: STATUS, RSSTAT updated
Flags: .C = 1 if error
EX A M P LE :
LDY # 0 ;index
MORE JSR $FFCF ;input a character
STA data,Y ;buffer it
INY
CMP #$0D ;carriage return?
BNE MORE
BASIN (alias CHRIN) reads a character from the current input device (DFLTN
$99) and returns it in .A. Input from devices other than the keyboard (the default
input device) must be OPENed and CHKINed. The character is read from the input
buffer associated with the current input channel:
a. Cassette data is returned a character at a time from the cassette buffer at $B00,
with additional tape blocks being read when necessary.
b. RS-232 data is returned a character at a time from the RS-232 input buffer at $C00,
waiting until a character is received if necessary. If RSSTAT ($A14) is bad from
a prior operation, input is skipped and null input (carriage return) is substituted.
c. Serial data is returned a character at a time directly from the serial bus, waiting
until a character is sent if necessary. If STATUS ($90) is bad from a prior
operation, input is skipped and null input (carriage return) is substituted.
d. Screen data is read from screen RAM starting at the current cursor position and
ending with a pseudo carriage return at the end of the logical screen line. The
way the BASIN routine is w ritten, the end of line (EOL) is not recog
nized. Users must therefore count characters themselves or otherwise detect
when the logical EOL has been reached.
e. Keyboard data is input by turning on the cursor, reading characters from the
keyboard buffer, and echoing them on the screen until a carriage return is
encountered. Characters are then returned one at a time from the screen until
all characters input have been passed, including the carriage return. Any calls
after the EOL will start the process over again.
The path to BASIN is through an indirect RAM vector at $324. Applications

may therefore provide their own BASIN procedures or supplement the system’s by
28. $FFD2 BSOUT ;output to channel
PREPARATION:
Registers: .A = character
Memory: system map
Flags: none
Calls: CKOUT (if necessary)
RESULTS:
E X A M P LE :
LDA #character
JSR $FFD2 ;output a character
BSOUT (alias CHROUT) writes the character in .A to the current output

device (DFLTO $9A). Output to devices other than the screen (the default output
device) must be OPENed and CKOUTed. The character is written to the output
buffer associated with the current output channel:
a. Cassette data is put a character at a time into the cassette buffer at $B00, with
tape blocks being written when necessary.
b. RS-232 data is put a character at a time into the RS-232 output buffer at $D00,
waiting until there is room if necessary.
c. Serial data is passed to CIOUT, which buffers one character and sends the
previous character.
d. Screen data is put into screen RAM at the current cursor position.
e. Keyboard output is illegal.
The path to BSOUT is through an indirect RAM vector at $326. Applications

may therefore provide their own BSOUT procedures or supplement the system’s by
434 COMMODORE 128
29. $FFD5 LOAD ;load from file
PREPARATION:
Registers: .A = 0 —> LOAD
.A > 0 —> VERIFY
.X = load adr lo (if SA = 0)
.Y = load adr hi (if SA = 0)
Memory: system map
Flags: none
RESULTS:
.X = ending adr lo
.Y = ending adr hi
Memory: per command
STATUS updated
E X A M P LE : L O A D “ p ro g ram ” ,8 ,1
LDA #length ;fnlen

LDX # < filen am e ;fnadr
LDY # > filen am e
JSR $FFBD SETNAM
LDA # 0 load/verify bank (RAM 0)

LDX # 0 fnbank (RAM 0)
JSR $FF68 SETBNK
LDA # 0 la (not used)

LDX # 8 fa
LDY # $ F F sa (SA X ) normal load)
JSR $FFBA SETLFS
LDA # 0 load, not verify

LDX # < lo a d adr (used only if SA = 0)
LDY # > lo a d adr (used only if SA = 0)
JSR $FFD5 LOAD
BCS error
STX end lo
STY end hi
filename .BYTE “ program”
length =7
This routine LOADs data from an input device into C l28 memory. It can also
be used to VERIFY that data in memory matches that in a file. LOAD performs
THE COMMODORE 128 OPERATING SYSTEM 43S
device-specific tasks for serial and cassette LOADs. You cannot LOAD from
RS-232 devices, the screen or the keyboard. While LOAD performs all the tasks of
an OPEN, it does not create any logical files as an OPEN does. Also note that LOAD
cannot “ wrap” memory banks. As with any I/O, the I/O status is updated appropriately
and can be read via READSS. LOAD has two options that the user must select:
a. LOAD vs. VERIFY: The contents of .A passed at the call to LOAD deter
mines which mode is in effect. If .A is zero, a LOAD operation will be
performed and memory will be overwritten. If .A is nonzero, a VERIFY
operation will be performed and the result passed via the error mechanism.
b. LOAD ADDRESS: the secondary address (SA) setup by the call to SETLFS
determines where the LOAD starting address comes from. If the SA is zero,
the user wants the address in .X and .Y at the time of the call to be used. If the
SA is nonzero, the LOAD starting address is read from the file header itself
and the file is loaded into the same place from which it was SAVEd.
The C128 serial LOAD routine automatically attempts to BURST load a file,
and resorts to the normal load mechanism (but still using the FAST serial routines)
if the BURST handshake is not returned.
The path to LOAD is through an indirect RAM vector at $330. Applications
may therefore provide their own LOAD procedures or supplement the system
procedures by redirecting this vector to their own routine.
30. $FFD8 SAVE ;save to file
PREPARATION:
Registers: .A = pointer to start adr
.X = end adr lo
.Y = end adr hi
Memory: system map
Flags: none
RESULTS:
.X = used
.Y = used
Memory: STATUS updated
E X A M P LE : SAVE “ program” ,8
LDA #length ;fnlen
LDX # < filen am e ;fnadr
LDY # > filen am e
JSR $FFBD ;SETNAM
436 COMMODORE 128
LDA # 0 ;save from bank (RAM 0)

LDX # 0 ;fnbank (RAM 0)
JSR $FF68 ;SETBNK
LDA # 0 ;la (not used)
LDX # 8 ;fa
LDY # 0 ;sa (cassette only)
JSR $FFBA ;SETLFS
LDA # start ;pointer to start address
LDX end ;ending address lo
LDY e n d + 1 ;ending adr hi
JSR $FFD8 ;SAVE
BCS error
filename .BYTE “ program’
length = 7
start .WORD address 1 ;page-0
end .WORD address2
This routine SAVEs data from C128 memory to an output device. SAVE per
forms device-specific tasks for serial and cassette SAVEs. You cannot SAVE from
RS-232 devices, the screen or the keyboard. While SAVE performs all the tasks of an
OPEN, it does not create any logical files as an OPEN does. The starting address of
the area to be SAVEd must be placed in a zero-page vector and the address of this
vector passed to SAVE in .A at the time of the call. The address of the last byte
to be SAVEd PLUS ONE is passed in .X and .Y at the same time. Cassette SAVEs
utilize the secondary address (SA) to specify the type of tape header(s) to be generated:
SA (bit 0) = 0 —* relocatable (blf)file
= 1 —> absolute (plf) file
SA (bit 1) = 0 —» normal end
= 1 —* write EOT header at end
There is no BURST save; the normal FAST serial routines are used. As with
any I/O, the I/O status will be updated appropriately and can be read via READSS.
The path to SAVE is through an indirect RAM vector at $332. Applications
may therefore provide their own SAVE procedures or supplement the system’s by
31. $FFDB SETTIM ;set internal clock
PREPARATION:
Registers: .A = low byte
.X = middle byte
.Y = high byte
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: TIME ($A0) updated
Flags: none
E X A M P LE :
LDA # 0 ;reset clock
TAX
TAY
JSR $FFDB ;SETTIM
SETTIM sets the system software (jiffic) clock, which counts sixtieths (1/60)
of a second. The timer is incremented during system IRQ processing (see UDTIM),
and reset at the 24-hour point. SETTIM disables IRQ’s, updates the three-byte timer
with the contents of .A, .X and .Y, and re-enables IRQ ’s.
32. $FFDE RDTIM ;read internal clock
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A = low byte
.X = middle byte
,Y = high byte
Memory: none
Flags: none
E X A M P LE :
JSR $FFDE ;RDTIM
RDTIM reads the system software (jiffie) clock, which counts sixtieths (1/60)
of a second. The timer is incremented during system IRQ processing (see UDTIM).
and reset at the 24-hour point. RDTIM disables IRQ ’s, loads .A, .X and . Y with the
contents of the 3-byte timer, and re-enables IRQ’s.
33. $FFE1 STOP ;scan STOP key
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
438 COMMODORE 128
RESULTS:
Registers: .A = last keyboard row
.X = used (if STOP key)
Memory: none
Flags: status valid
EXAM PLE:
JSR $FFE1 ;scan STOP key

BEQ stop ;branch if down
STOP checks a Kemal variable STKEY ($91), which is updated by UDTIM

during normal IRQ processing and contains the last scan of keyboard column C7.
The STOP key is bit 7, which will be 0 if the key is down. If it is, default I/O
channels are restored via CLRCH and the keyboard queue is flushed by resetting
NDX ($D0). The keys on keyboard line C7 are:
Bit: 7 6 5 4 3 2 1 0
Key: STOP Q O SPACE 2 CTRL 1
The path to STOP is through an indirect RAM vector at $328. Applications
may therefore provide their own STOP procedures or supplement the system ’s by
34. $FFE4 GETIN ;read buffered data
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: CHKIN (if necessary)
RESULTS:
Registers: .A = character (or error code)
.X used
.Y used
EXAM PLE:
wait JSR $FFE4 ;get any key

BEQ wait
STA character
GETIN reads a character from the current input device (DFLTN ($99)) buffer
and returns it in .A. Input from devices other than the keyboard (the default input
device) must be OPENed and CHKINed. The character is read from the input buffer
associated with the current input channel:
a. Keyboard input: A character is removed from the keyboard buffer and passed
in .A. If the buffer is empty, a null ($00) is returned.
b. RS-232 input: A character is removed from the RS-232 input buffer at $C00
and passed in .A. If the buffer is empty, a null ($00) is returned (use READSS
to check validity).
c. Serial input: GETIN automatically jumps to BASIN. See BASIN serial I/O.
d. Cassette input: GETIN automatically jumps to BASIN. See BASIN cassette
I/O.
e. Screen input: GETIN automatically jumps to BASIN. See BASIN serial I/O.
The path to GETIN is through an indirect RAM vector at $32A. Applications

may therefore provide their own GETIN procedures or supplement the system’s by
35. $FFE7 CLALL ;close all files and channels
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
ULTS:
Registers: .A used
.X used
Memory: LDTND, DFLTN, DFLTO updated
Flags: none
EXAMPLE:
JSR $FFE7 ;close files
C LA LL deletes all logical file table entries by resetting the table index,
LDTND ($98). It clears current serial channels (if any) and restores the default I/O
channels via CLRCH.
The path to CLALL is through an indirect RAM vector at $32C. Applications
may therefore provide their own CLALL procedures or supplement the system’s by
36. $FFEA UDTIM increm ent internal clock
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
440 COMMODORE 128
RESULTS:
Registers: .A used
.X used
Memory: TIME, TIMER, STKEY updated
Flags: none
EX A M P LE :
SEI
JSR $FFEA ;UDTIM
CLI
UDTIM increments the system software (jiffie) clock, which counts sixtieths
(1/60) of a second when called by the system 60Hz IRQ. TIME, a 3-byte counter
located at $A0, is reset at the 24-hour point. UDTIM also decrements TIMER, also
a 3-byte counter, located at $A1D (BASIC uses this for the SLEEP command, for
example). You should be sure IRQ ’s are disabled before calling UDTIM to prevent
system calls to UDTIM while you are modifying TIME and TIMER.
UDTIM also scans key line C7, on which the STOP key lies, and stores the
result in STKEY ($91). The Kemal routine STOP utilizes this variable.
37. $FFED SCRORG ;get current screen window size
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A = screen width
.X = window width
.Y = window height
Memory: none
Flags: none
E X A M P LE :
JSR $FFED ;SCRORG
SCRORG is an Editor routine that has been slightly changed from previous
CBM systems. Instead of returning the maximum SCREEN dimensions in .X and
.Y, the C l 28 SCRORG returns the current WINDOW dimensions. It does return
the maximum SCREEN width in .A. These changes make it possible for applica
tions to “ fit” themselves on the current screen window. SCRORG is also an Editor
jum p table entry ($C00F).
38. $FFF0 PLOT ;read/set cursor position
PREPARATION:
Registers: .X = cursor line
.Y = cursor column
Memory: system map
Flags: ,C = 0 —» set cursor position
.C = 1 —> get cursor position
Calls: none
RESULTS:
Registers: .X = cursor line
.Y = cursor column
Memory: TBLX, PNTR updated
Flags: .C = 1 —> error
E X A M P LE :
SEC
JSR $FFF0 ;read current position
INX ;move down one line
INY ;move right one space
CLC
JSR $FFF0 ;set cursor position
BCS error ;new position outside window
P L O T is an Editor routine that has been slightly changed from previous CBM
systems. Instead of using absolute coordinates when referencing the cursor position,
PLOT now uses relative coordinates, based upon the current screen window. The
following local Editor variables are useful:
SCBOT $E4 -> window bottom

SCTOP $E5 —> window top
SCLF $E6 window left side
SCRT $E7 window right side
TBLX $EB cursor line
PNTR SEC -> cursor column
LINES $ED maximum screen height
COLUMNS SEE ->• maximum screen width
When called with the carry status set, PLOT returns the current cursor
position relative to the current window origin (not screen origin). When called with
the carry status clear, PLOT attempts to move the cursor to the indicated line and
column relative to the current window origin (not screen origin). PLOT will return a
clear carry status if the cursor was moved, and a set carry status if the requested
position was outside the current window (no change has been made).
442 COMMODORE 128
39. $FFF3 IOBASE ;read base address of I/O block
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .X = lsb of I/O block
.Y = msb of I/O block
Memory: none
Flags: none
E X A M P LE :
JSR $FFF3 ;find the I/O block
IO BA SE is not used in the C l28 but is included for compatibility and complete
ness. It returns the address of the I/O block in .X and .Y.
NEW C l28 KERNAL CALLS

The following system calls are a set of extensions to the standard CBM jum p table.
They are specifically for the C l 28 and as such should not be considered as permanent
additions to the standard jum p table. With the exception of C64 MODE, they are all true
subroutines and will terminate via an RTS. As with all Kemal calls, the system
configuration (high ROM, RAM-0 and I/O) must be in context at the time of the
call.
1. $FF47 SPIN SPOUT ;setup fast serial ports for I/O
PREPARATION:
Registers: none
Memory: system map
Flags: .C = 0 -> select SPINP
,C = 1 —» select SPOUT
Calls: none
RESULTS:
Registers: .A used
Memory: CIA-1, MMU
Flags: none
E X A M P LE :
CLC
JSR $FF47 ;setup for fast serial input
The C 128/1571 fast serial protocol utilizes CIA 1 (6526 at $DC00) and a
special driver circuit controlled in part by the MMU (at $D500). SPINP and
SPOUT are routines used by the system to set up the CIA and fast serial driver
circuit for input or output. SPINP sets up CRA (CIA 1 register 14) and clears the
FSDIR bit (MMU register 5) for input. SPOUT sets up CRA, ICR (CIA 1 register
13), timer A (CIA 1 registers 4 and 5), and sets the FSDIR bit for output. Note the
state of the TODIN bit of CRA is always preserved, but the state of the GAME,
EXROM and SENSE40 outputs of the MMU are not (reading these ports return the
state of the port and not the register values— consequently they cannot be pre
served). These routines are required only by applications driving the fast serial bus
themselves from the lowest level.
2. $FF4A CLOSE ALL ;close all files on a device
PREPARATION:
Registers: .A —» device # (FA: 0-31)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: none
Flags: none
E X A M P LE :
LDA #$08
JSR $FF4A ;close all files on device
The FA T is searched for the given FA. A proper CLOSE is performed for all
matches. If one of the CLOSEd channels is the current I/O channel, then the default
channel is restored.
This call is utilized, for example, by the BASIC command DCLOSE. It is
also called by the Kemal BOOT routine.
444 COMMODORE 128
3. $FF4D C64MODE ;reconfigure system as a C64
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: none
Flags: none
E X A M P LE :
JMP $FF4D ;switch to C64 mode
There is no return from this routine. The 8502 DDR and port are initialized,
and the VIC is set to 1MHz (slow) mode. Control is passed to code in common
(shared) RAM, which sets the MMU mode register (# 5 ) to C64 mode. From this
point on, the MMU and C128 ROMs are not accessible. The routine exits via an
indirect jum p through the C64 RESET vector.
Since C64 operation does not allow for MMU access, all MMU registers must
be configured for proper operation before the C64 mode bit is set. Similarly,
because the start-up of the C64 operating system is not from a true hardware reset,
there is the possibility that unusual I/O states in effect prior to C64MODE calls can
cause unpredictable and presumably undesirable situations once in C64 mode.
There is no way to switch from C64 mode back to C l28 mode; only a
hardware reset or power off/on will restore the C l28 mode of operation. A reset
will always initiate C l 28 mode, although altering the SYSTEM vector beforehand
is one way to automatically “ throw” a system back to C64 mode.
4. $FF50 DMA CALL ;send command to DMA device
PREPARATION:
Registers: .X = bank (0-15)
.Y = DMA controller command
Memory: DMA registers set up system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
M em o ry : changed as per command
Flags: none
EX A M P LE :
LDA # $ 0 0 ;setup C l28 base address

STA $DF02 ;low
LDA # $ 2 0
STA $DF03 ;high
LDA # $ 0 0 ;setup expansion RAM address

STA $DF04 ;low
STA $DF05 ;high
STA $DF06 ;bank (0-n, where n = 3 if 256K)
LDA # $ 4 0 ;setup number of bytes

STA $DF07 ;low
LDA # $ 1 F
STA $DF08 ;high
LDX # $ 0 0 ;C 128 bank

LDY # $ 8 4 ;DMA command to “ STASH”
JSR $FF50 ;execute DMA command
DMA CALL is designed to communicate with an external expansion car

tridge capable of DMA and mapped into system memory at 102 ($DFxx). The
DMA CALL converts the logical C l 28 bank parameter to MMU configuration via
GETCFG, O R ’s in the I/O enable bit, and transfers control to RAM code at $3F0.
Here the C128 bank specified is brought into context, and the user’s command is
issued to the DMA controller. The actual DMA transfer is performed at this point,
with the 8502 kept off the bus in a wait state. As soon as the DMA controller
releases the processor, memory is reconfigured to the state it was in at the time of
the call and control is returned to the caller. The user must analyze the completion
status by reading the DMA status register at $DF00.
Care should be taken in the utilization of the C l28 RAM expansion product
by any application using the built-in Kernal interface. This includes especially the
use of the C128 BASIC commands FETCH, STASH and SWAP. In the routine that
prepares a DMA request for the user, the Kernal forces the I/O block to be always
in context. Consequently, data from the DMA device is likely to corrupt sensitive
I/O devices. Users should either bypass the Kernal DMA routine by providing
their own interface, or limit the DMA data transfers to the areas above and below the
I/O block. Only strict observance of the latter will guarantee proper utilization of
the BASIC commands. The following code, used instead of the DMA CALL in the
above example, illustrates a work-around:
LDX # $ 0 0 ;C 128 bank

LDY # $ 8 4 ;DMA command to ‘STASH’
JSR $FF6B ;GETCFG
TAX
JSR $3F0 ;execute DMA command
446 COMMODORE 128
5. $FF53 BOOT CALL ;boot load program from disk
PREPARATION:
Registers: .A = drive number (ASCII)
.X = device number (0-31)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: changed as per command
Flags: .C 1 if I/O error
E X A M P LE :
LDA # $ 3 0 ;drive 0
LDX #$08 ;device 8
JSR $FF53 ;BOOT
BCS 10 ERROR
BCC NO BOOT SECTOR
BOOT attempts to load and execute the boot sector from an auto-boot disk in
the given drive and device. The BOOT protocol is as follows:
a. Close all open files on boot device.

b. Read track 1 sector 0 into TBUFFR ($B00).
c. Check for auto-boot header, RTS if not.
d. If (b lk # > 0), BLOCK READ sequential sectors into RAM at given (adrl,
adrh, bank) location.
e. If LEN(filename) > 0, LOAD file into RAM-0 (normal load).
f. JSR to user code at location C above.
On any error, the BOOT operation is aborted and the UI command is issued
to the disk. A return may or may not be made to the caller depending upon the
completion status and the BOOTed code. The BOOT sector has the following
layout:
$00 $01 $02 $03 $04 $05 $06 A B C

C B M adrl adrh bank b lk # title 0 file 0 code
where: A= $07 + LEN(title)

B= A + LEN(filename)
C= B + 1
The following examples illustrate the flexibility of this layout. This loads and runs a
BASIC program:
$00 CBM :key

$03 — » $00,$00,$00.$00 :no other BOOT sector
$07 -> NAM E,$00 :message “ NAM E”
$0C $00 :no filename
$0D $A2,$13,$A0,$0B :code
$4C,$A5,$AF
$14 -> RUN“ PROGRAM” :data (BASIC stmt)
$20 $00
This results in the message Booting N A M E... being displayed and, utilizing a
C128 BASIC jum p table entry that finds and executes a BASIC statement, loads
and runs the BASIC program named “ PROGRAM .” The same header can be used
to load and execute a binary (machine code) program by simply changing RUN to
BOOT. (While the file auto-load feature of the boot header could be used to load
binary files simply by furnishing a filename, to execute it you must know the
starting address and JMP to it. BASIC’s BOOT command does that, and allows a
more generic mechanism.) In the next example, a menu is displayed and you are
asked to select the operating mode. Nothing else is loaded in this “ configure’’-type
header:
$00 - h* CBM :key

$03 —» $00,$00,$00,$00 :no other BOOT sector
$07 —> $00 :no message
$0C —» $00 :no filename
$0D -» $20,$7D ,$FF,$0D , $53, $45,$4C, $45
$43, $54, $20,$4D, $4F, $44, $45,$3A
$0D,$0D, $20, $31,$2E , $20, $43, $36
$34, $20, $20, $42, $41, $53, $49, $43
$ 0 D ,$20, $ 3 2 ,$ 2 E ,$20, $43,$31, $32
$38, $20, $42, $41, $53, $49, $43,$0D
$20, $33, $2E, $20, $43, $31, $32, $38
$20,$4D, $4F, $4E, $49, $54, $4F. $52
$0D ,$0D , $00, $20, $E4, $FF,$C9, $31
$D0, $03,$4C,$4D , $FF,$C9, $32,$D0
$03,$4C, $03, $40,$C9, $33,$D0, $E3
$4C, $00,$B0
The loading of sequential sectors is designed primarily for specialized applications

(such as CP/M or games) that do not need a disk directory entry.
448 COMMODORE 128
6. $FF56 PHOENIX ;init function cartridges
PREPARATION:
Registers: none
Memory: cartridge map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: changed as per command
Flags: none
E X A M P LE :
JSR $FF56 ;PHOENIX
The C l 28 Kernal initialization routine POLL creates a Physical Address

Table (PAT) containing the ID ’s of all installed function ROM cartridges. PHOENIX
calls each logged cartridge’s cold-start entry in the order: external low/high, and
internal low/high. After calling the cartridges (if any), PHOENIX calls the Kemal
BOOT routine to look for an auto-boot disk in drive 0 of device 8 (see BOOT CALL).
Control may or may not be returned to the user. PHOENIX is called by BASIC at
the conclusion of its cold initialization.
7. $FF59 LKUPLA ;search tables for given LA
$FF5C LKUPSA ;search tables for given SA
PREPARATION:
Registers: .A = LA (logical file number)
if LKUPLA
.Y = SA (secondary address)
if LKUPSA
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A = LA (only if found)
.X = FA (only if found)
.Y = SA (only if found)
M em o ry : none
Flags: .C = 0 if found
.C = 1 if not found
E X A M P LE :
LDY #$60 ;find an available SA

AGAIN INY
CPY #$6F
BCS TOO MANY ;too many files open
JSR $FF5C ;LKUPSA
BCC AGAIN ;get another if in use
LKUPLA and LKUPSA are Kernal routines used primarily by BASIC DOS
commands to work around a user’s open disk channels. The Kernal requires unique
logical device numbers (LA ’s), and the disk requires unique secondary addresses
(SA’s); therefore BASIC must find alternative unused values whenever it needs to
establish a disk channel.
9. $FF5F SWAPPER ;switch between 40 and 80 columns
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: local variables swapped
Flags: none
E X A M P LE :
LDA $D7 check display mode

BMI is_80 branch if 80-column
JSR $FF5F switch from 40 to 80
SWAPPER is an Editor utility used to switch between the 40-column VIC

(composite) video display and the 80-column 8563 (RGBI) video display. The
routine simply swaps local (associated with a particular screen) variables, TAB
tables and line wrap maps with those describing the other screen. The MSB of
MODE, location $D7, is toggled by SWAPPER to indicate the current display
mode: $ 8 0 = 80-column, $ 0 0 = 40-column.
450 COMMODORE 128
10. $FF62 DLCHR ;init 80-col character RAM
PREPARATION:
Registers: none
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
Memory: 8563 character RAM initialized
Flags: none
E X A M P LE :
JSR $FF62 initialize 8563 char, defns.
DLCHR (alias INIT80) is an Editor utility to copy the VIC character

definitions from ROM ($D000-$DFFF, bank 14) to 8563 display RAM ($2000-
$3FFF, local to 8563-not in processor address space). The 8 by 8 VIC character
cells are padded with nulls ($00) to fill out the 8 by 16 8563 character cells.
Refer to Chapter 10, Programming the 80-Column (8563) Chip for details concerning
the 8563 font layout.
11. $FF65 PFKEY ;program a function key
PREPARATION:
Registers: .A = pointer to string adr
(lo/hi/bank)
.Y = string length
.X = key number (1-10)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A used
.X used
.Y used
M em o ry : PKYBUF, PKYDEF tables updated
F lags: .C = 0 if successful
.C = 1 if no room available
EX A M P LE :
LDA #$F A pointer to string address

LDY #$06 length
LDX #$0A key # (HELP key)
JSR $FF65 install new key def’n
BCS NO ROOM
>000FA 00 13 00 :ptr to $1300 bank 0

>01300 53 54 52 49 4E 47 :“ string”
PFKEY (alias KEYSET) is an Editor utility to replace a C l 28 function key

string with a user’s string. Keys 1-8 are F1-F8, 9 is the S H I F T RUN string,
and 10 is the HELP string. The example above replaces the “ help” RETURN
string assigned at system initialization to the HELP key with the string “ string.”
Both the key length table, PKYBUF ($1000-$1009), and the definition area,
PKYDEF ($100A-$10FF) are compressed and updated. The maximum length of all
ten strings is 246 characters. No change is made if there is insufficient room for a
new definition.
12. $FF68 SETBNK ;set bank for I/O operations
PREPARATION:
Registers: .A = BA, memory bank (0-15)
.X = FNBANK, filename bank
Memory: system map
Flags: none
Calls: SETNAM
RESULTS:
Registers: none
Memory: BA, FNBANK updated
Flags: none
EX A M P LE :
See OPEN
SETBNK is a prerequisite for any memory I/O operations, and must be used
along with SETLFS and SETNAM prior to OPENing files, etc. BA ($C6) sets the
current 64KB memory bank for LOAD/SAVE/VERIFY operations. FNBANK ($C7)
indicates the bank in which the filename string is found. The Kernal routine
GETCFG is used to translate the given logical bank numbers (0-15). SETBNK is
often used along with SETNAM and SETLFS calls prior to O PEN’s. See the Kernal
452 COMMODORE 128
13. $FF6B GETCFG ;lookup MMU data for given bank
PREPARATION:
Registers: .X = logical bank # (0-15)
Memory: system map
Flags: none
Calls: none
RESULTS:
Registers: .A = MMU configuration data
Memory: none
Flags: none
E X A M P LE :
LDX # $ 0 0 ;logical bank 0 (RAM O)

JSR $FF6B ;GETCFG
STA $FF01 ;setup MMU pre-config #1
G E T C F G allows a universal, logical approach to physical bank numbers by

providing a simple lookup conversion for obtaining the actual MMU configuration
data. In all cases where a bank number 0-15 is required, you can expect GETCFG
to be called to convert that number accordingly. There is no error checking; if the
given logical bank number is out of range the result is invalid. Refer to the Memory
Management Unit in the Commodore 128 section later in this chapter for details concern
ing memory configuration. The C128 Kemal memory banks are assigned as follows:
o. %oo l i n n RAM 0 only

1. %01111111 RAM 1 only
2. %10111111 RAM 2 only
3. % 1 1111111 RAM 3 only
4. %00010110 INT ROM, RAM 0, I/O
5. %01010110 INT ROM, RAM 1, I/O
6. %10010110 INT ROM, RAM 2, I/O
7. % 1 1010110 INT ROM, RAM 3, I/O
8. %00101010 EXT ROM, RAM 0, I/O
9. % 0 1 101010 EXT ROM, RAM 1, I/O
10. %10101010 EXT ROM, RAM 2, I/O
11. % 1 1101010 EXT ROM, RAM 3, I/O
12. %00000110 KERNAL, INT LO, RAM 0, I/O
13. %00001010 KERNAL, EXT LO, RAM 0, I/O
14. %00000001 KERNAL, BASIC, RAM 0, CHAR ROM
15. %00000000 KERNAL, BASIC, RAM 0, I/O
14. $FF6E JSRFAR ;gosub in another bank
15. $FF71 JMPFAR ;goto another bank
PREPARATION:
Registers: none
Memory: system map, also:
$02 bank (0-15)
$03 PC high
$04 -> PC low
$05 .S (status)
$06 .A
$07 .X
$08 .Y
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: as per call, also:
$05 —> .S (status)
$06 -> .A
$07 .X
$08 .Y
Flags: none
The two routines, JSRFAR and JMPFAR, enable code executing in the
system bank of memory to call (or JMP to) a routine in any other bank. In the case
of JSRFAR, a return will be made to the caller’s bank. It should be noted that
JSRFAR calls JMPFAR, which calls GETCFG. When calling a non-system bank,
the user should take necessary precautions to ensure that interrupts (IRQ’s and NM I’s)
will be handled properly (or disabled beforehand). Both JSRFAR and JMPFAR are
RAM-based routines located in common (shared) RAM at $2CD and $2E3
respectively.
The following code illustrates how to call a subroutine in the second RAM
bank from the system bank. Note that we need not worry about IRQ ’s and N M I’s in
this case because the system will handle them properly in any configuration that has
the Kernal ROM or any valid RAM bank in context at the top page of memory.
STY $08 ;assumes registers and status

STX $07 ;already setup for call
STA $06
PHP
PLA
STA $05
454 COMMODORE 128
LDA #1 ;want to call $2000 in bank 1

LDY #$20
LDX #$00
STA $02
STY $03
STX $04
JSR $FF6E ;JSRFAR

LDA $05 ;restore status and registers
PHA
LDA $06
LDX $07
LDY $08
PLP
16. $FF74 INDFET ;LDA (fetvec),Y from any bank
PREPARATION:
Registers: .A = pointer to address
.X = bank (0—15)
.Y = index
Memory: setup indirect vector
Flags: none
Calls: none
RESULTS:
Registers: .A = data
.X used
Memory: none
Flags: status valid
E X A M P LE :
LDA #$00 ;setup to read $2000

STA $FA
LDA #$20
STA $FB
LDA #$ F A
LDX #$01 ;in bank 1
LDY #$00
JSR $FF74 ;LDA ($FA,RAM 1),Y
BEQ etc
INDFET enables applications to read data from any other bank. It sets up
FETVEC ($2AA), calls GETCFG to convert the bank number, and JMPs to code in
common (shared) RAM at $2A2 which switches banks, loads the data, restores the
user’s bank, and returns. When calling a non-system bank, the user should take
necessary precautions to ensure that interrupts (IRQ’s and N M I’s) will be handled
properly (or disabled beforehand).
17. $FF77 INDSTA ;STA (stavec),Y to any bank
PREPARATION:
.X = bank (0-15)
.Y = index
setup STAVEC ($2B9) pointer
Flags: none
Calls: none
RESULTS:
Registers: .X used
Memory: changed per call
Flags: status invalid
E X A M P LE :
LDA #$00 ;setup write to $2000

STA $FA
LDA #$20
STA $FB
LDA #$ F A
STA $2B9
LDA data
LDX #$01 ;in bank 1
LDY #$00
JSR $FF77 ;STA ($FA,RAM 1),Y
INDSTA enables applications to write data to any other bank. After you set
up STAVEC ($2B9), it calls GETCFG to convert the bank number and JMPs to
code in common (shared) RAM at $2AF which switches banks, stores the data,
restores your bank, and returns. When calling a nonsystem bank, the user should
take necessary precautions to ensure that interrupts (IRQ's and N M I’s) will be
handled properly (or disabled beforehand).
456 COMMODORE 128
18. $FF7A INDCMP ;CMP (cmpvec),Y to any bank
PREPARATION:
.X = bank (0-15)
.Y = index
setup CMPVEC ($2C8) pointer
Flags: none
Calls: none
RESULTS:
Registers: .X used
Memory: none
Flags: status valid
EX A M P LE :
LDA #$00 ;setup to verify $2000

STA $FA
LDA #$20
STA $FB
LDA #$F A
STA $2C3
LDA data
LDX #$01 ;in bank 1
LDY #$00
JSR $FF7A ;CMP ($FA,RAM 1),Y
BEQ same
CMPSTA enables applications to compare data to any other bank. After you
set up CMPVEC ($2C8), it calls GETCFG to convert the bank number and JM P’s to
code in common (shared) RAM at $2BE which switches banks, compares the data,
restores your bank, and returns. When calling a nonsystem bank, the user should
take necessary precautions to ensure that interrupts (IRQ’s and N M I’s) will be
handled properly (or disabled beforehand).
19. $FF7D PRIMM ;print immediate utility
PREPARATION:
Registers: none
Memory: none
Flags: none
Calls: none
RESULTS:
Registers: none
Memory: none
Flags: none
E X A M P LE :
JSR $FF7D ;display following text
.BYTE “ m essage”
.BYTE $00 ;terminator
JMP continue ;execution resumes here
PRIMM is a Kernal utility used to print (to the default output device) an
ASCII string which immediately follows the call. The string must be no longer than
255 characters and is terminated by a null ($00) character. It cannot contain any
embedded null characters. Because PRIMM uses the system stack to find the string
and a return address, you must not JMP to PRIMM. There must be a valid address
on the stack.
C l28 DEVICE NUMBERS

The following are the device numbers for the Commodore 128:
0 —* Keyboard
1 —» Cassette
2 -> RS-232
3 —» Screen (current)
4 —» Serial bus device:
4—7 usually printers
8-30 usually disks
Device number 31 should not be used. While it is specified to be a valid serial bus
address, when it is ORed with certain serial commands it results in a bad command,
hanging the bus and the serial drivers.
458 COMMODORE 128
MEMORY MANAGEMENT IN
THE COMMODORE 128
COMMODORE 128 MODE

In Commodore 128 mode, all memory management organization depends on the cur
rently selected memory configuration. In C128 BASIC and the Machine Language
M onitor, the memory is organized into sixteen default memory configurations. Different
portions of memory are present depending on the memory configuration. Figure 13-3
lists the default memory configurations of the C l28 on system power-up.
BANK CONFIGURATION
0 RAM(O) only
1 R A M (l) only
4 Internal R ()M ,R A M (0),I/0
5 Internal R 0 M ,R A M (l),I/0
6 Internal R 0 M ,R A M (2 ),I/0 (same as 4)
7 Internal R 0 M ,R A M (3 ),I/0 (same as 5)
8 External RC)M ,RAM (0),I/0
9 External R 0 M ,R A M (l),I/0
10 External R 0 M ,R A M (2 ),I/0 (same as 8 )
11 External R 0 M ,R A M (3 ),I/0 (same as 9)
12 Kernal and Internal ROM (LOW ), R A M (0),I/0
13 Kernal and External ROM (LOW), R A M (0),I/0
14 Kernal and BASIC ROM, RAM (0), Character ROM
Figure 13-3. M em o ry C onfiguration (B ank) T ab le
THE MEMORY MANAGEMENT

U N IT (MMU)
In Commodore 128 mode, all memory management is controlled through a series of

I/O registers called the Memory Management Unit (MMU). The Memory Manage
ment Unit consists of memory locations that reside between $D500 and $D50B and
$FF00 and $FF04. The configuration register appears twice, once at $D500 and again at
$FF00. This is done in case I/O is switched out of the $D000-$DFFF memory range,
which makes the MMU registers from $D500-$D50B inaccessible.
When this occurs, some features of the MMU are no longer available to the
programmer. So before switching out I/O in the $D000-$DFFF range, make sure you
have made all the manipulations you need on the MMU (particularly, preselecting the
preconfiguration register values for the load configuration registers). See Figure 13-4 for
a graphic depiction of the way the MMU registers map into memory.
$ F F 0 4 ----- LCRD - LOAD C O N F I G U R A T I O N REG D

$ F F 0 3 ----- LCRC - LORD C O N F I G U R A T I O N REG C
$ F F 0 2 ------ LCRB - LOAD C O N F I G U R A T I O N REG B
$ F F 0 1 ------ LCR A - LOAD C O N F I G U R A T I O N REG A
$ F F 0 0 ------ CR / - CONFIGURATION REGISTER
$ D 5 0 B ---- VR -VERSION REGISTER

$ D 5 Q H ---- P1H - PAGE 1 POINTER HIGH
$ 0 5 0 9 ---- P1L - P A G E 1 POINTER LOW
$ D 5 0 8 ---- P0H - P A G E 0 POINTER HIGH
$ 0 5 0 7 ---- POL - P A G E 0 P O I N T E R LON
$ D 5 0 6 ---- RCR -RAM C0NFIG0RATI0N REGISTER
$ 0 5 0 5 ---- MCR -MODE CONFIGURATION REGISTER
$ D 5 0 4 ---- PCRD -PRECONFIGURATION REGISTER 0
$ 0 5 0 3 ---- PCRC -PRECONFIGURATION REGISTER C
$ 0 5 0 2 ---- PCRB -PRECONFIGURATION REGISTER B
$ D 5 0 1 ---- PCRH -PRECONFIGURATION REGISTER A
$ 0 5 0 0 ---- CR ^-CONFIGURATION REGISTER
Figure 13-4. M M U Register M ap
H O W TO SW ITCH BANKS
In BASIC, type the BANK command to switch from one bank to another as follows:
BANK n
where n is a digit between 0 and 15. When your application program needs to access a
bank in which the microprocessor is not in context (for SYS, PEEK, POKE, and WAIT
commands) use the BANK command to reach layers of the computer’s memory that are
not accessible in the current bank. For instance, suppose you want to access the VIC
chip when you are doing graphics in the first part of your program. The next segment
460 COMMODORE 128
of your program needs to access the character ROM, which is only visible to the 8502 in
bank 14. Change to bank 14; then read the character ROM data, which makes up the
images of the characters in the character sets. While the microprocessor is “ looking” in
bank 14, the VIC chip is not available to the microprocessor, so you must issue another
BANK command in order to return to (a configuration containing I/O and) processing
VIC video information.
In machine language, switching banks is a little more difficult. You must change
the value of the registers [in particular the Configuration Register (CR), Load Config
uration Register (LCR) and Preconfiguration Register (PCR)], either directly or
indirectly. The Kernal routine GETCFG allows you to change configurations and
maintains the same ones used by BASIC and the Monitor as they appear in Figure 13-3.
There are four PCR’s and four LCR’s (as shown in Figure 13-4). Each PCR corresponds
directly to a LCR. PCR A pertains to LCR A, PCR B corresponds to LCR B, and so on.
To change the value of the Configuration Register directly, perform a write
operation (STA, STX, STY) to the Configuration Register.
To change the value of the Configuration Register indirectly, write a value
specifying a memory configuration to the Preconfiguration Register. When a subsequent
store instruction is performed to the corresponding Load Configuration Register, the
value previously stored in the corresponding PCR is loaded into the CR. When a store
instruction is executed on an LCR, the value in the corresponding PCR is loaded into the
CR and the memory management organization conforms to the values associated with
that memory management scheme. The value written to the LCR is of no consequence;
any value triggers the preconfiguration mechanism.
NOTE: Basic expects the LCR’s and PCR’s to be left alone. If you use
them to manage memory in your application, do not plan on using BASIC.
THE C O N FIG U R A TIO N REGISTER

The Configuration Register (CR) is the most important register in the MMU. It specifies
and organizes the ROM, RAM and input/output configurations for the entire Commo
dore 128 memory in C l 28 mode. (The MMU is not present at all in the C64 memory
map.) The CR is located at address $D500 when I/O is available and at address $FF00 at
all times. When I/O functions are disabled, the MMU memory disappears between
$D500 and $D50B. The MMU memory is always present between $FF00 and $FF04.
Each of the eight bits in the CR controls a separate memory function.
Bit 0 (zero) in the Configuration Register specifies whether I/O Registers are
available, or whether ROM (High) is present in the memory range $D000 through
$DFFF. The I/O Registers consist of the registers of the VIC chip, SID chip, MMU
(from $D500 through $D50B); CIA number 1, which controls the joystick port; and CIA
number 2, which controls the serial bus and user port. If bit 0 is high (equal to 1), ROM
or RAM is present in the range $D000 through $DFFF, depending upon the values of
the ROM HIGH bits (4 and 5) in this register. If bit 0 is low (equal to 0), I/O is present
in this range. The value of bit 0 on power-up is 0.
When I/O is switched out of (not present in) this range, the registers in the MMU
disappear from the memory map in the range $D500 through $D50B. The memory manage
ment is then controlled through the MMU registers at locations $FF00 through $FF04. Any
write operation to an LCR ($FF01-$FF04) loads the corresponding PCR value (currently
invisible to the 8502) into the CR. Reading an LCR returns the value stored in the
currently inaccessible PCR. The MMU registers ranging from $FF00 through $FF04 are
always present in the Commodore 128 memory, regardless of the memory configuration.
Bit 1 in the CR specifies how the microprocessor accesses the address range $4000
through $7FFF, called ROM LOW memory. If bit 1 is high, the microprocessor
accesses RAM in this range. If bit 1 is low (equal to 0), the microprocessor maps in the
BASIC LOW ROM in that range. Upon power-up or reset, this bit is set low, so
BASIC is available to the user as soon as the computer is turned on.
Bits 2 and 3 determine the type of memory that resides in the midrange of memory,
the address range $8000 through $BFFF. If both bits 2 and 3 are set high, RAM is
placed in this range. If bit 2 is high and 3 is low, INTERNAL FUNCTION ROM is
placed here. If bit 2 is low and 3 is high, EXTERNAL FUNCTION ROM appears. If
bits 2 and 3 are low, the BASIC HIGH ROM is placed here. Upon power-up or reset,
the MMU sets both bits 2 and 3 low, so BASIC is available to the user immediately.
Bits 4 and 5 work similarly to bits 2 and 3 and specify memory in the range $C000
through $FFFF, referred to as HIGH memory. If bits 4 and 5 are set high, RAM is
placed in this range. If bit 4 is high and 5 is low, INTERNAL FUNCTION ROM is
placed. If bit 4 is low and 5 is high, EXTERNAL FUNCTION ROM appears. If bits 4
and 5 are low, the Kernal and character ROMs are placed here. Upon power-up or reset,
bits 4 and 5 are set low, so the Kemal and character ROM are available to the user at once.
Note that bit 0 in the Configuration Register, the bit that switches in and out I/O in
address range $D000 through $DFFF, overrides the memory organization for bits 4 and 5.
If bit 0 is set high (1), the I/O Registers are not in place in the address range $D000
through $DFFF. Either the character ROM, internal or external function ROM or RAM
is located in this range, depending on the value of bits 4 and 5 of the configuration
register. If bit 0 in the Configuration Register is set low (0), the Input/Output registers
are present between $D000 through $DFFF, regardless of the value of bits 4 and 5 of the
configuration register. This means no matter what memory configuration is chosen
between $C000 and $FFFF with bits 4 and 5, if bit 0 is set low (0), the character ROM
(or whatever was originally in this address range) is overlaid by the I/O registers and
becomes unavailable to the microprocessor. This is why the character ROM and the I/O
registers are never available at the same time.
Finally, the last two bits of the MMU, 6 and 7, determine the RAM BANK
selection. For the base system of 128K, only bit 6 is significant; bit 7 is not imple
mented. When bit 6 is high (1), RAM bank 1 is selected. When bit 6 is low (0), RAM
bank 0 is selected.
The 128K of RAM is organized into two 64K RAM banks. The microprocessor
only addresses 64K at a time, but since the two 64K RAM banks can be switched in and
out so quickly, the computer acts as though it addresses 128K at the same time. When
one RAM bank is being addressed by the microprocessor, the other bank stores
information to be processed once it is banked in. Portions of both banks can be shared in
memory at the same time. This is called Common RAM, and is discussed in the section
462 COMMODORE 128
The RAM Configuration Register. RAM bank 0 is used typically for BASIC text area,
while RAM bank 1 is used for BASIC arrays and variable storage.
As indicated above, the MMU has a feature that allows a portion of RAM to be
common for the two RAM banks. The RAM Configuration Register controls the amount
of common RAM. This is discussed in detail later in this section under the description of
the RAM Configuration Register.
Figure 13-5 is a diagram of the Configuration Register, showing how the bits
control each memory organization.
7 6 5 4 3 2 1 0
BANK SEL ECT HI SPACE MID-SPACE LO -SPACE I/O SPACE

($C000-$FFFF) ($8000-$BFFF) ($4000- ($D000-
$7FFF) $DFFF)
00 = RAM BANK 0 00 = KERNAL ROM 00 = BASIC ROM HI 0 = BASIC 0 = 1/0

01 = RAM BANK 1 01 = INT. FUNCTION ROM 01 = INT. FUNCTION ROM RO M LO REGISTERS
10 = EXPANSION BANK 2 10 = EXT. FUNCTION ROM 10 = EX T. FUNCTION ROM 1 = RAM 1 = RAM/ROM
11= EXPANSION BANK 3 11 = RAM 11= RAM
Figure 13-5. C o nfiguration Register
PRECONFIGURING COMMODORE 128 MEMORY

As noted, the Configuration Register in the MMU is the means by which the Commo
dore 128 organizes the memory layout. Commodore 128 memory management has a
mechanism that allows the programmer to set up four predefined memory configura
tions besides the one already operating. The four Preconfiguration Registers (PCR)
and the four Load Configuration Registers (LCR) are designed to provide this feature.
W hen a different memory management structure is desired, a previously defined
configuration in one of the Preconfiguration Registers can be chosen easily and
instantly.
This mechanism allows the programmer to preset several memory configurations
and, with a single store instruction to an LCR, reorganize the entire memory layout
instantly. This enables the programmer to use several memory organizations inter
changeably; thus, if one part of your program requires one memory configuration and
another part requires a different configuration, you can switch back and forth between
the two, each with a single store instruction, once you predefine the alternate memory
setup in one of the Preconfiguration Registers.
To alter memory management through this preconfiguration mechanism, you
must change the value of the Preconfiguration Register. The four PCR’s and four
LCR ’s of the MMU are shown in the MMU Register Map in Figure 13-6. Each
PCR corresponds directly to an LCR (i.e., PCR A pertains to LCR A, PCR B
corresponds to LCR B, and so on). The format for the Load Configuration' and
Preconfiguration Registers is the same as for the Configuration Register. The four LCRs
and four PCRs are all initialized to zero.
$ F F 0 4 ----- LCRD
$ F F 0 3 ----- LCRC
$ F F 0 2 ----- LCRB
$ F F Q 1 ----- LCRR LORD C O N F I G U R A T I O N REG R
$ F F 0 0 ----- CR CONFIGURATION REGISTER
$ D 5 0 B ----- VR
$ D 5 0 A ----- P1H
$ 0 5 0 9 ----- P1L
$ 0 5 0 8 ----- P0H
$ D 5 0 7 ----- POL
$ D 5 0 6 ----- RCR
$ 0 5 0 5 ----- MCR
$ D 5 0 U ----- PCRD
$ 0 5 0 3 ----- PCRC
$ 0 5 0 2 ----- PCRB
$ 0 5 0 1 ----- PCRR PREC0NFIGURATION REGISTER A
$ 0 5 0 0 ----- CR CONFIGURATION REGISTER
Figure 13-6. M M U Register Map
To directly change the value of the Configuration Register (therefore bypassing

the preconfiguration mechanism), perform a write operation (STA, STX, STY) directly
to the Configuration Register at either $D500 or $FF00.
To indirectly change the value of the Configuration Register (therefore utilizing
the preconfiguration mechanism), perform a write operation (STA, STX, STY) to the
Preconfiguration Register. This loads a value into the PCR. When a subsequent store
instruction is performed to the corresponding Load Configuration Register, the value
that was previously stored in the corresponding PCR is loaded into the CR. The store
instruction acts as a triggering mechanism that passes the contents of a PCR into the CR.
W hen a store instruction is executed upon an LCR, the value in the corresponding PCR
is loaded into the CR and memory management organization conforms to the value
associated with that PCR.
For example, to alter the memory management organization specified on power-up,
use the following machine language segment:
464 COMMODORE 128
PART 1 LDA # $ 0 E Load accumulator with 14

This selects
I/O ($D000-$DFFF)
RAM ($4000-$7FFF,$8000-$BFFF)
Kernal and Character ROM ($C000-$FFFF)
RAM Bank 0
STA $D501 Store in PCR A— no immediate results
Place interim part of the program here

PART 2 STA $FF01 Write to LCR A, selects above configuration
In this program segment, PART 1 initializes PCR A ($D501) with the value 14
($0E), which performs no immediate result. When PART 2 is encountered, the STA
instruction performs a write operation to location $FF01, which triggers the preconfiguration
mechanism and loads the value from PCR A into the Configuration Register. The store
instruction value is not significant; it must operate only on the address, and any store
instruction works. Once this instruction is executed, the memory management organiza
tion is immediately changed according to the value in the appropriate PCR, in this case
PCR A ($D501).
The value 14 ($0E) loaded into PCR A selects the following memory organization:
TYPE ADDRESS RANGE BIT(S) AFFECTED IN CR
I/O $D000-$DFFF 0
RAM $4000-$7FFF 1
RAM $8000-$BFFF 2,3
Kernal, Char $C000-$FFFF 4,5
RAM BANK 0 6,7
14 = $0E = 0000 1110 (binary)
It is good practice to initialize the PCR’s in the beginning of your program, as in

PART 1 of the above program segment. Then place the interim portion of your program
in memory. PART 2 marks the place in your program where you are actually going
to alter the memory management setup. For the utmost speed and processor efficiency,
use absolute addressing for the MMU registers, as in the above example.
As previously noted, when the Input/Output Registers are switched out, the registers
of the MMU that appear in memory range $D500-$D50B are unavailable to the micro
processor. When this occurs, a write operation to an LCR still loads the corresponding PCR
value (currently invisible to the 8502) into the CR, even though the PCR’s are not acces
sible to the microprocessor when I/O is switched out. If you switch out I/O, make sure to
set up the PCR’s first, which are present when bit 0 of the Configuration Register is equal
to 0. However, the Load Configuration Registers are available at all times since they
appear in the address range $FF00-$FF04. If I/O is switched out, the PCR’s must be
set prior to switching out I/O in order for them to be of any service. Note that BASIC uses
the PCRs; if you alter them while BASIC is resident, you may obtain unpredictable results.
The MMU registers control the memory organization for RAM, BASIC, Kemal
and character ROM, internal and external function ROM, and I/O. The MMU has
additional registers that determine the mode (C l28, C64 or CP/M) in which the
Commodore 128 operates, the common RAM configurations, and the location of pages
0 and 1. The registers in the MMU that control these operations are the Mode
Configuration Register (MCR), the RAM Configuration Register (RCR), and the
Page Pointers respectively. The following sections explain how these additional regis
ters of the Memory Management Unit operate.
THE MODE C O N FIG U R ATIO N REGISTER

The Mode Configuration Register (MCR) specifies which microprocessor is .currently in
operation (8502, Z80A) and which operating system mode is currently invoked ( C l28 or
C64). The MCR is located at address $D505. As in the other registers in the MMU,
each bit in the Mode Configuration Register controls a separate and independent
operation.
Bit 0 determines which microprocessor is in control of the Commodore 128. Bit 0 is
powered-up low so the Z80 microprocessor initiates control of the computer. The Z80
performs a small start-up procedure, then bit 0 is set to a 1 and the 8502 takes over
if no CP/M system disk is present in the disk drive.
When the Commodore 128 is first powered-up or reset and the disk drive detects
the CP/M operating system diskette in the drive, the Z80A microprocessor BOOTs the
CP/M operating system from disk. The value of bit 0 in the Mode Configuration
Register in this case is 0. When the Z80A takes control of the Commodore 128, all
memory references from $0000 through $0FFF are translated to $D000 through $DFFF,
where the CP/M BIOS exists in ROM. For memory accesses in the range $0000 through
$0FFF in the Z80 BIOS, the memory status lines MS0 and MSI are brought low to
reflect ROM; otherwise they are high. Note that C64 mode and Z80A mode is an
undefined configuration.
Bits 1 and 2 are not used. They are reserved for future expansion. IF0, bit 3 sets
an input for FAST serial. It is not used as an input port at all.
Bit 3 is the fast serial (FSDIR) disk drive control bit. It acts like a bit in a
bidirectional 6529 port, which means it acts differently depending upon whether the bit
is used for input or output operation. As an output signal, bit 3 controls the direction of
the data in the disk drive data buffer. The MMU pin FSDIR reflects the status of bit 3,
which is reset to zero upon power-up. If bit 3 is equal to 1, an output operation occurs;
selecting a data direction for the data in the serial bus buffer. If zero, bit 3 sets
an input for FAST serial. It is not used as an input port at all.
Bits 4 and 5 are the /GAME and /EXROM sense bits respectively. These cartridge
control lines initiate Commodore 64 mode and act as the /GAME and /EXROM
hardware lines as in the Commodore 64. When these control lines detect a cartridge in
the Commodore 128 expansion port, C64 mode is instantly enabled, the computer acts
as a Commodore 64 and takes its instructions from the software built into the cartridge.
Upon power-up, /GAME and /EXROM are pulled as inputs. If either one is low, C64
466 COMMODORE 128
mode is selected. Thus, a C l28 cartridge should not pull these lines low on power-up.
In C128 mode, these lines are active as I/O lines (latched) to the expansion port. These
can be used as input or output lines, but make sure they are not brought low upon
power-up in C l 28 mode.
Bit 6 selects the operating system that takes over the Commodore 128. Upon
power-up or reset, this bit is cleared (0) to enable all of the MMU registers and
Commodore 128 mode features. Setting this bit high (1) initiates C64 mode.
Bit 7, a read-only bit, indicates whether the 40/80 D I S P L A Y key is in
the up (40-column) or down (80-column) position. The value of bit 7 is high (1) if
the 4 0 / 8 0 key is in the up position. The value of bit 7 is low (0) if the 4 0 / 8 0
D I S P L A Y key is in the down position.
This is useful in certain application programs that utilize both the 40- and
80-column displays. For instance, check the value of this bit to see if the user is viewing
the VIC screen. If so, carry on with the program; otherwise display a message telling the
user to switch from the 80-column screen to the VIC (40-column) screen in order to
display a VIC bit map. This may be invalid if the user typed < E S C > X to switch
screens. See the SCORG Kernal routine.
See Figure 13-7 for a summary of Mode Configuration Register activities.
$D505 7 6 5 4 3 2 1 0
Mode Configuration Register
VALUE
BIT FUNCTION DESCRIPTION HIGH LOW
0 /Select microprocessor 8502 Z80A (inverted)

1 Unused — —
2 Unused — —
3 Fast serial DD control Fast serial out Fast serial in

4 /GAME Access game cartridge C64 mode set
(C64 mode only) on power up
5 /EXROM Access external C64 mode set
software cartridge (C64 mode on power up
only)
6 Select operating system C64 mode C128 mode,
(MMU disappears assert MMU
from memory map) registers
7 Read 4 0 / 8 0 key 40/80 column 40/80 column
position key is UP key is DOWN
Figure 13-7. M ode Configuration Register S u m m a ry

THE RAM C O N FIG U R A TIO N REGISTER

The RAM Configuration Register (RCR) in the MMU specifies the amount of Common
RAM shared between the two 64K RAM banks, how the RAM is shared, and which
bank is delegated for the VIC chip. The value of the bits in the RCR determine how
each RAM bank is allocated for specific purposes. The RAM Configuration Register is
located within the I/O block at address $D506.
Bits 0 and 1 determine the amount of shared RAM between banks. If both bits 0
and 1 equal 0, IK of RAM becomes common. If bit 0 is equal to 1 (high) and bit 1 is
low (0), then 4K of common RAM is shared between banks. If bit 0 is low (0) and bit 1
is high (1), then 8K is common, and if both bits 2 and 3 are high (1), then 16K of RAM
becomes common. (These bits have no effect in Commodore 64 mode). The reset values
of these bits are both 0. See Figure 13-8 to understand how the two RAM banks share
common RAM.
468 COMMODORE 128
BANKO BANK 1 BANKO BANK 1
Hi Common RAM Hi Common RAM

$F000 $F000 •
$0FFF
S03FF Low Common RAM

Low Common RAM
1KCOMMON RAM 4KCOMMON RAM
BANKO BANK 1 BANKO BANK 1
Hi Common
RAM Hi
$E000 - Common
RAM
$C000 -
$0FFF -
Low
$03FF -
RAM
Low Common
RAM
8KC O M M O N RAM 16K COMMON RAM
Figure 13 -8 . C o m m o n RAM
Bits 2 and 3 specify which portions of the two RAM banks are common, if at all. If
both bits are low (0), no RAM sharing occurs. If bit 2 is set high (1) and bit 3 is low (0),
a section of the bottom of RAM bank 0 replaces the corresponding section of RAM bank
1 for all RAM address accesses. If bit 3 is set high (1) and bit 2 is set low (0), a section
of the top RAM bank 0 replaces the corresponding section of RAM bank 1 for all RAM
address accesses. If both bits 2 and 3 equal 1 (high), RAM is common to both the top
and bottom of the RAM banks. Upon power-up or reset, bits 2 and 3 are set to 0, and no
RAM is shared between banks. From a hardware standpoint, the 128K MMU selects the
common RAM by forcing the CASO enable line low and CAS 1 enable line high for all
common memory accesses.
Bits 4 and 5 have no assigned function. They are reserved for future expansion.
Bits 6 and 7 in the RCR operate as a RAM bank pointer to tell the VIC chip which
portion of RAM to use. At the present time, bit 7 is ignored. It too is reserved for
future RAM expansion. When bit 6 is low (0) (driving CASO low), the VIC chip is told
to look in RAM bank 0. When bit 6 is set high (1) (driving CAS1 low), the VIC chip is
steered into RAM bank 1. Either setup allows the VIC chip RAM bank to be selected
from the microprocessor RAM bank independently.
When the microprocessor speed is increased to 2MHz, the VIC chip is disabled
and the 80-column (8563) chip takes over the video processing. The VIC chip is
affected by holding the AEC hardware line high. The disabling of the VIC chip is not
directly affected by the actions of the MMU.
Figure 13-9 summarizes the RAM Configuration Register activities.
$D506 7 6 5 4 3 2 1 0
RAM Configuration Register
BITS FUNCTIONAL DESCRIPTION
1 0 - Determine amount of shared RAM between RAM banks
0 0 = IK common RAM
0 1 = 4K common RAM
1 0 = 8K common RAM
1 1 = 16K common RAM
3 2 - Determine which portions o f RAM are shared
0 0 = No common RAM
0 1 = Bottom o f RAM bank 0 is common
1 0 = Top o f RAM bank 0 is common
1 1 = Both top and bottom o f RAM bank 0 are common
5 4 - Not implemented; reserved for future expansion
7 6 - Selects RAM bank for the VIC chip (Bit 7 is ignored)
X 0 = Select Bank 0 for VIC chip
X 1 = Select Bank 1 for VIC chip
Note: X = not implemented
Figure 13-9. RA M Configuration Register S u m m ary

470 COMMODORE 128
THE PAGE POINTERS

The Commodore 128 has a feature that allows you to relocate page 0 ($00-$FF) and
page 1 ($100-$1FF) of memory. Certain applications may require you to keep page 0
intact while running BASIC, switch out the BASIC ROM, resume processing within the
control of your machine language program, and then switch 0 page and the BASIC
ROM back in. Instead of transferring 0 page with the machine language monitor, the
page pointers make it easy to relocate pages 0 and 1. Four registers within the MMU,
called the page pointers, allow you to do this. The page pointers are located within the
I/O block in the range $D507 through $D50A. The page pointers follow the standard
8502 low-byte/high-byte format. Here are the actual addresses of the high and low bytes
of the page pointers:
$D507 Page 0 Pointer (low)

$D508 Page 0 Pointer (high)
$D509 Page 1 Pointer (low)
$D50A Page 1 Pointer (high)
Bit 0 of the high byte page pointers corresponds to the RAM bank number
for any page 0 address access. Bit 0 controls the generation of the CASO hardware
control line if it is low and the CAS1 line if it is high, and bits 1 and 3 are
ignored.
To relocate page 0, perform a write operation on the high-byte 0 page pointer.
This is stored in the high-byte page pointer location and has no direct result until a write
operation is performed on the low-byte 0 page pointer. When this occurs, bits 0 through
7 of the low-byte page pointer correspond to the Translated Address lines TA8 through
TA15 for any 0 page address reference, which relocates the 0 page. Any subsequent 0
page address is relayed to the new page zero.
The page 1 pointer works the same way. Both pairs of pointers are initialized to 0
upon power-up, placing pages 0 and 1 in actual page 0 and page 1 locations.
It is important to note that memory addresses 0 and 1 are always available at those
absolute address references, regardless of whether pages zero or one are relocated.
The page zero low-byte pointer directly replaces the high-order address of zero
page (normally 00). When pages zero and one are directed to locations in RAM memory
above page one, the MMU translates the addresses back to the normal locations of pages
zero and one, effectively swapping those two pages of memory. This address translation
applies only to RAM; ROM and I/O registers are not back-translated. VIC chip
addresses are not translated back to their original memory locations, so you must ensure
that you do not place page zero or one in the address of the VIC chip. The ROM
appearing in these address ranges still overlays the RAM regardless of whether the RAM
is zero page, page one or any other page of RAM memory. If you need to use the
Kemal, the necessary variables required for the Kemal routine must be placed in
memory where the Kernal is in context, Machine Language Monitor bank 15 ($0F) for
example.
Bit 0 of both high-byte page pointers (0 and 1) corresponds to the RAM bank
number for any address access in page zero or one. The page zero high byte (bit 0)
normally overrides the RAM bank set by the configuration register. However, if
common RAM (at the bottom of RAM bank 0) is specified, the high-byte pointer for
pages zero and one is ignored and pages zero and one appear in common RAM. Other
wise, if common RAM is not allocated, pages zero and one appear where you specify
according to the contents of the page pointers. In other words, common RAM takes priority
over the page pointers, if common RAM is allocated by the RAM configuration register.
This feature of relocatable page zero and page one provides many benefits to the
programmer. This allows machine language programs to create several pages of zero
page variables or several different stacks. When you need to access the additional zero
page variables or extra 256 bytes of the stack, simply change the pointer to look at the
next page. This provides additional speed in your programs since you may use zero page
addressing for subsequent zero pages. You can even place zero page in screen memory
for extra fast writing to the screen. In addition, it gives you a way to implement deeper
levels of subroutines since you have a larger stack area. Remember, though, to leave
three bytes on the top of the stack for interrupt requests and servicing.
THE SYSTEM VERSION REGISTER

The System Version Register, located in address $D50B, contains a value that tells the
user which version of the MMU is inside the C128 and how large the memory is. Bits 0
through 3 contain the MMU version number. Bits 4 through 7 contain a value (in
memory blocks) pertaining to the size of the C l 28 memory. This allows the programmer
to check the version of the C l28 and the memory size, and make it compatible with
systems that will be expanded in the future. The current version of the C l 28 contains the
value $20, signifying two 64K blocks.
A U T O STARTING A ROM
APPLICATION CARTRIDGE
Many of you may want to place your application program in a cartridge which plugs into
the expansion port. In order to automatically start the program as soon as you turn on
the computer, you must place a particular coded sequence in the first 6 bytes where the
external (or internal) ROM cartridge maps into memory. Here’s the auto start sequence
in both C l28 and C64 modes.
COMMODORE 128 MODE

Any C64 cartridge is asserted automatically if the system recognizes the /GAME or
/EXROM as being pulled low.
If any C l 28 cartridges are installed into the expansion port:
1. Log the cartridge I.D. into the Physical Address Table (PAT).
2. If the I.D. equals 1, call the cold start vector (which may RTS).
472 COMMODORE 128
The first 9 bytes of the Commodore 128 auto start sequence are:
BYTE
$X000 Cold Start Vector

$X003 Warm Start Vector (not used but must be specified)
$X006 Cartridge I.D . (The I.D . = 1 for an auto start card)
$X007 The ASCII Character “ C ” with the high bit set
$X008 The ASCII Character “ B ” with the high bit set
$X009 The ASCII Character “ M ” with the high bit set
where X is the hexadecimal digit “ 8 ” for $8000 or “ C ” for $C000.

There are four slots where cartridges (ROM’s) may plug in (two internal, two
external). They must follow the sequence described above, whether they are internal or
external.
COMMODORE 64 MODE
The first 10 bytes of the Commodore 64 auto start sequence are:
BYTE
$XOOO Cold Start Vector

$X003 Warm Start Vector
$X006 The ASCII Character “C” with the high bit set
$X007 The ASCII Character “B” with the high bit set
$X008 The ASCII Character “M” with the high bit set
$X009 The ASCII Character “8 ” with the high bit set
$X00A The ASCII Character “ 0” with the high bit set
where X is the hexadecimal digit “ 8” for $8000 or “ C ” for $C000.

The recommended cartridge header for both operational modes is as follows:
SEI
JMP START
NOP
NOP
This header is recommended so that the interrupt disable status bit is set when
control is passed to the software in the cartridge ROM.
THE COMMODORE 128

SCREEN EDITOR
The Commodore screen editor is among the easiest to use of all screen editors. As soon
as you turn on the computer, the screen editor is available to you. You don’t have to call
any additional text editors. Using the keys for manipulating text, the screen editor gives
you more freedom than most other editors.
C l28 EDITOR ESCAPE CODES

To use the following ESCAPE functions, press the ESCAPE key and then press
the key for the function you want.
KEY FUNCTION
A Enable auto-insert mode

B Set bottom right o f screen window at cursor position
C Disable auto-insert mode
D Delete current line
E Set cursor to nonflashing mode
F Set cursor to flashing mode
G Enable bell (control-G)
H Disable bell
I Insert line
J Move to start of current line
K Move to end o f current line
L Enable scrolling
M Disable scrolling
N Return screen to normal (nonreverse video) state (80-column screen only)
0 Cancel insert, quote, underline, flash and reverse modes
P Erase to start of current line
Q Erase to end of current line
R Set screen to reverse video (80-column screen only)
S Change to block cursor (80-column screen only)
T Set top left of screen window at cursor position
U Change to underline cursor (80-column screen only)
V Scroll up
w Scroll down
X Swap 40/80-column display output device
Y Set default tab stops ( 8 spaces)
Z Clear all tab stops
@ Clear to end of screen
474 COMMODORE 128
C l 28 EDITOR CONTRO L CODES

The following control characters in the CBM ASCII table have been added or changed
from those found in the C64. Codes not shown in this table have the same function as
found in the C64.
CHR$ KEYBOARD
VALUE CONTROL CHARACTER FUNCTION
2 B Underline ON (80-column screen only)

7 G Produces bell tone
9 I Tab character
10 J Line feed character
11 K Disable shift Commodore key (formerly code 9)
12 L Enable shift Commodore key (formerly code 8 )
IS 0 Turn ON flash on (80-column screen only)
24 X Tab set/clear
27 [ Escape character
130 Underline OFF (80-column screen only)
143 Turn flash OFF (80-column screen only)
C l28 EDITOR JUMP TABLE

The editor calls listed below are a set of extensions to the standard CBM jump table.
They are specifically for the C l 28 and should not be considered as permanent additions
to the standard jum p table. They are all true subroutines and terminate with the RTS
instruction. As with all Kernal calls, the system configuration (high ROM, RAM-0 and
I/O) must be in context at the time of the call.
1. $C000 CINT in itialize editor and screen

2. $C003 DISPLY ;display .A = char, .X = color
3. $C006 LP2 ;get a key from irq buffer in .A
4. SC009 LOOP5 ;get a chr from screen line in .A
5. $C00C PRINT ;print character in .A
6 . $C00F SCRORG ;get size of current window
7. $C012 SCNKEY ;scan keyboard subroutine
8. $C015 REPEAT ;repeat key logic and CKIT2
9. $C018 PLOT ;read or set cursor position
10. $C018 CURSOR ;move 8563 cursor subroutine
11. $C01E ESCAPE ;execute escape function
12. $C021 KEYSET ;redefine a programmable key
13. $C024 IRQ ;irq entry
14. $C027 INIT80 in itialize 80-column char set
15. $C02A SW APPER ;40/80 mode change
16. $C02D WINDOW ;set UL or BR of window
17. $C033 LDTB2 ;screen lines low byte table
18. $C04C LDTB1 ;screen lines high byte table
19. $334 CONTRL ;print C T R L indirect
20. $336 SHIFTD ;print S H F T indirect
21. $338 ESCAPE ;print E S C indirect
22. $33A KEYVEC ;keyscan logic indirect
23. $33C KEYCHK ;keyscan store indirect
24. $33E DECODE ;keyboard decode table vectors
Entries 17, 18, and 24 are table pointers, and are not callable routines. Entries 19-23
are considered indirect vectors, not true entry points.
This chapter has presented the Commodore 128 operating system. Chapter 16
provides information on CP/M on the Commodore 128 and Commodore 64 memory
maps.
*4
CP/M 3.0
ON THE
COMMODORE 128
477
478 COMMODORE 128
CP/M® is a microprocessor operating system produced by Digital Research, Inc. (DRI).

The version of CP/M used on the Commodore 128 is CP/M Plus® Version 3.0. In this
chapter, CP/M is generally referred to as CP/M 3.0, or simply CP/M.
This chapter summarizes the non-C 128-dependent aspects of CP/M on the Commodore
128. For detailed information on C l 28-dependent CP/M 3.0, see Appendix K of this
guide. For detailed information on non-C 128-dependent CP/M 3.0, see the Commodore-
produced volume that includes DRI’s CPIM Plus User’s Guide, Programmer’s Guide,
and System Guide.
REQUIREMENTS FOR A CP/M 3.0 SYSTEM
The general requirements for CP/M 3.0 are:
■ A computer containing a Z80 microprocessor.

■ A console consisting of a keyboard and a display screen.
■ At least one floppy disk drive.
■ CP/M system software on disk.
CP/M 3.0 on the Commodore 128 Personal Computer normally consists of the
following elements:
■ A built-in Z80 microprocessor.

■ A console consisting of the full Commodore 128keyboard and an 80-column
monitor.
■ The Commodore 1571 fast disk drive.
■ The CP/M system disk, which includes the CP/M 3.0 system, an extensive
HELP utility program and a number of other utility programs.
NOTE: CP/M can also be used with a 40-column monitor. To view all
80 columns of display, you must scroll the screen horizontally by press
ing the C O N T R O L key and the appropriate CURSOR key
(left or right).
CP/M 3.0 can also be used with the 1541 disk drive. In this case, only single
sided GCR disks can be used, and the speed of operation will be approximately
one-tenth the speed achieved using the 1571 disk drive. (See the discussion of disk
formats later in this chapter for more details.)
CP/M 3.0 ON THE COMMODORE 128 479
COMMODORE ENHANCEMENTS
T O CP/M 3.0
Commodore has added a number of enhancements to CP/M 3.0. These enhancements
tailor the capabilities of the Commodore 128 to those of CP/M 3.0. They include such
things as a selectively displayed disk status line, a virtual disk drive, local/remote
handling of keyboard codes, programmable function keys (strings) and a number of
additional functions/characters that are assigned to various keys. These enhancements
are described at appropriate points in this chapter.
CP/M FILES
There are two types of CP/M files:
■ PR O G R A M or C O M M A N D files, consisting of a series of instructions that

the computer follows to achieve a specified result.
■ D A T A files, consisting usually of a collection of related information (e.g., a
list of customer names and addresses; inventory records; accounting records; the
text of a document).
FILE SPECIFICATION
A CP/M file is identified by a file specification that can consist of up to four individual
elements, as follows:
■ D rive Specifier (optional), consisting of a single letter, followed by a colon.

Each disk drive is assigned a letter, in the range A through E. (E denotes a
virtual drive, as explained later in this chapter.)
■ Filenam e (mandatory), which can be from one to eight characters long. (Note
that this is the only mandatory element of the file specification.)
a Filetype (optional), consisting of one to three characters. It must be separated
from the filename by a period.
■ Password (optional), which can be from one to eight characters. Must be
separated from the filetype (or filename, if no filetype is included) by a
semicolon.
E X A M P LE :
The following file specification contains all four possible elements, all separated by the
appropriate symbols:
A:DOCUMENT.LAW;FIREBIRD
480 COMMODORE 128
USER NUMBER
CP/M 3.0 further identifies all files by assigning each one a user number, which can
range from 0 to 15. CP/M 3.0 assigns the user number to a file when the file is created.
User numbers allow you to separate your files into sixteen file groups.
Any user number other than 0 must precede the drive specifier. User 0, which is
the default user number, is not displayed in the prompt.
If a file resides in user 0 and is marked with a system file attribute, that file can
be accessed from any user number. Otherwise, a command can access only those files
that have the current user number.
CREATING A FILE
There are several ways to create a CP/M file, including:
■ Using the CP/M text editor (ED).

■ Using the PIP command to copy and rename a file.
■ Using a program such as MAC (a CP/M machine language assembler program),
which creates output files as it processes input files.
USING W ILDCARD CHARACTERS TO

ACCESS MORE T H A N ONE FILE
A wildcard is a character that can be used in a filename or filetype in place of some other
characters. CP/M 3.0 uses the question mark (?) and asterisk (*) as wildcards. A ? stands
for any character that may be encountered in that position. An * tells CP/M to fill the
filename with question mark characters as indicated. A file specification containing a wild
card can refer to more than one file and is therefore called an ambiguous file specification.
RESERVED CHARACTERS
The characters in Table 14-1 are reserved characters in CP/M 3.0. Use only as indicated.
CHARACTER MEANING
<$,:!>[] ]
TAB,SPACE, File specification delimiters
CARRIAGE RETURN]
: Drive delimiter in file specification
Filetype delimiter in file specification
; Password delimiter in file specification
; Comment delimiter at the beginning of a command line
* ? Wildcard characters in an ambiguous file specification
< > & ! I \ H— Option list delimiters
[] Option list delimiters for global and local options
() D elim iters for m ultiple m odifiers inside square
brackets for options that have modifiers
/ $ Option delimiters in a command line
T a b le 14-1. C P /M 3.0 Reserved C haracters

CP/M 3.0 O N THE COMMODORE 128 481
RESERVED FILETYPES
The filetypes defined in Table 14—2 are reserved for system use.
FILETY PE MEANING
ASM Assembler source file

BAS BASIC source program
COM Z80 or equivalent machine language program
HEX Output file from MAC (used by HEXCOM)
HLP HELP message file
$$$ Temporary file
PRN Print file from MAC or RMAC
REL Output file from RMAC (used by LINK)
SUB List of commands to be executed by SUBMIT
SYM Symbol file from M AC, RMAC or LINK
SYS System file
RSX Resident System Extension (a file automatically loaded by a command
file when needed)
T ab le 14-2. C P /M 3.0 Reserved Filetypes
CP/M COMMANDS
There are two types of commands in CP/M 3.0:
■ Built-in commands, which identify programs in memory.

■ Transient utility commands, which identify program files.
CP/M 3.0 has six built-in commands and over twenty transient utility commands.
Utilities can be added by purchasing CP/M 3.0-com patible application programs. In
addition, experienced programmers can write utilities that operate with CP/M 3.0.
BUILT-IN COMMANDS
Built-in commands are entered in the computer’s memory when CP/M 3.0 is loaded,
and are always available for use, regardless of which disk is in which drive. Table 14—3
lists the Commodore 128 CP/M 3.0 built-in commands.
Some built-in commands have options that require support from a related transient
utility. The related transient utility command has the same name as the built-in com
mand and has a filetype of COM.
482 COMMODORE 128
COMMAND FUNCTION
DIR Displays filenames of all files in the directory except those marked with
the SYS attribute.
DIRSYS Displays filenames o f files marked with the SYS (system) attribute in the
directory.
ERASE Erases a filename from the disk directory and releases the storage space
occupied by the file.
RENAME Renames a disk file.
TYPE Displays contents of an ASCII (TEXT) file at your screen.
USER Changes to a different user number.
T a b le 14-3. B uilt-In Com m ands
TRANSIENT U T IL IT Y COMMANDS
The CP/M 3.0 transient utility commands are listed in Table 14-4. When a command key
word identifying a transient utility is entered, CP/M 3.0 loads the program file from the
disk and passes that file any filenames, data or parameters specified in the command tail.
USING CONTROL CHARACTERS

FOR LINE EDITING
Table 14—5 lists the line-editing control characters for CP/M 3.0 on the Commodore 128.
H O W TO MAKE COPIES OF
CP/M 3.0 DISKS A N D FILES
NOTE: The Digital Research Inc. COPYSYS command, used in many

CP/M systems in formatting a disk, is not implemented on the Commo
dore 128 computer. Instead, the Commodore 128 uses a special FORMAT
command.
To make backups of the CP/M system disks, use the FORMAT and PIP utility
programs. FORMAT formats the disk as either a C l28 single-sided or C l 28 double-sided
diskette.
NAME FUNCTION
DATE Sets or displays the date and time.
DEVICE Assigns logical CP/M devices to one or more physical devices, changes
device driver protocol and baud rates, or sets console screen size.
DIR Displays directory with files and their characteristics.
DUM P Displays a file in ASCII and hexadecimal format.
ED Creates and alters ASCII files.
ERASE Used for wildcard erase.
GENCOM Creates a special COM file with attached RSX file.
GET T em porarily gets console input from a disk file rather than the
keyboard.
FORMAT Initializes disk in GCR format for CP/M use.
HELP Displays information on how to use CP/M 3.0 commands.
INITDIR Initializes a disk directory to allow time and date stamping.
KEYFIG Allows redefinition o f keys.
PATCH Displays or installs patches to CP/M system.
PIP Copies files and combines files.
PUT Temporarily directs printer or console output to a disk file.
RENAM E Changes the name of a file, or a group of files, using wildcard characters.
SAVE Saves a program in memory to disk.
SET Sets file options including disk labels, file attributes, type o f time and
date stamping and password protection.
SETDEF Sets system options including the drive search chain.
SHOW Displays disk and drive statistics.
SUBMIT Automatically executes multiple commands.
TYPE Displays contents of text file (or group o f files, if wildcard characters are
used) on screen (and printer if desired).
T ab le 14-4. T ran sie n t U tility Com m ands

484 COMMODORE 128
CHARACTER MEANING
CTRL-A or Moves the cursor one character to the left.

SHIFT-LEFT CURSOR
CTRL-B Moves the cursor to the beginning of the command line without
having any effect on the contents of the line. If the cursor is
at the beginning, CTRL-B moves it to the end o f the line.
CTRL-E Forces a physical carriage return but does not send the
command line to CP/M 3.0. M oves the cursor to the begin
ning of the next line without erasing the previous input.
CTRL-F or Moves the cursor one character to the right.
RIGHT CURSOR
CTRL-G Deletes the character at current cursor position. The cursor
does not move. Characters to the right o f the cursor shift
left one place.
CTRL-H Deletes the character to the left of the cursor and moves
the cursor left one character position. Characters to the right
of the cursor shift left one place.
CTRL-I Moves the cursor to the next tab stop. Tab stops are auto
matically set at each eighth column. Has the same effect
as pressing the T A B key.
CTRL-J Sends the command line to CP/M 3.0 and returns the cursor
to the beginning of a new line. Has the same effect as a
R E T U R N or a CTRL-M keystroke.
CTRL-K Deletes to the end of the line from the cursor.
CTRL-M Sends the command line to CP/M 3.0 and returns the cursor
to the beginning of a new line. Has the same effect as a
R E T U R N or a CTRL-J keystroke.
CTRL-R Retypes the command line. Places a # character at the
current cursor location, moves the cursor to the next line,
and retypes any partial command you typed so far.
CTRL-U Discards all the characters in the command line, places a
# character at the current cursor position, and moves the
cursor to the next line. However, you can use a CTRL-W to
recall any characters that were to the left of the cursor
when you pressed CTRL-U.
CTRL-W or f CRSR | Recalls and displays previously entered command line both
at the operating system level and within executing programs,
if the CTRL-W is the first character entered after the
prompt. CTRL-J, CTRL-M , CTRL-U and RETURN define
the command line you can recall. If the command line
contains characters, CTRL-W moves the cursor to the end
of the command line. If you press R E T U R N , CP/M
3.0 executes the recalled command.
CTRL-X Discards all the characters left of the cursor and moves the
cursor to the beginning of the current line. CTRL-X saves
any characters to the right of the cursor.
T ab le 14-5. C P /M 3.0 Line-Editing C o n tro l Characters
MAKING COPIES W ITH A

SINGLE DISK DRIVE
You can copy the contents of a disk with a single Commodore disk drive (1541 or
1571). For example, use the following sequence of commands to create a bootable
CP/M system disk. First type:
A >FO RM A T
and follow the instructions given on the screen. When the copy disk is formatted,
type:
A > P IP E: = A :C PM + .SYS
When the CPM + SYS file is copied, you type:
A > P IP E: = A:CCP.COM
To copy everything on a disk, use the following command sequence:
A >FO RM A T
A > P IP E: = A:*.*
The system prompts the user to change disks as required.

Use drive A as the source drive and drive E as the destination drive. Drive E is
referred to as a virtual drive; that is, it does not exist as an actual piece of hardware— it is
strictly a logical drive.
MAKING COPIES W ITH TW O DISK DRIVES

You can back up disks in CP/M by using two drives: drive A and drive B. The drives
can be named with other letters from A through D. To make a copy of your CP/M 3.0
system disk, first use the FORMAT utility to copy the operating system loader. Make
sure that the CP/M system disk is in drive A, the default drive, and the blank disk is in
drive B. Then enter the following command at the system prompt:
A > P IP B: = A:CPM + .SYS
When you have copied the CPM + .SYS file, use the PIP command to copy the
CCP.COM file. This provides a copy of the operating system only. To copy all of
the files from the system disk, enter the following PIP command:
A > P IP B: = A:*.*
This PIP command copies all the files on the disk in drive A to the disk in drive B. PIP
displays the message COPYING, followed by each filename as the copy operation
proceeds. When PIP finishes copying, the system prompt (A > ) is displayed.
486 COMMODORE (28
GENERAL CP/M 3.0 SYSTEM LA YO U T
The Commodore 128 computer is a two-processor system, with the 8502 as the primary
processor and the Z80 as the secondary processor. The 8502 has the same instruction set
as the 6502. The Z80’s primary function is to run CP/M 3.0. This section describes the
general requirements and methods for implementing CP/M 3.0 on the C l28.
When CP/M is running, the normal functions of the C l 28 are not supported
(CP/M and BASIC cannot run at the same time). Also, CP/M does not directly support
all the display modes of the VIC chip. (An application could be written to run under
CP/M that could use additional graphics capabilities, but the application would have to
keep track of all the details, such as memory maps).
CP/M 3.0 OPERATING SYSTEM

COM PONENTS
The CP/M 3.0 operating system consists of the following three modules: the Console
Command Processor (CCP), the BASIC Disk Operating System (BDOS), and the Basic
Input Output System (BIOS).
The CCP is a program that provides the basic user interface to the facilities of the
operating system. The CCP supplies the six built-in commands: DIR, DIRS, ERASE,
RENAME, TYPE, and USER. The CCP executes in the Transient Program Area
(TPA), the region of memory where all application programs execute. The CCP contains
the Program Loader Module, which loads transient (applications) programs from disk
into the TPA for execution. On the Commodore 128, a 58K to 59K TPA area is provided
for CP/M.
The BDOS is the logical nucleus and file system of CP/M 3.0. The BDOS
provides the interface between the application program and the physical input/output
routines of the BIOS.
The BIOS is a hardware-dependent module that interfaces the BDOS to a particu
lar hardware environment. The BIOS performs all physical I/O in the system. The BIOS
consists of a number of routines that are configured to support the specific hardware of
the target computer system (in this case, the Commodore 128).
The BDOS and the BIOS modules cooperate to provide the CCP and other
transient programs with hardware-independent access to CP/M 3.0 facilities. Because
the BIOS can be configured for different hardware environments and the BDOS remains
constant, you can transfer programs that run under CP/M 3.0 to systems with different
hardware configurations.
C O M M U N IC A T IO N BETWEEN MODULES
The BIOS loads the CCP into the TPA at system cold and warm starts. The CCP moves
the Program Loader Module to the top of the TPA and uses the Program Loader Module
to load transient programs.
The BDOS contains a set of functions that the CCP and applications programs call
to perform disk and character input and output operations.
The BIOS contains a Jump Table with a set of thirty-three entry points that the
BDOS calls to perform hardware-dependent primitive functions, such as peripheral
device I/O. For example, CO N IN is an entry point of the BIOS called by the BDOS to
read the next console input character.
Similarities exist between the BDOS functions and BIOS functions, particularly
for simple device I/O. For example, when a transient program makes a console output
function call to the BDOS, the BDOS makes a console output call to the BIOS. In the
case of disk I/O, however, this relationship is more complex. The BDOS file may make
many BIOS function calls to perform a single BDOS file I/O function. BDOS disk I/O is
in terms of 128-byte logical records. BIOS disk I/O is in terms of physical sectors and
tracks.
The System Control Block (SCB) is a 100-byte (decimal) CP/M 3.0 data structure
that resides in the BDOS system component. The BDOS and the BIOS communicate
through fields in the SCB. The SCB contains BDOS flags and data, CCP flags and data,
and other system information, such as console characteristics and the current date and
time. You can access some of the System Control Block fields from the BIOS.
Note that the SCB contains critical system parameters that reflect the current state
of the operating system. If a program modifies these parameters, the operating system
can crash. See Section 3 of the DRI CPIM Plus System Guide and the description of
BDOS Function 49 in the DRI CPIM Plus Programmer’s Guide for more information on
the System Control Block.
Page zero is a region of memory that acts as an interface between transient
programs and the operating system. Page zero contains critical system parameters, includ
ing the entry to the BDOS and entry to the BIOS Warm BOOT routine. At system
start-up, the BIOS initializes these two entry points in page zero. All linkage between
transient programs and the BDOS is restricted to the indirect linkage through page zero.
CP/M 3.0 BIOS OVERVIEW

This section describes the organization of the CP/M 3.0 BIOS and the BIOS jump
vector. It provides an overview of the System Control Block, followed by a discussion
of system initialization procedures, character I/O, clock support, disk I/O, and memory
selects and moves.
O R G A N IZA TIO N OF THE BIOS

The BIOS is the CP/M 3.0 module that contains all hardware-dependent input and
output routines. To configure CP/M 3.0 for a particular hardware environment, the
sample BIOS supplied with the CPIM Plus System Guide must be adapted to the specific
hardware of the target system (the Commodore 128).
The BIOS essentially is a set of routines that performs system initialization,
488 COMMODORE 128
character-oriented I/O to the console and printer devices, and physical sector I/O to disk
devices. The BIOS also contains routines that manage block moves and memory selects
for bank-switched memory. The BIOS supplies tables that define the layout of the disk
devices and allocate buffer space that the BDOS uses to perform record blocking and
deblocking. The BIOS can maintain the system time and date in the System Control Block.
Table 14-6 describes the entry points into the BIOS from the Cold Start Loader
and the BDOS. Entry to the BIOS is through a set of jum p vectors. The jum p table is
a set of thirty-three jump instructions that pass program control to the individual BIOS
subroutines.
NO. INSTRUCTION DESCRIPTION
0 JM P BOOT Perform cold start initialization.

1 JM P WBOOT Perform warm start initialization.
2 JM P CONST Check for console input character ready.
3 JM P CONIN Read console character in.
4 JM P CONOUT Write console character out.
5 JM P LIST Write list character out.
6 JM P AUXOUT Write auxiliary output character.
7 JM P AUXIN Read auxiliary input character.
8 JM P HOME Move to Track 00 on selected disk.
9 JM P SELDSK Select disk drive.
10 JM P SETTRK Set track number.
11 JM P SETSEC Set sector number.
12 JM P SETDMA Set DMA address.
13 JM P READ Read specified sector.
14 JM P WRITE Write specified sector.
15 JM P LISTST Return list status.
16 JM P SECTRN Translate logical to physical sector.
17 JM P CONOST Return output status of console.
18 JM P AUXIST Return input status of aux. port.
19 JM P AUXOST Return output status o f aux. port.
20 JM P DEVTBL Return address of char. I/O table.
21 JM P DEVINI Initialize char. I/O devices.
22 JM P DRVTBL Return address of disk drive table.
23 JM P MULTIO Set number o f logically consecutive sectors to be read or
written.
24 JM P FLUSH Force physical buffer flushing for user-supported deblocking.
25 JM P MOVE Memory to memory move.
26 JM P TIME Time set/get signal.
27 JM P SELMEM Select bank of memory.
28 JM P SETBNK Specify banks for inter-bank MOVE.
29 JM P XMOVE Set banks for an inter-bank MOVE.
30 JM P USERF Commodore CP/M system functions.
31 JM P RESERV1 Reserved for future use.
32 JM P RESERV2 Reserved for future use.
T ab le 14-6. C P /M 3.0 B IO S Jump Vectors

All the entry points in the BIOS jum p vector are included in the C l28 CP/M 3.0
BIOS.
Each jum p address in Table 14-6 corresponds to a particular subroutine that
performs a specific system operation. Note that two entry points are reserved for future
versions of CP/M, and one entry point is provided for the Commodore system functions.
Table 14-7 shows the five categories of system operations and the function calls
that accomplish these operations.
OPERATION FUNCTION
System BOOT, W BOOT, DEVTBL, DEVINI, DRVTBL

Initialization
Character I/O CONST, CONIN, CONOUT, LIST, AUXOUT, AUXIN, LISTST,

CONOST, AUXIST, AUXOST
Disk I/O HOME, SELDSK, SETTRK, SETSEC, SETDM A, READ, W RITE,

SECTRN, M ULTIO, FLUSH
Memory Selects M OVE, SELM EM , SETBNK, XMOVE

and Moves
Clock Support TIME
T a b le 14-7. C P /M 3 B IO S Functions
Appendix K illustrates how to call a Commodore CP/M system function in Z80

machine language.
SYSTEM MEMORY
O R G A N IZA TIO N
Figure 14—1 shows the general memory organization of CP/M 3.0.
490 128
F000
C O M M O N SYSTEM M E M O R Y
EOOO
DOOO
COOO
BOOO B A N K E D SYSTEM
AOOO
9000
8000
T
7000 p
A
6000
5000
4000
C C P BUFFER
3000
VIC SCREEN
8502 BIOS
2000 40-COLUMN
LOGICAL SCREEN
1000 KEY TABLES
ROM C O DE
0000
14-1. C P /M 3.0 G eneral M em o ry O rg an izatio n

The memory map is limited to 64K at any one point in time. However, the RAM
bank can be selected and then different ROM areas can overlay the RAM (with
bleed-through on write operations). The actual memory map is controlled by the MMU. The
MMU can be accessed in the I/O area (only when I/O is enabled in Z80 space) or
through the Load Configuration Register located at FF00 through FF04.
If the Load Configuration Registers are read, then the current value is read. A
write to FF00 changes the configuration after completing the current instruction. A
write to FF01 to FF04 updates the current configuration to the value stored in the
Preconfiguration Registers (the data written is not used). The MMU page pointers have
both a low (page) and a high (page) pointer. The high is written first and latched in the
MMU; the high value is updated from the latch when the low byte is written. The MMU
Control Registers are listed in Chapter 13.
DISK O R G A N IZA TIO N
CP/M 3.0 supports a number of different disk formats, including three Commodore
formats and a number of MFM formats. (MFM is the industry standard format.) The
first Commodore format is single-sided Commodore GCR, which is compatible with the
CP/M 2.2 that runs on the Commodore 64. With this format, the File Control Block
(FCB) is set up as 32 tracks of 17 sectors each and a track offset of 2. The BIOS routine
adds a 1 to tracks greater than 18 (this is the C64 directory track).
The second format, known as the C128 CP/M Plus disk format, is new and is also
single-sided. This format, which is also a GCR format, takes advantage of the full disk
capacity by setting up the FCB with 638 tracks of 1 sector each and a track offset of 0.
This has the effect of having CP/M set the track to the block number relative to
the beginning of the disk, with the sector always set to 0. The following algorithm is
used to convert the requested TRACK to a real track and sector number.
REQUESTED TRACK ACTUAL TRACK
000 > = TRACK > 355 ((TRACK + 1)121) + 1

355 > = TRACK > 487 ((TRACK - 354)/19) + 18
487 > = TRACK > 595 ((TRACK - 487V18) + 25
595 > = TRACK > 680 ((TRACK - 595)/17) + 31
680 > = TRACK > 1360 SET SIDE 2 TRACK = TRACK - 680
492 COMMODORE 128
The effective sector is then translated to provide a skew that speeds up operations.
The skew is used only with the new larger format. A different skew table is used for
each region of the disk.
The third Commodore format is a GCR double-sided format. The disk is treated as
1276 sectors of data with a track offset of 0. Side 1 is used first; then side 2 is used.
N O TE: This is not the usual way to handle a two-sided disk; however,
allocating the disk in this manner, the user with a 1541 may still be able
to read data written at the start of a two-sided disk.
The third Commodore format and all MFM formats require that the user have the
new 1571 disk drive. This disk drive supports both single- and double-sided diskettes
and both the Commodore GCR and industry standard MFM data coding formats.
The following table summarizes 1541/1571 disk drive capabilities with regard to
the various disk formats.
DISK FORMAT 1541 DRIVE 1571 DRIVE
C64 GCR
single-sided
J /
C128 GCR
single-sided
J J
C128 GCR
d o ub le-sided
J
MFM
fo r m a t
J
C64 CP/M DISK FORMAT

(SINGLE-SIDED)
This format, shown in Figure 14—2, is provided to allow the user to read/write files
created using the C64 CP/M 2.2 cartridge. (However, do not use the CP/M 2.2 cartridge
with the C128.) Notice the unused space in this format.
SECTO R
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
0 B B B B B B B B B B B B B B B B B X X X X
1 B B B B B B B B B B B B B B B B B x X X X
2 x X X X
3 X X X X
4 X X X X
5 X X X X
6 X X X X
7 X X X X
8 X X X X
9 X X X X
10 X X X X
11 X X X X
12 X X X X
13 X X X X
14 X X X X
15 X X X X
16 X X X X
17 D D D D D D D D D D D D D D D D D D D
18 X X
19 X X
20 X X
21 X X
22 X X
23 X X
24 X
25 X
26 X
27 X
28 X
29 X
3 0 ...............................................................
3 1 ......................................................
3 2 ...............................................................
3 3 ...............................................................
3 4 ...............................................................
. U sed by C P /M .
B Boot S ector (System ).
D D irecto ry S ector (Disk DOS),
x N ot used by C P /M .
Figure 14-2. C 64 C /P M Disk F o rm a t
C l28 CP/M DISK FORMATS

(SINGLE- A N D DOUBLE-SIDED)
F ig u re 1 4 -3 show s th e C l 28 C P /M P lus fo rm a t fo r a sin g le -sid e d disk.
494 COMMODORE 128
SECTOR
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
0 B . . . . X ...........................................................................................................................
1 ...................................................................................................................
2
...........................................................................................................................................................................
3 ................................................................................................................................................................
4 ................................................................................................................................................................
5 ................................................................................................................................................................
6 ................................................................................................................................................................
7 ................................................................................................................................................................
8 ................................................................................................................................................................
9 ...........................................................................................................................................................................
1 0 ...........................................................................................................................................................................
1 1 ...........................................................................................................................................................................
1 2 ...........................................................................................................................................................................
1 3 ..................................................................................................................................................................
1 4 ..................................................................................................................................................................
1 5 ...................................................................................................................................................................
1 6 .................................................................................................................................
1 7 ................................................................................................................................................
18 D .................................................................................................................................................
1 9 ................................................................................................................................................
2 0 ...................................................................................................................
2 1 .........................................................................................................................................................
2 2 .........................................................................................................................................................
2 3 .................................................................................................................................................
2 4 .......................................................................................................................................
2 5 .......................................................................................................................................
2 6 .................................................................................................................
2 7 .......................................................................................................................................
2 8 ...................................................................................................................
2 9 .......................................................................................................................................
3 0 ..............................................................................................................................
3 1 ...............................................................................................................................
3 2 ..............................................................................................................................
3 3 ..............................................................................................................................
3 4 ..............................................................................................................................
Used by C P /M , reg . = region (1-4).

B B oot S ector (System ).
D D irecto ry S ector (Disk DOS),
x N ot used by C P /M .
SKEW TABLE
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1 00 17 13 09 05 01 18 14 10 06 02 19 15 11 07 03 20 16 12 08
2 00 04 08 12 16 01 05 09 13 17 02 06 10 14 18 03 07 11 15
3 00 11 04 15 08 01 12 05 16 09 02 13 06 17 10 03 14 07
4 00 07 14 04 11 01 08 15 05 12 02 09 16 06 13 03 10
Figure 14-3. C l 28 C P/M Plus Disk F o rm a t

N O TE: This format is duplicated on the second side of a two-sided disk

(with the exception of B, which is unused and therefore becomes X).
MFM DISK FORMATS

A number of MFM disk formats are built into CP/M on the C l 28. These formats, which
can be selected at the time a program is to be run, include Osborne, Kaypro, Epson and
IBM CP/M-86. The IBM CP/M-86 capability is provided so that data can be transferred
between machines. However CP/M-86 programs cannot be run on the C128, since on
the C l 28, CP/M Plus runs on a Z80, not an 8088.
When used with the 1571 Disk Drive, the Commodore 128 supports a variety
of double-density MFM disk formats (for reading and/or writing), including:
Epson QX10 (512 byte sectors, double-sided, 10 sectors per track)
IBM-8 SS (CP/M 86) (512 byte sectors, single-sided, 8 sectors per track)
IBM-8 DS (CP/M 86) (512 byte sectors, double-sided, 8 sectors per track)
KayPro II (512 byte sectors, single-sided, 10 sectors per track)

KayPro IV (512 byte sectors, double-sided, 10 sectors per track)
Osborne DD (1024 byte sectors, single-sided, 5 sectors per track)
When you insert one of these disks into the disk drive and try to access it, the system
senses the type of disk with respect to the number of bytes per sector and the number of
sectors per track. If the disk format is not unique, a box is displayed near the bottom-left
corner of the screen, showing which disk type you are accessing. The system requires
you to select the specific disk type by scrolling through the choices given in this
window.
N O TE: The choices are given one at a time; scroll through using the
right and left arrow keys. Press R E T U R N when the disk type that
you know is in the disk drive is displayed. Press C O N T R O L
R E T U R N to lock this disk format so that you will not need to
select the disk type each time you access the disk drive.
496 COMMODORE 128
KEYBOARD SCANNING
The keyboard scan routine that is called to get a keyboard character returns the key code
of the pressed key, or a code indicating that no key is currently being pressed. The
keyboard scan code is also responsible for handling programmable keys, programmable
function keys, setting character and background colors, selecting MFM disk formats and
selecting current screen emulation type.
Any key on the keyboard can be defined to generate a code or function except the
following keys:
LEFT S H I F T
RIGHT SHIFT
S H IF T LOCK
G
CONTROL
R E ST OR E (8502 N M I )
40/ 80 D I S P L A Y
C A P S LOCK KEY
The keyboard recognizes the following special functions:
Cursor left key-Used to define a key
Cursor right key-Used to define a string

(points to function keys)
ALT key-Used as toggle key filter
To indicate these functions, hold down the C O N T R O L key and the

RIGHT SHIFT key and simultaneously press the desired function key.
DEFINING A KEY
The K E Y FIG utility program allows the user to define the code that a key can produce.
Each key has four modes of use for this function:
■ Normal
■ Alpha shift
■ Shift
■ Control
The alpha shift mode is toggled on/off by pressing the Commodore (O ) key. When
this mode is turned on, a small white box appears on the bottom of the screen. The first
key that is pressed thereafter is the key to be defined. The current hex value assigned to
this key is displayed, and the user can then type the new hex code for the key, or abort
by typing a non-hex key. The following is a definition of the codes that can be assigned
to a key. See KEYFIG HELP for more information.
CODE FUNCTION
00h Null (same as not pressing a key)

Olh to 7Fh Normal ASCII codes
80h to 9Fh String assigned
AOh to AFh 80-column character color
BOh to BFh 80-column background color
COh to CFh 40-column character color
DOh to DFh 40-column background color
EOh to EFh 40-column border color
FOh Toggle disk status on/off
F lh System Pause
F2h (Undefined)
F3h 40-column screen window right
F4h 40-column screen window left
F5h to FFh (Undefined)
DEFINING A STRING
The D E FIN E ST R IN G function allows the user to assign more than one key code to a
single key. Any key that is typed in this mode is placed in the string. The user can see
the results of typing in a long box at the bottom of the screen. Note, however, that some
keys may not display what they are.
To allow the user control of entering data, five special edit functions are provided.
To access any of these functions, you first press the C ON T R OL and RI GHT
S H I F T keys. (This allows the user to enter any key into the buffer.)
The functions assigned to the five string edit keys are as follows:
KEY FUNCTION
RE T U R N End string definition.
4- symbol* Insert space in string.
— symbol* Delete cursor character.
Right Cursor Move cursor to right.
Left Cursor Move cursor to left.
*On main keyboard only.
ALT MODE
This function is a toggle (on/off) and is provided to allow the user to send 8-bit
codes to an application without the keyboard driver “ eating” the code from 80h to
FFh.
498 COMMODORE 128
UPDA TING THE

40/80 COLUM N DISPLAY
As noted elsewhere in this book, there are two different display systems within
the C128. The first, which is controlled by the VIC chip, produces a 25-line by
40-column display, has many graphics modes of operation, and can be used with a
standard color (or black-and-white) television or color monitor. (See Chapters 8 and 9
for details.) The only VIC-controlled display mode used by CP/M is standard character
mode, with each character and screen background having up to sixteen colors.
The second display system available in C l28 CP/M is controlled by the 8563
display controller. The display format of this controller is 25 lines by 80 columns, with
character color attributes. The VIC chip is a memory-mapped display, and the 8563 is
I/O-controlled. The two display subsystems are treated as two separate displays. CP/M
3.0 can assign one or both to the console output device.
Both displays are controlled by a common terminal emulation package, a Lear
Siegler ADM-31 (ADM-3A is a subset of this) driver. The terminal driver is divided
into two parts: terminal emulation and terminal functions. Terminal emulation is
handled by the Z80 BIOS, and the terminal function is handled primarily in the
Z80 ROM.
The following section shows the various terminal emulation protocols supported
by Commodore 128 CP/M.
TERMINAL EM ULATION PROTOCOLS
FUNCTION CHARACTER SEQUENCE HEX CHAR CODE
Position Cursor ESC (ro w # + 32) (c o l# + 32) IB 3D 2 0 + 20 +

Cursor Left Control H 08
Cursor Right Control L OC
Cursor Down Control J OA
Cursor Up Control K OB
Home and Clear Screen Control Z 1A
Carriage Return Control M 0D
Escape Control [ IB
Bell Control G 07
NOTE: Display is 24 (1—24) by 80 (1-80); cursor origin is always 1/1.
Figure 14-4. L ear S iegler ADM-3A P rotocol

N O TE: The following have been added to allow the system to emulate
the KayPro II display more closely.
CHARACTER
FUNCTION SEQUENCE
Home Cursor Control t

CEL (Clear to End of Line) Control-X
CES (Clear to End of Screen) Control-W
CHARACTER SEQUENCE HEX CHARACTER CODES
Clear to end of line ESC T IB 54

ESC t IB 74
Clear to end of screen ESC Y IB 59
ESC y IB 79
Home cursor and clear screen ESC I IB 3A
ESC * IB 2A
Half intensity on ESC ) IB 29
H alf intensity off ESC ( IB 28
Reverse video on ESC G4 IB 47 34
Blinking on ESC G2 IB 47 32
Underline on * ESC G3 IB 47 33
Select alternate character set* ESC G1 IB 47 31
Reverse video and blinking off ESC GO IB 47 30
Insert line ESC E IB 45
Insert character ESC Q IB 51
Delete line ESC R IB 52
Delete character ESC W IB 57
Set screen colors* ESC ESC ESC color #
where color # = 20h to 2Fh haracter color
30h to 3Fhi— background color Physical
40h to 4Fh border color (40 Colors
columns only)
50h to 50Fh character color
60h to 60Fh — background color Logical
70h to 70Fhi— border color (40 Colors
columns only)
*Indicates this is not a normal ADM31 sequence.
Note: Display is 24 (1-24) by 80 (1-80); cursor origin is always 1/1.
Figure 14—5. L ear Siegler ADM-31 P rotocol

500 COMMODORE 128
SYSTEM OPERATIONS
SETTING SYSTEM TIME

The time of day is set with this function. The time of day is stored in packed BCD
format in the System Control Block (SCB) in three locations (hours, minutes, seconds).
This routine reads the SCB time and writes that time to the time of day clock within the
6526. This time is updated on the chip and is used by CP/M. The Z80 is able to
read/write the 6526 directly.
U PD A TIN G SYSTEM TIME

The SCB time is updated from the time of day clock on the 6526 by doing a system call.
8502 BIOS O R G A N IZA TIO N
The 8502 is responsible for most of the low-level I/O functions. The request for these
functions is made through a set of mailboxes. Once the mailboxes are set up, the Z80
shuts down and the 8502 starts up (BIOS85). The 8502 looks at the command in the
mailbox and performs the required task, sets the command status and shuts down. The
Z80 is re-enabled; it then looks at the command status and takes the appropriate actions.
The 8502 BIOS commands are defined in Appendix K.
This concludes the summary explanation of the Commodore 128 CP/M system.
However, the Commodore 128 CP/M System includes many additional Commodore
128-dependent routines and functions that are performed by the Z80 microprocessor.
Since most of these routines, system calls and functions are tabular information, they are
covered in Appendix K. For information on any of the following topics, refer to
Appendix K.
8502 BIOS commands

All Commodore 128 Z80 system-dependent user functions
Calling a CP/M BIOS, 8502 BIOS and CP/M User system functions in Z80 machine
language
More information on MFM disk formats
Commodore 128 CP/M (Z80) Memory Map
For more information on the general (non-Commodore 128-dependent) CP/M 3.0

system, see the offer in section 15 of the C128 System Guide for the Digital Research
CP/M Plus User’s Guide, Programmer’s Guide and System Guide.
15
THE
COMMODORE 128
AND
COMMODORE 64
MEMORY MAPS
502 COMMODORE 128
This chapter provides the memory maps for both C l28 and C64 modes. A memory map
tells you exactly how memory is laid out internally in both RAM and ROM. It tells you
exactly what resides in each memory location. The memory map directs you in finding
address vectors for routines and entry points and provides information about the general
layout of the computer. The memory map is probably the most vital programming tool.
Refer to the memory map whenever you need directions throughout the memory of your
Commodore 128. Addresses listed with more than one address label are used for more
than one purpose. To BASIC, the variable has one purpose; to the Machine Language
Monitor, it may have another.
The conventions used for the memory maps are as follows;
Column 1 Column 2 Column 3 Column 4

MEMORY
ADDRESS HEXADECIMAL DECIMAL
LABEL ADDRESS ADDRESS DESCRIPTION
See Appendix K for the Z80 memory map for CP/M on the Commodore 128.
C l28 MEMORY MAP
MEMORY
LABEL ADDRESSADDRESS DESCRIPTION
D6510 0000 0 6510 DATA DIRECTION REGISTER

R6510 0001 1 6510 DATA REGISTER
BANK 0002 2 TOKEN ‘SEARCH’ LOOKS FOR, OR
BANK #
PC—HI 0003 3 ADDRESS FOR BASIC SYS
COM MAND OR M ONITOR AND
LONG CALL/JUMP ROUTINES
P C _L O 0004 4 ADDRESS, STATUS, A-REG, X-REG,
Y-REG
S_R EG 0005 5 STATUS REG TEMP
AREG 0006 6 .A REG TEMP
X _R E G 0007 7 .X REG TEMP
Y _R E G 0008 8 .Y REG TEMP
STKPTR 0009 9 STACK POINTER TEMP
BASIC ZERO PAGE STORAGE

INTEGR 0009 9
CHARAC SEARCH CHARACTER
ENDCHR 000A 10 FLAG: SCAN FOR QUOTE AT END
OF STRING
TRM POS 000B 11 SCREEN COLUMN FROM LAST TAB
VERCK oooc 12 FLAG: 0 = LQAD, 1= VERIFY
THE COMMODORE 128 A N D COMMODORE 64 MEMORY MAPS 503
C128 Memory M ap (continued)
M EM ORY
ADDRESS HEXADECIM AL DECIM AL
COUNT OOOD 13 INPUT BUF.PTR / # OF SUBSCRIPTS

DIMFLG OOOE 14 FLAG: DEFAULT ARRAY DIMENSION
VALTYP 000F 15 DATA TYPE: $FF = STRING,
$00 = NUMERIC
INTFLG 0010 16 DATA TYPE: $00 = FLO AT.PT,
$80 = INTEGER
GARBFL 0011 17 FLAG: DATA SCAN / LIST QUOTE /
GARBAGE COLLECTION
DORES
SUBFLG 0012 18 FLAG: SUBSCRIPT REF. / USER
FUNC. CALL
INPFLG 0013 19 FLAG: $00 = INPUT, $40 = GET,
$98 = READ
DOMASK 0014 20
TANSGN FLAG: TAN SIGN / COM PARISON
RESULT
CHANNL 0015 21
POKER 0016 22
LINNUM TEMP INTEGER VALUE
TEMPPT 0018 24 POINTER: TEMP STRING STACK
LASTPT 0019 25 LAST TEM P STRING ADDRESS
TEMPST 001B 27 STACK FOR TEMP STRINGS
INDEX 0024 36 UTILITY POINTER AREA
INDEX 1
INDEX2 0026 38
RESHO 0028 40 FLOAT.PT. PRODUCT OF MULTIPLY
RESMOH 0029 41
ADDEND 002A 42
RESMO
RESLO 002B 43
TXTTAB 002D 45 POINTER: START OF BASIC TEXT
VARTAB 002F 47 POINTER: START OF BASIC
VARIABLES
ARYTAB 0031 49 POINTER: START OF BASIC ARRAYS
STREND 0033 51 POINTER: END OF BASIC ARRAYS
4- 1
FRETOP 0035 53 POINTER: BOTTOM OF STRING
STORAGE
FRESPC 0037 55 UTILITY STRING POINTER
M A X _M E M _1 0039 57 TOP OF STRING/VARIABLE BANK
(BANK 1)
CURLIN 003B 59 CURRENT BASIC LINE NUMBER
TXTPTR 003D 61 POINTER TO BASIC TEXT USED BY
CHRGET,ETC.
FORM 003F 63 USED BY PRINT USING
FNDPNT POINTER TO ITEM FOUND BY
SEARCH
504 COMMODORE 128
C l28 Memory M ap (continued)
M EM ORY
ADDRESS HEXADECIM AL DECIMAL

DATLIN 0041 65 CURRENT DATA LINE NUMBER
DATPTR 0043 67 CURRENT DATA ITEM ADDRESS
INPPTR 0045 69 VECTOR: INPUT ROUTINE
VARNAM 0047 71 CURRENT BASIC VARIABLE NAME
FDECPT 0049 73
VARPNT POINTER: CURRENT BASIC
VARIABLE DATA
LSTPNT 004B 75
FORPNT POINTER: INDEX VARIABLE FOR
FOR/NEXT
ANDMSK
EORMSK 004C 76
VARTXT 004D 77
OPPTR
OPMASK 004F 79
GRBPNT 0050 80
TEM PF3
DEFPNT
DSCPNT 0052 82
0054 84
HELPER 0055 85 FLAGS ‘HELP’ OR ‘LIST ’
JMPER 0056 86
0057 87
OLDOV 0058 88
TEMPF1 0059 89
PTARG1 M ULTIPLY DEFINED FOR INSTR
PTARG2 005B 91
STR1 005D 93
STR2 0060 96
POSITN 0063 99
MATCH 0064 100
ARYPNT 005A 90
HIGHDS
HIGHTR 005C 92
TEMPF2 005E 94
DECCNT 005F 95 NUMBER OF DIGITS AFTER THE
DECIMAL POINT
TENEXP 0060 96
TO ML M ONITOR Z .P. STORAGE IN
FAC
GRBTOP 0061 97
DPTFLG DECIMAL POINT FLAG
LOWTR
EXPSGN 0062 98
FAC 0063 99
MEMORY
DSCTMP
LEFT__FLAG PAINT-LEFT FLAG
FACEXP F A C #1E X P O N E N T
TI M ONITOR Z .P. STORAGE IN FAC
R IG H T_FLAG 0064 100 PAINT-RIGHT FLAG
FACHO F A C #1 MANTISSA
FACMOH 0065 101
INDICE 0066 102
FACMO
T2 M ONITOR Z .P. STORAGE IN FAC
FACLO 0067 103
FACSGN 0068 104 F A C #1 SIGN
DEGREE 0069 105
SGNFLG POINTER: SERIES-EVAL. CONSTANT
ARGEXP 006A 106 F A C #2E X P O N E N T
ARGHO 006B 107 F A C # 2 MANTISSA
ARGMOH 006C 108
IN IT _A S_0 JUST A COUNT FOR INIT
ARGMO 006D 109
ARGLO 006E 110
ARGSGN 006F 111 F A C # 2 SIGN
STRNG1 0070 112
ARISGN SIGN COM PARISON RESULT:
F A C #1 VS # 2
FACOV 0071 113 F A C #1 LOW -ORDER (ROUNDING)
STRNG2 0072 114
POLYPT
CURTOL
FBUFPT POINTER: CASSETTE BUFFER
AUTINC 0074 116 INC. VAL FOR AUTO (0 = OFF)
MVDFLG 0076 118 FLAG IF 10K HIRES ALLOCATED
Z _P _T E M P _1 0077 119 PRINT USING’S LEADING ZERO
COUNTER
M OVSPR & SPRITE TEM PORARY
M ID$ TEM PORARY
HULP 0078 120 COUNTER
KEYSIZ
SYNTMP 0079 121 USED AS TEM P FOR INDIRECT
LOADS
DSDESC 007A 122 DESCRIPTOR FOR DS$
TXTPTR MONITOR Z .P. STORAGE
TOS 007D 125 TOP OF RUN TIM E STACK
RUNMOD 007F 127 FLAGS RUN/DIRECT MODE
PARSTS 0080 128 DOS PARSER STATUS WORD
POINT USING’S POINTER TO DEC.PT
506 COMMODORE 128
C128 Memory Map (continued)
MEMORY

PARSTX 0081 129
OLDSTK 0082 130
BASIC Z-P STORAGE FOR GRAPHIC COMMANDS

COLSEL 0083 131 CURRENT COLOR SELECTED
M U L T IC O L O R _l 0084 132
MULTICOLOR—2 0085 133
FOREGROUND 0086 134
SC ALE_X 0087 135 SCALE FACTOR IN X
SC ALE_Y 0089 137 SCALE FACTOR IN Y
STOPNB 008B 139 STOP PAINT IF NOT BACKGROUND/
NOT SAME COLOR
GRAPNT 008C 140
VTEMP1 008E 142
VTEMP2 008F 143
KERNAL/EDITOR STORAGE
STATUS 0090 144 I/O OPERATION STATUS BYTE
STKEY 0091 145 STOP KEY FLAG
SVXT 0092 146 TAPE TEM PORARY
VERCK 0093 147 LOAD OR VERIFY FLAG
C3P0 0094 148 SERIAL BUFFERED CHAR FLAG
BSOUR 0095 149 CHAR BUFFER FOR SERIAL
SYNO 0096 150 CASSETTE SYNC #
XSAV 0097 151 TEMP FOR BASIN
LDTND 0098 152 INDEX TO LOGICAL FILE
DFLTN 0099 153 DEFAULT INPUT DEVICE #
DFLTO 009A 154 DEFAULT OUTPUT DEVICE #
PRTY 009B 155 CASSETTE PARITY
DPSW 009C 156 CASSETTE DIPOLE SWITCH
M SGFLG 009D 157 OS M ESSAGE FLAG
PTR1 009E 158 CASSETTE ERROR PASS1
TI TEMPORARY 1
PTR2 009F 159 CASSETTE ERROR PASS2
T2 TEMPORARY 2
TIME 00A0 160 24 HOUR CLOCK IN 1/60TH SECONDS
R2D2 00 A3 163 SERIAL BUS USAGE
PCNTR CASSETTE
BSOUR1 00A4 164 TEMP USED BY SERIAL ROUTINE
FIRT
COUNT 00A5 165 TEMP USED BY SERIAL ROUTINE
CNTDN CASSETTE SYNC COUNTDOW N
BUFPT 00A6 166 CASSETTE BUFFER POINTER
M EM ORY
INBIT 00A7 167 RS-232 RCVR INPUT BIT STORAGE

SHCNL CASSETTE SHORT COUNT
BITCI 00A8 168 RS-232 RCVR BIT COUNT IN
RER CASSETTE READ ERROR
RINONE 00A9 169 RS-232 RCVR FLAG FOR START BIT
CHECK
REZ CASSETTE READING ZEROES
RIDATA 00AA 170 RS-232 RCVR BYTE BUFFER
RDFLG CASSETTE READ MODE
RIPRTY 00AB 171 RS-232 RCVR PARITY STORAGE
SHCNH CASSETTE SHORT CNT
SAL OOAC 172 POINTER: TAPE BUFFER / SCREEN
SCROLLING
SAH OOAD 173
EAL OOAE 174 TAPE END ADDRESSES / END OF
PROGRAM
EAH 00 AF 175
CMPO 00B0 176 TAPE TIMING CONSTANTS
TEMP 0 0 B1 177
TAPE1 00B2 178 ADDRESS OF TAPE BUFFER
BITTS 00B4 180 RS-232 TRNS BIT COUNT
SNSW1
NXTBIT 00B5 181 RS-232 TRNS NEXT BIT TO BE SENT
DIFF
RODATA 00 B6 182 RS-232 TRNS BYTE BUFFER
PRP
FNLEN 00B7 183 LENGTH CURRENT FILE N STR
LA 00B8 184 CURRENT FILE LOGICAL ADDR
SA 00B9 185 CURRENT FILE 2ND ADDR
FA 00BA 186 CURRENT FILE PRIMARY ADDR
FNADR 00BB 187 ADDR CURRENT FILE NAM E STR
ROPRTY 00BD 189 RS-232 TRNS PARITY BUFFER
OCHAR
FSBLK 00BE 190 CASSETTE READ BLOCK COUNT
DRIVE OOBF 191
MYCH SERIAL WORD BUFFER
CAS1 00C0 192 CASSETTE M ANUAL/CNTRLED
SW ITCH (UPDATED DURING IRQ)
TRACK 00C1 193
STAL I/O START ADDRESS (LO)
SECTOR 00C2 194
STAH I/O START ADDRESS (HI)
MEMUSS 00C3 195 CASSETTE LOAD TEM PS (2 BYTES)
TMP2
DATA 00C5 197 TAPE READ/W RITE DATA
508 COMMODORE 128
MEMORY
BA 00 C6 198 BANK FOR CURRENT LOAD/SAVE/
VERIFY OPERATION
FNBANK 00C7 199 BANK WHERE CURRENT FN IS
FOUND (AT ‘FN A D R ’)
RIBUF 00C8 200 RS-232 INPUT BUFFER POINTER
ROBUF 00CA 202 RS-232 OUTPUT BUFFER POINTER
GLOBAL SCREEN EDITOR VARIABLES

KEYTAB oocc 204 KEYSCAN TABLE POINTER
IMPARM 00CE 206 PRIMM UTILITY STRING POINTER
NDX 00D0 208 INDEX TO KEYBOARD QUEUE
KYNDX 00D1 209 PENDING FUNCTION KEY FLAG
KEYIDX 00D2 210 INDEX INTO PENDING FUNCTION
KEY STRING
SHFLAG 00D3 211 KEYSCAN SHIFT KEY STATUS
SFDX 00D4 212 KEYSCAN CURRENT KEY INDEX
LSTX 00D5 213 KEYSCAN LAST KEY INDEX
CRSW 00D6 214 < C R > INPUT FLAG
MODE 00D7 215 40/80 COLUMN M ODE FLAG
GRAPHM 00D8 216 TEXT/GRAPHIC MODE FLAG
CHAREN 00D9 217 RAM/ROM VIC CHARACTER FETCH
FLAG (BIT-2)
THE FOLLOW ING LOCATIONS ARE SHARED BY SEVERAL EDITOR ROUTINES.

SEDSAL 00DA 218 POINTERS FOR MOVLIN
BITM SK 00DA 218 TEMPORARY FOR TAB & LINE
WRAP ROUTINES
SAVER 00DB 219 ANOTHER TEM PORARY PLACE TO
SAVE A REG.
SEDEAL 00DC 220
SEDT1 00DE 222 SAVPOS
SEDT2 00DF 223
KEYSIZ 00DA 218 PROGRAMMABLE KEY VARIABLES
KEYLEN 00DB 219
KEYNUM 00DC 220
KEYNXT 00DD 221
KEYBNK 00DE 222
KEYTMP 00DF 223
LOCAL SCREEN EDITOR VARIABLES.

THESE ARE SW APPED OUT TO $0A40
WHEN SCREEN (40/80) MODE CHANGES.
PNT 00E0 224 POINTER TO CURRENT LINE (TEXT)
USER 00E2 226 POINTER TO CURRENT LINE
(ATTRIBUTE)
MEMORY
LABEL ADDRESSADDRESS DESCRIPTION
LOCAL SCREEN EDITOR VARIABLES.

THESE ARE SWAPPED OUT TO $0A40
WHEN SCREEN (40/80) MODE CHANGES
SCBOT 00E4 228 WINDOW LOWER LIMIT

SCTOP 00E5 229 WINDOW UPPER LIMIT
SCLF 00E6 230 WINDOW LEFT MARGIN
SCRT 00E7 231 WINDOW RIGHT MARGIN
LSXP 00E8 232 CURRENT INPUT COLUMN START
LSTP 00E9 233 CURRENT INPUT LINE START
INDX 00EA 234 CURRENT INPUT LINE END
TBLX 00EB 235 CURRENT CURSOR LINE
PNTR 00EC 236 CURRENT CURSOR COLUMN
LINES OOED 237 MAXIMUM NUMBER OF SCREEN
LINES
COLUMNS 00EE 238 MAXIMUM NUMBER OF SCREEN
COLUMNS
DATAX 00EF 239 CURRENT CHARACTER TO PRINT
LSTCHR 00F0 240 PREVIOUS CHAR PRINTED (FOR
<ESC > TEST)
COLOR 00F1 241 CURR ATTRIBUTE TO PRINT
(DEFAULT FGND COLOR)
TCOLOR 00F2 242 SAVED ATTRIB TO PRINT (‘INSERT’
& ‘DELETE’)
RVS OOF3 243 REVERSE MODE FLAG
QTSW 00F4 244 QUOTE MODE FLAG
INSRT 00F5 245 INSERT MODE FLAG
INSFLG 00F6 246 AUTO-INSERT MODE FLAG
LOCKS 00F7 247 DISABLES < S H IF T > < C'>,
<CTRL>S
SCROLL 00F8 248 DISABLES SCREEN SCROLL, LINE
LINKER
BEEPER 00F9 249 DISABLES <CTRL> G
FREKZP 00FA 250 FREE ZERO PAGE RESERVED FOR
APPLICATIONS SOFTWARE
($FA-$FE)
LOFBUF 00FF 255
BASIC/DOS INTERFACE VARS
BAD 0100 256 TAPE READ ERRORS

FBUFFR AREA TO BUILD FILENAME IN
(16 BYTES)
XCNT 0110 272 DOS LOOP COUNTER
510 COMMODORE 128
MEMORY
DOSF1L 0111 273 DOS FILENAM E 1 LEN

DOSDS1 0112 274 DOS DISK DRIVE 1
DOSF2L 0113 275 DOS FILENAM E 2 LEN
DOSDS2 0114 276 DOS DISK DRIVE 2
DOSF2A 0115 277 DOS FILENAME 2 ADDR
DOSOFL 0117 279 BLOAD/BSAVE STARTING ADDRESS
DOSOFH 0119 281 ....A N D ENDING ADDRESS
DOSLA 011B 283 DOS LOGICAL ADDR
DOSFA one 284 DOS PHYS ADDR
DOSSA 011D 285 DOS SEC. ADDR
DOSRCL 011E 286 DOS RECORD LENGTH
DOSBNK 011F 287
DOSDID 0120 288 DOS DISK ID
DIDCHK 0122 290 DOS DSK ID FLG
SPACE USED BY PRINT USING
BNR 0123 291 POINTER TO BEGIN. NO.
ENR 0124 292 POINTER TO END NO.
DOLR 0125 293 DOLLAR FLAG
FLAG 0126 294 COMMA FLAG
SWE 0127 295 COUNTER
USGN 0128 296 SIGN EXPONENT
UEXP 0129 297 POINTER TO EXPONENT
VN 012A 298 # OF DIGITS BEFORE DECIMAL
POINT
CHSN 012B 299 JUSTIFY FLAG
VF 012C 300 # OF POS BEFORE DECIMAL POINT
(FIELD)
NF 012D 301 # OF POS AFTER DECIM AL POINT
(FIELD)
POSP 012E 302 + / - FLAG (FIELD)
FESP 012F 303 EXPONENT FLAG (FIELD)
ETOF 0130 304 SWITCH
CFORM 0131 305 CHAR COUNTER (FIELD)
SNO 0132 306 SIGN NO
BLFD 0133 307 BLANK/STAR FLAG
BEGFD 0134 308 POINTER TO BEGIN OF FIELD
LFOR 0135 309 LENGTH OF FORMAT
ENDFD 0136 310 POINTER TO END OF FIELD
SYSTK 0137 311 SYSTEM STACK ($0137-$01FF)
BUF 0200 512 INPUT BUFFER: BASIC & MONITOR
($ 0 2 0 0 - $ 0 2 A l)
THE COMMODORE 128 AN D COMMODORE 64 MEMORY MAPS 511
MEMORY
FETCH 02A2 674 L D A (-),Y FROM ANY BANK

FETVEC 02AA 682
STASH 02AF 687 ST A (-),Y TO ANY BANK
STAVEC 02B9 697
CMPARE 02BE 702 C M P(-),Y TO ANY BANK
CMPVEC 0 2 C8 712
JSRFAR 02CD 716 JSR XXXX TO ANY BANK & RETURN
JMPFAR 02E3 739 JM P XXXX TO ANY BANK
VECTORS
E SC _F N _V E C 02FC 764 VECTOR FOR ADDITIONAL

FUNCTION ROUTINES
BNKVEC 02FE 766 VECTOR FOR FUNCTION CART.
USERS
IERROR 0300 768 VECTOR FOR PRINT BASIC ERROR
(ERR IN .X)
IMAIN 0302 770 VECTOR TO MAIN (SYSTEM DIRECT
LOOP)
ICRNCH 0304 772 VECTOR TO CRUNCH
(TOKENIZATION ROUTINE)
IQPLOP 0306 774 VECTOR TO LIST BASIC TEXT
(CHAR LIST)
IGONE 0308 776 VECTOR TO GONE (BASIC CHAR
DISPATCH)
IEVAL 030A 778 VECTOR TO BASIC TOKEN
EVALUATION
IESCLK 030C 780 VECTOR TO ESCAPE-TOKEN
CRUNCH,
IESCPR 030E 782 ...L IST ,
IESCEX 0310 784 ...A N D EXECUTE.
IIRQ 0314 788 IRQ RAM VECTOR
CINV
IBRK 0316 790 BRK INSTR RAM VECTOR
CBINV
INMI 0318 792 NMI VECTOR
IOPEN 031A 794 KERNAL OPEN ROUTINE VECTOR
ICLOSE 031C 796 KERNAL CLOSE ROUTINE VECTOR
ICHKIN 031E 798 KERNAL CHKIN ROUTINE VECTOR
ICKOUT 0320 800 KERNAL CHKOUT ROUTINE
VECTOR
ICLRCH 0322 802 KERNAL CLRCHN ROUTINE
VECTOR
512 COMMODORE 128
MEMORY
VECTORS
IBASIN 0324 804 KERNAL CHRIN ROUTINE VECTOR

IBSOUT 0326 806 KERNAL CHROUT ROUTINE
VECTOR
ISTOP 0328 808 KERNAL STOP ROUTINE VECTOR
IGETIN 032A 810 KERNAL GETIN ROUTINE VECTOR
ICLALL 032C 812 KERNAL CLALL ROUTINE VECTOR
EXMON 032E 814 M ONITOR COMMAND VECTOR
ILOAD 0330 816 KERNAL LOAD ROUTINE VECTOR
ISAVE 0332 818 KERNAL SAVE ROUTINE VECTOR
EDITOR INDIRECT VECTORS
CTLVEC 0334 820 EDITOR: PRINT ‘CONTRL’ INDIRECT

SHFVEC 0336 822 EDITOR: PRINT ‘SHIFTD’ INDIRECT
ESCVEC 0338 824 EDITOR: PRINT ‘ESCAPE’ INDIRECT
KEYVEC 033A 826 EDITOR: KEYSCAN LOGIC INDIRECT
KEYCHK 033C 828 EDITOR: STORE KEY INDIRECT
DECODE 033E 830 VECTORS TO KEYBOARD MATRIX
DECODE TABLES
KEYD 034A 842 IRQ KEYBOARD BUFFER (10 BYTES)
TABMAP 0354 852 BITM AP OF TAB STOPS (10 BYTES,

$0354-D)
BITABL 035E 862 BITM AP OF LINE WRAPS
TABM AP AND BITABL GET
SW APPED TO $0A60 W HEN SCREEN
40/80 MODE IS CHANGED.
LAT 0362 866 LOGICAL FILE NUMBERS
FAT 036C 876 PRIM ARY DEVICE NUM BERS
SAT 0376 886 SECONDARY ADDRESSES
CHRGET 0380 896
CHRGOT 0386 902
QNUM 0390 912
INDIRECT LOAD SUBROUTINE AREA
INDSUB_RAMO 039F 927 SHARED ROM FETCH SUB

INDSUB_RAM 1 03AB 939 SHARED ROM FETCH SUB
INDIN1_RAM 1 03B7 950 INDEX 1 INDIRECT FETCH
INDIN2 03C0 959 INDEX2 INDIRECT FETCH
INDTXT 03C9 968 TXTPTR
MEMORY
INDIRECT LOAD SUBROUTINE AREA
ZERO 03D2 977 NUMERIC CONSTANT FOR BASIC

C UR RENT_ 03D5 979 CONTEXT FOR SY S,PO K E,PEEK
BANK FROM BANK CMMD
TMPDES 03D6 980 TEMP FOR INSTR
FIN _BA N K 03DA 984 BANK POINTER FOR STRING/
NUM BER CONVERT RTN
SAVSIZ 03DB 985 TEMP WORK LOCATIONS FOR
SSHAPE
BITS 03DF 989 F A C #1 OVERFLOW DIGIT
SPRTM P_1 03E0 990 TEMP FOR SPRSAV
SPR TM P_2 03E1 991
FG _B G 03E2 992 PACKED FOREGROUND/
BACKGROUND COLOR NYBBLES
FG _M C 1 03E3 993 PACKED FOREGROUND/
M ULTICOLOR 1 COLOR NYBBLES
PAGE FOUR & HIGHER DECLARATIONS

(BEGINNING OF BANKABLE RAM )
VICSCN 0400 1024 VIDEO MATRIX # 1: VIC 40-COLUMT>
TEXT SCREEN
$0400-$07FF
0800 2048 BASIC RUN-TIM E STACK (512
BYTES)
$0800-$09FF
ABSOLUTE KERNAL VARIABLES
SY ST E M _ VECTOR 0A00 2560 VECTOR TO RESTART SYSTEM

(BASIC WARM)
DEJAVU 0A02 2562 KERNAL W ARM /COLD IN IT ’N
STATUS BYTE
PALNTS 0A03 2563 PAL/NTSC SYSTEM FLAG
IN IT_STATUS 0A04 2564 FLAGS RESET VS. NMI STATUS
FOR IN IT’N RTNS
MEMSTR 0A05 2565 PTR TO BOTTOM OF AVAIL.
M EMORY IN SYSTEM BANK
MEMSIZ 0A07 2567 PTR TO TOP OF AVAILABLE
M EMORY IN SYSTEM BANK
IRQTMP 0A09 2569 TAPE HANDLER PRESERVES IRQ
INDIRECT HERE
CASTON 0A0B 2571 TOD SENSE DURING TAPE
OPERATIONS
KIKA26 0A0C 2572 TAPE READ TEM PORARY
STUPID 0A0D 2573 TAPE READ D1IRQ INDICATOR
TIMOUT 0A0E 2574 FAST SERIAL TIM EOUT FLAG
ENABL 0A0F 2575 RS-232 ENABLES
514 COMMODORE 128
M EM ORY
ABSOLUTE KERNAL VARIABLES

M51CTR 0A10 2576 RS-232 CONTROL REGI STER
M51CDR 0A11 2577 RS-232 COMMAND REGISTER
M51AJB 0A12 2578 RS-232 USER BAUD RATE
RSSTAT 0A14 2580 RS-232 STATUS REGISTER
BITNUM 0A15 2581 RS-232 NUMBER OF BITS TO SEND
BAUDOF 0A16 2582 RS-232 BAUD RATE FULL BIT TIME
(CREATED BY OPEN)
RIDBE 0A18 2584 RS-232 INPUT BUFFER INDEX TO
END
RIDBS 0A19 2585 RS-232 INPUT BUFFER INDEX TO
START
RODBS 0A1A 2586 RS-232 OUTPUT BUFFER INDEX TO
START
RODBE 0A1B 2587 RS-232 OUTPUT BUFFER INDEX TO
END
SERIAL 0A1C 2588 FAST SERIAL INTERNAL/EXTERNAL
FLAG
TIMER 0A1D 2589 DECREMENTING JIFFIE REGISTER
GLOBAL ABSOLUTE SCREEN EDITOR DECLARATIONS

XMAX 0A20 2592 KEYBOARD QUEUE MAXIMUM SIZE
PAUSE 0A21 2593 <C TR L>SFLA G
RPTFLG 0A22 2594 ENABLE KEY REPEATS
KOUNT 0A23 2595 DELAY BETW EEN KEY REPEATS
DELAY 0A24 2596 DELAY BEFORE A KEY STARTS
REPEATING
LSTSHF 0A25 2597 DELAY BETW EEN < 0 > < S H F T >
TOGGLES
BLNON 0A26 2598 VIC CURSOR M ODE (BLINKING,
SOLID)
BLNSW 0A27 2599 VIC CURSOR DISABLE
BLNCT 0A28 2600 VIC CURSOR BLINK COUNTER
GDBLN 0A29 2601 VIC CURSOR CHARACTER BEFORE
BLINK
GDCOL 0A2A 2602 VIC CURSOR COLOR BEFORE BLINK
CURMOD 0A2B 2603 VDC CURSOR M ODE (WHEN
ENABLED)
VM1 0A2C 2604 VIC TEXT SCREEN/CHARACTER
BASE POINTER
VM2 0A2D 2605 VIC BIT MAP BASE POINTER
VM3 0A2E 2606 VDC TEXT SCREEN BASE
VM4 0A2F 2607 VDC ATTRIBUTE BASE
LINTM P 0A30 2608 TEM PORARY POINTER TO LAST
LINE FOR LOOP4
MEMORY
GLOBAL ABSOLUTE SCREEN EDITOR DECLARATIONS
SAV80A 0A31 2609 TEMPORARY FOR 8 O-COL ROUTINES

SAV80B 0A32 2610 TEMPORARY FOR 8 O-COL ROUTINES
CURCOL 0A33 2611 VDC CURSOR COLOR BEFORE
BLINK
SPLIT 0A34 2612 VIC SPLIT SCREEN RASTER VALUE
FNADRX 0A35 2613 SAVE .X DURING BANK
OPERATIONS
PALCNT 0A36 2614 COUNTER FOR PAL SYSTEMS
(JIFFIE ADJUSTMENT)
SPEED 0A37 2615 SAVE SYSTEM SPEED DURING TAPE
AND SERIAL OPS
SPRITES 0A38 2616 SAVE SPRITE ENABLES DURING
TAPE AND SERIAL OPS
BLANKING 0A39 2617 SAVE BLANKING STATUS DURING
TAPE OPS
H O LD_O FF 0A3A 2618 FLAG SET BY USER TO RESRV
FULL CNTRL OF VIC
LDTB1_SA 0A3B 2619 HI BYTE: SA OF VIC SCRN (USE
W/VM1 TO MOVE SCRN)
C L R _E A _L O 0A3C 2620 8563 BLOCK FILL
C L R _E A _H I 0A3D 2621 8563 BLOCK FILL
0A40 2624 $0A4ft-$0A7F RESERVED SW AP
AREA FOR SCREEN VARIABLES
WHEN (40/80) M ODE CHANGES
M ONITOR’S DOMAIN
XCNT OA8 O 2688 COMPARE BUFFER (32 BYTES)
HULP OAAO 2720
FORM AT OAAA 2730
LENGTH OAAB 2731 ASM/DIS
MSAL OAAC 2732 FOR ASSEMBLER
SXREG OAAF 2735 1 BYTE TEMP USED ALL OVER
SYREG OABO 2736 1 BYTE TEMP USED ALL OVER
WRAP 0AB1 2737 1 BYTE TEMP FOR ASSEM BLER
XSAVE 0AB2 2738 SAVE .X HERE DURING INDIRECT
SUBROUTINE CALLS
DIRECTION 0AB3 2739 DIRECTION INDICATOR FOR
TRANSFER’
COUNT 0AB4 2740 PARSE NUMBER CONVERSION
NUMBER 0AB5 2741 PARSE NUMBER CONVERSION
SHIFT 0AB6 2742 PARSE NUMBER CONVERSION
TEMPS 0AB7 2743
FUNCTION KEY ROM CARD TABLES
CURBNK OACO 2752 CURRENT FUNCTION KEY ROM

BANK BEING POLLED
516 COMMODORE 128
MEMORY
FUNCTION KEY ROM CARD TABLES
PAT 0AC1 2753 PHYSICAL ADDRESS TABLE(IDS OF

LOGGED-IN CARDS)
D K _FLA G 0AC5 2757 RESERVED FOR FOREIGN SCREEN
EDITORS
0AC6 2758 $0AC6-$0AFF RESERVED FOR
SYSTEM
TBUFFR OBOO 2810 CASSETTE BUFFER (192 BYTES)
$0B 00-$0BC 0, THIS PAGE ALSO
USED AS A BUFFER FOR THE DISK
AUTO-BOOT
RS232I 0C00 3072 RS-232 INPUT BUFFER
R S 2320 0D00 3328 RS-232 OUTPUT BUFFER
0E00 3584 SPRITE DEFINITION AREA (MUST
BE BELOW $1000)
$0E00-$0FFF, 512 BYTES
PKYBUF 1000 4096 PROGRAM M ABLE FUNCTION KEY
LENGTHS TABLE FOR 10 KEYS
(F 1-F 8, <SH IF T R U N > , HELP)
PKYDEF 100 A 4106 PROGRAM M ABLE FUNCTION KEY
STRINGS
DOS/VSP AREA
DOSSTR 1100 4352 DOS OUTPUT STR. BUF
48 BYTES TO BUILD DOS STRING
VW ORK 1131 4401 GRAPHICS VARS
XYPOS 1131 4401
XPOS 1131 4401 CURRENT X POSITION
YPOS 1133 4403 CURRENT Y POSITION
XDEST 1135 4405 X-COORDINATE DESTINATION
YDEST 1137 4407 Y-COORDINATE DESTINATION
XYABS 1139 4409 LINE DRAW ING VARIABLES
XABS 1139 4409
YABS 113B 4411
XYSGN 113D 4413
XSGN 113D 4413
YSGN 113F 4415
FCT 1141 4417
ERRVAL 1145 4421
LESSER 1147 4423
GREATR 1148 4424
ANGLE ROUTINE VARIABLES
ANGSGN 1149 4425 SIGN OF ANGLE
SINVAL 114A 4426 SINE OF VALUE OF ANGLE
MEMORY
DOS/VSP AREA
COSVAL 114C 4428 COSINE OF VALUE OF ANGLE
ANGCNT 114E 4430 TEMPS FOR ANGLE DISTANCE
ROUTINES
BASIC GRAPHIC VARIABLES.

THE FOLLOW ING 24 BYTES ARE MULTIPLY DEFINED.
CIRCLE DRAW ING VARIABLES
XCIRCL 1150 4432 CIRCLE CENTER, X COORDINATE
YCIRCL 1152 4434 CIRCLE CENTER, Y COORDINATE
XRADUS 1154 4436 X RADIUS
YRADUS 1156 4438 Y RADIUS
ROTANG 1158 4440 ROTATION ANGLE
ANGBEG 115C 4444 ARC ANGLE START
ANGEND 115E 4446 ARC ANGLE END
XRCOS 1160 4448 X RADIUS * COS(ROTATION ANGLE)
YRSIN 1162 4450 Y RADIUS * SIN(ROTATION ANGLE)
XRSIN 1164 4452 X RADIUS * SIN(ROTATION ANGLE)
YRCOS 1166 4454 Y RADIUS * COS(ROTATION ANGLE)
BASIC GENERAL USE PARAMETERS

XCENTR 1150 4432
YCENTR 1152 4434
XDIST1 1154 4436
YDIST1 1156 4438
XDIST2 1158 4440
YDIST2 115A 4442
DISEEND 115C 4444 PLACEHOLDER
COLCNT 115E 4446 CHAR’S COL. COUNTER
ROW CNT 115F 4447
STRCNT 1160 4448
BOX-DRAW ING VARIABLES
XCORD1 1150 4432 POINT 1 X-COORD.
YCORD1 1152 4434 POINT 1 Y-COORD.
BOXANG 1154 4436 ROTATION ANGLE
XCOUNT 1156 4438
YCOUNT 1158 4440
BXLENG 115A 4442 LENGTH OF A SIDE
XCORD2 115C 4444
YCORD2 115E 4446
SHAPE AND MOVE-SHAPE VARIABLES

KEYLEN 1151 4433
KEYNXT 1152 4434
STRSZ 1153 4435 STRING LEN
518 COMMODORE 128
C128 M emory M ap (continued)
M EM ORY
SHAPE AND M OVE-SHAPE VARIABLES
GETTYP 1154 4436 REPLACE SHAPE MODE

STRPTR 1155 4437 STRING PO S’N COUNTER
OLDBYT 1156 4438 OLD BIT M AP BYTE
NEW BYT 1157 4439 NEW STRING OR BIT M AP BYTE
1158 4440 PLACEHOLDER
XSIZE 1159 4441 SHAPE COLUMN LENGTH
YSIZE 115B 4443 SHAPE ROW LENGTH
XSAVE 115D 4445 TEM P FOR COLUM N LENGTH
STRADR 115F 4447 SAVE SHAPE STRING DESCRIPTOR
BITIDX 1161 4449 BIT INDEX INTO BYTE
BASIC GRAPHIC VARIABLES
CHRPAG 1168 4456 HIGH BYTE: ADDR OF CHARROM

FOR ‘CHAR’ CMD.
BITCNT 1169 4457 TEMP FOR GSHAPE
SCALEM 116A 4458 SCALE MODE FLAG
WIDTH 116B 4459 DOUBLE W IDTH FLAG
FILFLG 116C 4460 BOX FILL FLAG
BITM SK 116D 4461 TEM P FOR BIT MASK
NUMCNT 116E 4462
TRCFLG 116F 4463 FLAGS TRACE MODE
R EN U M _T M P_1 1170 4464 A TEM P FOR RENUMBER
R EN U M _T M P_2 1172 4466 A TEM P FOR RENUM BER
T3 1174 4468
T4 1175 4469
VTEMP3 1177 4471 GRAPHIC TEM P STORAGE
VTEMP4 1178 4472
VTEMP5 1179 4473
ADRAY1 117 A 4474 PTR TO ROUTINE: CONVERT
FLOAT -* INTEGER
ADRAY2 117C 4476 PTR TO ROUTINE: CONVERT
INTEGER -> FLOAT
SPRITE_DATA 117E 4478 SPRITE SPEED/DIRECTION TABLES
($117E-D5)
V IC_SA VE 11D6 4566 COPY OF VIC R EG ’S, USED TO
UPDATE CHIP DURING RETRACE
(21 BYTES, $11D6--EA)
UPPER_LOW ER 11EB 4587 POINTER TO UPPER/LOW ER CHAR
SET FOR CHAR
UPPER_G RAPHIC 11EC 4588 PTR. TO UPPER/GRAPHIC CHAR. SET
DOSSA 11ED 4589 TEMP STORAGE FOR FILE SA
DURING RECORD CMD
MEMORY
BASIC GENERAL NON-ZP STORAGE

OLDLIN 1200 4608 PREVIOUS BASIC LINE NUMBER
OLDTXT 1202 4610 POINTER: BASIC STATEM ENT FOR
CONTINUE
PRINT USING DECLARATIONS

PUCHRS 1204 4612
PUFILL 1204 4612 PRINT USING FILL SYMBOL
PUCOMA 1205 4613 PRINT USING COM M A SYMBOL
PUDOT 1206 4614 PRINT USING D .P. SYMBOL
PUMONY 1207 4615 PRINT USING MONETARY SYM BOL
ERRNUM 1208 4616 USED BY ERROR TRAPPING
ROUTINE-LAST ERR NO
ERRLIN 1209 4617 LINE # OF LAST ERROR— $FFFF IF
NO ERROR
TRAPNO 120B 4619 LINE TO GO TO ON ERROR. $FFXX
IF NONE SET
TM PTRP 120D 4621 HOLD TRAP # TEM POR.
ERRTXT 120E 4622
T EX T _TO P 1210 4624 TOP OF TEXT POINTER
M A X _M E M _0 1212 4626 HIGHEST ADDRESS AVAILABLE
TO BASIC IN RAM 0
TMPTXT 1214 4628 USED BY DO-LOOP. COULD BE
M ULT. ASSIGNED
TM PLIN 1216 4630
USRPOK 1218 4632
RNDX 121B 4635
CIRCLE . SEGMENT 1220 4640 DEGREES PER CIRCLE SEGM ENT
DEJAVU 1221 4641 ‘COLD’ OR ‘W A RM ’ RESET STATUS
BASIC STORAGE FOR MUSIC VECTORS

TEM PO _RA TE 1222 4642
VOICES 1223 4643
NTIME 1229 4649
OCTAVE 122B 4651
SHARP 122C 4652
PITCH 122D 4653
VOICE 122F 4655
WAVE0 1230 4656
DNOTE 1233 4659
FLTSAV 1234 4660
FLTFLG 1238 4664
NIBBLE 1239 4665
TONNUM 123A 4666
520 COMMODORE 128
MEMORY
BASIC STORAGE FOR MUSIC VECTORS

TONVAL 123B 4667
PARCNT 123E 4668
ATKTAB 123F 4669
SUSTAB 1249 4681
WAVTAB 1253 4691
PULSLW 125D 4701
PULSHI 1267 4711
FILTERS 1271 4721
INTERRUPT VECTORS
IN T _T R IP _F L A G 1276 4726
IN T _A D R _L O 1279 4729
IN T _A D R _H I 127C 4732
INTVAL 127F 4735
COLTYP 1280 4736
BASIC SOUND COMMAND VARS

SO U N D _V OICE 1281 4737
SO U N D _T IM E _L O 1282 4738
SO U N D _TIM E_H I 1285 4741
SO U N D _M A X _L O 1288 4744
SO U N D _M A X _H I 128B 4747
SO U N D _M IN _L O 128E 4750
SO U N D _M IN _H I 1291 4753
SOUND_DIRECTION 1294 4756
SO U N D _ST E P_L O 1297 4759
SO U N D _ST E P_H I 129 A 4762
SO U N D _FR E Q _LO 129D 4765
SO U N D _FR E Q _H I 12A0 4768
T EM P _T IM E _L O 12A3 4771
TEMP_TIMF, HI 12A4 4772
T E M P_M A X _L O 12A5 4773
T E M P _M A X _H I 12A6 4774
T EM P_M IN _L O 12A7 4775
T E M P _M IN _H I 12A8 4776
TEM P_DIRECTION 12A9 4777
T EM P_ST E P_L O 12AA 4778
T EM P_ST E P_H I 12AB 4779
T E M P_FR E Q _L O 12AC 4780
T E M P _F R E Q _H I 12AD 4781
T E M P_PU L SE _L O 12AE 4782
T EM P_PU L SE _H I 12AF 4783
TEM P_W AVEFORM 12B0 4784
MEMORY
BASIC SOUND COMMAND VARS

P O T _T E M P _l 12B1 4785 TEM PORARIES FOR ‘PO T ’
FUNCTION
P O T _T E M P _2 12B2 4786
W INDO W _TEM P 12B3 4787
SAVRAM 12B7 4791 USED BY SPRDEF & SAVSPR
DEFMOD 12FA 4858 USED BY SPRDEF & SAVSPR
LINCNT 12FB 4859 USED BY SPRDEF & SAVSPR
SPRITE_NUM BER 12FC 4860 USED BY SPRDEF & SAVSPR
IR Q _W R A P_FL A G 12FD 4861 USED BY BASIC IRQ TO BLOCK
ALL BUT ONE IRQ CALL
1300 4864 APPLICATION PROGRAM AREA
$1300-$1BFF
RAMBOT 1C00 7168 START OF BASIC TEXT $1C 00-
$EFFF (KERNAL SETS MEMBOT
HERE)
OR
1C00 7168 VIDEO MATRIX # 2 (1KB OF
COLORS FOR BITM AP, IF
ALLOCATED) $1C 00-$1FFF
2000 8192 VIC BITM AP (8KB, IF ALLOCATED)
$2000-$3FFF
BEGINNING OF ROM OVER RAM

4000 16384 C128 BASIC LO ROM
START OF BASIC TEXT IF BIT MAP
IS ALLOCATED (RAM)
$4000-$EFFF
8000 32768 C128 BASIC HI ROM (OR FUNCTION
ROM ) $8000-$BFFF
BASIC JUMP TABLE

MEMORY
BASIC ENTRY
JM P H A RD_RESET 4000 16384 COLD ENTRY
JM P SOFT_RESET 4003 16387 WARM ENTRY
JM P BASIC_IR Q 4006 16390 IRQ ENTRY
FORMAT CONVERSIONS
JM P AYINT AFOO 44800 CONVERT F.P. TO INTEGER
JM P GIVAYF AF03 44803 CONVERT INTEGER TO F.P.
522 COMMODORE 128
Basic Jump Table (continued)
M EM ORY
FORMAT CONVERSIONS
JM P FOUT AF06 44806 CONVERT F .P . TO ASCII STRING

JM P VAT, I AF09 44809 CONVERT ASCII STRING TO F.P.
JM P GETADR AFOC 44812 CONVERT F .P . TO AN ADDRESS
JM P FLOATC AFOF 44815 CONVERT ADDRESS TO F.P.
MATH FUNCTIONS
JM P FSUB AF12 44818 MEM - FACC
JM P FSUBT AF15 44821 ARG - FACC
JM P FADD AF18 44824 MEM + FACC
JM P FADDT AF1B 44827 ARG - FACC
JM P FMULT AF1E 44830 M EM * FACC
JM P FMULTT AF21 44833 ARG * FACC
JM P FDIV AF24 44836 M EM / FACC
JM P FDIVT AF27 44839 A R G / FACC
JM P LOG AF2A 44842 COMPUTE NATURAL LOG OF FACC
JM P INT AF2D 44845 PERFORM BASIC INT ON FACC
JM P SQR AF30 44848 COMPUTE SQUARE ROOT OF FACC
JM P NEGOP AF33 44851 NEGATE FACC
JM P FPWR AF36 44854 RAISE ARG TO THE MEM POWER
JM P FPWRT AF39 44857 RAISE ARG TO THE FACC POW ER
JM P EXP AF3C 44860 COM PUTE EXP OF FACC
JM P COS AF3F 44863 COM PUTE COS OF FACC
JM P SIN AF42 44866 COMPUTE SIN OF FACC
JM P TAN AF45 44869 COMPUTE TAN OF FACC
JM P ATN AF48 44872 COM PUTE ATN OF FACC
JM P ROUND AF4B 44875 ROUND FACC
JMP ABS AF4E 44878 ABSOLUTE VALUE OF FACC
JM P SIGN AF51 44881 TEST SIGN OF FACC
JMP FCOM P AF54 44884 COMPARE FACC WITH MEM
JM P RND 0 AF57 44887 GENERATE RANDOM F.P. NUMBER
MOVEMENT
JM P CONUPK AF5A 44890 M OVE RAM MEM TO ARG
JM P ROMUPK AF5D 44893 M OVE ROM MEM TO ARG
JM P MOVFRM AF60 44896 M OVE RAM MEM TO FACC
JM P MOVFM AF63 44899 M OVE ROM MEM TO FACC
JM P MOVMF AF66 44902 M OVE FACC TO MEM
JM P MOVFA AF69 44905 M OVE ARG TO FACC
JM P MOVAF AF6C 44908 M OVE FACC TO ARG
OTHER BASIC ROUTINES

JM P OPTAB AF6F 44911
JM P DRAWLN AF72 44914
M EM ORY
ADDRESS HEXADECIM AL D ECIM AL
OTHER BASIC ROUTINES
JM P GPLOT AF75 44917

JM P CIRSUB AF78 44920
JM P RUN AF7B 44923
JM P RUNC AF7E 44926
JM P CLEAR AF81 44929
JM P NEW AF84 44932
JM P LNKPRG AF87 44935
JM P CRUNCH AF8A 44938
JM P FNDLIN AF8D 44941
JM P NEWSTT AF90 44944
JM P EVAL AF93 44947
JM P FRMEVL AF96 44950
JM P R U N _A _
PROGRAM AF99 44953
JM P SETEXC AF9C 44956
JM P LINGET AF9F 44959
JM P GARBA2 AFA2 44962
JM P E X E C U T E _A _
LINE AFA5 44965
M ONITOR ENTRY
JMP CALL BOOO 45056 MONITOR CALL ENTRY
JMP BREAK B003 45059 M ONITOR BREAK ENTRY
JMP MONCMD B006 45062 MONITOR COM M AND PARSER
ENTRY
cooo 49152 KERNAL (OR FUNCTION) ROM
$C 000-$FFFF
EDITOR JUMP TABLE

M EM ORY
JM P CINT COOO 49152 INITIALIZE EDITOR & SCREEN

JM P DISPLY C003 49155 DISPLAY CHARAC IN .A , COLOR
IN .X
JM P LP2 C006 49158 GET KEY FROM IRQ BUFFER
INTO .A
JM P LOOP5 C009 49161 GET A CHR FROM SCRN LINE
INTO.A
JM P PRINT COOC 49164 PRINT CHARACTER IN .A
JM P SCRORG COOF 49167 GET # OF SCRN ROW S, COLS
INTO X&Y
524 COMMODORE 128
Editor Jump Table (continued)
M EM ORY
JM P SCNKEY C012 49170 SCAN KEYBOARD SUBROUTINE

JM P REPEAT C015 49173 HANDLE REPEAT KEY&STORE
DECODED KEY
JM P PLOT C018 49176 READ OR SET CRSR POSITION IN
.X, .Y
JM P CURSOR C01B 49179 MOVE 8563 CURSOR SUBROUTINE
JM P ESCAPE C01E 49182 EXECUTE ESC FUNCTION USING
CHR IN .A
JM P KEYSET C021 49185 REDEFINE A PROGRAM M ABLE
FUNC’N KEY
JM P IRQ C024 49188 IRQ ENTRY
JM P INIT80 C027 49191 INITIALIZE 8 O-COLUMN
CHARACTER SET
JM P SW APPER C02A 49194 SW AP EDITOR LOCALS (40/80
CHANGE)
JM P W INDOW C02D 49197 SET TOP-LEFT OR BOT-RIGHT OF
W INDOW
DOOO 53248 VIC CHARACTER ROM
($D 000-$D FFF)
VIC CHIP REGISTERS
VICREGO DOOO 53248 SPRITE 0, X-LOCATION

VICREG1 D001 53249 SPRITE 0, Y-LOCATION
VICREG2 D002 53250 SPRITE 1, X-LOCATION
VICREG3 D003 53251 SPRITE 1, Y-LOC ATION
VICREG10 DOOA 53258 SPRITE 5, X-LOCATION
VICREG11 DOOB 53259 SPRITE 5, Y-LOCATION
VICREG12 DOOC 53260 SPRITE 6, X-LOCATION
VICREG13 DOOD 53261 SPRITE 6, Y-LOCATION
VICREG14 DOOE 53262 SPRITE 7, X-LOCATION
VICREG15 DOOF 53263 SPRITE 7, Y-LOCATION
VICREG16 D010 53264 MSBIT OF X-LOCATION FOR
SPRITES 0 -7
M EM ORY
VIC CHIP REGISTERS
VICREG17 D O ll 53265 VIC CONTROL REGISTER 1
7 RASTER COMPARE BIT

6 EXTENDED COLOR TEXT
MODE (1 = ON)
BIT MAP MODE (1 = ENABLE)
BLANK SCREEN TO BORDER
CLR (0 = BLANK)
SELECT 24/25 ROW TEXT
DISPLAY (1 = 25)
2-0 SMOOTH SCROLL TO Y
DOT POSITION
VICREG18 D012 53266 READ/W RITE RASTER VALUE FOR
COM PARE IRQ
VICREG19 D013 53267 LIGHT PEN LATCH X-POSITION
VICREG20 D014 53268 LIGHT PEN LATCH Y-POSITION
VICREG21 D015 53269 SPRITES 0 -7 DISPLAY ENABLE
(1 = ENABLE)
VICREG22 D016 53270 VIC CONTROL REGISTER 2
BITS
7-6 UNUSED
5 RESET
4 M ULTI-COLOR MODE
(1 = ENABLE)
3 SELECT 38/40 COLUMN
DISPLAY (1 = 40 COLS)
2 -0 SMOOTH SCROLL TO
X-POSITION
VICREG23 DO17 53271 SPRITES 0 -7 Y EXPAND

VICREG24 D018 53272 VIC MEMORY CONTROL REGISTER
BITS ($D018)
7 -4 VIDEO MATRIX BASE

ADDRESS
3 -0 CHARACTER DOT-DATA
BASE ADDRESS
526 COMMODORE 128
M EM ORY
VIC CHIP REGISTERS
VICREG25 D019 53273 VIC INTERRUPT FLAG REGISTER

(1 = IRQ OCCURRED)
7 SET ON ANY ENABLED VIC

IRQ CONDITION
6 -4 NOT USED
3 LIGHT PEN TRIGGERED
IRQ FLAG
2 SPRITE TO SPRITE
COLLISION IRQ FLAG
1 SPRITE TO BACKGROUND
COLLISION IRQ FLAG
0 RASTER COM PARE IRQ
FLAG
VICREG26 D01A 53274 RQ ENABLE (1 = ENABLED) BITS
7 -4 NOT USED
3 LIGHT PEN
2 SPRITE TO SPRITE
1 SPRITE TO BACKGROUND
0 IRQ
VICREG27 D01B 53275 SPRITES 0 -7 BACKGROUND

PRIORITY (1 = SPRITE)
VICREG28 D01C 53276 SPRITES 0-7 MULTI-COLOR MODE
(1 = MULTI-COLOR)
VICREG29 D01D 53277 SPRITES 0 -7 X EXPAND
VICREG30 D01E 53278 SPRITE TO SPRITE COLLISION
LATCH
VICREG31 D01F 53279 SPRITE TO BACKGROUND
COLLISION LATCH
VICREG32 D020 53280 BORDER COLOR
VICREG33 D021 53281 BACKGROUND COLOR 0
VICREG37 D025 53285 SPRITE MULTI-COLOR REGISTER 0
VICREG38 D026 53286 SPRITE MULTI-COLOR REGISTER 1
VICREG39 D027 53287 SPRITE 0 COLOR
V1CREG42 D02A 53290 SPRITE 3 COLOR
M EM ORY
VIC CHIP REGISTERS
VICREG43 D02B 53291 SPRITE 4 COLOR

VICREG44 D02C 53292 SPRITE 5 COLOR
VICREG45 D02D 53293 SPRITE 6 COLOR
VICREG46 D02E 53294 SPRITE 7 COLOR
VICREG47 D02F 53295 KEYBOARD LINES
BITS
7 -3 NOT USED
2-0 K2, K1 AND KO
VICREG48 D030 53296 CLOCK SPEED

BITS
7 -2 NOT USED
1 TEST
0 2 MHZ
SID REGISTERS
SIDREGO D400 54272 VOICE 1 FREQUENCY LO
SIDREG1 D401 54273 VOICE 1 FREQUENCY HI
SIDREG2 D402 54274 VOICE 1 PULSE W IDTH LO
SIDREG3 D403 54275 VOICE 1 PULSE W IDTH HI (0-15)
SIDREG4 D404 54276 VOICE 1 CONTROL REGISTER
NOISE NOISE)
(1 =
PULSE PULSE)
(1 =
SAW SAW TOOTH)
(1 =
TRI TRIANGLE)
(1 =
TEST DISABLE
(1 =
OSCILLATOR)
RING (1 = RING MODULATE
OSC 1 WITH OSC
3 OUTPUT)
SYNC (1 = SYNCHRONIZE
OSC 1 WITH OSC
3 FREQ)
GATE (1 = START ATTACK/
DECAY/SUSTAIN
0 = START RELEASE)
SIDREG5 D405 54277 VOICE 1 ATTACK/DECAY
7 -4 ATTACK (0-15)
3 -0 DECAY (0-15)
528 COMMODORE 128
M EM ORY
SID REGISTERS
SIDREG6 D406 54278 VOICE 1 SUSTAIN/RELEASE
7 -4 SUSTAIN (0-15)
3 -0 RELEASE (0-15)
SIDREG7 D407 54279 VOICE 2 FREQUENCY LO

SIDREG8 D408 54280 VOICE 2 FREQUENCY HI
SIDREG10 D40A 54282 VOICE 2 PULSE W IDTH HI (0-15)
SIDREG11 D40B 54283 VOICE 2 CONTROL REGISTER
7 NOISE (1 =
NOISE)
6 PULSE (1 =
PULSE)
5 SAW (1 =
SAW TOOTH)
4 TRI (1 =
TRIANGLE)
3 TEST (1 =
DISABLE
OSCILLATOR)
2 RING (1 = RING MODULATE
OSC 2 WITH OSC
1 OUTPUT)
1 SYNC (1 = SYNCHRONIZE
OSC 2 WITH OSC
1 FREQ)
0 GATE (1 = START ATTACK/
DECAY/SUSTAIN
0 = START RELEASE)
SIDREG12 D40C 54284 VOICE 2 ATTACK/DECAY
7 -4 ATTACK (0-15)
3 -0 DECAY (0-15)
SIDREG13 D40D 54285 VOICE 2 SUSTAIN/RELEASE
7 -4 SUSTAIN (0-15)
3 -0 RELEASE (0-15)
SIDREG14 D40E 54286 VOICE 3 FREQUENCY LO

SIDREG15 D40F 54287 VOICE 3 FREQUENCY HI
SIDREG17 D411 54289 VOICE 3 PULSE W IDTH HI (0-15)
M EM ORY
SID REGISTERS
SIDREG18 D412 54290 VOICE 3 CONTROL REGISTER
NOISE (1 NOISE)
PULSE (1 PULSE)
SAW (1 SAW TOOTH)
TRI (l = rTRIANGLE)
TEST (1 = DISABLE
OSCILLATOR)
RING (1 = RING MODULATE
OSC 3 WITH OSC
2 OUTPUT)
SYNC (1 = SYNCHRONIZE
OSC 3 WITH OSC
2 FREQ)
GATE (1 = START ATTACK/
DECAY/SUSTAIN
0 = START RELEASE)
SIDREG19 D413 54291 VOICE 3 ATTACK/DECAY
7 -4 ATTACK (0-15)
3-0 DECAY (0-15)
SIDREG20 D414 54292 VOICE 3 SUSTAIN/RELEASE
7-4 SUSTAIN (0-15)

3 -0 RELEASE (0-15)
SIDREG21 D415 54293 FILTER CUTOFF FREQUENCY LO

SIDREG22 D416 54294 FILTER CUTOFF FREQUENCY HI
SIDREG23 D417 54295 RESONANCE/FILTER
7-4 FILTER RESONANCE (0-15)

3 FILTER EXTERNAL INPUT
(1 = YES)
2 FILTER VOICE 3 OUTPUT
(1 = YES)
(1 = YES)
(1 = YES)
530 COMMODORE 128
M EM ORY
SID REGISTERS
SIDREG24 D418 54296 MODE/VOLUME
7 CUTOFF VOICE 3 OUTPUT

(1 = OFF)
6 SELECT HI-PASS FILTER
( l = ON)
5 SELECT BAND-PASS FILTER
(1 = ON)
4 SELECT LO-PASS FILTER
(1 = ON)
3 -0 OUTPUT VOLUM E (0-15)
SIDREG25 D419 54297 POT X, A/D CONVERTER, PADDLE 1

SIDREG26 D41A 54298 POT Y, A/D CONVERTER, PADDLE 2
SIDREG27 D41B 54299 OSCILLATOR 3, RANDOM NUMBER
GENERATOR
SIDREG28 D41C 54300 ENVELOPE GENERATOR 3 OUTPUT
C128 M EMORY M ANAGEM ENT UNIT, PRIMARY REGISTERS

IM PLEM ENTS C128, C64, & CP/M 3.0 MODES
MMUCR1 D500 54528 CONFIGURATION REGISTER
PCRA D501 54529 PRECONFIGURATION REGISTER A
PCRB D502 54530 PRECONFIGURATION REGISTER B
PCRC D503 54531 PRECONFIGURATION REGISTER C
PCRD D504 54532 PRECONFIGURATION REGISTER D
BITS ($D 500-$D504)
7 -6 RAM BANK (0-3)

5 -4 ROM HI (SYSTEM , INT,
EXT, RAM)
3 -2 ROM MID (SYSTEM , INT,
EXT, RAM)
1 ROM LO (SYSTEM , RAM)
0 I/O (I/O BLOCK,
ELSE ROM-HI)
MMUMCR D505 54533 MODE CONFIGURATION REGISTER

7 40/80 KEY SENSE
6 OS M ODE, 0 = C128/1 = C64
5 /EXROM LINE SENSE
4 /GAME LINE SENSE
3 FSDIR
2-1 NOT USED
0 PROCESSOR, 0 == Z80/1 = 8502
M EM ORY
C128 M EM ORY M ANAGEM ENT UNIT, PRIM ARY REGISTERS

IMPLEMENTS C128, C64, & CP/M 3.0 MODES
MMURCR D506 54534 RAM CONFIGURATION REGISTER
7 -6 VIC RAM BANK (VA17 &

VA16)
5 -4 RAM BLOCK (FOR FUTURE
EXPANSION) /
3 -2 RAM SHARE STATUS (NONE,
BOT, TOP, BOTH)
1-0 RAM SHARE AM OUNT (IK ,
4K, 8K, 16K)
MMUPOL D507 54535 PAGE 0

POINTER LOW SWAPS PAGEO
MMUPOH D508 54536 PAGE 0 AND/OR PAGE1
POINTER HIGH W ITH ANY
MMUP1L D509 54537 PAGE 1 OTHER PAGE IN
POINTER LOW THE 256K
M MUP1H D50A 54538 PAGE 1 ADDRESS SPACE
POINTER HIGH
BITS ($D508 & $D50A)
7 -4
3 -2 A 19-A 18 (USED IN 1MB
SYSTEM)
1-0 A 17-A 16 (256K SYSTEM)
BITS ($D507 & $D509)
7 -0 A 15-A 8
M M UVER D50B 54539 M M U VERSION NUMBER
7 -4 BANK VERSION
3 -0 MM U VERSION
C128 8 O-COLUMN VIDEO DISPLAY CONTROLLER
VDCADR D600 54784 8563 ADDRESS REGISTER

532 COMMODORE 128
M EM ORY
ADDRESS AND DATA REGISTER BITS
7 6 5 4 3 2 1 0
D600------ WRITE R5 R4 R3 R2 R l R0
READ STATUS LP VBLANK — — — —
VDCDAT D601 54785 8563 DATA REGISTER
D601- -DATA D7 D6 D5 D4 D3 D2 Dl DO
ADDITIONAL 8563 REGISTERS ARE NOT VISIBLE TO THE 8502
D700 55040 RESERVED I/O BLOCK

VICCOL D800 55296 VIC COLOR M ATRIX, 1 KB
($D800-$DBFF)
6526 C IA # 1 , COMPLEX INTERFACE ADAPTER # 1

KEYBOARD, JOYSTICK, PADDLES, LIGHTPEN, FAST DISK
D1PRA DC00 56320 PORT A (OUTPUT KEYBOARD

COLUMNS)
BITS ($DC00)
PRAO KEYBD O/P CO/JOY #1 DIRECTION

PRA1 KEYBD O/P C l/JO Y #1 DIRECTION
PRA2 KEYBD O/P C2/JOY #1 DIRECTION/PADDLE FIRE BUTTON
PRA3 KEYBD O/P C3/JOY #1 DIRECTION/PADDLE FIRE BUTTON
PRA4 KEYBD O/P C4/JOY #1 FIRE BUTTON
PRA5 KEYBD O/P C5/
PRA6 KEYBD O/P C6/ /SELECT PORT # 1 PADDLES
PRA7 KEYBD O/P C7/ /SELECT PORT # 2 PADDLES
D1PRB DC01 56321 PORT B (INPUT KEYBOARD ROWS)

BITS ($DC01)
PRBO KEYBD I/P RO/JOY # 2 DIRECTION

PRB1 KEYBD I/P R l/JO Y # 2 DIRECTION/PADDLE FIRE BUTTON
PRB2 KEYBD I/P R2/JOY # 2 DIRECTION/PADDLE FIRE BUTTON
PRB3 KEYBD I/P R3/JOY # 2 DIRECTION/
PRB4 KEYBD I/P R4/JOY # 2 FIRE BUTTON
PRB5 KEYBD I/P R5/
PRB6 KEYBD I/P R6/TIM ER B: TOGGLE /PULSE OUTPUT
PRB7 KEYBD I/P R7/TIM ER A: TOGGLE /PULSE OUTPUT
D1DDRA DC02 56322 DATA DIRECTION PORT A

D1DDRB DC03 56323 DATA DIRECTION PORT B
D1T1L DC04 56324 TA LO
M EM ORY
6526 C IA # 1 , COMPLEX INTERFACE ADAPTER # 1

KEYBOARD, JOYSTICK, PADDLES , LIGHTPEN, FAST DISK
D1T1H DC05 56325 TA HI (TIMER A)

D1T2L DC06 56326 TB LO
D1T2H DC07 56327 TB HI (TIMER B)
D1TOD1 DC08 56328 TOD (TENTHS)
D1TODS DC09 56329 TOD (SECONDS)
D1TODM DCOA 56330 TOD (MINUTES)
D1TODH DCOB 56331 TOD (HOURS)
D1SDR DCOC 56332 SERIAL DATA REGISTER
D1ICR DCOD 56333 INTERRUPT CONTROL REGISTER
D1CRA DCOE 56334 CONTROL REGISTER A
D1CRB DCOF 56335 CONTROL REGISTER B
6526 C IA # 2 COMPLEX INTERFACE ADAPTER # 2

USER PORT,RS-232,SERIAL BUS,VIC M EM ORY,NM I
D2PRA DDOO 56576 PORT A, SERIAL BUS, RS-232, VA14

& VA15
BITS
0 PRAO : VA14
1 PRA1 : VA15
2 PRA2 : RS232 DATA
OUTPUT
3 PRA3 : SERIAL ATN
OUTPUT
4 PRA4 : SERIAL CLK
OUTPUT
5 P R A 5 : SERIAL DATA
OUTPUT
6 PRA6 : SERIAL CLK INPUT
7 P R A 7 : SERIAL DATA INPUT
534 COMMODORE 128
M EM ORY
D2PRB D D 01 56577 PORT B, USER PORT, RS-232

BITS
PRBO : USER PORT / RS-232

RECEIVED DATA
PRB1 : USER PORT / RS-232
REQUEST TO SEND
DATA TERM INAL READY
RING INDICATOR
CARRIER DETECT
PRB5 : USER PORT
CLEAR TO SEND
DATA SET READY
D2DDRA DD02 56578 DATA DIRECTION PORT A

D2DDRB DD03 56579 DATA DIRECTION PORT B
D2T1L DD04 56580 TA LO
D2T1H DD05 56581 TA HI (TIMER A)
D2T2L DD06 56582 TB LO
D2T2H DD07 56583 TB HI (TIMER B)
D2TOD1 DD08 56584 TOD (TENTHS)
D2TODS DD09 56585 TOD (SECONDS)
D2TODM DDOA 56586 TOD (MINUTES)
D2TODH DDOB 56587 TOD (HOURS)
D2SDR DDOC 56588 SERIAL DATA REGISTER
D2ICR DDOD 56589 INTERRUPT CONTROL REGISTERS
(NM I’S)
D2CRA DDOE 56590 CONTROL REGISTER A
D2CRB DDOF 56591 CONTROL REGISTER B
101 DEOO 56832 EXPANSION I/O SLOT (RESERVED)
102 DFOO 57088 EXPANSION I/O SLOT (RESERVED)

C128 DMA CONTROLLER FOR
EXPANSION RAM ACCESS
OPTIONAL DEVICE M APPED INTO
1 0 2 BLOCK VIA SYSTEM
EXPANSION PORT (PRELIMINARY)
DMA ST DFOO 57088 DMA CONTROLLER STATUS
REGISTER (READ ONLY)
M EM ORY
INTERRUPT
PENDING (1 = INT.
W AITING TO BE
SERVICED)
END OF BLOCK
(1 = TRANSFER
COMPLETE)
FAULT (1 = BLOCK
VERIFY ERROR)
SIZE (0 = EXP.
M EM ORY =
128K)
(1 = EXP.
M EM ORY =
512K)
VERSION
DMA CMD DFOl 57089 DMA CONTROLLER COM MAND

REGISTER
EXECUTE
RESERVED
LOAD (1 = ENABLE
AUTO = LOAD)
$FF00 (1 = DISABLE $FF00
DECADES)
3 RESERVED
2 RESERVED
1-0 MODE (00 = TRANSFER
FROM INTERNAL TO
EXTERNAL,
0 1 = FROM EXT TO INT,
10 = SW AP, 11 = VERIFY)
DMA ADL DF02 57090 LSB OF INTERNAL (C128) ADDRESS

TO ACCESS
DMA ADH DF03 57091 MSB OF INTERNAL (C128) ADDRESS
TO ACCESS
DMA LO DF04 57092 LSB OF EXTERNAL EXPANSION
RAM TO ACCESS
DMA HI DF05 57093 MSB OF EXTERNAL EXPANSION
RAM TO ACCESS
DMA BNK DF06 57094 64K EXTERNAL RAM BANK
BITS ($DF06)
536 COMMODORE 128

M EM ORY
7-3 NOT USED

2-0 EXPANSION BANK NUMBER
DMA DAL DF07 57095 LSB OF BYTE COUNT

DMA DAH DF08 57096 MSB OF BYTE COUNT (BLOCK
COUNT)
DMA SUM DF09 57097 INTERRUPT MASK REGISTER
INTERRUPT ENABLE
(1 = INTERRUPTS ENABLED)
END OF BLOCK MASK
(1 = INTERRUPT ON END
OF BLOCK)
VERIFY ERROR
(1 = INTERRUPT ON VERIFY
ERROR)
DMA VER DFOA 57098 ADDRESS CONTROL REGISTER
BITS 7 AND 6
0,0 = INCREMENT BOTH ADDRESS
(DEFAULT)
0,1 = FIX EXPANSION ADDRESS.
1.0 = FIX C128 ADDRESS.
1.1 = FIX BOTH ADDRESSES.
EOOO 57344 KERNAL ROM (8K OPERATING
SYSTEM , $E000-$FFFF)
MMU SECONDARY REGISTERS

M MUCR FFOO 65280 CONFIGURATION REGISTER
(SECONDARY)
LCRA FFOl 65281 LOAD CONFIGURATION
REGISTER A
LCRB FF02 65282 LOAD CONFIGURATION
REGISTER B
LCRC FF03 65283 LOAD CONFIGURATION
REGISTER C
LCRD FF04 65284 LOAD CONFIGURATION
REGISTER D
BITS ($FFOO-$FF04)
7 -6 RAM BANK (0-3)
5 -4 ROM HI (SYSTEM , INT,
EXT, RAM)
3 -2 ROM MID (SYSTEM , INT,
EXT, RAM)
1 ROM LO (SYSTEM, RAM)
0 I/O (I/O ELSE ROM-
HI)
KERNAL JUMP TABLE

NEW ENTRIES FOR C l 28
JM P SPIN SPOUT FF47 65351 SET UP FAST SERIAL PORT
FOR I/O
JM P CLOSE ALL FF4A 65354 CLOSE ALL LOGICAL
FILES FOR A DEVICE
JM P C64M ODE FF4D 65357 RECONFIGURE SYSTEM
AS A C64 (NO RETURN)
JM P DMA CALL FF50 65360 INITIATE DMA REQUEST
TO EXTERNAL RAM
EXPANSION, SEND
COM MAND TO DMA
DEVICE
JM P BOOT CALL FF53 65363 BOOT LOAD PROGRAM
FROM DISK
JM P PHOENIX FF56 65366 CALL ALL FUNCTION
CARDS’ COLD START
ROUTINES, INITIALIZE
JM P LKUPLA FF59 65369 SEARCH TABLES FOR
GIVEN LA
JM P LKUPSA FF5C 65372 SEARCH TABLES FOR
GIVEN SA
JM P SW APPER FF5F 65375 SWITCH BETW EEN 40 AND
80 COLUMNS (EDITOR)
JM P DLCHR FF62 65378 INIT 8 O-COL CHARACTER
RAM (EDITOR)
JM P PFKEY FF65 65381 PROGRAM FUNCTION KEY
(EDITOR)
JM P SETBNK FF68 65384 SET BANK FOR I/O
OPERATIONS
JM P GETCFG FF6B 65387 LOOKUP M M U DATA FOR
GIVEN BANK
JM P JSRFAR FF6E 65390 JSR TO ANY BANK, RTS
TO CALLING BANK
JM P JMPFAR FF71 65393 JMP TO ANY BANK
JM P INDFET FF74 65396 LDA (FETVEC), Y FROM
ANY BANK
JM P INDSTA FF77 65499 STA (STAVEC), Y TO ANY
BANK
JM P INDCMP FF7A 65402 CMP (CM PVEC),Y TO ANY
BANK
JM P PRIMM FF7D 65405 PRINT IMMEDIATE UTILITY
(ALWAYS JSR TO THIS
ROUTINE)
STANDARD KERNAL JUMP TABLE

FF80 65408 RELEASE NUM BER OF
KERNAL
JM P CINT FF81 65409 INIT EDITOR & DISPLAY
CHIPS (EDITOR)
538 COMMODORE 128

STANDARD KERNAL JUM P TABLE
JM P IOINIT FF84 65412 INIT I/O DEVICES (PORTS,

TIM ERS, ETC.)
JM P RAMTAS FF87 65415 INITIALIZE RAM AND
BUFFERS FOR SYSTEM
JM P RESTOR FF8A 65418 RESTORE VECTORS TO
INITIAL SYSTEM
JM P VECTOR FF8D 65421 CHANGE VECTORS FOR
USER
JM P SETMSG FF90 65424 CONTROL O .S. MESSAGE
JM P SECND FF93 65427 SEND SA AFTER LISTEN
JM P TKSA FF96 65430 SEND SA AFTER TALK
JM P M EMTOP FF99 65433 SET/READ TOP OF SYSTEM
RAM
JM P M EMBOT FF9C 65436 SET/READ BOTTOM OF
SYSTEM RAM
JM P KEY FF9F 65439 SCAN KEYBOARD
(EDITOR)
JM P SETTMO FFA2 65442 SET TIM EOUT IN IEEE
(RESERVED)
JM P ACPTR FFA5 65445 HANDSHAKE SERIAL BYTE
IN
JM P CIOUT FFA8 65448 HANDSHAKE SERIAL BYTE
OUT
JM P UNTLK FFAB 65451 SEND UNTALK OUT SERIAL
JM P UNLSN FFAE 65454 SEND UNLISTEN OUT
SERIAL
JM P LISTN FFB1 65457 SEND LISTEN OUT SERIAL
JM P TALK FFB4 65460 SEND TALK OUT SERIAL
JM P READSS FFB7 65463 RETURN I/O STATUS BYTE
JM P SETLFS FFBA 65460 SET LA, FA, SA
JM P SETNAM FFBD 65469 SET LENGTH AND FILE
NAME ADDRESS
JM P (IOPEN) FFCO OPEN 65472 OPEN LOGICAL FILE
JM P (ICLOSE) FFC3 CLOSE 65475 CLOSE LOGICAL FILE
JM P (ICHKIN) FFC6 CHKIN 65478 SET CHANNEL IN
JM P (ICKOUT) FFC9 CKOUT 65481 SET CHANNEL OUT
JM P (ICLRCH) FFCC CLRCH 65484 RESTORE DEFAULT I/O
CHANNEL
JM P (IBASIN) FFCF BASIN 65487 INPUT FROM CHANNEL
JM P (IBSOUT) FFD2 BSOUT 65490 OUTPUT TO CH ANNEL
JM P LOADSP FFD5 65493 LOAD FROM FILE
JM P SAVESP FFD8 65496 SAVE TO FILE
JM P SETTIM FFDB 65599 SET INTERNAL CLOCK
JM P RDTIM FFDE 65502 READ INTERNAL CLOCK
JM P (ISTOP) FFE1 STOP 65505 SCAN STOP KEY
JM P (IGETIN) FFE4 GETIN 65508 READ BUFFERED DATA
JM P (ICLALL) FFE7 CLALL 65511 CLOSE ALL FILES AND
CHANNELS
STANDARD KERNAL JUM P TABLE
JM P UDTIM FFEA CLOCK 65514 INCREM ENT INTERNAL

CLOCK
JM P SCRORG FFED 65517 RETURN SCREEN WINDOW
SIZE (EDITOR)
JM P PLOT FFFO 65520 READ/SET X ,Y CURSOR
COORD (EDITOR)
JM P IOBASE FFF3 65523 RETURN I/O BASE
SYSTEM FFF8 65528 OPERATING SYSTEM
VECTOR (RAM I)
NMI FFFA 65530 PROCESSOR NMI VECTOR
RESET FFFC 65532 PROCESSOR RESET
VECTOR
IRQ FFFE 65534 PROCESSOR IRQ/BRK
VECTOR
KERNAL/EDITOR FLAGS AND

SHADOW REGISTERS
The following symbols are used by the C128 Editor. Note that the Editor IRQ VIC
screen handler depends upon them. In most cases the contents of these locations will be
placed directly into the appropriate register and should be used instead of the actual
register. For example, to change the location of the character set used by VIC, use VM1
($0A2C) instead of VIC register 24 ($D018). VM1 will be used by the editor to update
VIC register 24.
ADDRESS/NAME EXPLANATION
S00D8/GRAPHM SEE BELOW. IF = $FF THEN EDITOR LEAVES VIC ALONE.

S00D9/CHAREN MASK FOR 8502 /CHAREN BIT.
$0A2C/VM1 VIC TEXT MODE VIDEO MATRIX & CHARACTER BASE
POINTER.
$0A2D/VM2 VIC GRAPHIC MODE VIDEO MATRIX & BIT M AP
POINTER.
$0A2E/VM3 8563 TEXT DISPLAY BASE ADDRESS.
$0A2F/VM 4 8563 ATTRIBUTE BASE ADDRESS.
$0A34/SPLIT IN SPLIT SCREEN M ODE, CONTAINS VALUE FOR
MIDDLE RASTER IRQ.
$OA2B/CURMOD 8563 CURSOR MODE.
$0A21/PAUSE CONTROL S FLAG (IN EFFECT = $13)
540 COMMODORE 128
E X P L A N A T IO N O F V A R IO U S K E R N A L /E D IT O R FLA G B Y T E S , E T C .
A D DRESS SY M B O L D E S C R IP T IO N
7 6 5 4 3 2 l 0
0000 D6510 — (IN) (OUT) (IN) (OUT) (OUT) (OUT) (OUT)

0001 R6510 — CAPKEY CASM TR CASSEN CASW RT CH AREN HIRAM LORAM
00F7 LOCKS CASE CTLS — ------ — — — —
00F8 SCRO LL OFF LINKER

00D3 SHFLAG — — — ALT ALPHA CTRL SHIFT
0A22 RPTFLG ALL NONE — — — — — —
0A26 BLNON ON BLNK — — — — — —
00F9 BEEPER ON — — — — — — —
00D8 G RAPHM MCM SPLIT BMM — — — — —
00D7 M O DE 40/80 — — — — — — —
0A04 IN IT _ CHRSET C IN T ■—~" BASIC

STATUS
Notes on Kernal Symbols:
lnit_S tatus. See also above. Lets system know what has been initialized and what
hasn’t. Set to $00 by a reset but untouched by NMI.
S ystem _V ector. Where the Kernal goes when it has to go somewhere. It’s set to
BASIC cold at reset. BASIC itself sets it to BASIC warm after it has initialized.
The monitor respects it too.
System. Vector in RAMI at $FFF8. Set at power-up to C128MODE, user may redirect
it to his code. Taken at reset always providing user with control (protection) from
reset.
COMMODORE 64 MEMORY MAP
HEX DECIM AL
LABEL ADDRESS LOCATION DESCRIPTION
D6510 0000 0 6510 On-Chip Data-Direction Register

R6510 0001 1 6510 On-Chip 8-Bit Input/Output
Register
0002 2 Unused
ADRAY1 0003-0004 3 -4 Jump Vector: Convert Floating—
Integer
ADRAY2 0005-0006 5 -6 Jump Vector: Convert Integer—
Floating
CHARAC 0007 7 Search Character
ENDCHR 0008 8 Flag: Scan for Quote at End of String
HEX DECIMAL
TRM POS 0009 9 Screen Column From Last TAB

VERCK 000A 10 Flag: 0 = Load, 1 = Verify
COUNT 000B 11 Input Buffer Pointer/No. o f Subscripts
DIMFLG OOOC 12 Flag: Default Array DIM ension
VALTYP 000D 13 Data Type: $FF = String, $00 =
Numeric
INTFLG 000E 14 Data Type: $80 = Integer, $00 =
Floating
GARBFL 000F 15 Flag: DATA scan/LIST quote/
Garbage Coll
SUBFLG 0010 16 Flag: Subscript R ef / User Function
Call
INPFLG 0011 17 Flag: $00 = INPUT, $40 = GET,
$98 = READ
TANSGN 0012 18 Flag: TAN sign / Comparison Result
0013 19 Flag: INPUT Prompt
LINNUM 0014-0015 20-21 Temp: Integer Value
TEMPPT 0016 22 Pointer: Temporary String Stack
LASTPT 0017-0018 23-24 Last Temp String Address
TEMPST 0019-0021 25-33 Stack for Temporary Strings
INDEX 0022-0025 34-37 Utility Pointer Area
RESHO 0026-002A 38-42 Floating-Point Product o f Multiply
TXTTAB 002B-002C 43-44 Pointer: Start of BASIC Text
VARTAB 002D -002E 45—46 Pointer: Start of BASIC Variables
ARYTAB 002F-0030 47-48 Pointer: Start of BASIC Arrays
STREND 0031-0032 49-50 Pointer: End of BASIC Arrays ( + 1)
FRETOP 0033-0034 51-52 Pointer: Bottom o f String Storage
FRESPC 0035-0036 53-54 Utility String Pointer
MEMSIZ 0037-0038 55-56 Pointer: Highest Address Used by
BASIC
CURLIN 0039-003A 57-58 Current BASIC Line Number
OLDLIN 003B-003C 59-60 Previous BASIC Line Number
OLDTXT 003D -003E 61-62 Pointer: BASIC Statement for CONT
DATLIN 003F-0040 63-64 Current DATA Line Number
DATPTR 0041-0042 65-66 Pointer: Current DATA Item Address
INPPTR 0043-0044 67-68 Vector: INPUT Routine
VARNAM 0045-0046 69-70 Current BASIC Variable Name
VARPNT 0047-0048 71-72 Pointer: Current BASIC Variable
Data
FORPNT 0049-004A 73-74 Pointer: Index Variable for FOR/NEXT
004B-0060 75-96 Temp Pointer / Data Area
FACEXP 0061 97 Floating-Point Accumulator # 1 :
Exponent
FACHO 0062-0065 98-101 Floating Accum. # 1 : M antissa
FACSGN 0066 102 Floating Accum. # 1 : Sign
542 COMMODORE 128
Commodore 64 Memory Map (continued)
HEX DECIMAL
SGNFLG 0067 103 Pointer: Series Evaluation Constant

BITS 0068 104 Floating Accum. # 1 : Overflow Digit
ARGEXP 0069 105 Floating-Point Accumulator # 2 :
Exponent
ARGHO 006A -006D 106-109 Floating Accum. # 2 : Mantissa
ARGSGN 006E 110 Floating Accum. # 2 : Sign
ARISGN 006F 111 Sign Comparison Result: Accum. # 1
vs # 2
FACOV 0070 112 Floating Accum. # 1 . Low-Order
(Rounding)
FBUFPT 0071-0072 113-114 Pointer: Cassette Buffer
CHRGET 0073-008A 115-138 Subroutine: Get Next Byte o f BASIC
Text
CHRGOT 0079 121 Entry to Get Same Byte o f Text
Again
TXTPTR 007A-007B 122-123 Pointer: Current Byte of BASIC Text
RNDX 008B-008F 139-143 Floating RND Function Seed Value
STATUS 0090 144 Kernal I/O Status Word: ST
STKEY 0091 145 Flag: STOP key / RVS key
SVXT 0092 146 Timing Constant for Tape
VERCK 0093 147 Flag: 0 = Load, 1 = Verify
C3PO 0094 148 Flag: Serial Bus— Output Char.
Buffered
BSOUR 0095 149 Buffered Character for Serial Bus
SYNO 0096 150 Cassette Sync No.
0097 151 Temp Data Area
LDTND 0098 152 No. of Open Files / Index to File
Table
DFLTN 0099 153 Default Input Device (0)
DFLTO 009A 154 Default Output (CMD) Device (3)
PRTY 009B 155 Tape Character Parity
DPSW 009C 156 Flag: Tape Byte-Received
MSGFLG 009D 157 Flag: $80 = Direct M ode, $00 =
Program
PTR1 009E 158 Tape Pass 1 Error Log
PTR2 009F 159 Tape Pass 2 Error Log
TIME OOAO-OOA2 160-162 Real-Time Jiffy Clock (approx) 1/60
Sec
00A 3-00A4 163-164 Temp Data Area
CNTDN 00A5 165 Cassette Sync Countdown
BUFPNT 00A6 166 Pointer: Tape I/O Buffer
INBIT 00A7 167 RS-232 Input Bits / Cassette Temp
BITCI 00A8 168 RS-232 Input Bit Count / Cassette
Temp
RINONE 00A9 169 RS-232 Flag: Check for Start Bit
HEX DECIMAL
RIDATA OOAA 170 RS-232 Input Byte Buffer/Cassette

Temp
RIPRTY OOAB 171 RS-232 Input Parity / Cassette Short
Cnt
SAL 00AC-00AD 172-173 Pointer: Tape Buffer/Screen Scrolling
EAL OOAE-OOAF 174-175 Tape End Addresses / End of Program
CMPO 00B0-00B1 176-177 Tape Timing Constants
TAPE1 00B2-00B3 178-179 Pointer: Start o f Tape Buffer
BITTS 00B4 180 RS-232 Out Bit Count / Cassette
Temp
NXTBIT 00B5 181 RS-232 Next Bit to Send / Tape EOT
Flag
RODATA 00B6 182 RS-232 Out Byte Buffer
FNLEN 00B7 183 Length of Current File Name
LA 00B8 184 Current Logical File Number
SA 00B9 185 Current Secondary Address
FA OOBA 186 Current Device Number
FNADR 00BB-00BC 187-188 Pointer: Current File Name
ROPRTY OOBD 189 RS-232 Out Parity / Cassette Temp
FSBLK OOBE 190 Cassette Read/W rite Block Count
MYCH OOBF 191 Serial W ord Buffer
CAS1 OOCO 192 Tape M otor Interlock
STAL 00C 1-00C2 193-194 I/O Start Address
MEMUSS 00C 3-00C4 195-196 Tape Load Temps
LSTX 00C5 197 Current Key Pressed: CHR$(n) 0 =
No Key
NDX 00C6 198 No. of Chars, in Keyboard Buffer
(Queue)
RVS 00C7 199 Flag: Print Reverse C hars.— 1 =
Yes, 0 = Not Used
INDX 00C8 200 Pointer: End of Logical Line for
INPUT
LXSP 00C 9-00CA 201-202 Cursor X-Y Pos. at Start of INPUT
SFDX OOCB 203 Flag: Print Shifted Chars.
BLNSW OOCC 204 Cursor Blink enable: 0 = Flash
Cursor
BLNCT OOCD 205 Timer: Countdown to Toggle Cursor
GDBLN OOCE 206 Character Under Cursor
BLNON OOCF 207 Flag: Last Cursor Blink On/Off
CRSW 00D0 208 Flag: INPUT or GET from Keyboard
PNT 00D 1-00D2 209-210 Pointer: Current Screen Line Address
PNTR 00D3 211 Cursor Column on Current Line
QTSW 00D4 212 Flag: Editor in Quote M ode, $00 =
NO
LNMX 00D5 213 Physical Screen Line Length
544 COMMODORE 128
C o m m o d o re 6 4 M e m o ry M a p (co n tin u ed )
HEX DECIMAL
TBLX 00D6 214 Current Cursor Physical Line

Number
00D7 215 Temp Data Area
INSRT 00D8 216 Flag: Insert M ode, > 0 = # INSTs
LDTB1 00D 9-00F2 217-242 Screen Line Link Table / Editor
Temps
USER 00F3-00F4 243-244 Pointer: Current Screen Color RAM
loc.
KEYTAB 00F5-00F6 245-246 Vector: Keyboard Decode Table
RIBUF 00F7-00F8 247-248 RS-232 Input Buffer Pointer
ROBUF 00F9-00FA 249-250 RS-232 Output Buffer Pointer
FREKZP 00FB-00FE 251-254 Free 0-Page Space for User Programs
BASZPT 00FF 255 BASIC Temp Data Area
0100-01FF 256-511 Micro-Processor System Stack Area
0100-010A 256-266 Floating to String W ork Area
BAD 0100-013E 256-318 Tape Input Error Log
BUF 0200-0258 512-600 System INPUT Buffer
LAT 0259-0262 601-610 KERNAL Table: Active Logical File
N o’s.
FAT 0263-026C 611-620 KERNAL Table: Device No. for Each
File
SAT 026D -0276 621-630 KERNAL Table: Second Address
Each File
KEYD 0277-0280 631-640 Keyboard Buffer Queue (FIFO)
MEMSTR 0281-0282 641-642 Pointer: Bottom of M emory for O .S.
M EMSIZ 0283-0284 643-644 Pointer: Top o f M emory for O .S.
TIMOUT 0285 645 Flag: Kernal Variable for IEEE
Timeout
COLOR 0286 646 Current Character Color Code
GDCOL 0287 647 Background Color Under Cursor
HIBASE 0288 648 Top of Screen M emory (Page)
XMAX 0289 649 Size of Keyboard Buffer
RPTFLG 028A 650 Flag: REPEAT Key Used, $80 =
Repeat
KOUNT 028B 651 Repeat Speed Counter
DELAY 028C 652 Repeat Delay Counter
SHFLAG 028D 653 Flag: Keyb’rd SHIFT Key / CTRL
Key / Key
LSTSHF 028E 654 Last Keyboard Shift Pattern
KEYLOG 028F-0290 655-656 Vector: Keyboard Table Setup
MODE 0291 657 Flag: $00 = Disable SHIFT Keys,
$80 = Enable SHIFT Keys
AUTODN 0292 658 Flag: Auto Scroll Down, 0 = ON
M51CTR 0293 659 RS-232: 6551 Control Register Image
M51CDR 0294 660 RS-232: 6551 Command Register
Image
HEX DECIMAL
M51AJB 0295-0296 661-662 RS-232 Non-Standard BPS

(Time/2-100) USA
RSSTAT 0297 663 RS-232: 6551 Status Register Image
BITNUM 0298 664 RS-232 Number o f Bits Left to Send
BAUDOF 0299-029A 665-666 RS-232 Baud Rate: Full Bit Time ( |xs)
RIDBE 029B 667 RS-232 Index to End o f Input Buffer
RIDBS 029C 668 RS-232 Start of Input Buffer (Page)
RODBS 029D 669 RS-232 Start of Output Buffer (Page)
RODBE 029E 670 RS-232 Index to End o f Output
Buffer
IRQTM P 029F-02A 0 671-672 Holds IRQ Vector During Tape I/O
ENABL 02 A 1 673 RS-232 Enables
02A2 674 TOD Sense During Cassette I/O
02A3 675 Temp Storage For Cassette Read
02A4 676 Temp D1IRQ Indicator For Cassette
Read
02A5 677 Temp for Line Index
02A6 678 PAL/NTSC Flag, 0 = NTSC, 1 =
PAL
02A 7-02FF 697-767 Unused
IERROR 0300-0301 768-769 Vector: Print BASIC Error Message
IMAIN 0302-0303 770-771 Vector: BASIC W arm Start
ICRNCH 0304-0305 772-773 Vector: Tokenize BASIC Text
IQPLOP 0306-0307 774-775 Vector: BASIC Text LIST
IGONE 0308-0309 776-777 Vector: BASIC Char. Dispatch
IEVAL 030A-030B 778-779 Vector: BASIC Token Evaluation
SAREG 030C 780 Storage for 6502 .A Register
SXREG 030D 781 Storage for 6502 .X Register
SYREG 030E 782 Storage for 6502 .Y Register
SPREG 030F 783 Storage for 6502 .SP Register
USRPOK 0310 784 USR Function Jump Instr (4C)
USRADD 0311-0312 785-786 USR Address Low Byte / High Byte
0313 787 Unused
CINV 0314-0315 788-789 Vector: Hardware IRQ Interrupt
CBINV 0316-0317 790-791 Vector: BRK Instr. Interrupt
NMINV 0318-0319 792-793 Vector: Non-M askable Interrupt
IOPEN 031A-031B 794-795 KERNAL OPEN Routine Vector
ICLOSE 031C -031D 796-797 KERNAL CLOSE Routine Vector
ICHKIN 031E-031F 798-799 KERNAL CHKIN Routine Vector
ICKOUT 0320-0321 800-801 KERNAL CHKOUT Routine Vector
ICLRCH 0322-0323 802-803 KERNAL CLRCHN Routine Vector
IBASIN 0324-0325 804-805 KERNAL CHRIN Routine Vector
IBSOUT 0326-0327 806-807 KERNAL CHROUT Routine Vector
ISTOP 0328-0329 808-809 KERNAL STOP Routine Vector
IGETIN 032A-032B 810-811 KERNAL GETIN Routine Vector
ICLALL 032C -032D 812-813 KERNAL CLALL Routine Vector
USRCMD 032E-032F 814-815 User-Defined Vector
546 COMMODORE 128
Commodore 64 Memory M ap (continued)
HEX DECIMAL
ILOAD 0330-0331 816-817 KERNAL LOAD Routine Vector

ISAVE 0332-0333 818-819 KERNAL SAVE Routine Vector
0334-033B 820-827 Unused
TBUFFR 033C-03FB 828-1019 Tape I/O Buffer
03FC-03FF 1020-1023 Unused
VICSCN 0400-07FF 1024-2047 1024 Byte Screen Memory Area
0400-07E7 1024-2023 Video Matrix: 25 Lines x 40
Columns
07F8-07FF 2040-2047 Sprite Data Pointers
0800-9FFF 2048-40959 Normal BASIC Program Space
8000-9FFF 32768-40959 VSP Cartridge ROM— 8192 Bytes
A000-BFFF 40960-49151 BASIC ROM— 8192 Bytes (or 8K
RAM)
C000-CFFF 49152-53247 RAM— 4096 Bytes
D000-DFFF 53248-57343 Input/Output Devices and Color
RAM
or Character Generator ROM
or RAM— 4096 Bytes
E000-FFFF 57344-65535 KERNAL ROM— 8192 Bytes (or 8K
RAM)
COMMODORE 64 IN P U T /O U T P U T ASSIGNMENTS
HEX DECIMAL BITS DESCRIPTION
0000 0 7-0 MOS 6510 Data Direction Register

( x x lO llll)
Bit = 1: Output, Bit = 0:
Input, x = Either
0001 1 MOS 6510 M icro-Processor On-Chip
I/O Port
0 /LORAM Signal (0 = Switch BASIC
ROM Out)
1 /HIRAM Signal (0 = Switch Kernal
ROM Out)
2 /CHAREN Signal (0 = Switch Char.
ROM In)
3 Cassette Data Output Line
4 Cassette Switch Sense
1 = Switch Closed
5 Cassette Motor Control
0 = ON, 1 = OFF
6 -7 Undefined
D 000-D02E 53248-54271 MOS 6566 VIDEO INTERFACE

CONTROLLER (VIC)
DOOO 53248 Spr te 0 X Pos
D001 53249 Spr te 0 Y Pos
D002 53250 Spr te 1 X Pos
DOOA 53258 Spr te 5 X Pos
DOOB 53259 Spr te 5 Y Pos
DOOC 53260 Spr te 6 X Pos
DOOD 53261 Spr te 6 Y Pos
DOOE 53262 Spr te 7 X Pos
DOOF 53263 Spr te 7 Y Pos
D010 53264 Spr tes 0 -7 X Pos (msb o f X coord.)
D011 53265 VIC Control Register
7 Raster Compare: (Bit 8) See 53266
6 Extended Color Text Mode: 1 =
Enable
5 Bit-Map Mode: 1 = Enable
4 Blank Screen to Border Color: 0 =
Blank
3 Select 24/25 Row Text Display: 1 =
25 Rows
2 -0 Smooth Scroll to Y Dot-Position (0-7)
DO12 53266 Read Raster / Write Raster Value for
Compare IRQ
D013 53267 Light-Pen Latch X Pos
0014 53268 Light-Pen Latch Y Pos
D015 53269 Sprite Display Enable: 1 = Enable
D016 53270 VIC Control Register
7 -6 Unused
5 ALWAYS SET THIS BIT TO 0
4 Multi-Color Mode: 1 = Enable (Text
or Bit-Map)
3 Select 38/40 Column Text Display: 1
= 40 Cols
2 -0 Smooth Scroll to X Pos
D017 53271 Sprites 0 -7 Expand 2 x Vertical (Y)
D018 53272 VIC Memory Control Register
7 -4 Video Matrix Base Address (inside
VIC)
S48 COMMODORE 128
Commodore 64 Input/Output Assignments (continued)
3 -1 Character Dot-Data Base Address

(inside VIC)
D019 53273 VIC Interrupt Flag Register (Bit =
1: IRQ Occurred)
7 Set on Any Enabled VIC IRQ
Condition
3 Light-Pen Triggered IRQ Flag
2 Sprite to Sprite Collision IRQ Flag
1 Sprite to Background Collision IRQ
Flag
0 Raster Compare IRQ Flag
D01A 53274 IRQ Mask Register: 1 = Interrupt
Enabled
D01B 53275 Sprite to Background
Display Priority: 1 = Sprite
D01C 53276 Sprites 0 -7 M ulti-Color M ode Select:
1 = M .C .M .
D01D 53277 Sprites 0 -7 Expand 2 x Horizontal
(X)
D01E 53278 Sprite to Sprite Collision Detect
D01F 53279 Sprite to Background Collision
Detect
D020 53280 Border Color
D021 53281 Background Color 0
D025 53285 Sprite M ulti-Color Register 0
D026 53286 Sprite Multi-Color Register 1
D027 53287 Sprite 0 Color
D02A 53290 Sprite 3 Color
D02B 53291 Sprite 4 Color
D02C 53292 Sprite 5 Color
D02D 53293 Sprite 6 Color
D02E 53294 Sprite 7 Color
D 400-D7FF 54272-55295 MOS 6581 SOUND INTERFACE
DEVICE (SID)
D400 54272 Voice 1: Frequency Control— Low-Byte
D401 54273 Voice 1: Frequency Control— High-
Byte
D402 54274 Voice 1: Pulse W aveform W idth—
Low-Byte
HEX DECIM AL BITS DESCRIPTION
D403 54275 7 -4 Unused

3 -0 Voice 1: Pulse W aveform Width—
High-Nybble
D404 54276 Voice 1: Control Register
7 Select Random Noise W aveform , 1 =
On
6 Select Pulse W aveform , 1 = On
5 Select Sawtooth W aveform , 1 = On
4 Select Triangle W aveform , 1 = On
3 Test Bit: 1 = Disable Oscillator 1
2 Ring M odulate Osc. 1 with Osc. 3
Output, 1 = On
1 Synchronize Osc. 1 with Osc. 3
Frequency, 1 = On
0 Gate Bit: 1 = Start Att/Dec/Sus, 0 =
Start Release
D4U5 54277 Envelope Generator 1: Attack / Decay
Cycle Control
7 -4 Select Attack Cycle Duration: 0-15
3 -0 Select Decay Cycle Duration: 0-15
D406 54278 Envelope Generator 1: Sustain /
Release Cycle Control
7 -4 Select Sustain Cycle Duration: 0-1 5
3 -0 Select Release Cycle Duration: 0-15
D407 54279 Voice 2: Frequency Control— Low-Byte
D40B 54280 Voice 2: Frequency Control—
High-Byte
D409 54281 Voice 2: Pulse W aveform Width—
Low-Byte
D40A 54282 7 -4 Unused
3 -0 Voice 2: Pulse W aveform Width—
High-Nybble
D40B 54283 Voice 2: Control Register
On
6 Select Pulse W aveform, 1 = On
5 Select Sawtooth W aveform , 1 = On
2 Ring Modulate Osc. 2 with Osc. 1
Output, 1 = On
Frequency, 1 = On
0 Gate Bit: 1 = Start Att/Dec/Sus, 0
= Start Release
D40C 54284 Envelope Generator 2: Attack / Decay
Cycle Control
550 COMMODORE 128

D40D 54285 Envelope Generator 2: Sustain /
7^» Select Sustain Cycle Duration: 0-15
D40E 54286 Voice 3: Frequency Control— Low-Byte
D40F 54287 Voice 3: Frequency Control—
High-Byte
D410 54288 Voice 3: Pulse W aveform W idth—
Low-Byte
D411 54289 7 -4 Unused
3 -0 Voice 3: Pulse W aveform W idth—
High-Nybble
D412 54290 Voice 3: Control Register
On
6 Select Pulse W aveform , 1 = On
5 Select Sawtooth W aveform, 1 = On
2 Ring M odulate Osc. 3 with Osc. 2
Output, 1 = On
Frequency, 1 = On
0 Gate Bit: 1 = Start Att/Dec/Sus, 0 =
Start Release
D413 54291 Envelope Generator 3: Attack / Decay
Cycle Control
D414 54292 Envelope Generator 3: Sustain /
7 -4 Select Sustain Cycle Duration: 0 -1 5
D415 54293 Filter Cutoff Frequency: Low-Nybble
(Bits 2-0)
D416 54294 Filter Cutoff Frequency: High-Byte
D417 54295 Filter Resonance Control / Voice
Input Control
7 -4 Select Filter Resonance: 0-15
3 Filter External Input: 1 = Y es, 0 =
No
2 Filter Voice 3 Output: 1 = Y es, 0 =
No
1 Filter Voice 2 Output: 1 = Y es, 0 =
No
0 Filter Voice 1 Output: 1 = Yes, 0 =

No
D418 54296 Select Filter Mode and Volume
7 Cut-Off Voice 3 Output: 1 = Off, 0
= On
6 Select Filter High-Pass Mode: 1 =
On
5 Select Filter Band-Pass Mode: 1 =
On
4 Select Filter Low-Pass Mode: 1 = On
3 -0 Select Output Volume: 0-15
D419 54297 Analog/Digital Converter: Game
Paddle 1 (0-255)
D41A 54298 Analog/Digital Converter: Game
Paddle 2 (0-255)
D41B 54299 Oscillator 3 Random Number
Generator
D41C 54300 Envelope Generator 3 Output
D500-D7FF 54528-55295 SID IMAGES
D 800-D BFF 55296-56319 Color RAM (Nybbles)
DCOO-DCFF 56320-56575 MOS 6526 Complex Interface
Adapter (CIA) # 1
DCOO 56320 Data Port A (K eyboard, Joystick,
Paddles, Light-Pen)
7 -0 Write Keyboard Column Values for
Keyboard Scan
7-6 Read Paddles on Port A / B (01 =
Port A, 10 = Port B)
4 Joystick A Fire Button: 1 = Fire
3-2 Paddle Fire Buttons
3 -0 Joystick A Direction (0-15)
DC01 56321 Data Port B (Keyboard, Joystick,
Paddles): Game Port 1
7 -0 Read Keyboard Row Values for
Keyboard Scan
7 Timer B: Toggle/Pulse Output
6 Timer A: Toggle/Pulse Output
4 Joystick 1 Fire Button: 1 = Fire
3 -2 Paddle Fire Buttons
3 -0 Joystick 1 Direction
DC02 56322 Data Direction Register— Port A
(56320)
DC 03 56323 Data Direction Register— Port B
(56321)
DC04 56324 Timer A: Low-Byte
DC05 56325 Timer A: High-Byte
DC 06 56326 Timer B: Low-Byte
552 COMMODORE 128
DC07 56327 Timer B: High-Byte

DC08 56328 Time-of-Day Clock: 1/10 Seconds
DC09 56329 Time-of-Day Clock: Seconds
DCOA 56330 Time-of-Day Clock: Minutes
DCOB 56331 Time-of-Day Clock: Hours +
AM/PM Flag (Bit 7)
DCOC 56332 Synchronous Serial I/O Data Buffer
DCOD 56333 CIA Interrupt Control Register (Read
IRQ’s/W rite Mask)
7 IRQ Flag (1 = IRQ Occurred) /
Set-Clear Flag
4 FLAG1 IRQ (Cassette Read / Serial
Bus SRQ Input)
3 Serial Port Interrupt
2 Time-of-Day Clock Alarm Interrupt
1 Timer B Interrupt
0 Timer A Interrupt
DCOE 56334 CIA Control Register A
7 Time-of-Day Clock Frequency: 1 =
50 Hz, 0 = 60 Hz
6 Serial Port I/O Mode: 1 = Output, 0
= Input
5 Timer A Counts: 1 = CNT Signals, 0
= System 02 Clock
4 Force Load Timer A: 1 = Yes
3 Timer A Run Mode: 1 = One-Shot,
0 = Continuous
2 Timer A Output M ode to PB6: 1 =
Toggle, 0 = Pulse
1 Timer A Output on PB6: 1 = Yes, 0
= No
0 Start/Stop Timer A: 1 = Start, 0 =
Stop
DCOF 56335 CIA Control Register B
7 Set Alarm/TOD-Clock: 1 = Alarm, 0
= Clock
6 -5 Timer B Mode Select:
00 = Count System 02 Clock Pulses
01 = Count Positive CNT
Transitions
10 = Count Timer A Underflow
Pulses
11 = Count Timer A Underflows
While CNT Positive
4 -0 Same as CIA Control Reg. A— for
Timer B
DDOO-DDFF 56576-56831 MOS 6526 Complex Interface

Adapter (CIA) # 2
DDOO 56576 Data Port A (Serial Bus, RS-232,
VIC Memory Control)
7 Serial Bus Data Input
6 Serial Bus Clock Pulse Input
5 Serial Bus Data Ouput
4 Serial Bus Clock Pulse Output
3 Serial Bus ATN Signal Output
2 RS-232 Data Output (User Port)
1-0 VIC Chip System M emory Bank
Select (Default = 11)
DD01 56577 Data Port B (User Port, RS-232)
7 User / RS-232 Data Set Ready
6 User / RS-232 Clear to Send
5 User
4 User / RS-232 Carrier Detect
3 User / RS-232 Ring Indicator
2 User / RS-232 Data Terminal Ready
1 User / RS-232 Request to Send
0 User / RS-232 Received Data
DD02 56578 Data Direction Register— Port A
DD03 56579 Data Direction Register— Port B
DD04 56580 Timer A: Low-Byte
DD05 56581 Timer A: High-Byte
DD06 56582 Timer B: Low-Byte
DD07 56583 Timer B: High-Byte
DD08 56584 Time-of-Day Clock: 1/10 Seconds
DD09 56585 Time-of-Day Clock: Seconds
DDOA 56586 Time-of-Day Clock: Minutes
DDOB 56587 Time-of-Day Clock: Hours +
AM/PM Flag (Bit 7)
DDOC 56588 Synchronous Serial I/O Data Buffer
DDOD 56589 CIA Interrupt Control Register (Read
N M I’s/W rite Mask)
7 NMI Flag (1 = NM I Occurred) /
Set-Clear Flag
4 FLAG1 NM I (User/RS-232 Received
Data Input)
3 Serial Port Interrupt
1 Timer B Interrupt
0 Timer A Interrupt
DDOE 56590 CIA Control Register A
7 Time-of-Day Clock Frequency: 1 =
50 Hz, 0 = 60 Hz
6 Serial Port I/O Mode: 1 = Output, 0
= Input
554 COMMODORE 128
5 Timer A Counts: 1 = CNT Signals, 0

= System 02 Clock
4 Force Load Timer A: 1 = Yes
3 Timer A Run Mode: 1 = One-Shot,
0 = Continuous
2 Timer A Output M ode to PB6: 1 =
Toggle, 0 = Pulse
1 Timer A Output on PB6: 1 = Yes, 0
= No
0 Start/Stop Timer A: 1 = Start, 0 =
Stop
DDOF 56591 CIA Control Register B
7 Set Alarm/TOD-CIock: 1 = Alarm, 0
= Clock
6 -5 Timer B Mode Select:
00 = Count System 02 Clock
Pulses
01 = Count Positive CNT
Transitions
10 = Count Timer A Underflow
Pulses
11 = Count Timer A Underflows
While CNT Positive
4 -0 Same as CIA Control Reg. A— for
Timer B
DEOO-DEFF 56832-57087 Reserved for Future I/O Expansion
DFOO-DFFF 57088-57343 Reserved for Future I/O Expansion
16____
C128 HARDWARE
SPECIFICATIONS
555
556 COMMODORE 128
This chapter describes the Commodore C128 hardware, VLSI integrated circuit require
ments, and the relationship between the hardware configuration and the operating
system.
The C l 28 personal computer is compatible with C64 software and peripherals. In
addition to C64 compatibility, a C l 28 mode exists in which 128K of RAM is available
for system/user use. BASIC version 7.0 is the default language. The Commodore Kernal
is supported in a compatible fashion.
The C l 28 also has a Z80 coprocessor that can make full use of system RAM and
Kernal utilities, intended for use with such powerful operating systems as CP/M version
3.0. The Banking ROM scheme allows function key software to be installed internal to
the system or added externally as an expansion cartridge.
Another major feature is the 8563 80-column display capability, available in C l 28
mode as an addition to the 40-column mode.
Peripheral support includes the Commodore mouse, joystick, paddle, light pen
interface (both 40- and 80-column light pens); the Commodore Datassette; the User Port,
which supports RS232; modems; the Expansion Bus, which supports external memory
expansion; and the Commodore standard Serial Bus, which supports all existing Serial
Bus components. There are also several features intended to reduce software overhead,
such as relocatable zero page and system stack.
In C64 mode, the standard sixty-six keys are available. In C l 28 mode, twenty-
six extra keys are available, including separate cursor keys, a H E L P key, addi
tional function keys, and a true C A PS L O C K key. The additional keys,
grouped into the alternate keypad, are user-definable, increasing the flexibility and user
friendliness of the system.
The following is a summary of C128 features:
■ C64 compatibility
■ 80-column display capability
■ Z80 coprocessor (CP/M version 3.0 (2 MHz))
a 2 MHz 8502 operation in 80-column mode
a 128K standard system RAM
■ 48K standard system ROM
■ 32K internal function ROM (optional)
■ 32K external cartridge ROM
■ Fast serial disk drive interface
■ Full keyboard, ninety-two keys with CAPS LOCK key, HELP key
and separate cursor-control keys.
C l 28 HARDWARE SPECIFICATIONS 557
SYSTEM ARCHITECTURE
The C l 28 computer utilizes a shared bus structure similar to that of the C64. The shared
bus emulates dual-port RAM and ROM, which allows the character ROM, color RAM
and system RAM to be shared by both the microprocessor and the 8564 VIC video
controller, with no interference to each other. This requires that the RAM be fast enough
to supply valid data in half the time of a normal microprocessor machine cycle. Normal
sharing is controlled by a coprocessor that will enable or disable the processor during
alternate halves of the machine cycle.
The C l28 system splits the address bus into shared and nonshared sections. All
normal 8502 I/O parts are on the nonshared address bus; the VIC chip and its associated
support chips are located on the shared bus. The VIC chip will gate processor addresses
onto the shared bus based upon its AEC control line. The data bus is common to both
sides of the address bus.
The processor interfaces with most of the system chips like a standard 6502 bus
cycle, where a machine cycle is equal to a clock cycle. This allows the use of 1 MHz
parts for a 1 MHz clock and eliminates the need to create valid address and data strobes,
as this information is now supplied by the edges of the master clock, <l>0. Chip selects
for the I/O and system ROM are generated by the PLA that tracks the microprocessor
addresses during (1>1.
For system RAM access, the row address is the address from the microproces
sor, and the column address is the MMU output address (called the T ra n sla ted A d d ress).
The Translated Address outputs are calculated by considering the contents of the
M M U’s Configuration Register and RAM Configuration Register. From these values,
the MMU generates either normal or translated addresses, a CAS selection. The
CAS-gating circuitry in the MMU enables either one of the two banks of 64K RAM in a
128K system. For VIC cycle access of RAM, the RAM bank is set independently of the
processor’s bank. A write to ROM will result in the write “ bleeding through” to RAM
underneath, while a read from ROM will always disable CAS in both banks. The MMU
allows custom arrangements of RAM for both banks 0 and 1.
Banking ROM is effected through the setting of bits in the Configuration Register
contained in the MMU and communicated to the PLA decoder. This allows Banking
Function ROM and any attached C128 cartridge to be included in the basic system
configuration.
The PLA generates chip selects for the color RAM, VIC control registers and
character ROM, which are used during processor and VIC cycles, as well as all chip
selects needed for the processor-only ROM and peripherals. To avoid bus contention,
the PLA must also generate CAS disable for any accesses to ROM or I/O devices.
The VIC chip generates the signals used to control Dynamic Memory and provides
macro-control functions such as RAM refresh. The VIC’s primary purpose is to fetch
screen data from memory, using either cycle sharing or DMA, and create an NTSC- or
PAL-compatible video output that is applied to a monitor or modulated and applied to a
TV set. The C128 provides outputs for Composite, Chroma/Luminance and RF video
558 COMMODORE 128
outputs from the VIC chip, as well as an edge-triggered light pen input latch going to
the VIC chip.
The output from the SID chip sound generator is buffered and applied directly to
an external amplifier, like that found in an external monitor, or modulated and repro
duced in the user’s television set. The SID chip also has an external input for mixing
another sound source.
The 8563 video control chip fetches screen data from a dedicated section of RAM
referred to as the display RAM and creates an RGBI (Red-Green-Blue-Intensity) output
for use with an external 80-column monitor. The 8563 also creates all needed signals for
dynamic refresh of its dedicated display RAM. The C l28 provides RGBI and composite
monochrome outputs from the 8563 chip.
The cassette port is implemented using the zero page ports available on the 8502
and software control of hardware handshaking. The Commodore serial bus port is imple
mented in a similar manner using a 6526 CIA for I/O. The serial bus works with
Commodore serial components, and in C64 mode is actually driven by the software
routines contained in the C64. The User Port is a multipurpose port comprising several
parallel port lines that support peripherals such as slow RS232, modem, etc. The
joystick ports are identical to those on the C64 and are implemented using a 6526 CIA
to read/write the port.
The video connector has composite video as well as separate chroma and lumi
nance outputs for use with monitors. The 1701 and 1702 Commodore monitors interface
directly to this connector. The RF output jack supplies an RF signal compatible with the
regulations for TV interface devices and is switch selectable between channels 3 and 4.
Both NTSC and PAL television standards are supported. The RGBI connector and
signal are similar to the ones used by IBM, and are compatible with most monitors
supporting Type I RGBI. Additionally, a composite monochrome signal is available on
the RGBI connector and is generally compatible with NTSC (or PAL) composite. Audio
is available only from the 40-column video/audio connector.
SYSTEM SPECIFICATION
This section discusses various features and constraints of the C128 system. Included are
descriptions of the system and its configurations and limiting factors such as power,
loading and environment.
Figure 16-1 shows the C l 28 system.
560 COMMODORE 128
SYSTEM BUS ARCHITECTURE

The buses described include:
■ Processor bus ■ Shared address bus

■ Translated address bus H Color data bus
■ Multiplexed address bus ■ Display bus
TH E PROCESSOR BUS
The Processor Bus includes the data and address buses that are connected directly to the
8502 processor. These buses are designated D0-D7 for the 8-bit data bus and A0-A 15 for the
16-bit address bus. These buses tie the processor to most of the system ROM and I/O
devices, including at least part of all system ROM, all built-in function ROM , the
M M U, the PLA, the 8563 video processor, the SID and both CIA chips.
The processor bus is in communication with the Z80 coprocessor as well. All
address lines are shared directly by both processors. In order to allow the Z80 to operate
on a 6502 family bus, it is necessary to latch data going into the Z80 and gate the data
leaving the Z80. Thus, the Z80 has a small local data bus, designated ZD0-ZD7. During
a write cycle, when AEC is high, Z80 data is gated to the processor bus. During a read
cycle, processor bus data is gated to the Z80 data bus. This read data is transparently
latched by the 1 MHz system clock.
NOTE: Read and write cycles referred to in this document, unless

otherwise specified, are 8502-type bus cycles. The Z80 Read Enable and
Write Enable outputs are conditioned using logic to interface with an
8502 bus cycle, so no distinction is made as to the differences between
cycles of the different processors. For more information on this logic,
consult the section on the Z80 processor and the C l 28 Schematic,
Commodore Part No. 310378.
As mentioned above, the Z80 is not in direct communication with the processor
data bus, because of the need to adapt the Z80 to 8502 bus protocol. Note, however,
that every other device and every other bus (except two that will be explained later)
shares the processor data bus as a common data bus.
TH E TRANSLATED ADDRESS BUS

Another C l 28 system bus is the Translated Address Bus produced by the MMU during
AEC high. This bus consists only of high-order addressing lines, designated TA8-TA 15.
These lines reflect the action of the MMU on the normal high-order address lines, which
may or may not include some sort of translation. The MMU can translate the address of
page zero or page one in normal operation, and it translates the Z80 address from $0000
through $0FFF in order to direct it to read the Z80 BIOS. A more complete description
of MMU translations can be found in the MMU section of this document.
Normally, the translated address bus indirectly drives the system RAM and the VIC
C128 HARDWARE SPECIFICATIONS 561
chip by driving the multiplexed address buses. It directly drives system ROM 4 address
line 12 to allow the Z80 ROM relocation. Finally, this bus becomes address lines 8
through 15 of the C64 compatible expansion port.
During a VIC cycle or a DMA, the MMU pulls TA 12-T A 15 high, while TAg-TAn
are tri-stated. This allows the VIC chip to drive TA8-T A n as VIC addresses VAg-V A u .
During an external DMA cycle, the TA lines of the MMU are tri-stated, and the TA
bus, presumably driven by the DMA source, in turn drives the processor address bus
from Ag to A 15. Thus allows the DMA source to access any peripheral chip except the
MMU. The action of the VIC during a VIC cycle is described below.
TH E M ULTIPLEXED ADDRESS BUS

This section describes two related address buses, the Multiplexed Address Bus and the
VIC Multiplexed Address Bus, known respectively as MA0-M A 7 and VMA0-V M A 7.
The VIC multiplexed address bus is created during AEC high by multiplexing the
high-order translated address bus (TA8-T A 15) with the low-order processor address bus
(A0- A 7), controlled via the MUX signal. This bus, driven by a hardware multiplexer
through series resistors, is called the Multiplexed Address Bus. The VIC multiplexed
address bus is used for processor access of the VIC chip registers. It is also used for VIC
access of system RAM.
During a VIC cycle (AEC low), the VIC chip address lines will be asserted. There
is no completely separate address bus for the VIC addresses, so it shares the VMA0-V M A 7
and address lines that are tri-stated during AEC high. Most of the VIC addresses come
out of the VIC chip already multiplexed, but two of them, VA6 and VA7, do not supply
column information, as the VIC chip supplies only 14 bits of addressing. The higher-
order address bits V A 14 and V A ,5 come from CIA-2, as in the C64. This means the VIC
supplies complete VMA0-V M A 7 for a VIC DRAM access or DRAM refresh. The
TA8- T A U supplied by VIC are used in conjunction with another addressing bus for
nonmultiplexed VIC cycle addresses, such as character ROM and color RAM accesses.
TH E SHARED ADDRESS BUS

The Shared Address Bus is a nonmultiplexed address bus used by both the processor
and the VIC chip to communicate with common resources, namely the character ROM
and color RAM (and the 8563 system RAM indirectly). During AEC high, the shared
address bus, designated Sa0-S A 7, is driven by A0- A 7, the low er-order processor
address bits. The higher-order bits needed are supplied by the translated address bus,
which is also a shared address bus. Thus, the processor is able to access both shared
items.
During AEC low, the VIC addresses VA0-V A 7 (VMA0-V M A 7) must come onto
the shared address bus. Since VA0-V A g are actually multiplexed, the row address only
must be sent to the shared address bus. Thus, the multiplexed VIC addresses are
transparently gated when either /RAS or MUX are high, but latched and held afterward
so that when combined with the column address, the full address is presented. The high-
order address bits here are supplied by the shared translated address bus. Note that the
shared address bus provides the lower 8 bits of the expansion port address, allowing VIC
access to cartridges and some additional drive capability by way of the TTL chips used to
562 COMMODORE 128
drive the shared address bus. During DMA, the SA lines, like the TA, are driven
backward to drive the processor bus. As noted above, this allows peripheral chips, ROM
and RAM to be accessed by a DMA source, like the RAM expansion module. Only the
MMU and the 8563 video controller cannot be accessed during DMA. See the auto-start
ROM section in Chapter 13 for more details on initializing both external and internal
expansion ROMS.
TH E COLOR D A TA BUS
The color RAM is written to or read from a nybble data bus called the Color Data Bus.
During AEC high, the color data bus is connected to the lower half of the processor data
bus via an analog switch, allowing the processor full access to the color FLAM. During
AEC low, that switch is opened, effectively isolating the color data bus from the
processor data bus. In this state, it is driven by the VIC extended data bus D 8- D n -
Since the color RAM is banked, only one-half appears here at a time.
TH E DISPLAY BUS
The Display Bus is a bus local to the 8563 video controller, consisting of the Display
Address (DA0-D A 7) and the Display Data Bus (DD0-D D 7). This local bus supports
the 8563 display RAM, which is completely isolated from the rest of the C l 28 system.
The display address bus is a multiplexed address bus providing addressing to the display
DRAM. The display data bus provides communication between this DRAM and the
8563. The 8563 also provides row and column strobes and dynamic refresh to this
DRAM.
SYSTEM MEMORY
O R G A N IZA TIO N
This section describes the C l 28 memory system. Figure 16-2 is a detailed diagram of
the C l 28 Memory Map.
CJ CD <->
*-< cr m cr
3=0_ siQ- —1 CL
CD cn CD
L,
Ll_ CD CD CD GD CD CD CD CD G3 CD CD CD C\J
$0000
Li_ CD CD CD CD CD CD CD CD CD CD CD CD CD
Ll_ CJ CD CD CD CD CD CD CD CD CD CD =r CD
li_ Li_ U_ LU □ CJ cr CO LO ZT OJ «—< CD CD
V* & -tA- ■W 4ft w- VI VI 4A V*
Figure 16-2. C l 28 M em o ry Map (C 64 M ode)

564 COMMODORE 128
C l 28 ROM MEMORY O R G A N IZA T IO N

C128 mode is achieved at system reset and is controlled by a bit in the MMU
Configuration Register. In C l 28 mode, the MMU asserts itself in the C l 28 memory
map at $FF00 and in the I/O space, starting at $D500. Use of MMU registers located at
$FF00 allows memory management without actually having the I/O block banked in at
the time and with a minimum loss of contiguous RAM. The MMU is completely
removed from the memory map in C64 mode. It is, however, still used by the hardware
to manage memory.
Figure 16-2 presents the standard map for the C64 mode. Some of the alternate modes
are shown in Figure 16-3. All C64 modes are compatible with the C64 computer, as the
C128 basically becomes a C64 when in C64 mode. The details of MMU register
location/operation are discussed in Chapter 13.
The ROMs in C64 mode look like Commodore 64 ROMs. The internal BASIC
and Kemal provide the C64 mode with the normal Commodore 64 operating system in
ROM. This ROM actually duplicates some of the ROM used in C l 28 mode, but it is
necessary, as it is not accessible from C128 mode. In C128 mode, up to 48K of the
operating system is present, with the exact amount being set by software control. This
allows quicker access to underlying RAM by turning off unneeded sections of the
operating system.
The external ROMs represented on the memory map are those used in C64 mode.
They obey the Commodore 64 rules for mapping; i.e., cartridges assert themselves in
hardware via the /EXROM and /GAME lines. External ROMs in C128 mode (i.e., C128
cartridges) are mapped as banked ROMs; when the system is initialized, all ROM slots
are polled for the existence of a ROM, and the ROM ’s priority, if one exists. This
allows much more flexibility than the hard-wired ROM substitution method, since the
Kemal and BASIC ROMs can be swapped for an application program, or for external
program control, or can be turned off altogether. This banking manipulation is accom
plished by writing to the Configuration Register at location $D500 or $FF00 in the
MMU.
The hardware also features the ability to store preset values for the configuration
and force a load of the Configuration Register by writing to one of the LCRs (Load
Configuration Registers). This allows the programmer to imply that ROM does not
appear in the bank (by default) any time an access occurs to a bank where data is stored.
=3 tr
a o)
Q .^
II II I II 030
x ~ -a
z r
d d liJ O
3 : rCEzCC UJ Oz
c c cc 3E X cc x x cc
O —■CE X o - a x
-J X O U J
151
0°= tr2
1 i
d J J d
GD CD CD GD
CD t_) CO
»A V>
■ -Q Q Q
II II II I acc o coc aa_ l
o
3: x ->
□ 0
d d liJ O ddUJO to cc c
x x s: x cc az 2: cc
o •—a: x O '- 'G X
_J X CD UJ
r\ \ \ \
J J
0 .- 0
II II II II O O
3
1O
es : z
aariua X CE UJ O
x x x x
o >-■a x
X XXX
o « a x
—I X LO UJ —I X CD UJ
5Z X l\ ■ \, \, \
o
J J
o O a
UJ CD CJ
V* W -IA
Figure 16-3. C64 A lte rn a te M em ory C onfigurations

566 COMMODORE 128
C l 28 RAM MEMORY O R G A N IZA T IO N

As shown in Figure 16-2, the RAM present in the system is actually composed of two
banks of 64K by 8 bytes of DRAM. The RAM is accessed by selecting one of the two 64K
banks, the maximum address range of the 8502 and Z80, according to the RAM banking
rules set in the RAM configuration register of the MMU. This area shown as RAM
represents what the processor would see if all ROM were disabled. Bank switching can
be accomplished in one of two ways.
The bank in use is a function of the value stored in the configuration register. A
store to this register will always take effect immediately. An indirect store to this
register, using programmed bank configuration values, can be accomplished by writing
to one of the indirect load registers known as LCRs, located in the $FF00 region of
memory. By writing to an LCR, the contents of its corresponding PCR (PreConfiguration
Register) will be latched into the configuration register. This allows the programmer to
set up four different preprogrammed configurations that allow each bank to be personal
ized ahead of time; e.g ., bank 1 as a data bank might be strictly a RAM bank with no
ROM or I/O enabled, while bank 0 as the system bank can have the system ROM and
I/O enabled by default. Additionally, reading any LCR will return the value of its
corresponding PCR.
W hen dealing with 64K banks of memory at once, it may be desirable to bank in
bank 1 but still retain the system RAM (stack, zero page, screen, etc.). The MMU has
provisions for what is referred to as com m on RAM. This RAM does not bank and is
programmable in size and position (top, bottom, or both) in the memory map. The size
is set by bits 0 and 1 in the RAM Configuration Register (RCR). If the value of the
bits is 0, IK will be common. Values of 1, 2 and 3 produce common areas of 4K, 8K
and 16K respectively. If bit 2 of the RCR is set, bottom memory is held common; if bit
3 is set, then top memory is common. In all cases, common RAM is physically located
in bank 0.
Zero page and page one can be located (or relocated) independently of the
RCR. When the processor accesses an address that falls within zero page or page
one, the MMU adds to the high-order processor address the contents o f the P0
register pair or the PI register pair, respectively, and puts this new address on the
bus, including the extended addressing bit A16. RAM banking will occur as appropriate
to access the new address. Writes to the P0 and PI registers will be stored in pre
latch until a write to the respective low byte page pointer occurs. This prevents a PxH
high byte page pointer from affecting the translated address until both high and low
bytes have been written. Common memory overrides the page pointers.
At the same time, the contents of the P0 and PI registers are applied to a digital
comparator, and a reverse substitution occurs if the address from the 8502 falls within
the page pointed to by the register. This results in a swapping of the zero page or page
one with the memory that it replaced. Swapping occurs only if the swapped area is
defined as RAM; i.e ., system or function ROM must always be at its assigned addresses
and thus should not be back-substituted but of course will not cause contention of any
kind. Note that upon system reset, the pointers are set to true zero page and true page
one.
N O TE: There are actually several memory modes that override parts of
the bank as selected here. These modes are mentioned below, and are
covered in detail in a later section.
For VIC-chip access, one bit in the MMU status register substitutes for extended
address line A 16, selecting the proper CAS enable to make it possible to steer the VIC
to anywhere in the 128K range. Note that AEC is the mechanism the MMU uses to steer
a VIC space address; i.e., when AEC is low, a VIC access is assumed. This results in
the VIC bank being selected as well for an outside DMA, since this, too, will pull the
AEC line to the MMU low.
PERFORMANCE SPECIFICATIONS
POWER C O N S U M P TIO N
Table 16-1 contains values for C l 28 power consumption, including various add-on
options.
I.C . 1 Typ I Max I Typ 1 Max I Typ I Max Unit

(5v) (9v) (12v)
SUBTOTAL 61 2.63 3.83 0.03 0.05 0 0 A
PERIPH ERA LS
Magic Desk 1 0.24 0.35 0 0 0 0 A
RS-232 1 0.07 0.10 0 0 0 0 A
Auto Modem 1 0 0 0.07 0.10 0 0 A
Cassette 1 0 0 0.28 0.40 0 0 A
PE RIPH ER A L TOTAL 4 0.31 0.45 0.35 0.50 0 0 A
C128 AND PERIPH ERA LS 2.94 4.28 0.38 0.55 0 0 A
TOTAL PO W ER , W ORST CASE (WATTS)
5V 12V
Voltage 5.25 9.45 12.60 total
Power 22.47 5.20 0.00 27.67
T able 16-1. C l 28 P o w er C on su m p tion
BUS LO AD IN G
Table 16-2 details AC and DC device loading for the C l28 system. All capacitances
are in picofarads and currents are in microampheres.
568 COMMODORE 128
Dev: DRAM ROM VIC 8563 4016 2332 LS MMU PLA 8502 Z80 4066 SID TOTAL
Signal TTL CIA LOADS
T A .-T A ,,
# ... 1 1 — 1 1 2
I ... 10 2.5 — 10 10 800 833
c ... 8 8 — 4 8 30 58
sa 0-s a 7
# ... ............................... 1 1 1
I ... ........................ 10 10 400 420
C ... ........................ 4 8 15 27
d „-d 7
# 2 5 1 1 ... 1 1 1 1 3
I 20 50 2.5 2.5 — — 400 2.5 — 2.5 — 0.1 7.5 488
C ... 40 8 8 — — 15 8 8 8 24 129
R/w
# ................ 1 1 .............. 2 1 1 1 .............. 3
I ................. 2.5 2.5 .............. 800 2.5 2.5 2.5 .............. 7.5 820
C ................. 8 8 .............. 30 8 8 8 .............. 24 94
RES
# ........................... 1 ........................ 1 — 1 1 — 3
I ........................... 2.5 ........................ 2.5 — 2.5 2.5 — 7.5 18
C ........................... 8 ........................ 8 — 8 8 — 24 56
1MHz
I .......................................................... 2000 ............................................. 7.5 2008

C .......................................................... 75 ............................................. 24 99
T ab le 16-2. Bus Loading Pow er Requirem ents (Capacitances and C u rren ts)
EN VIR O N M EN TAL SPECIFICATIONS

The C128 is rated to operate at from 10 to 40 degrees Celsius (50 to 104 degrees
Fahrenheit) and at a noncondensing relative humidity of 5 to 95 percent.
C l28 HARDWARE SPECIFICATIONS 569
THE 8502
MICROPROCESSOR
GENERAL DESCRIPTION
The 8502 is an HMOSII Technology microprocessor, similar to the 6510/6502. It is the
normal operating processor used in C64 and C l28 modes. Software-compatible with
the 6510, hence the 6502, the 8502 also features a zero page port used in memory
management and cassette implementation.
The 8502 is also specified for operation at 2 MHz. The 2 MHz operation is made
possible by removing the VIC from the system as a display chip. (The VIC chip is never
completely removed from the C128 system, as it continues to function as clock
generator and refresh controller.) What this refers to is that the VIC is removed as a display
chip and co-processor; thus the full clock cycle can be devoted to processor functioning
instead of sharing the cycle with the VIC.
The task of the video display processor is taken over by the 8563, which can
function without the need for bus sharing. Since the I/O devices, SID, etc., are rated at
1 MHz only, stretching of the 2 MHz clock is used to allow these parts to be accessed
directly by the 2 MHz processor and still keep throughput to a maximum.
The I/O devices are not affected by the 2 MHz operation, as they are still driven
by a 1 MHz source (as such, all timer operations remain unchanged), and clock
stretching is used only to synchronize the 2 MHz machine cycle to the 1 MHz CK) high
time. The clock sources and clock-stretching capabilities are generated by the VIC chip.
ELECTRICAL SPECIFICATION
This section describes some of the electrical constraints and specifications of the system.
M A X IM U M RATINGS
Table 16-3 gives the absolute maximum ratings of the 8502 microprocessor.
RATING SYMBOL VALUE UNIT
Supply Voltage v cc -0.5 to +7.0 Vdc

Input Voltage Vin -0.5 to +7.0 Vdc
Storage Temperature Tstg -55 to + 150 °C
Operating Temperature Ta 0 to +70 °C
T ab le 16-3. 8502 Absolute M ax im u m Ratings
ELECTRICAL CHARACTERISTICS
Table 16-4 gives the 8502’s basic electrical specifications for minimum, typical and
worst-case operation, valid over the range of operation T.
570 COMMODORE 128
CH ARACTERISTIC SYMBOL MIN TYP MAX UNIT
Input High Voltage VlH

<t>o (in) Vss + 2.4 Y cc Vdc
/RES, P0- P 7,/IRQ, Data Vss + 2.2 — — Vdc
Input Low Voltage VlL

4>o (in) Vss-0 .3 Vss + 0.5 Vdc
/RES, P0- P 7,/IRQ, Data — — Yss + 0.8 Vdc
Input Leakage Current IlN

(Vin = 0 to 5.25V , Vcc = 5.25V)
Logic 2.5 |xA
<l>o (in) — — 10.0 |xA
3-State (Off) Inp. Cur. Its i

(Vin = 0.4 to 2.4V , Vcc = 5.25V)
Data Lines 10.0 jxA
Output High Voltage VOH

(IOH = -100|xA dc, Vcc = 4.75V)
Data, A0-A is , R/W , P 0- P 7 Vss + 2.4 Vdc
Output Low Voltage

( I 0 L = l-6m A dc, Vcc = 4.75V)
Data, A o-A] 5 , R/W , Po-P7 V ss + 0.4 Vdc
Power Supply Current Lc — 125 — mA
Capacitance c
(Vin = 0, Ta = 25 C, f = 1MHz)
Logic, Po-P7 r'-'in
. 10 pF
Data C0ut — — 15 pF
A0- A 7 C0ut — — 12 pF
4>o C<})0 — 30 50 pF
T a b le 16-4. 8502 Basic Electrical Specifications
SIGNAL DESCRIPTION
Below is a description of all the 8502 signals from a functional and electrical point of
view:
C L O C K (<£0)— This is the dual speed system clock. Note that the input level required is
above worst-case TTL; thus, extra precautions must be taken when attempting to
drive this input from a standard TTL level input.
A D D R E S S B U S (A 0- A | 5) — TTL output. Capable of driving 2 TTL loads at
130 pF.
D A T A BUS (D o- D 7)— Bidirectional bus for transferring data to and from the device
and the peripherals. The outputs are tri-state buffers capable of driving 2 standard
TTL loads at 130 pF.
RESET — This input is used to reset or start the processor from a power down
condition. During the time that this line is held low, writing to or from the
processor is inhibited. When a positive edge is detected on the input, the
processor will immediately begin the reset sequence. After a system initialization
time of 6 cycles, the mask interrupt flag will be set and the processor will load the
program counter from the contents of memory locations $FFFC and $FFFD. This is
the start location for program control. After Vcc reaches 4.75 volts in a power up
routine, reset must be held low for at least 2 cycles. At this time the R/W line will
become valid.
IN T E R R U P T R E Q U E S T (IR Q ) — TTL input; requests that the processor initiate an
interrupt sequence. The processor will complete execution of the current instruc
tion before recognizing the request. At that time, the interrupt mask in the Status
Code Register will be examined. If the interrupt mask is not set, the processor will
begin an interrupt sequence. The Program Counter and the Processor Status
Register will be stored on the stack and the interrupt disable flag is set so that no
other interrupts can occur. The processor will then load the program counter from
the memory locations $FFFE and $FFFF.
N O N -M A S K A B L E IN T E R R U P T R E Q U E S T (N M I)— TTL input, negative edge
sensitive request that the processor initiate an interrupt sequence. The processor
will complete execution of the current instruction before recognizing the request.
The Program Counter and the Processor Status Register will be stored on the
stack. The processor will then load the program counter from the memory
locations $FFFA and $FFFB. Since NMI is non-maskable, care must be taken to
insure that the NMI request will not result in system fatality.
A D DR ESS E N A B LE C O N T R O L (A E C ) — The Address Bus is only valid when the
AEC line is high. When low, the address bus is in a high impedance state. This
allows D M A ’s for dual processor systems.
I/O P O R T (Po-P6)— Bidirectional port used for transferring data to and from the
processor directly. The Data Register is located at location $0001 and the Data
Direction Register is located at location $0000.
R/W — TTL level output from processor to control the direction of data transfer between
the processor and memory, peripherals, etc. This line is high for reading memory
and low for writing.
R D Y — Ready. TTL level input, used to DMA the 8502. The processor operates
normally while RDY is high. When RDY makes a transition to the low state, the
processor will finish the operation it is on, and any subsequent operation if it
is a write cycle. On the next occurrence of read cycle the processor will halt,
making it possible to gain complete access to the system bus.
PROCESSOR TIM IN G
This section explores the timing considerations of the 8502 processor unit. Table 16-5 is
a processor timing chart. Figure 16-4 presents timing diagrams that show both general
timing and the particular method of clock stretching used in 2 MHz mode.
572 COMMODORE 128
Electrical Characteristics Vcc = 5v ± 5% , Vss = Ov, Ts = 0 °C to 70 °C
CHARACTERISTIC SYMBOL MIN MAX UNITS
AEC setup time T aec 25 60 ns

Up data setup from <|jo T mds 100 ns
Up write data hold T hw 40 ns
Data bus to tri-state from AEC T aedt 120 ns
Read data stable T dsu 40 ns
Read data hold T hr 40 ns
Address setup from <j>o T ads 40 75 ns
Address hold T ha 40 ns
Address setup from AEC T aads 60 ns
Address tri-state from AEC T aeat 120 ns
Port input setup TpDSU 105 ns
Port input half TpDH 65 ns
Port output data valid Tpow 195 ns
Cycle time T cyc 489 ns
()>o (in) pulse width @ 1.5v (crystal clock) P wh4>0 235 265 ns
4 ) 0 (in) rise time TR<t>o 10 ns
4>o (in) fall time TF4>o 10 ns
RDY setup time T rdy 80 ns
T a b le 16-5. Processor T im in g C h a rt
CLOCK STRETCHING
When running in 2 MHz mode, the processor clock sometimes must be stretched. This
is handled by the VIC chip, the processor and the PLA working together. When an I/O
operation is decoded during a 2 MHz cycle, the phase relationship between the 2 MHz
and 1 MHz clocks must be considered. If the 2 MHz access occurs during 1MHz <E>1, the
access to a clocked I/O chip would be out of synchronization with the 1 MHz clock that
drives all I/O chips. Thus, during this phase relationship, /IOACC from the PLA signals
the VIC chip to extend the 2 MHz clock. Should the 2 MHz cycles take place during the
1 MHz 4>2 cycle, no special attention is necessary.
1 MHz
Clock “ 1
2 MHz
Clock
I/O fl ee PHI 1
Pr o c I/O
H cce ss
Clock
I/O fl ee PHI 2
I/O
P r o c __ Access
Cl ock L
Figure 16—4. Clock S tretching in 2 M H z Mode
Note the speed implications of this. In 2 MHz mode, half the I/O access given will
occur at an effective speed of 1MHz.
Figure 16-5 is a diagram of the 8502 microprocessor pin configuration.
574 COMMODORE 128
VUD
PH I _ 0 R 15 2 3
R 14 22
R 13 2jf
R 12
£4 P6 R11 ?
25
P5 RIG 7
26
P4 R9 ]|
27
P3 R8 j|
U P2 R7 ]I
2_9_
PI R6 4
3? P0 R5 _2
R4 7
R3 ~0
O J
R2
CD R1
LD R0
03
J? RDT
/NMI
□7 3^
DB
40 /RES D5 3 3
D4 3J4
. 3 D3 3 5
/IR Q
D2 H
39 R/ W D1 37,
D0
x 5 REC
v ss
21
Figure 16-5. T h e 8502 Microprocessor Pin C onfiguration
For information on programming 8502 Machine Language, see Chapter 5, Machine

Language on the C l 28.
Z80 MICROPROCESSOR
HARDWARE SPECIFICATION
The Z80 microprocessor is used as a secondary processor in the C l28 to run CP/M
based programs. Covered in this section is not only the operation of the Z80 as part of
the C128 system, but some important electrical and timing specifications of the Z80. For
more information on Z80 bus interfacing, consult the Zilog Z80 Data Book.
SYSTEM DESCRIPTION
The Z80A, a 4MHz version of Zilog’s standard Z80 processor, is included as an
alternate processor in the C l 28 system. This allows the C l28 to run the CPM 3.0
operating system at an effective speed of 2 MHz. The Z80 is interfaced to the 8502 bus
interface and can access all the devices that the Z80 can access. The bus interface for the
Z80 (the most complex part of the Z80 implementation) is described in this section,
along with Z80’s operation as a coprocessor in the C128 system.
N O TE: See the Signal Description section later in this chapter for
definitions of the signals mentioned in the following paragraphs.
BUS INTERFACE
Because a Z80 bus cycle is much different than a 65xx family bus cycle, a certain
amount of interfacing is required for a Z80 to control a 65xx-type bus. Since the Z80
has built-in bus arbitration control lines, it is possible to isolate the Z80 by tri-stating its
address lines. Thus, both the Z80 and the 8502 share common address lines.
The interfacing of the data lines is more complex. Because of the shared nature of
the bus during Z80 mode, the Z80 must be isolated from the bus during AEC low. Thus,
a tri-statable buffer must drive the processor bus during Z80 data writes. The reverse
situation occurs during a Z80 read— the Z80 must not read things that are going on
during AEC low; it must latch the data that was present during AEC high. Thus, a
transparent latch drives the data input to the Z80. It is gated by the Z80 read-enable
output, and latched when the 1 MHz clock is low. It will be seen that the Z80 actually
runs during AEC low, but that the data bus interfaces with it only during AEC high.
C O N TR O L INTERFACE
The Z80 control read-enable interfacing must provide useful clock pulses to the Z80,
and must tailor the Z80 and write-enable signals for the 8502-type bus protocol. The
Z80 clock is provided by the VIC chip, and is basically a 4MHz clock that occurs only
during AEC low, as seen in Figure 16-6. This ensures that the Z80 is clocked only when
it is actively on the bus. One additional consideration in clocking the Z80 is that while
all of the 8502 levels and most of the Z80 levels are TTL-compatible, the Z80 clock
input expects levels very close to 5 volts. For that reason, the output from the VIC chip
is processed by the 9-volt supply; thus, the 9-volt circuit must be operational for the
Z80, and the system, to function. The most common power-up failure for the C128 is a
blown 9-volt fuse.
576 COMMODORE 128
Sy s ten
Clock
Z-8 0
Clock
Z-8 0
0ddr Address Valid
Z-80
0a t a D a ta V a l i d
Read
0a t a
Latch
0a t a Va l i d
Figure 16-6. Z 8 0 Bus T im in g
PROCESSOR SW ITC H IN G
It is important in normal operation for the Z80 and the 8502 to operate as coprocessors,
communicating with each other. Since only one processor may have the bus at any one
time, this is only serial coprocessing, not parallel coprocessing or multiprocessing. This
is important in several ways:
First, the C l 28 system must power-up with the Z80 as master processor. This is to
prevent the Z80 from accidentally accessing the bus when powering up. Thus, the Z80
is made master on power-up and can do anything it likes to the bus. Also, because the
Z80 can start-up certain C64 applications that would cause the 8502 to crash, the Z80 is
the logical choice as start-up processor. After some initializations, the Z80 starts-up the
8502 in either C l 28 or C64 mode, depending upon whether a cartridge is present, and
upon the type of cartridge, if one is present. The operating system also allows C64 mode
to be forced on power-up.
Second, processor switching allows the Z80 to access 8502 Kernal routines. For
standardized programs or for any I/O operation not supported in the Z80 BIOS, the Z80
can pass on the task of I/O to the 8502. Since the Z80 sees BIOS ROM where the 8502
sees its pages 0 through F, the Z80 can operate without fear of disrupting any 8502
pointers or the stack in RAM Bank 0. The Z80 ROM BIOS physically overlays that
critical section of RAM Bank 1.
The Z80 can receive a bus grant request from the MMU via /Z80EN, or from the
VIC chip via BA. Since the VIC control line is used for DMAs, the latter request is not
of immediate concern. The /Z80EN action, however, is important, since it is the
mechanism by which the two processors exchange control.
When the /Z80EN line goes high, it triggers a Z80 /BUSRQ. The Z80 then relin
quishes the bus by pulling /BUSACK low. This action drives the 8502 AEC high and
(providing VIC does not request a DMA) also drives the 8502 RDY line high, enabling
the 8502. To switch back, a low on the Z80 /BUSRQ will result in Z80 /BUSACK
going high, tri-stating and halting the 8502.
See Appendix K on CP/M for interchip communication details.
SIGNAL DESCRIPTION
The list below defines each Z80 signal. The Z80 pin configuration is shown in Figure 16-7.
Address Bus (A o -A |5): 16-bit tri-stating address bus. Used for 16-bit I/O cycles. This
allows up to 256 input or 256 output ports. During refresh time, the lower 7 bits
contain a valid refresh address. (This signal is not used in the C128 system.)
D a ta Bus (D o -D 7): Input/output bus capable of tri-stating; used for 8-bit exchanges
with memory and I/O devices.
Machine Cycle O n e (/M |): Output, active low. This signal indicates that the current
machine cycle is the operation code fetch of an instruction execution. During
execution of a two-byte opcode, /Mi is generated, as each byte is fetched. /Mj also
occurs with an input/output request (/IORQ) to indicate an interrupt acknowledge
cycle. The M, line is used to disable the I/O decoder during an interrupt acknowl
edge cycle (See Input/Output Request).
M em o ry Request (/M R E Q ): Active low, tri-state output that indicates that the address
bus holds a valid address for a memory read or write operation.
In p u t/O u tp u t Request (/IN R Q ): Active low, tri-state output. The /INRQ signal
indicates that the lower half of the address bus holds a valid address for an I/O
read or write operation. An /INRQ signal is also generated with a /M] signal when
an interrupt is being acknowledged to indicate that an interrupt response vector
can be placed on the data bus. An interrupt can acknowledge during /M j; I/O
operations never occur during /M ,.
M em o ry Read (/R D): Active low, tri-state output. /RD indicates that the CPU wants
to read data from memory or from an I/O device. This signal is generally used to
gate-read data onto the data bus.
M em o ry W rite (/W R): Active low tri-state output. /WR indicates that the data bus
holds valid data to be processed by memory or by an I/O device.
Refresh (/R FSH ): Active low output used to indicate that the address bus holds a
refresh address in its lower 7 bits. Thus, the current /MREQ signal should be used
to do a refresh read to all dynamic memories not refreshed from an alternate source.
A7 is set to 0 and the upper 8 bits contain the I register at this time.
H a lt State (/H A L T ): Active low output, indicating that the Z80 has executed a halt
instruction and is awaiting some kind of interrupt before execution can continue.
W hile in the halt state, the Z80 continuously executes NOPs to continue refresh
activity.
W a it (/W A IT ): Active low input, used to drive the Z80 into wait states. As long as this
signal is low, the Z80 executes wait states; thus, this signal can be used to access
slow memory and I/O devices.
578 COMMODORE 128
N O TE: While the Z80 is in either a wait state or a bus acknowledge

(/BUSAK) state, a dynamic memory refresh cannot be performed. See Bus
Acknowledge.
In te rru p t Request (/IN T ): Active low input, driven by external devices. If the
interrupt flag IFF is enabled and the bus request (/BUSRQ) line is not active, the
processor honors the request interrupt at the end of the current instruction. When
the Z80 acknowledges an interrupt, it generates an interrupt acknowledge signal
(/INRQ during /M i) at the beginning of the next instruction cycle. There are three
different modes of response to a given interrupt. See Bus Request.
N on-M askable In te rru p t (/N M I): Active low input. This interrupt is edge-triggered
and cannot be masked against. It is always recognized at the end of the current
instruction, forcing the Z80 to restart at location $0066. The program counter is
automatically saved in the stack to allow a return from the interrupt program. Note
that continuous cycles can delay an /NMI by preventing the end of the current
cycle, and that /BUSRQ will override /NMI.
Reset (/RESET): Active low input that forces the program counter to zero and
initializes the Z80, which will set interrupt mode 0, disable interrupts, and set
register I and R to 0. During /RESET, the address and data buses go tri-state and
all other signals go inactive.
Bus Request (/B U S R Q ): Active low input that requests the CPU address, data and
tri-statable output control signals to go tri-state for sharing and DM A’s. The lines
go tri-state upon termination of the current machine cycle.
Bus Acknowledge (/B U S A K ): Active low output, used to indicate to any device
taking over the bus that the Z80 has gone into tri-state and the bus has been
granted.
Clock (4>): Single phase system clock.
27 /Ml 7 AG 30
L 8 G A1 31
J^ -C /MEMREQ A? 32
/INRQ AS 33
/RD A4 34
/WR ' AS 35
AR 36
/RFSH A7 37
^ < 3
AR 38
39
-c /HALT AS
A 1n 40
/WHIT A11 1
A1? 2
/ INT A 1S 3
/NMI A14 4
A 15 5
/RESET
/BUSRQ DG 14
/BUSRK ni 15
n? 12
PHI ns 8
11 +5V A4 7
29 GND ns 9
ns 10
D7 13
Figure 16-7. Z 80 Microprocessor Pin Configuration
ELECTRICAL SPECIFICATIONS
ABSOLUTE M A X IM U M RATINGS
Table 16-6 gives the absolute maximum temperature, voltage and power dissipation
ratings for the Z80. Permanent damage is likely to occur if these ratings are exceeded.
580 COMMODORE 128
PARAM ETER SYMBOL RANGE UNITS
Operating Temperature Ta 0 to + 7 0 °C
Storage Temperature Tst -6 5 to + 1 5 0 °c
Input Voltage Vin -0 .3 to 7.0 Vdc
Power Dissipation Pcc 1.5 W
T a b le 16-6. Z 80 A bsolute M a x im u m Ratings
DC OPERATING CHARACTERISTICS
Table 16-7 shows the maximum DC operating ratings for the Z80. Except as noted,
these ratings apply over the full rated temperature and voltage ranges.
PARAM ETER SYMBOL RANGE UNITS
Power Supply Variance vcc 5 ± 5% Vdc

Clock Input Low Voltage VILC -0 .3 to 0.8 Vdc
Clock Input High Voltage V,HC Vcc-0 .6 to Vcc + 0.3 Vdc
Input Low Voltage V,L - 0 .3 to 0.8 Vdc
Input High Voltage VlH 2.0 to Vcc Vdc
Output Low Voltage V0 L Vss to 0.4 Vdc
(IOL= 1.8mA)
Output High Voltage VoH 2.4 to Vcc Vdc
(Ioh = -250(j.A)
Power Supply Current Icc 200 mA
Input Leakage Current I li 10 (JtA
(Vjn = 0 tO Vcc)
Tri-State Leakage Current I lo +10, -1 0 |jiA
(V o U T V()H> VOUT = VoiJ
Data Bus Input Leakage Current In) ±10 (j,A
(V
V < V-
T ss — T in <
— V¥ cc/)
T a b le 16-7. Z80 D C O p e ra tin g Characteristics
C A PA C ITA N CE
The line capacitance values for the Z80 are given in Table 16-8. All measurements are
at T = 25 degrees, F = 1 MHz.
PARAM ETER SYMBOL MAXIMUM UNIT
Clock Capacitance C4> 35 pF

Input Capacitance Cin 5 pF
Output Capacitance Cout 10 pF
T ab le 16-8. Z80 Capacitance Values

THE PROGRAMMED
LOGIC ARRAY (PLA)
The 8721 C128 PLA is a programmed version of the Commodore 48 Pin Programmable
Logic Array (Commodore Part #315011). It provides all the chip selects and other
decoded signals that are necessary for the C64, along with a number of such signals
new in the C128 system. Figure 16-8 shows the PLA chip.
The PLA does a number of things vital to the operation of the C128, including:
■ All ROM selects (Kernal, BASIC, function, external) in all operating modes.
■ VIC chip select.
■ Color RAM chip select.
■ Character RAM chip select.
■ Write enable to color RAM.
■ Latched write enable to DRAMs.
■ Z80 select decoding.
■ Z80 I/O decoding, for Z80 I/O cycle and Z80 memory mapping.
■ Data bus direction signal.
■ I/O group chip select (includes I/O -1, 1/0-2, CIA-1, CIA-2, SID, 8563).
■ I/O access signal indicating an I/O operation isoccurring,
a CAS ENaBle for DRAM enable.
CHIP SELECT GENERATION

This PLA device is responsible for defining the banking rules for ROM and RAM that
the system will follow. This chip generates chip selects for all ROM and the VIC chip.
It generates an enable for any other I/O device in the map, and can enable or disable
CAS based upon what else is enabled. In C128 mode, decisions are made using the
processor addresses and the four mode status lines: ROMBANKLO, ROMBANKH1,
I/O SELECT, and C128/64. The C128 mode banking scheme is quite straightforward
and simple. In Z80 mode, the selection mechanism becomes even simpler, thanks to the
I/O cycle of the Z80 processor.
C64 chip selects account for the bulk of the PLA font. The C64 selects I/O, RAM
and ROM based upon internal control lines: BA, HIRAM, LORAM, and CHAREN.
The status of these lines, and decoded addresses, determine for any given time which (if
any) chip is selected. When a cartridge is inserted, two additional control lines come
into play: /EXROM and /GAME. Various combinations of these lines cause different
memory maps to be asserted, all based upon the PLA font.
O THER FU N C TIO N S
The PLA performs a variety of functions other than chip selects. It creates the write
enable strobes for both DRAM and Color RAM. In C128 mode, the C64 control lines
(HIRAM, LORAM and CHAREN) are not needed, since the MMU controls the more
sophisticated C l 28 method of banking. Thus, these lines are used to extend the
functionality of the C l 28 at little or no additional cost in hardware. The CHAREN line is
582 COMMODORE 128
used in C l 28 mode to turn the Character ROM on and off in the VIC bank selected; in
C l 28 mode the ROM can appear or disappear in any VIC bank.
The second of the new functions uses LORAM and HIRAM to select one of two
Color RAM banks. The level of LORAM selects the bank that will be seen during
processor time; the level of HIRAM selects the bank that will be seen during VIC time.
Thus, a program can swap between two full-color pictures very cleanly, or the processor
can modify one full-color picture while displaying another.
el
v c c
R 1 5 / R 0 M 4
|2 H 1 4 / R 0 M 3
y +
3 -->.36
H 1 3 / R 0 M 2
y +
li / R O M 1
^37
0 1 2
5
f i l l
6 / F R O M ■ ji
A 1 0
21
B A P L A
47 / R O M H
C L K
10 R / W / R O M L
19 M A 4
20 M A 5 / C H A R O M
26
/ V A 1 4
9 / C O L R A M
A E C
1 R 0 M B A N K L 0 / V I C
17 R O M B A N K H I
IS / 1 O S E / 1 0 A C C
15 / I O C S
C 1 2 8 / 6 4
11 / G A M E / S D E N
12 / E X R O M D I R
39
1 2 8 / 2 5 6 / ' D W E
/ G W E
L O R A M
H I R A M / C A S E N B
C H A R E N
32
/ D M A A C K C L R B N K
/ Z 8 0 _ E N
/ Z 8 0 _ I 0
G N D
Figure 16-8. T he C l 28 PLA Chip

THE MEMORY
MANAGEMENT UNIT (MMU)
The MMU is designed to allow complex control of the C l28 system memory resources.
Because of the way it handles all the standard C64 modes of operation, it is completely
compatible with the C64. Additionally, it controls the management of particular C l 28
modes including the Z80 mode. The M M U’s features include:
■ Generation of translated address bus (TA8-TA15).

■ Generation of control signals for different processor modes (C l28, C64, Z80).
■ Generation of CAS selects lines for RAM banking.
■ Generation of ROMBANK lines for ROM banking.
Programming the MMU is described in detail in Chapter 13.
PHYSICAL DESCRIPTION
Many of the MMU input and output signals have been discussed informally so far.
This section contains descriptions of the MMU as a physial 48-pin device, includ
ing a description of all pin requirements, input and output signals, and electrical
requirements.
PIN REQUIREMENTS
Table 16-9 lists the required MMU pins, indicating the number in each signal category
and the total number.
584 COMMODORE 128
SIGNAL NAMES DESCRIPTION NUMBER O F PINS
A 0- A 3 , A 8- A 15 Address Lines In 12
A 4/5, A 6/7 Combined Address Lines In 2
D 0- D 7 Data Lines In/Out 8
T A s-T A 15 Translated Addr. Lines Out 8
Vee + 5V 1
GND Ground 1
PHI„ 2 M Hz 4>„ Clock In 1
RESET System Reset In 1
R/W Read/W rite Line In 1
/CAS 0-/C A S , DRAM CAS, 64K Bank Out
AEC Address Enable Control In 1
/Z 8 OEN Z80 Enable Out 1
/GAME Game ROM Enable In, Control Out 1
/EXROM External ROM Enable In, Control Out 1
M So-M S, Memory Status Out
I/O (M S2) I/O Select Out 1
C128/64 (M S3) C128 or C64 M ode Out 1
40/80 40/80 Status In 1
/FSDIR Fast Serial Direction Out 1
MUX M emory Multiplex In 1
TOTAL 48
Table 16-9. M M U Pin Requirements
PIN DESCRIPTION
The following comprises a signal-by-signal description of the MMU input and output
signals. Figure 16—9 shows the MMU pin configuration. Included here are any available
bond options.
The MMU input signals are:
■ A0- A 3, A8- A 15: Addresses from the microprocessor. Used to derive chip
selects as well as multiplexed address lines. A0- A 3 are found at pins 18-21 on
the MMU, while A8- A i 5 are located at pins 24—31.
b A4/5, A6/7: Combined addresses from the microprocessors. Used along with
simple addresses, combined in this fashion to lower the pin count of the
MMU. Located at pins 22 and 23 respectively.
h PHIo: Presystem clock. Used for early transition of gated signals on write
operations. Processor address is valid on the rising edge, and data is valid on
the falling edge. This is found at pin 33.
■ R/W: System Read/Write control line. This input is high for a processor read,
low for a processor write. This signal is located at pin 32 on the MMU chip.
■ RESET: System Reset. This input initializes internal registers on a power-up
or hardware reset. It can be found at pin 2.
■ AEC: Address Enable Control. Indicates whether the 8502 processor or the
VIC has access to the shared bus. When low, VIC or an external DMA has the
CI28 HARDWARE SPECIFICATIONS 585
bus and V A 16 have the processor bus, and no pointer or BIOS translation takes
place. This signal occupies pin 16.
■ MUX: The memory multiplex signal, used to clock various sections of the
MMU. It is located at pin 17.
■ Vdd: System + 5 Vdc supply, connected at pin 1.
■ Vss: System Ground, connected at pin 34.
The following represents the MMU bidirectional lines. Some of the port bits
detailed here are left for future expansion in a one-directional sense.
■ D0-D 7: Data inputs from microprocessor. Used for writing to internal regis
ters. Located at pins 35 to 42.
■ /EXROM: This signal is used to sense the /EXROM line on the expansion
connector in C64 mode and as an expansion control line in C l28 mode.
Located at pin 46. This line will drive one TTL load on output, and has a
passive depletion mode pull-up on input. This signal can be pulled down, but
not up, by an external driver.
n /GAME: This signal is used to sense the /GAME line on the expansion
connector in C64 mode and as the color RAM bank control line in C l 28 mode.
passive depletion mode pull-up on input. External hardware can pull this line
down, but not up.
■ 40/80: This port in input mode senses the 40/80 column switch. It detects
whether or not this switch is closed. Its output function is open for expansion.
passive depletion mode pull-up on input. External hardware can pull this line
down, but not up.
■ FSDIR: This port in output mode is used to control the data direction of the fast
serial disk interface. It is a general-purpose port signal, and is connected at pin
44. This line will drive one TTL load on output, and has a passive depletion
mode pull-up on input. External hardware can pull this line down, but not up.
The following list represents the MMU output signals, their physical locations on
the MMU and their logical levels if applicable.
a TA8- T A 15: Translated address outputs. Tri-stated for VIC cycles during AEC,
they provide translated physical addresses for use on the Multiplexed Address
Bus and the Shared Static Bus. TA 12 to T A )5 are each defined to have an
internal, depletion mode pull-up with an equivalent resistance of 3.3KO. TAg
to T A n each go tri-state during VIC time (AEC low). These are located on the
MMU at pins 10 to 3.
■ MS q-M S i : Also called ROMBANKO and ROMBANK1, these outputs control
ROM banking for all ROM slots. They are located on pins 15 and 14. These
lines are used to decode ROM bank selection for any ROM access in C l 28
mode. If they are both low, a system ROM has been selected. If MS] alone is
586 COMMODORE 128
high, then a built-in function ROM has been selected. If MS0 alone is high,
then an external function ROM has been selected. Finally, if both are high, the
RAM that occupies the particular slot has been selected. In C64 mode, the
PLA completely ignores these lines.
■ I/O: This output is used to select memory mapped I/O in C128 mode. It is on
pin 13, and is also known as MS2. In C128 mode, this line always reflects the
polarity of the I/O bit. It is ignored by the PLA in C64 mode, and remains
high throughout C64 mode.
■ C128: This output directs the system to act in either C128 or C64 mode. It is
located on pin 47, and is also known at MS3. It goes low to indicate C64
mode, high for C l 28 mode.
■ /Z80EN: This output is used to enable the Z80 processor and disable the
normal operation of the 8502 processor. It can be found at pin 43. It goes low
to indicate Z80 mode, high for all other modes.
■ /CAS q-C A S ,: CAS enables to control RAM banking. CAS0 enables the first
bank of 64K; CAS) enables the second bank of 64K. These are pins 12 and
11, respectively.
ABSOLUTE M A X IM U M RATINGS
IT E M SY M B O L RANGE U N ITS
Irtput Voltage v in -2.0 to +7.0 Vdc

Supply Voltage Vcc -2.0 to +7.0 Vdc
Operating Temperature Ta 0 to 70 °C
Storage Temperature Ts, -55 to 150 °C
Table 16-10. Absolute Maximum Ratings for the MMU
M A X IM U M OPERATING C O N D IT IO N S
IT E M SY M B O L RANGE U N IT S
Voltage Variance Vcc 5.0 ± 5% Vdc

Input Leakage Current Ii -1 .0 |xA
Input High Voltage V ,H Vss + 2.4 to Vcc+ 1 .0 Vdc
Input Low Voltage V lL Vss—2.0 to Vss + 0.8 Vdc
Output High Voltage VoH VS5+ 2.4 Vdc
( I o h = -200(xA. Vcc = 5V ± 5%)
Output Low Voltage VoL Vss + 0.4 Vdc
(IOL = -3 .2 m A , Vcc = 5V ± 5%)
Maximum Input Current Icc 250 mA
T able 16 - II. M axim um O p e ra tin g C onditions fo r th e MMU

j l
V O D
R 1 5 M U X
R 1 4 R / W
fl 1 3 R E C
H 1 2
H I 1
fl 1 G / Z 8 0 E N
A 9
R 8 T R 1 5
T R 1 4
A 6 / 7 T R 1 3
F m / 5 T R 1 2
T R I 1
R 3 T R 1 0
R 2 T H 9
H I T R 8
MMU
D 7
D 6
0 5 C 1 2 8
D 4 I / O
D 3 R O M B R N K H I
D 2 R O M B R N K L O
D 1
D 0 / C R S 0
/ C H S 1
P H I __ 0
/ E X R O M R E S E T
/ G R M E
4 0 / 8 0 F S D I R
V S S
34[
Figure 16-9. T h e M em ory M an ag em en t U nit Pin C onfiguration

588 COMMODORE 128
THE 8564
VIDEO INTERFACE CHIP
The 8564 VIC chip used in the C l 28 is an updated version of the VIC chip used in
current C64 systems. It contains all the video capabilities of the earlier 6567 VIC chip,
including high-resolution bit-mapped graphics and movable image blocks. It also sup
ports new features used by the C128 system, including extended keyboard scanning. Its
register map is upwardly compatible with the old VIC, allowing compatibility in C64
mode. It is powered by a single, 5V DC source instead of the two sources required by
the old VIC chip. The 8564 pin configuration is shown in Figure 16-10.
GENERAL DESCRIPTION
The 8564 VIC chip is similar to the 6567 VIC chip, yet it supports many new
features unique to the C128 system requirements. It will run on a single 5V DC supply,
and is packaged in a 48-pin dual-in-line package.
8564 VIC OPERATIONS

The 8564 VIC supports all the operations of the previous VIC chip. These functions,
quickly summarized, are:
■ Standard color character display mode.

■ M ulticolor character display mode.
■ Extended color character display mode.
■ Standard bit map mode.
■ M ulticolor bit map mode.
■ Movable image blocks.
■ Movable image block magnification.
■ Movable image block priority
■ Movable image block collision detection
■ Screen blanking
■ Row/column display select
■ Smooth scrolling
■ Light pen
■ Raster compare interrupt
As these functions exist in the previous VIC, their description is purposely kept to
a minimum here, while the VIC programming information is contained in Chapter 8. The
new functions, however, are described in detail below.
EXTEND ED KEYBOARD SC AN N IN G
The 8564 contains a register called the K ey b o a rd C o n tro l R e g iste r . This register
allows scanning of three additional keyboard control lines on the C128 keyboard. Thus,
the C128 keyboard can have advanced additional keys in C128 mode while still
retaining C64 keyboard compatibility in C64 mode. In this register (register 53295
($D02F)), bits 0 -2 are directly reflected in output lines K0 to K2, while bits 3-7 are
unused, returning high when read.
2 M Hz OPERATION
The VIC chip contains a register that allows the C l28 system to operate at 2 MHz
instead of the standard 1 MHz speed of the C64. This operating speed, however,
disallows the use of the VIC chip as a display processor. This is bit 0 in 53296 ($D030),
and setting this bit enables 2 MHz mode. During 2 MHz operations, the VIC is disabled
as a video processor. The processor spends the full time cycle on the bus, while VIC is
responsible only for dynamic RAM refresh. Clearing this bit will bring back 1 MHz
operation and allow the use of the VIC as a video display chip. During refresh and I/O
access, the system clock is forced to 1 MHz regardless of the setting of this bit.
The 2MHz speed is available in C64 mode by setting bit 0 of location 53296
($D030). Prior to this, blank the screen by clearing bit 4 of location 53265 ($D011).
You can then process at 2MHz (to perform number crunching, for example); however,
you will have no visible VIC screen. To return, set bit 4 of 53265 ($D011) and clear bit
0 of 53296 ($D030).
Bit 1 of this register contains a chip-testing facility. For normal operation, this bit
must be clear. None of the other bits in this register is connected.
SYSTEM CLOCK CO NTR O L

The new VIC chip generates several clocks used by the C128 system. The main clock is
the 1 MHz clock, which operates at approximately 1 MHz at all times. Most bus
operations and all I/O operations take place in reference to this clock. The next clock to
consider is the 2 MHz clock. This clock clocks selected system components, such as the
processor, at 2 MHz when in 2 MHz mode. The VIC chip monitors the /IOACC input,
which indicates the access of an I/O chip, and when asserted, will stretch the 2 MHz
clock to synchronize all 2 MHz parts with the 1 MHz I/O parts. Finally, the last clock is
the Z80 clock, which is a 4 MHz clock that takes place only during the low half of the 1
MHz clock. Since all timed I/O parts look only at the 1 MHz clock, all I/O timings
remain the same no matter what the 2 MHz clock is doing.
SIGNAL DESCRIPTION
The VIC chip is mounted in a 48-pin dual-in-line package. The following lists describe
the electrical signals that it generates.
The signals used in the system interface are:
■ D 0 - D 7: These are the bidirectional data bus signals. They are for communica
tion between the VIC and the processor, and can be accessed only during AEC
high. They occupy pins 7 through 1 and 47, respectively.
■ D g- D n : These are the extended data bus signals. They are used for VIC
communication with the color RAM. They occupy pins 47 to 43 in order.
■ /CS: Chip select, used by the processor to select the VIC chip. Found at pin 13.
■ R /W : Standard 8502 bus read/write for interface between the processor and
the various VIC registers. Pin 14.
590 COMMODORE 128
■ A0- A 6: Multiplexed address lines, pins 32 through 38. During row address
time, A0- A 8 are driven on A0- A 5. During column address time, A8- A 13 are
driven on A0- A5 and A6 is held at 1. During a processor read or write, A0- A 5
serve as address inputs that latch on the low edge or /RAS.
■ A -A i0: Static address lines, pins 39 through 42. These address lines are used
for non-multiplexed VIC memory accesses, such as to character ROM and
color RAM.
■ 1MHz: The 1 MHz system clock, pin 18. All system bus activity is referenced
to this clock.
■ 2MHz: This is the changing system clock, which will be either 1 MHz or 2
MHz. If the 2 MHz bit is clear, no VIC or external DMA is taking place, and
no I/O operation is occurring, the clock will be 2 MHz; otherwise it will be 1
MHz. Found at pin 23.
B Z80 Phi: The special 4 MHz Z80 clock, pin 25.
■ /IOACC: From PLA, indicating an I/O chip access for clock stretching. Pin 22
of the VIC.
■ /RAS: Row address strobe for DRAMs, pin 19.
■ /CAS: Column address strobe for DRAMs, pin 20.
■ MUX: Address multiplexing control for DRAMs, pin 21.
■ /IRQ: Interrupt output, used to signal one of the various internal interrupt
sources has taken place. Requires a pull-up, found at pin 8 on the chip.
® AEC: Address Enable Control, high for processor enable on the shared bus,
low for VIC cycle and VIC or external DMA. Found at pin 12 of the VIC.
■ BA: Bus Available signal, used to DMA processor. Pin 10.
■ K 0- K 2: Extended keyboard strobe bits, pins 26 to 28.
■ LP: Edge triggered latch for light pen input. Pin 9 of the VIC chip.
The signals comprised in the video interface, i.e ., all the signals required to create
a color video image, are:
■ PH IN: The fundamental shift rate clock, also called the dot clock. Used as the
reference for all system clocks. Located at pin 30.
■ PH CL: The color clock, used to derive the chroma signal. Pin 29.
® SYNC: Output containing composite sync information, video data and lumi
nance information. Requires a pull-up, pin 17.
■ COLOR: Output containing all color-based video information. Open source
output, should be tied through a resistor to ground. Found at pin 16.
ELECTRICAL SPECIFICATION
Tables 16-12 and 16-13 specify the electrical operation of the VIC chip in its new form.
These specs include absolute maximum ratings and maximum operating conditions.
ITEM SYMBOL RANGE UNIT
Input Voltage Vin -0 .5 to + 7 .0 Vdc

Supply Voltage VCC -0 .5 to + 7 .0 Vdc
Operating Temperature Ta 0 to 75 °C
Storage Temperature Ts, -6 5 to 150 °C
Table 16-12. Absolute Maximum Ratings for V IC Chip
Below is a list of the maximum operating specifications for the new VIC chip:
ITEM SYMBOL RANGE UNIT
Power Supply Variance V cc 5.0 ± 5% Vdc

Input Leakage Current II -1 .0 |xA
Input High Voltage V ,H Vss + 2.0 to Vcc Vdc
Input Low Voltage V IL Vss-0 .5 to Vss + 0.8 Vdc
Output High Voltage VoH V ss + 2.4 Vdc
( I o h = -2 0 0 | xA , V cc= 5.0 ± 5% Vdc)
Output Low Voltage VOL V ss + 0.4 Vdc
( I 0 L = -3 .2 m A , Vcc = 5.0 ± 5% Vdc)
M ax Power Supply Current Lc 200 mA
Table 16-13. Maxim um Operating Specifications for V IC Chip
RASTER REGISTER
The Raster Register is a dual function register. A read of the raster register 53266 ($D012)
returns the lower 8 bits of the current raster position [the MSB-RC8 'is located in register
53265 ($D011)]. A write to the raster bits (including RC8) is latched for use in an internal
raster compare. When the current raster matches the written value, the raster interrupt
latch is set. The raster register should be interrogated to prevent display flicker by
delaying display changes to occur outside the visible area. The visible area of the
display is from raster 51 to raster 251 ($033-$0FB).
INTERRUPT REGISTER
The Interrupt Register indicates the status of the four sources of interrupts. An interrupt
latch in register 53273 ($D019) is set to 1 when an interrupt source has generated an
interrupt request.
592 COMMODORE 128
LATCH ENABLE
BIT BIT W HEN SET
IRST ERST Actual raster count = stored raster count (bit 0)

IMDC EMDC M OB-DATA collision (first bit only) (bit 1)
IMMC EMMC MOB-MOB collision (first bit only) (bit 2)
ILP ELP First negative transition of LP per frame (bit 3)
IRQ When IRQ/ output low (bit 7)
Table 16-14. Interrupt Register Definitions
To enable an interrupt request to set the IRQ/ output to 0, the corresponding interrupt
enable bit in register 53274 ($D01A) must be set to “ 1” . Once an interrupt latch has
been set, the latch may be cleared only by writing a “ 1” to the associated bit in the
interrupt register. This feature allows selective handling of video interrupts without
software storing the active interrupts.
SCREEN BLANKING
The display screen can be blanked by clearing the BLN K bit (bit 4) in register 53265
($D011) to zero (0). When the screen is blanked, the entire screen displays the exterior
color specified by register 53280 ($D020). When blanking is enabled, only transparent
(phase 1) memory accesses are required, permitting full processor utilization of the
system bus. However, sprite data will be accessed if the sprites are not also disabled.
DISPLAY ROW /COLUMN SELECT

The normal display screen consists of 25 rows of 40 character regions per row. For
special display purposes, the display window is reduced to 2.4 rows of 38 characters.
There is no change in the format of the display information, except that characters (bits)
adjacent to the exterior border area are covered by the border.
RSEL NUMBER OF ROWS CSEL NUMBER OF COLUMNS
0 24 rows 0 38 columns
1 25 rows 1 40 columns
The R SEL bit (bit 3) is in register 53265 ($D011), and the C SEL bit (bit 3) is in
register 53270 ($D016). For standard display, the larger display window is normally
used, while the smaller display window is normally used in conjunction with scrolling.
SCROLLING
The display data may be scrolled up to one character in both the horizontal and vertical
directions. When used in conjunction with the smaller display window (above), scrolling
can be used to create a smooth panning motion of display data while updating the
system memory only when a new character row (or column) is required. Scrolling is
also used for centering a display within the border. An example of horizontal smooth
scrolling is found in Chapter 8.
BITS REG ISTER FUNCTION
0 -2 53270 ($D016) Horizontal Position

0 -2 53265 ($D011) Vertical Position
LIG H T PEN
The light pen input latches the current screen position into a pair of registers (LPX,
LPY) on a low-going edge. Since the X position is defined by a 9-bit counter (53267
($D013)), resolution to 2 horizontal dots is provided. Similarly, the Y position is latched
in register 53268 ($D014) with 8 bits providing unique raster resolution within the
visible display. The light pen latch may be triggered only once per frame, and subse
quent triggers within the same frame will have no effect.
For more information on programming the VIC (8564) chip, see Chapter 8, The
Power Behind Commodore 128 Graphics.
594 COMMODORE 128
48
VuU
7
D0 PH_I N30^
6
Dl PH_CL 29^
o 5
D2
4
D3 KG2 6
^3 D U K1 2 7
2
D5 K2 2 8
o 1
D6
47 D7 LP
46 D8
^ 5 D9 Z80_PH 25^
DIO 1MHZ■iô
43
Dll 2MHZ 23^
/ 1ORCC 22^>
3±
Rl 1 CO /RRS 1 9 ,
3 2
RQ LO /CRS 20,
^33 Rl MUX 2_1^
3}l_ R2 CO
035 R3 R/W14
036 R U /CS 13
31_
R5 /IRQ
3Q R6
.39 R7 BR 10^
o4 0
R8 REC 12c
41
R9
; 42
RIO 11
15
a16 COLOR
,17
SYNC
vss
I24
Figure 16-10. The 8564 V IC Chip
THE 8563
VIDEO CONTROLLER
The 8563 is a HMOSII technology custom 80-column, color video display controller.
The 8563 supplies all necessary signals to interface directly to 16K of DRAM, including
refresh, and generated RGBI for use with an external RGBI monitor. For more informa
tion on the 8563 video controller, see Chapter 10, Programming the 80-Column (8563) Chip.
GENERAL DESCRIPTION
The 8563 is a text display chip designed to implement an 80-column display system with
a minimum of parts and cost. The chip contains the high-speed pixel frequency logic
necessary for 80-column RGBI video. It can drive loads directly, though some buffering
is desirable in most real-world applications. The chip can address up to 64K of DRAM
for character font, character pointer, and attribute information. The chip provides RAS,
CAS, write enable, address, data and refresh for its subordinate DRAMs. A program
mable bit selects either two 4416 DRAMs (16K total) or eight 4164 DRAMs (64K total)
for the display RAM. The C l 28 system uses the 4416 DRAMs.
EXTERNAL REGISTERS
The 8563, which sits at $D600 in the C128, appears to the user as a device consisting of
only two registers. These two registers are indirect registers that must be programmed to
access the internal set of thirty-seven programming registers. The first register, located
at $D600, is called the Address Register. Bit 7 of $D600 is the Update Ready Status Bit.
W hen written to, the five least significant bits convey the address of an internal register
to access in some way. On a read of this register, a status byte is returned. Bit 7 of this
register is low while display memory is being updated, and goes high when ready for the
next operation. The sixth bit will return low for an invalid light pen register condition and
high for a valid light pen address. The final register indicates with a low that the scan is
not in vertical blanking, and with a high that it is in vertical blanking.
The other register is the Data Register. It can be read from and written to. Its
purpose is to write data to the internal register selected by the address register. All
internal registers can be read from and written to through this register, though not all of
them are a full 8 bits wide.
IN TERNAL REGISTERS
There are thirty-seven internal registers in the 8563, used for a variety of operations.
They fall into two basic groups: setup registers and display registers. Setup registers are
used to define internal counts for proper video display. By varying these registers, the
user can configure the 8563 for NTSC, PAL or other video standards.
The display registers are used to define and manipulate characters onthe screen.
Once a character set has been downloaded to this chip, it is possible to display
80-column text in 4-bit digital color. There are also block movement commands that
remove the time overhead needed to load large amounts of data to the chip through the
two levels of indirection. Figure 16-11 is a display of the 8563 internal register map.
596 COMMODORE 128
Bit 7 Bit 6 Bit 5 Bit 4 Bi t 3 | Bi t 2 | Bi t 1 | Bi t 0

R0 Q H o r i z o n t o l T o to 1
R01 Horizontal Displayed
R02 Horizontol Sync Position
R03 Vertical Sync Width Horizontol Sync Width
R04 V e r t ica 1 To ta l
R05 Vertical Total Adjust
R06 V er t ica 1 3 is p 1a y e d
R07 V e r t i c a l Sy nc P o s i t i o n
R08
nvwWwwivww'wwu y YViy\y \y w \y \v \y\y \vu Y\Y\Y\V\Y\Y\YIVIYIY\YU y \y;y w \y \y \y\y \y \y \v \yiy\y \y \V\y \y \yu '\y \y \y \yi Interlace Mode
R3 9 U1.I1.IU1.II.IU1.I1.I1.II.II L n e r e c t e r lots 1 V er ti c al
R 10 C u r s o r Mode' Cursor Stort Scon _lne
Rl 1
nv\v\v\v\Y\Y\v\v\v\v\vivmv\YW\vw\v\viv\v\nv\Y\v\v\Yw\Ymvumi Cursor End Scon Line
R 12 Display Stort Hddress (high)
R 13 D is p 1oy S t o r t H d d r e s s (Low)
R14 Cursor Position (High)
R 15 Cursor Position ( L ow)
R 16 Light Pen Vertical
R 17 L ig ht Pen H o r i z o n t a l
R 18 U p d a t e L o c o t io n (Hig h )
R 19 Update Location ( L ow)
R20 Attribute Start Hddress (High)
R2i H t tr b u t e S t a r Hddress (Low)
R22 C h a r a c t e r Io t o 1 - H o r i z o n t a 1 C h o r o c t e r U i s p 1o y e d - H o " i z o n t o 1
\wn w wwwwivvvwo axxwn wvvwiYwivvvn VYY\V\Y\Y\V\V\Y\\VYYYYY Chorocter Uispieyed-Vsrticol
R23 ui.IJ.ll.IJ.ll.lljj.II.IJ.IJ.IJ UJ.IJ.IJ.IJ.IJ.IJ.IJ.IJ.tUJ.il 1.I1.II.IUUUJ.IUJ.IJ.1J.IJ
R24 C o p y / F iI 1 R e v S c r e e n B l i n k R a t e Verticol Smooth Scroll
R25 G r e p h / T e x t fltrb £ n b Sen igraph P i x Ubl Horizontol Smooth Scroll
R26 Foreground Color Background Colo"
R27 Address Increnent per Row
R28 Character Set Address U 1 R U / L I U 1 R |Y\Y!Y\Y\Y\Y\Y\Y\Y\Y\YYYl\'\Y\YlY\Y\Y\YlYu\Y\Y\Y[Y\Y\Y\Y\Y\YW\YIVU'IYVYIYIYIY\Y\Y\Y\Y\Y\V\YYYIY
R29 nVVrtVWWlYto'lVlYlYWlVWVYWWVYWlVlVri'lVH'KYWWWn'WWWlYVYlYVl Underline Scan Line
uui.ii.ii.ii.ii.ii.ii.ii.ij.ii.ij.ii.ij.ii.ii.ii.ii.iui.ii.ii.iijiui.ii.ii.ii.ii.iuui.ii.ii.ii
R30 Word Count ( c o u n t - 1)
R31 CPU Head/Write Uata
R32 block C o p y S o u r c e H d d r e s s (High)
R33 block Copy Source Hddress (L ow)
R3U Display Lnable begin
R35 D i s p l a y En a b l e End
R36 LUJUlJUUUIMlJlJWJUUlJUl.ll.ll.ll.ll.ll.lllll.lUULII.U.II.IUl.lLll.lih.lLI/JLII.IUIJUIJIJIJl URHM Refresh per Scan Line
Figure 16-11. 8563 Register Map
SIGNAL DESCRIPTION
There are many different signals involved with the 8563 chip, but they can be divided
into three general categories. The CPU interface signals serve as an interface to the 8502
bus. The local bus management signals serve to maintain the local memory bus. Finally,
the video interface signals are the signals that are necessary to provide an RGBI image
on an RGBI monitor. The 8563 pin configuration is shown in Figure 16-12.
TH E CPU INTERFACE
The 8563 chip interfaces to the 8502 bus using a minimum of signals. This is due
mainly to the local memory used by the 8563. The CPU interface signals are as follows:
■ D0- D 7: Bidirectional data bus allowing data to be passed between the 8563
and 8502. Found on pins 18 to 13, 11 and 10.
■ CS: Chip select input. This input must be high for selection and proper
operation of the chip. Located at pin 4.
■ /CS: Chip select input. This input must be low for selection and proper
operation of the chip. Located at pin 7.
■ /RS: Register Select input. A high allows reads and writes to the selected data
register. A low allows reads of the status register and writes to the address
register. In the system, this line is tied to AO. It is located at pin 8.
■ R/W: This line controls the data direction for the data bus. This is a typical
8502 control signal. Found at pin 9.
■ INIT: Active low input for clearing internal control latches, allowing the chip
to begin operation following initial power-on. Connect to /RES in the C l 28,
at pin 23.
■ DISPEN: Display Enable output signal, unused in the C l 28. Found at pin 19.
■ RES: This input initializes all internal scan counters, but not control registers.
It is not actively used in the C l 28 circuit, and is not found at pin 22.
■ TST: Pin used for testing only, tied to ground in the C128. Located at pin 24
of the chip.
TH E LOCAL BUS M ANAG EM ENT INTERFACE

The local bus management interface is a group of signals generated by the 8563for the
management of local video DRAM. This local DRAM both simplifies the addition of an
80-column video display to a system and enables a computer system with a limited
address space to support an 80-column display without taxing its limited memory
resources.
■ DD0-D D 7: Bidirectional local display DRAM data bus, comprising pins 35-36
and 38-42.
■ DA0-D A 7: Local display DRAM multiplexed address bus. Takes up pins
26-33.
■ DR/W: Local display DRAM Read/Write, pin 21.
■ /RAS: Row Address Strobe for local DRAM, pin 47.
■ /CAS: Column Address Strobe for local DRAM, pin 48.
TH E V ID EO INTERFACE
The final set of 8563 signals are the video interface signals. These signals are directly
related to the display video image.
■ DCLK: Video Dot Clock, determines the pixel width and is used internally as
the timing basis for all synchronized signals, such as character clock and
DRAM timing. Found at pin 2.
■ CCLK: The character clock output, unused in the C l28 system, and found at pin 1.
■ LP2: Input for light pen; a positive going transition on this input latches the
vertical and horizontal position of the character being displayed at that time.
Found at pin 25.
■ HSYNC: The horizontal sync signal, fully programmable via internal 8563
registers, and found at pin 20.
■ R ,G ,B,I: The pixel data outputs. They form a 4-bit code associated with each
pixel, containing color/intensity information, allowing a total of sixteen colors
or gray shades to be displayed. Located at pins 46, 45, 44, 43, respectively.
598 COMMODORE 128
I 37 u
■ ,12 D D 7 VDD C S D R 7
33g
; 141 D D 6 32;
D R 6
oiS. D D 5 D R 5
31
3 39 3Q ^
D D 4 D n ^
y 38 29g
D D 3 D R 3
a 36 28
D D 2 D H 2
a 35 21
D D 1 D H 1
0 314 _26^
D D 0 D R 0
3 IB D7 o n / c n s
. 11 D6 H7o
/ R f l S
3 13
D5
CD D R / H
21
3m D 4
LO
0 15
D3
0 16
D2 0 0 R 46
45^
a 17
D1 G
. 18
DO B
13^
I
; 8
/RS
7
/CS VSTN 20g
0 9
R/W HSYN
3
3 23
/RES
1
CCLK
DISPEN 19
a 25
/LP2
2i4p
TST
^ 2
/DCLK INIT
2gg
VSS
112
Figure 16-12. T he 8563 Chip Pin C onfiguration

6581 SOUND INTERFACE DEVICE (SID)

CH IP SPECIFICATIONS
CONCEPT
The 6581 Sound Interface Device (SID) is a single-chip, three-voice electronic music
synthesizer/sound effects generator compatible with the 8502 and similar microprocessor
families. SID provides wide-range, high-resolution control of pitch (frequency), tone
color (harmonic content), and dynamics (volume). Specialized control circuitry mini
mizes software overhead, facilitating use in arcade/home video games and low-cost
musical instruments.
FEATURES
■ 3 TONE OSCILLATORS
Range: 0 -4 kHz
■ 4 WAVEFORMS PER OSCILLATOR
Triangle, Sawtooth,
Variable Pulse, Noise
B 3 AMPLITUDE MODULATORS
Range: 48 dB
a 3 ENVELOPE GENERATORS
Exponential response
Attack Rate: 2m s-8 s
Decay Rate: 6 m s-24 s
Sustain Level: 0-peak volume
Release Rate: 6 m s-24 s
a OSCILLATOR SYNCHRONIZATION
■ RING MODULATION
■ PROGRAMMABLE FILTER
Cutoff range: 30 H z-12 kHz
12 dB/octave Rolloff
Low pass, Bandpass,
High pass, Notch outputs
Variable Resonance
» MASTER VOLUME CONTROL
■ 2 A/D POT INTERFACES
■ RANDOM NUMBER/MODULATION GENERATOR
■ EXTERNAL AUDIO INPUT
600 COMMODORE 128
.. ^ ..... .. ’
C A P 1a 1 28 V DD
c a p 1b 2 27 AUDIO OUT
c a p 2A 3 26 EXT IN
c a p 2B 4 25
<
o
o
RES 5 24 POTX
<t>2 6 23 POTY
R/W 7 6581 22 Dj
SID
CS 8 21 06
Ao 9 20 d5
Ai 10 19
a2 11 18 d3
12 17 d2
A3
a4 13 16 Di
GND 14 15 Do
Figure 16-13. 6581 SID Pin Configuration

a CL CL CL
POT X
M ---------
POTS
POT Y ------------------- H
Figure 16-14. 6581 Block D iag ram

602 COMMODORE 128
DESCRIPTION
The 6581 consists of three synthesizer “ voices” which can be used independently or in
conjunction with each other (or external audio sources) to create complex sounds. Each
voice consists of a Tone Oscillator/Waveform Generator, an Envelope Generator and an
Amplitude Modulator. The Tone Oscillator controls the pitch of the voice over a wide
range. The Oscillator produces four waveforms at the selected frequency, with the
unique harmonic content of each waveform providing simple control of tone color. The
volume dynamics of the oscillator are controlled by the Amplitude Modulator under the
direction of the Envelope Generator. When triggered, the Envelope Generator creates an
amplitude envelope with programmable rates of increasing and decreasing volume. In
addition to the three voices, a programmable Filter is provided for generating complex,
dynamic tone colors via subtractive synthesis.
SID allows the microprocessor to read the changing output of the third Oscillator
and third Envelope Generator. These outputs can be used as a source of modulation
information for creating vibrato, frequency/filter sweeps and similar effects. The third
oscillator can also act as a random number generator for games. Two A/D converters are
provided for interfacing SID with potentiometers. These can be used for ‘ paddles” in a
game environment or as front panel controls in a music synthesizer. SID can process
external audio signals, allowing multiple SID chips to be daisy-chained or mixed in
complex polyphonic systems. For full register descriptions, see Chapter 11, Sound and
Music on the Commodore 128.
SID PIN DESCRIPTION

CAP I A, CAP IB, (PINS 1,2)/ CAP2A.CAP2B (PINS 3,4)
These pins are used to connect the two integrating capacitors required by the program
mable filter. C l connects between pins 1 and 2, C2 between pins 3 and 4. Both
capacitors should be the same value. Normal operation of the filter over the audio range
(approximately 30 H z-12 kHz) is accomplished with a value of 2200 pF for C l and C2.
Polystyrene capacitors are preferred and in complex polyphonic systems, where many
SID chips must track each other, matched capacitors are recommended.
The frequency range of the filter can be tailored to specific applications by the
choice of capacitor values. For example, a low-cost game may not require full high-
frequency response. In this case, large values for C l and C2 could be chosen to provide
more control over the bass frequencies of the filter. The maximum cutoff frequency of
the filter is given by:
FCmax = 2.6E-5/C
where C is the capacitor value. The range of the filter extends nine octaves below the
maximum cutoff frequency.
RES (PIN 5)
This TTL-level input is the reset control for SID. When brought low for at least ten 4>2
cycles, all internal registers are reset to 0 and the audio output is silenced. This pin is
normally connected to the reset line of the microprocessor or a power-on-clear circuit.
4>2 (PSN 6)
This TTL-level input is the master clock for SID. All oscillator frequencies and
envelope rates are referenced to this clock. <J>2 also controls data transfers between SID
and the microprocessor. Data can only be transferred when <|>2 is high. Essentially, 4>2 acts
as a high-active chip select as far as data transfers are concerned. This pin is normally
connected to the system clock, with a nominal operating frequency of 1.0 MHz.
R/W (PIN 7)
This TTL-level input controls the direction of data transfers between SID and the
microprocessor. If the chip select conditions have been met, a high on this line allows
the microprocessor to Read data from the selected SID register and a low allows the
microprocessor to Write data into the selected SID register. This pin is normally
connected to the system Read/Write line.
CS (PIN 8)
This TTL-level input is a low active chip select which controls data transfers between
SID and the microprocessor. CS must be low for any transfer. A Read from the selected
SID register can only occur if CS is low, <j>2 is high and R/W is high. A Write to the
selected SID register can only occur if CS is low, <j>2 is high and R/W is low. This pin is
normally connected to address decoding circuitry, allowing SID to reside in the memory
map of a system.
A0-A4 (PINS 9-13)

These TTL-level inputs are used to select one of the 29 SID registers. Although enough
addresses are provided to select one of thirty-two registers, the remaining three register
locations are not used. A Write to any of these three locations is ignored and a Read
returns invalid data. These pins are normally connected to the corresponding address
lines of the microprocessor so that SID may be addressed in the same manner as
memory.
G N D (PIN 14)
For best results, the ground line between SID and the power supply should be separate
from ground lines to other digital circuitry. This will minimize digital noise at the audio
output.
D 0-D 7 (PINS 15-22)

These bidirectional lines are used to transfer data between SID and the microprocessor.
They are TTL compatible in the input mode and capable of driving two TTL loads in the
output mode. The data buffers are usually in the high-impedance off state. During a
W rite operation, the data buffers remain in the off (input) state and the microprocessor
supplies data to SID over these lines. During a Read operation, the data buffers turn on
and SID supplies data to the microprocessor over these lines. The pins are normally
connected to the corresponding data lines of the microprocessor.
604 COMMODORE 128
P O T X ,P O T Y (PINS 24,23)
These pins are inputs to the A/D converters used to digitize the position of potentiome
ters. The conversion process is based on the time constant of a capacitor tied from the
POT pin to ground, charged by a potentiometer tied from the POT pin to + 5 volts. The
component values are determined by:
RC = 4.7 E -4
Where R is the maximum resistance of the pot and C is the capacitor.

The larger the capacitor, the smaller the POT value jitter. The recommended
values of R and C are 470 kO and 1000 pF. Note that a separate pot and cap are required
for each POT pin.
v cc (PIN 25)
As with the GND line, a separate -1-5 VDC line should be run between SID Vcc and the
power supply in order to minimize noise. A bypass capacitor should be located close to
the pin.
E X T IN (PIN 26)
This analog input allows external audio signals to be mixed with the audio output of SID
or processed through the Filter. Typical sources include voice, guitar, and organ. The
input impedance of this pin is on the order of 100 kfl. Any signal applied directly to the
pin should ride at a DC level of 6 volts and should not exceed 3 volts p-p. In order to
prevent any interference caused by DC level differences, external signals should be
AC-coupled to EXT IN by an electrolytic capacitor in the 1-10 (xf range. As the direct
audio path (FILTEX = 0) has unity gain, EXT IN can be used to mix outputs of many
SID chips by daisy-chaining. The number of chips that can be chained in this manner is
determined by the amount of noise and distortion allowable at the final output. Note that
the output volume control will affect not only the three SID voices, but also any external
inputs.
A U D IO O U T (PIN 27)
This open-source buffer is the final audio output of SID, comprised of the three
SID voices, the filter and any external input. The output level is set by the output
volume control and reaches a maximum of 2 volts p -p at a DC level of 6 volts. A
source resistor from Audio Out to ground is required for proper operation. The recom
mended resistance is 1 kO for a standard output impedance.
As the output of SID rides at a 6-volt DC level, it should be AC-coupled to any
audio amplifier with an electrolytic capacitor in the 1-10 |xf range.
V DD (PIN 28)
As with Vcc, a separate + 12 VDC line should be run to SID VDD and a bypass capacitor
should be used.
C l28 HARDWARE SPECIFICATIONS 60S
6581 SID CHARACTERISTICS
ABSOLUTE M AXIM UM RATINGS
RATING SYMBOL VALUE UNITS
Supply Voltage V dd -0 .3 to + 1 7 VDC

Supply Voltage Vcc -0 .3 to + 7 VDC
Input Voltage (analog) v ina -0 .3 to + 1 7 VDC
Input Voltage (digital) Vind -0 .3 to + 7 VDC
Operating Temperature TA 0 to + 7 0 °C
Storage Temperature T stg -5 5 to + 1 5 0 °C
(VDD = 12 V D C ± 5 % , Vcc = 5 VDC ± 5 % , T A = 0 to 70° C)
CHARACTERISTIC SYMBOL MIN TYP MAX UNITS
Input High Voltage (RES, 4>2, R/W , CS VIH 2 — v cc VDC

Input Low Voltage A 0-A 4, D 0-1)7) VIL -0 .3 — 0.8 VDC
Input Leakage Current (RES, 4>2, R/W , CS, lin — — 2.5 |xA
A 0-A4; Vj„ = 0 -5 VDC)
Three-State (Off) (D0-D7; Vcc = max) Itsi 10 |xA
Input Leakage Current Vin = 0 .4 -2 .4 VDC
Output High Voltage (D0-D7; Vcc = min, VoH 2.4 — Vcc-0 .7 VDC
1 load = 200 (xA)
Output Low Voltage (D0-D7; Vcc = max, VoL GND — 0.4 VDC
1 load = 3.2 mA)
Output High Current (D0-D7; Sourcing, IoH 200 — — (JlA

V oh = 2.4 VDC)
Output Low Current (D 0-D 7; Sinking lo 3.2 — — mA

V ol = 0.4 VDC)
Input Capacitance (RES, 4>2, R/W , CS, Cin — — 10 pF

A 0-A 4, D 0-D7)
Pot Trigger Voltage (POTX, POTY) vP„, — Vcc/2 — VDC

Pot Sink Current (POTX, POTY) Ipot 500 — — jxA
606 COMMODORE 128
CHARA CTERISTIC SYMBOL MIN TYP MAX UNITS
Input Impedance (EXT IN) R„ 100 150 — k fl
Audio Input Voltage (EXT IN) Vin 5.7 6 6.3 VDC

— 0.5 3 VAC
Audio Output Voltage (AUDIO OUT; 1 kH

load, volume = max) v ou, 5.7 6 6.3 VDC
One Voice on: 0.4 0.5 0.6 VAC
All Voices on: 1.0 1.5 2.0 VAC
Power Supply Current (VD0) I dd — 20 25 mA
Power Supply Current (Vcc) l Cc — 70 100 mA
Power Dissipation (Total) Pd — 600 1000 mW
6581 SID TIMING

t cyc
SYMBOL NAME MIN TYP MAX UNITS
T cyc Clock Cycle Time 1 — 20 JJ.S

Tc Clock High Pulse Width 450 500 10,000 ns
T r ,T f Clock Rise/Fall Time — — 25 ns
T rs Read Set-Up Time 0 — — ns
T rh Read Hold time 0 — — ns
T acc Access Time — — 300 ns
T ah Address Hold Time 10 — — ns
T ch Chip Select Hold Time 0 — — ns
T dh Data Hold Time 20 — — ns
Figure 16-15. Read Cycle

1
---------------- T w ‘-------------- -------- ------------ 4 ~ t W H
-
R,w WWWW_____________ J~ ~
TA W S 4 ~ j | (-------- t ah
wmr~ n~x~
« wxxxw_______ L r
k- t vd — | |-------* - t dh
D°
-°
7 t T
* T W Is m e a s u re d fr o m th e la te s t o c c u r r in g o f o 2, C S , R /W
SYMBOL NAME MIN TYP MAX UNITS
Tw Write Pulse Width 300 -- -- ns

Twh Write Hold Time 0 -- -- ns
T aw s Address Set-up Time 0 -- -- ns
T ah Address Hold Time 10 -- -- ns
Tch Chip Select Hold Time 0 -- -- ns
T vd Valid Data 80 -- -- ns
T dh Data Hold Time 10 — — ns
Figure 16-16. W rite Cycle
EQUAL-TEMPERED
MUSICAL SCALE VALUES
The table in Chapter 11 lists the numerical values which must be stored in the SID
Oscillator frequency control registers to produce the notes of the equal-tempered musical
scale. The equal-tempered scale consists of an octave containing twelve semitones
(notes): C ,D ,E ,F,G ,A ,B and C # ,D # ,F # ,G # ,A # . The frequency of each semitone is
exactly the 12th root of 2 C ^ 2 ) times the frequency of the previous semitone. The table
is based on a <j>2 clock of 1.02 MHz. Refer to the equation given in the Register
Description in Chapter 11 for use of other master clock frequencies. The scale selected
is concert pitch, in which A -4 = 440 Hz. Transpositions of this scale and scales other
than the equal-tempered scale are also possible.
Although the table in Chapter 11 provides a simple and quick method for generat
ing the equal-tempered scale, it is very memory inefficient as it requires 192 bytes for
the table alone. Memory efficiency can be improved by determining the note value
algorithmically. Using the fact that each note in an octave is exactly half the
frequency of that note in the next octave, the note look-up table can be reduced from
ninety-six entries to twelve entries, as there are twelve notes per octave. If the twelve
608 COMMODORE 128
entries (24 bytes) consist of the 16-bit values for the eighth octave (C-7 through B-7),
then notes in lower octaves can be derived by choosing the appropriate note in the eighth
octave and dividing the 16-bit value by two for each octave of difference. As division by
two is nothing more than a right-shift of the value, the calculation can easily be
accomplished by a simple software routine. Although note B-7 is beyond the range of
the oscillators, this value should still be included in the table for calculation purposes
(the MSB of B-7 would require a special software case, such as generating this bit in the
CARRY before shifting). Each note must be specified in a form which indicates which
of the twelve semitones is desired, and which of the eight octaves the semitone is in.
Since 4 bits are necessary to select one of twelve semitones and 3 bits are necessary to
select one of eight octaves, the information can fit in one byte, with the lower nybble
selecting the semitone (by addressing the look-up table) and the upper nybble being used
by the division routine to determine how many times the table value must be right-shifted.
SID ENVELOPE GENERATORS

The four-part ADSR (ATTACK, DECAY, SUSTAIN, RELEASE) envelope generator
has been proven in electronic music to provide the optimum trade-off between flexibility
and ease of amplitude control. Appropriate selection of envelope parameters allows the
simulation of a wide range of percussion and sustained instruments. The violin is a good
example of a sustained instrument. The violinist controls the volume by bowing the
instrument. Typically, the volume builds slowly, reaches a peak, then drops to an
intermediate level. The violinist can maintain this level for as long as desired,
then the volume is allowed to slowly die away. A “ snapshot” of this envelope is
shown below:
Figure 16-17. ADSR Envelope
This volume envelope can be easily reproduced by the ADSR as shown below,
with typical envelope rates:
ATTACK: 10 ($A) 500 ms

DECAY: 8 300 ms
SUSTAIN: 10 ($A)
RELEASE: 9 750 ms
Note that the tone can be held at the intermediate SUSTAIN level for as long as
desired. The tone will not begin to die away until GATE is cleared. With minor
alterations, this basic envelope can be used for brass and woodwinds as well as strings.
An entirely different form of envelope is produced by percussion instruments such
as drums, cymbals and gongs, as well as certain keyboards such as pianos and
harpsichords. The percussion envelope is characterized by a nearly instantaneous attack,
immediately followed by a decay to zero volume. Percussion instruments cannot be
sustained at a constant amplitude. For example, the instant a drum is struck, the sound
reaches full volume and decays rapidly regardless of how it was struck. A typical
cymbal envelope is shown below
ATTACK: 0 2 ms
DECAY: 9 750 ms
SUSTAIN: 0
RELEASE: 9 750 ms
Note that the tone immediately begins to decay to zero amplitude after the peak is
reached, regardless of when GATE is cleared. The amplitude envelope of pianos and
harpsichords is somewhat more complicated, but can be generated quite easily with the
ADSR. These instruments reach full volume when a key is first struck. The amplitude
immediately begins to die away slowly as long as the key remains depressed. If the key
is released before the sound has fully died away, the amplitude will immediately drop to
zero. This envelope is shown below:
ATTACK: 2 ms
DECAY: 750 ms
SUSTAIN:
RELEASE: 6 ms
Note that the tone decays slowly until GATE is cleared, at which point the
amplitude drops rapidly to zero.
The most simple envelope is that of the organ, When a key is pressed, the tone
immediately reaches full volume and remains there. When the key is released, the tone
drops immediately to zero volume. This envelope is shown below:
610 COMMODORE 128
ATTACK: 0 2 ms
DECAY: 0 6 ms
SUSTAIN: 15 ($F)
RELEASE: 0 6 ms
The real power of SID lies in the ability to create original sounds rather than
simulations of acoustic instruments. The ADSR is capable of creating envelopes which
do not correspond to any “ real” instruments. A good example would be the “ back
wards” envelope. This envelope is characterized by a slow attack and rapid decay
which sounds very much like an instrument that has been recorded on tape then played
backwards. This envelope is shown below:
ATTACK 10 ($A) 500 ms

DECAY 0 6 ms
SUSTAIN 15 ($F)
RELEASE 3 72 ms
GATE j
Many unique sounds can be created by applying the amplitude envelope of one
instrument to the harmonic structure of another. This produces sounds similar to familiar
acoustic instruments, yet notably different. In general, sound is quite subjective and
experimentation with various envelope rates and harmonic contents will be necessary in
order to achieve the desired sound.
Figure 16-18. Typical 6581/SID A pplication

6526 COMPLEX INTERFACE ADAPTER

(CIA) CHIP SPECIFICATIONS________
DESCRIPTION
The 6526 Complex Interface Adapter (CIA) is an 8502 bus compatible peripheral
interface device with extremely flexible timing and I/O capabilities. Figure 16-19 shows
the 6526 pin configuration. Figure 16-20 shows the 6526 block diagram.
FEATURES
■ 16 Individually programmable I/O lines

■ 8 or 16-bit handshaking on read or write
■ 2 independent, linkable 16-bit interval timers
■ 24-hour (AM/PM) time of day clock with programmable alarm
■ 8-bit shift register for serial I/O
■ 2TTL load capability
■ CMOS compatible I/O lines
■ 1 or 2 MHz operation available
ORDERING IN FO R M A TIO N
MXS 6526
FREQUENCY RANGE
NO SUFFIX = lMHz
A = 2MHz
PACKAGE DESIGNATOR
C = CERAMIC
P = PLASTIC
612 COMMODORE 128
Figure 16-19. 6526 C IA Pin Configuration

C128 HARDWARE SPECIFICATIONS 6 13
D0-D7
DATA BUS BUFFERS

----------- 7 \ -------------
PA
PRA BUFFERS < > PA 0 -PA7
SP SERIAL
SP
BUFFER P ORT
DORA
CNT PC
CNT PC
BUFFER BUFFER
PB
BUFFERS ])> PB 0-PB7
TOD TOD/
TOD
BUFFER ALARM
DDRB
TIMER I
CRB
FLAG
BUFFER
TIMER A
IRQ INT /
IRQ
BUFFER MASK \_
CRA
C H IP A C C ES S C O N T R O L
R/W 02 CS R S 3 RS2 RS1 RSO RES
Figure 16-20. 6526 Block Diagram
ABSOLUTE M A XIM U M RATINGS

Supply Voltage, Vcc -0 .3 V to + 7.0V
Input/Output Voltage, VIN -0 .3 V to + 7 .0 V
Operating Temperature, Tqp 0° C to 70° C
Storage Temperature, T STg -55° C to 150° C
614 COMMODORE 128
All inputs contain protection circuitry to prevent damage due to high static
discharges. Care should be exercised to prevent unnecessary application o f voltages in
excess of the allowable limits.
C O M M EN T
Stresses above those listed under “ Absolute Maximum Ratings” may cause permanent
damage to the device. These are stress ratings only. Functional operation of this device
at these or any other conditions above those indicated in the operational sections of this
specification is not implied and exposure to absolute maximum rating conditions for
extended periods may affect device reliability.
(V cc ± 5%, V ss - O V, T a = Q-70°C)
CHARA CTERISTIC SYMBOL MIN. TYP. MAX. UNIT
Input High Voltage VIH + 2.4 — V cc V

Input Low Voltage VIL -0 .3 — — V
Input Leakage Current; IlN 1.0 2.5 (xA
V ,N = VSS + 5 V
(TOD, R/W , FLAG,4>2,
RES, R S 0-R S 3, CS)
Port Input Pull-up Resistance R pi 3.1 5.0 —■ Kn

Output Leakage Current for I tsi ± 1 .0 ± 1 0 .0 |xA
High Im pedance State (Three
State); VIN = 4V to 2.4V;
(D B 0-D B 7, SP, CNT, IRQ)
Output High Voltage V qH + 2.4 V cc V
Vc c = M IN, I l o a d <
- 2 0 0 A (PA 0-PA 7, PC
PB 0-PB 7, D B0-DB7)
Output Low Voltage V oL — — + 0.40 V
Vc c = M IN, I l o a d < 3.2 mA
Output High Current (Sourcing); Io H -2 0 0 -1 0 0 0 jjlA
VOH > 2.4V (PA 0-PA 7,

PB 0-PB 7, PC, D B0-DB7
Output Low Current (Sinking); Io L 3.2 --- --- mA
VOL < -4V (PA 0-PA 7, PC,
P B 0-PB 7, D B0-DB7)
Input Capacitance ClN — 7 10 Pf
Output Capacitance C oU T — 7 10 pf
Power Supply Current Ic c — 70 100 mA
0 2 IN P U T
PERIPHERAL
DATA OUT
DATA IN
D B 7 -D B 0
Figure 16-21. 6526 W rite Tim ing Diagram
Figure 16-22. 6526 Read Timing Diagram
6526 INTERFACE SIGNALS

4)2— CLOCK IN P U T
The 4>2 clock is a TTL-compatible input used for internal device operation and as a timing
reference for communicating with the system data bus.
CS—C HIP SELECT IN P U T _

The CS input controls the activity of the 6526. A low level on CS while <J>2 is high causes
the device to respond to signals on the R/W and address (RS) lines. A high on CS
prevents these lines from controlling the 6526. The CS line is normally activated (low)
at <)>2 by the appropriate address combination.
616 COMMODORE 128
R/W— READ/WRITE IN P U T
The R/W signal is normally supplied by the microprocessor and controls the direction of
data transfers of the 6526. A high on R/W indicates a read (data transfer out of the
6526), while a low indicates a write (data transfer into the 6526).
RS3-RS0— ADDRESS INPUTS

The address inputs select the internal registers as described by the Register Map.
DB7-DB0— D A TA BUS IN PUTS/O U TPU TS

The eight data bus pins transfer information between the 6526 and the system data bus.
These pins are high impedance inputs unless CS is low and R/W and 4>2 are high to read
the device. During this read, the data bus output buffers are enabled, driving the data
from the selected register onto the system data bus.
IRQ— INTERRUPT REQUEST O U T P U T

IRQ is an open drain output normally connected to the processor interrupt input. An
external pullup resistor holds the signal high, allowing multiple IRQ outputs to be
connected together. The IRQ output is normally off (high impedance) and is activated
low as indicated in the functional description.
RES— RESET IN P U T
A low on the RES pin resets all internal registers. The port pins are set as inputs and
port registers to zero (although a read of the ports will return all highs because of
passive pullups). The timer control registers are set to 0 and the timer latches to all ones.
All other registers are reset to 0.
6526 TIM ING CHARACTERISTICS
1 MHz 2MHz
SYM BOL CH ARACTERISTIC MIN MAX M IN MAX UNIT
(|>2 Clock
T cyc Cycle Time 1000 20,000 500 20,000 ns
T r, t f Rise and Fall Time — 25 — 25 ns
T chw Clock Pulse Width (High) 420 10,000 200 10,000 ns
T clw Clock Pulse Width (Low) 420 10,000 200 10,000 ns
Write Cycle
T pd Output Delay From ct>2 1000 500 ns
Twcs CS low while <\>2 high 420 — 200 — ns
T ads Address Setup Time 0 — 0 — ns
T adh Address Hold Time 10 — 5 — ns
T rw s R/W Setup Time 0 — 0 — ns
T rw h R/W Hold Time 0 — 0 — ns
T ds Data Bus Setup Time 150 — 75 — ns
T dh Data Bus Hold Time 0 — 0 — ns
6526 TIMING CHARACTERISTICS (continued)
1 MHz 2MHz
SYMBOL CHARACTERISTIC MIN MAX MIN MAX UNIT
Read Cycle
Tps Port Setup Time 300 — 150 — ns
Tyvcs(2) CS low while <j>2 high 420 — 20 — ns
T ads Address Setup Time 0 — 0 — ns
T aoh Address Hold Time 10 — 5 — ns
T rws R/W Setup Time 0 — 0 — ns
T rw h R/W Hold Time 0 — 0 — ns
T acc Data Access from R S3-RS0 — 550 — 275 ns
Ico(3) Data Access from CS — 320 — 150 ns
T dr Data Release Time 50 — 25 — ns
1. All timings are referenced from VtL max and V1Hmin on inputs and VOL max and
VDH min on outputs.
2- TWCs is measured from the later of c{>2 high or CS low. CS must be low at least
until the end of <J>2 high.
3. TCo is measured from the later of <j>2 high or CS low. Valid data is available only
after the later of TACC or Tc o .
REGISTER MAP
RS3 RS2 RSI RS0 REG NAME DESCRIPTION
0 0 0 0 0 PRA PERIPHERAL DATA REG A

0 0 0 1 1 PRB PERIPHERAL DATA REG B
0 0 1 0 2 DDRA DATA DIRECTION REG A
0 0 1 1 3 DDRB DATA DIRECTION REG B
0 1 0 0 4 TA LO TIMER A LOW REGISTER
0 1 0 1 5 TA HI TIMER A HIGH REGISTER
0 1 1 0 6 TB LO TIMER B LOW REGISTER
0 1 1 1 7 TB HI TIMER B HIGH REGISTER
1 0 0 0 8 TOD 10THS 10THS OF SECONDS REGISTER
1 0 0 1 9 TOD SEC SECONDS REGISTER
1 0 1 0 A TOD MIN MINUTES REGISTER
1 0 1 1 B TOD HR HOURS— AM/PM REGISTER
1 1 0 0 C SDR SERIAL DATA REGISTER
1 1 0 1 D ICR INTERRUPT CONTROL REGISTER
1 1 1 0 E CRA CONTROL REG A
1 1 1 1 F CRB CONTROL REG B
618 COMMODORE 128
6526 FU N C TIO N A L DESCRIPTION

I/O PARTS (PRA, PRB, DDRA, DDRB).
Parts A and B each consist of an 8-bit Peripheral Data Register (PR) and an 8-bit Data
Direction Register (DDR). If a bit in the DDR is set to one, the corresponding bit in the PR
is an output', if a DDR bit is set to a 0, the corresponding PR bit is defined as an input.
On a READ, the PR reflects the information present on the actual port pins (PA0-PA7,
PB0-PB7) for both input and output bits. Port A and Port B have passive pull-up
devices as well as active pull-ups, providing both CMOS and TTL compatibility. Both
parts have two TTL load drive capability. In addition to normal I/O operation, PB6 and
PB7 also provide timer output functions.
H A N D SH A K IN G
Handshaking on data transfers can be accomplished using the PC output pin and the
FLAG input pin. PC will go low for one cycle following a read or write of PORT B.
This signal can be used to indicate “ data ready” at PORT B or “ data accepted” from
PORT B. Handshaking on 16-bit data transfers (using both PORT A and PORT B) is
possible by always reading or writing PORT A first. FLAG is a negative edge sensitive
input which can be used for receiving the PC output from another 6526, or as a general
purpose interrupt input. Any negative transition of FLAG will set the FLAG interrupt
bit.
REG NAME d7 d6 d5 d4 d ., d2 d, D„
0 PRA pa7 pa6 pa5 pa4 pa3 PA2 PA! PAo

1 PRB pb7 pb6 pb5 pb4 PBj PB, PBj PB0
2 DDRA d p a 7 d p a 6 DPA s d p a 4 d p a 3 DPA2 DPA j DPA|
3 DDRB d p b 7 d p b 6 d p b 5 d p b 4 d p b , DPB2 DPB! DPB(
INTERVAL TIMERS (TIMER A, TIMER B)

Each interval timer consists of a 16-bit read-only Timer Counter and a 16-bit write-
only Timer Latch. Data written to the timer are latched in the Timer Latch, while
data read from the timer are the present contents of the Time Counter. The timers
can be used independently or linked for extended operations. The various timer modes
allow generation of long time delays, variable width pulses, pulse trains and variable
frequency waveforms. Utilizing the CNT input, the timers can count external pulses
or measure frequency, pulse width and delay times of external signals. Each timer
has an associated control register, providing independent control of the following
functions:
START/STOP
A control bit allows the timer to be started or stopped by the microprocessor at any
time.
PB ON/OFF
A control bit allows the timer output to appear on a PORT B output line (PB6 for
TIMER A and PB7 for TIMER B). This function overrides the DDRB control bit and
forces the appropriate PB line to an output.
TOGGLE/PULSE
A control bit selects the output applied to PORT B . On every timer underflow the output
can either toggle or generate a single positive pulse of one cycle duration. The toggle
output is set high whenever the timer is started and is set low by RES.
O N E -S H O T /C O N T IN U O U S
A control bit selects either timer mode. In one-shot mode, the timer will count down
from the latched value to 0, generate an interrupt, reload the latched value, then stop. In
continuous mode, the timer will count from the latched value to 0, generate an interrupt,
reload the latched value and repeat the procedure continuously.
FORCE LOAD
A strobe bit allows the timer latch to be loaded into the timer counter at any time,
whether the timer is running or not.
IN P U T MODE
Control bits allow selection of the clock used to decrement the timer. TIMER A can
count c)>2 clock pulses or external pulses applied to the CNT pin. TIMER B can count 4>2
pulses, external CNT pulses, TIMER A underflow pulses or TIMER A underflow pulses
while the CNT pin is held high.
The timer latch is loaded into the timer on any timer underflow, on a force load or
following a write to the high byte of the prescaler while the timer is stopped. If the timer
is running, a write to the high byte will load the timer latch, but not reload the counter.
READ (TIMER)
REG NAME
4 TA LO TAL7 tal6 TALS TAL4 t a l 3 tal2 TAL i TAL o

5 TA HI tah7 tah6 tah5 tah4 tah3 tah2 TAH! TAH o
6 TB LO tbl7 t b l 6 TBLS t b l 4 TBL, t b l 2 TBL! TBLo
7 TB HI tbh7 t b h 6 TBHS t b h 4 t b h 3 t b h 2 TBH, TBH0
WRITE (PRESCALER)
REG NAME
4 TA LO pal7 pal6 pal5 PAL4 pal3 pal2 PAL] PAL0

5 TA HI pah7 pah6 pah5 pah4 pah3 PAH, PAH] PAH0
6 TB LO p b l 7 p b l 6 PBLS p b l 4 p b l 3 PBL2 PBL, PBL0
7 TB HI p b h 7 p b h 6 PBHS p b h 4 pbh3 pbh2 PBHi PBH0
620 COMMODORE 128
TIM E OF D A Y CLOCK (T O D )
The TOD clock is a special purpose timer for real-time applications. TOD consists of a
24-hour (AM/PM) clock with l/10th second resolution. It is organized into four regis
ters: lOths of seconds, Seconds, Minutes and Hours. The AM/PM flag is in the MSB of
the Hours register for easy bit testing. Each register reads in BCD format to simplify
conversion for driving displays, etc. The clock requires an external 60 Hz or 50 Hz
(programmable) TTL level input on the TOD pin for accurate time keeping. In addition
to timekeeping, a programmable ALARM is provided for generating an interrupt at a
desired time. The ALARM registers are located at the same addresses as the correspond
ing TOD registers. Access to the ALARM is governed by a Control Register bit. The
ALARM is write-only; any read of a TOD address will read time regardless of the state
of the ALARM access bit.
A specific sequence of events must be followed for proper setting and reading of
TOD. TOD is automatically stopped whenever a write to the Hours register occurs. The
clock will not start again until after a write to the lOths of seconds register. This assures
TOD will always start at the desired time. Since a carry from one stage to the next can
occur at any time with respect to a read operation, a latching function is included to
keep all Time O f Day information constant during a read sequence. All four TOD
registers latch on a read of Hours and remain latched until after a read of lOths of
seconds. The TOD clock continues to count when the output registers are latched. If
only one register is to be read, there is no carry problem and the register can be read
“ on the fly ,” provided that any read of Hours is followed by a read of lOths of seconds
to disable the latching.
READ
REG NAME
8 TOD 10THS 0 0 0 0 t8 t4 t2 T,
9 TOD SEC 0 sh 4 sh 2 SH! sl 8 sl4 sl2 SL,
A TOD MIN 0 mh4 mh2 mh. m l8 m l4 m l2 ML!
B TOD HR PM 0 0 hh hl8 hl4 hl2 HLj
WRITE
CRB7 = 0 TOD
CRB7 = 1 ALARM
(SAME FORMAT AS READ)
SERIAL PORT (SDR)

The serial port is a buffered, 8-bit synchronous shift register system. A control bit
selects input or output mode. In input mode, data on the SP pin is shifted into the shift
register on the rising edge of the signal applied to the CNT pin. After eight CNT pulses,
the data in the shift register is dumped into the Serial Data Register and an interrupt is
generated. In the output mode, TIMER A is used for the baud rate generator. Data is
shifted out on the SP pin at one half the underflow rate of TIMER A. The maximum
baud rate possible is (J>2 divided by 4, but the maximum useable baud rate will be
determined by line loading and the speed at which the receiver responds to input data.
Transmission will start following a write to the Serial Data Register (provided TIMER A
is running and in continuous mode). The clock signal derived from TIMER A appears as
an output on the CNT pin. The data in the Serial Data Register will be loaded into the
shift register then shift out to the SP pin when a CNT pulse occurs. Data shifted out
becomes valid on the falling edge of CNT and remains valid until the next falling edge.
A fter eight CNT pulses, an interrupt is generated to indicate more data can be sent. If
the Serial Data Register was loaded with new information prior to this interrupt, the new
data will automatically be loaded into the shift register and transmission will continue. If
the microprocessor stays one byte ahead of the shift register, transmission will be
continuous. If no further data is to be transmitted, after the 8th CNT pulse, CNT will
return high and SP will remain at the level of the last data bit transmitted. SDR data is
shifted out MSB first and serial input data should also appear in this format.
The bidirectional capability of the Serial Port and CNT clock allows many 6526
devices to be connected to a common serial communication bus on which one 6526 acts
as a master, sourcing data and shift clock, while all other 6526 chips act as slaves. Both
CNT and SP outputs are open drain to allow such a common bus. Protocol for
master/slave selection can be transmitted over the serial bus, or via dedicated handshaking
lines.
R EG NAME
C SDR S7 Sg S5 S4 S3 S2 Sj So
INTERRUPT C O N TR O L (ICR)
There are five sources of interrupts on the 6526: underflow from TIM ER A, underflow
from TIMER B, TOD ALARM, Serial Port full/empty and FLAG. A single register
provides masking and interrupt information. The interrupt Control Register consists of a
write-only MASK register and a read-only DATA register. Any interrupt will set the
corresponding bit in the DATA register. Any interrupt which is enabled by the MASK
register will set the IR bit (MSB) of the DATA register and bring the IRQ pin low. In a
multi-chip system, the IR bit can be polled to detect which chip has generated an
interrupt request. The interrupt DATA register is cleared and the IRQ line returns high
following a read of the DATA register. Since each interrupt sets an interrupt bit
regardless of the MASK, and each interrupt bit can be selectively masked to prevent the
generation of a processor interrupt, it is possible to intermix polled interrupts with true
interrupts. However, polling the IR bit will cause the DATA register to clear, therefore,
it is up to the user to preserve the information contained in the DATA register if any
polled interrupts were present.
The MASK register provides convenient control of individual mask bits. When
writing to the MASK register, if bit 7 (SET/CLEAR) of the data written is a ZERO, any
622 COMMODORE 128
mask bit written with a one will be cleared, while those mask bits written with a 0
will be unaffected. If bit 7 of the data written is a ONE, any mask bit written with a one
will be set, while those mask bits written with a 0 will be unaffected. In order for an
interrupt flag to set IR and generate an Interrupt Request, the corresponding MASK bit
must be set.
READ (INT DATA)
REG NAME
D ICR IR 0 0 FLG SP ALRM TB TA
WRITE (INT MASK)
REG NAME
D ICR S/C X X FLG SP ALRM TB TA
C O N TRO L REGISTERS
There are two control registers in the 6526, CRA and CRB. CRA is associated
with TIMER A and CRB is associated with TIMER B. The register format is as
follows:
CRA:
BIT NAME FUNCTION
0 START 1 = START TIMER A , 0 = STOP TIMER A , This bit is autom at

ically reset when underflow occurs during one-shot mode.
1 PBON 1 = TIM ER A output appears on PB6, 0 = PB6 normal operation.
2 OUTMODE 1 = TOGGLE, 0 = PULSE
3 RUNMODE 1 = ONE-SHOT, 0 = CONTINUOUS
4 LOAD 1 = FORCE LOAD (this is a STROBE input, there is no data
storage, bit 4 will always read back a 0 and writing a 0 has no
effect).
5 INMODE 1 = TIMER A counts positive CNT transitions, 0 = TIMER A counts
4)2 pulses.
6 SPMODE 1 = SERIAL PORT output (CNT sources shift clock). 0 = SERIAL
PORT input (external shift clock required).
7 TODIN 1 = 50 Hz clock required on TOD pin for accurate time, 0 = 60 Hz
clock required on TOD pin for accurate time.
CRB:
BIT NAME FUNCTION
(Bits C RB0-C RB4 are identical to C RA 0-CR A 4 for TIM ER B

with the exception that bit 1 controls the output o f TIMER B on
PB7).
5,6 INMODE Bits CRB5 and CRB6 select one of four input modes for TIM ER
B as:
CRB6 CRB5
0 0 TIMER B counts 4>2 pulses.
0 1 TIM ER B counts positive CNT transitions.
1 0 TIM ER B counts TIMER A underflow pulses.
1 1 TIMER B counts TIM ER A underflow pulses
while CNT is high.
7 ALARM 1 = writing to TOD registers sets ALARM , 0 = writing to TOD
registers sets TOD clock.
TOD SP LN RUN OUT

REG NAME IN MODE M ODE LOAD M ODE MODE PS ON ST A R T
CRA 0 = 60H z 0 = IN P U T 0 = <})2 1 = FO R CE 0 = CONT 0 = PU LSE o = p b 6o f f 0= S T O P

LOAD
1 = 50H z 1 = OUTPUT 1= CNT (S T R O B E ) 1 = 0 .5 1 = TO G G LE 1 = p b 6 o n 1 = STA RT
L -TA- J
R EG NAME A LARM IN MODE LOAD M ODE MODE PS ON ST A RT
F CRS 0 = TOD 0 0 = 4 )2 1 = FO RCE 0 = CONT. 0 = PU LSE 0 = P S 7O F F G =STO P
1 1 = CNT LOAD
1 0 = TA
1= 1 1= CN T TA (S T R O B E ) i=o.s. 1 = TO G G LE i = p s 7o n 1 = STA RT
A LARM
All unused register bits are unaffected by a write and are forced to zero on a
read.
624 COMMODORE 128
DYNAMIC RANDOM ACCESS

MEMORY
This section discusses the characteristics of the C l28 dynamic RAMs, which are
currently the 4164 64K-bit RAM. This RAM device is in the 64K by 1-bit configura
tion. This section also contains information on the 4416 dynamic RAMs used by
the 8563 video controller. This type of RAM is a 64K RAM in the 16K by 4 bit
configuration.
SYSTEM RAM DESCRIPTION

The C l 28 system contains 128K of processor-addressable DRAM in the 64K by 1
configuration, organized into two individual 64K banks. Additionally, the system
contains 16K of video display DRAM local to the 8563 video controller, not directly
accessable by the processor.
RAM banking, described in detail in the MMU section in Chapter 13, is controlled
by several MMU registers: the Configuration Register, the RAM Configuration Regis
ter, and the Zero Page and Page One Pointers. Simply put, the Configuration Register
controls which 64K bank of RAM is selected; the RAM Configuration Register controls
if and how much RAM is kept in common between banks, and the Pointer Registers
redirect the zero and one pages to any page in memory, overriding the effect of the two
Configuration Registers. In the system, RAM bank select is achieved via gated CAS
control.
PHYSICAL CHARACTERISTICS
This section covers some of the characteristics of the 64K by 1-bit RAM and 16K by
4-bit RAM that are used in the C l 28 system. A pinout table and a figure are given for
both the 4164 and the 4416 packages (See Tables 16-15 and 16-16 and Figures 16-23 and
16-24).
C l 28 HARDWARE SPECIFICATIONS 62S
PIN NAME DESCRIPTION
1 NC No Connection
2 Din Data In
3 /WE Write Enable (Active Low)
4 /RAS Row Address Strobe (Active Low)
5 Ao Address Bit 0
6 a2 Address Bit 2
7 Ai Address Bit 1
8 V cc Power Supply + 5 Vdc
9 a7 Address Bit 7
10 a5 Address Bit 5
11 a4 Address Bit 4
12 a3 Address Bit 3
13 A<s Address Bit 6
14 Dout Data Out
15 /CAS Column Address Strobe (Active Low)
16 v’ ss Power Supply Ground
Table 16-15. 4164 Pinout
Figure 16-23. T h e 4164 Chip

626 COMMODORE 128
1 /OE Output Enable (Active Low)

2 D, Data Bit 1
3 d2 Data Bit 2
4 /WE Write Enable (Active Low)
5 /RAS Row Address Strobe (Active Low)
6 ^6 Address Bit 6
7 As Address Bit 5
8 a4 Address Bit 4
9 Vdd Power Supply + 5 Vdc
10 a7 Address Bit 7
11 a3 Address Bit 3
12 a2 Address Bit 2
13 A, Address Bit 1
14 Ao Address Bit 0
15 d3 Data Bit 3
16 /CAS Column Address Strobe (Active Low)
17 d4 Data Bit 4
18 Vss Power Supply Ground
Table 16-16. 4416 Pinout
Figure 16-24. T h e 4464 Chip

READ ONLY MEMORY (ROM)

This section describes the C l 28 system ROM, both from a logical and a hardware point
of view. It mentions aspects of the ROM banking structure, the management of Kernal
and BASIC, and explains the physical specifications of the ROM devices themselves.
SYSTEM ROM DESCRIPTION

In C64 mode, the operating system resides in 16K of ROM, which includes approxi
mately 8K for Kemal and 8K for BASIC. In C128 mode, the operating system resides in
48K of ROM and includes advanced Kemal and BASIC features. The Kernal, by
definition, is the general operating system of the computer, with fixed entry points into
usable subroutines to facilitate ROM update transparent use by higher-level programs.
There is also a character ROM that resides on the Shared Bus, shared by the VIC chip
and the processor. This ROM is a 4016 8K by 8, NMOS ROM. The C64 OS ROM is
wired so as to appear as two chunks of noncontiguous ROM, copying the actual C64
ROM memory map. Provision is included to handle system ROM as either four 16K X
8 ROMs or as two 32K X 8 ROMs. All internal C128 function ROMs will be the 32K
x 8 variety.
ROM BANKING
Refer back to the MMU register map, Figure 13-4 in Chapter 13. Note that the
Configuration Register (CR) controls the type of ROM or RAM seen in a given address
location. Dependent on the contents of the CR, ROM may be enabled and disabled to
attain the most useful configuration for the application at hand. ROM is enabled in three
memory areas in C l 28 mode, each consisting of 16K of address space. The lower ROM
may be defined as RAM or System ROM, the upper two ROMs may be System ROM,
Function ROM, Cartridge ROM or RAM. In C64 mode, the C64 memory mapping rules
apply, which are primitive compared to those used in C l 28 mode. C64 ROM is banked
as two 8K sections, BASIC and Kernal, according to the page zero port and the
cartridge in place at the time. No free banking can occur when a cartridge is in
place.
In the C l 28, if an address falls into the range of an enabled ROM, the MMU will
communicate the status of ROM to the PLA decoder via the memory status lines.
Essentially, the MMU looks up in the Configuration Register which ROM or RAM is
set. See Chapter 13. The way the banking scheme is implemented, it allows up to 32K
of internal, bankable ROM for use in such programs as Function Key Applications,
and will support 32K of internal bankable ROM. Various combinations of ROM are
possible, and can be noted by studying the configurations for the Configuration
Register. Type 23128 (16K by 8) and 23256 (32K by 8) ROMs are used by the
system.
628 COMMODORE 128
T IM IN G SPECIFICATION
IN TER N A L ROMs
This section specifies timing parameters for both the 23128 and the 27256 Read Only
Memories. This timing spec applies to internal ROMs and for external ROMs run at 1
MHz. For external ROMs run at 2 M Hz, see Table 16-18.
PARAM ETER SYMBOL MIN MAX UNIT
Address Valid to Output Delay T acc 300 — ns

Chip Enable to Output Delay T ce 300 — ns
Output Enable to Output Delay T oe 120 — ns
Table 16-17. Internal ROM Timing
RDDRES5-
CHIP
ENABLE
OUTPUT
ENABLE
DATA
KM
Figure 16-25. ROM Timing Diagram
EXTERNAL ROMS
All C64 mode external ROMs and many C l 28 mode external ROMs can be of the type
mentioned above. Any external ROM that is to run at 2 MHz must be faster, as
specified in Table 16-18.
PARAM ETER SYMBOL MIN MAX UNIT
Address Valid to Output Delay T acc 250 — ns

Chip Enable to Output Delay T ce 200 — ns
Output Enable to Output Delay T oe 100 — ns
T ab le 16-18. E xtern al 2 M H z RO M T im in g
THE 23128 ROM
PIN NAM E D E S C R IP T IO N
1 v pp Programming Voltage
2 An Address Bit 12 (AJ3 on the C64 OS ROM)
3 a7 Address Bit 7
4 A6 Address Bit 6
5 A5 Address Bit 5
6 a4 Address Bit 4
7 a3 Address Bit 3
8 A2 Address Bit 2
9 A, Address Bit 1
10 Ao Address Bit 0
11 D„ Data Bit 0
12 Dj Data Bit 1
13 d2 Data Bit 2
14 GND Power Supply Ground
15 d3 Data Bit 3
16 d4 Data Bit 4
17 d5 Data Bit 5
18 d6 Data Bit 6
19 d7 Data Bit 7
20 /CE Chip Enable (Active Low)
21 A io Address Bit 10
23 A n Address Bit 11
24 a 9 Address Bit 9
25 As Address Bit 8
26 A i3 Address Bit 13
27 /PGM Program Enable (Active Low)
28 V Power Supply + 5 Vdc
cc
Table 16-19. 23128 ROM Pinout

630 COMMODORE 128
V P P v c c
0 ------------ ------------ 0
H 1 P / P G M
o -------------
o
n 7 f l 1 3
o ------------- ------------ ©
f l f i f l 8
o ------------- ------------ ©
R 5 2 3 1 2 8 f l 9
0 — ------------ 0
n 4 f l l i
o ------------- ------------ 0
H 3 / 0 E
o -------------
o
r p f l 1 0 — ©
o —
f l l / C E
0 -------------
O
f l n D 7
o ------------ — ©
n n D 6
6 ------------- ------------ 0
m D 5
o ------------- -------------©
n p D 4 :
< 5 ------------ -------------©
G N D D 3
-------------©
4 —
Figure 16-26. The 23128 ROM Chip (BASIC, Kernal, Editor and External Function
ROMs)
THE 23256 ROM
1 v pp Programming Voltage
2 a 12 Address Bit 12
3 A7 Address Bit 7
4 A6 Address Bit 6
5 As Address Bit 5
6 A4 Address Bit 4
7 a3 Address Bit 3
8 a2 Address Bit 2
9 A, Address Bit 1
10 Ao Address Bit 0
11 D0 Data Bit 0
12 D, Data Bit 1
13 d2 Data Bit 2
14 GND Power Supply Ground
15 d3 Data Bit 3
16 d4 Data Bit 4
17 D5 Data Bit 5
18 d6 Data Bit 6
19 d7 Data Bit 7
20 /C E -/PG M Chip Enable-Program Enable (Active Low)
21 A 10 Address Bit 10
23 Ah Address Bit 11
24 A9 Address Bit 9
25 A8 Address Bit 8
26 A i3 Address Bit 13
27 A 14 Address Bit 14
28 V CC Power Supply + 5 Vdc
Table 16-20. 23256 ROM Pinout (Internal or External Function ROMs)

632 COMMODORE 128
V P P ^ V C C
0 — — e
R 1 2 R 1 4
o ------------ — o
R 7 R 1 3
0 ------------ — o
R G R 8
o — — o
R E 2 3 2 5 6 R 9
0 ------------ ------------ 0
R 4 R 1 1
o ------------ — e
R 3 / O E
o ------------
o
R P R 1 0
0 — — o
0 ________ R 1 / C E - / P G M
R R D 7
o ------------- — e
n n D 6
0 ------------- ------------ 0
n i D 5
o — — 0
n p D 4
o ------------ — o
G N D D 3
— 0
- 4 —
Figure 16-27. The 23256 ROM Chip

THE SERIAL BUS

The C l 28 Serial Bus is an improved version of the C64 serial bus. This bus uses
IEEE-488-type addressing, thus maintaining software compatibility between the con
sumer market serial devices and the CBM IEEE-488 parallel devices. The C128
improves this bus by allowing communication at much greater speeds with specially
designed peripherals, the most important being the disk drive, while still maintaining
compatibility with older, slower peripherals used by the C64. This section describes the
hardware and some of the software aspects of both the old and the new serial transmis
sion scheme.
BUS OPERATIONS
There are three basic bus operations that take place on the serial bus, in both fast and
slow modes. The first of these is called Control. The C l28 is the controller in most
circumstances. The controller of the bus is always the device that initiates protocol on
the bus, requesting peripheral devices to do one of the two other serial operations, either
Talk or Listen.
All serial bus devices can listen. A listening device is a device that has been
ordered by the controller to receive data. Some devices, such as disk drives, can talk. A
talking device is sending data to the controller. Both hardware and software drive this
bus protocol.
BUS SIGNALS
The Commodore serial bus is composed of the following signals:
■ SRQ (pin 1): This signal is called Service Request. The slow serial bus does
not use this line, the fast bidirectional clock line. (Not used in C64 M ode.)
■ GND (pin 2): Chassis ground.
■ ATN (pin 3): This signal is called Attention. It is used to address a device on
the bus. When requesting a device to talk to or listen, the controller brings this
signal low, creating some sort of interrupt on all serial bus devices. It then
sends out an address that will select one device on the bus. It is the controller’s
responsibility to time out if a device on the bus does not respond within a
reasonable amount of time.
■ CLK (pin 4): This is the slow serial clock. It is used by slow serial devices,
which are software-clocked, to clock data transmitted on the serial bus.
■ DATA (pin 5): This is the serial data line. It is used by both slow and fast
serial devices to transmit data in sync with clock signal.
■ RESET (pin 6): This is the reset line, used to reset all peripherals when the
host resets.
634 COMMODORE 128
FAST SERIAL BUS

FAST PROTOCOL
To function as a fast talker, the system must be addressing a fast device, such as the
1571 disk drive. When addressing any device, the C128 sends a fast byte, toggling the
SQR line eight times, with the ATN line low. If the device is a fast device, it will record
the fact that a fast controller accessed it and respond with a fast acknowledge. If the
device is a slow device, no response is delivered and the C l28 then assumes it is talking
with a slow device. The status of drive speed is retained until the device is requested to
untalk or unlisten, if an error occurs, or if a system reset occurs.
FAST HARDWARE
The fast serial bus, in order to achieve its speed increase, uses different hardware from
that of the slow serial bus. The slow serial bus uses several 6526 port lines to drive
ATN, CLK and DATA. Thus, clocking of the data transfers must be software-driven.
The fast serial method is to use the serial port line of a 6526 (CIA-1) to actually transfer
the serial data. This increases the transfer rate dramatically, to the point where the
transfer becomes limited more by software overhead than anything else. The actual
speed of transfer is set by the 6526 timer. Current 6526’s have a minimum serial timer
value of 4, though in actual use this value is closer to 6, owing to loading. Any
advances in the 6526 would make a faster data transfer possible.
This scheme could interfere with slow serial transmissions, since the DATA line is
shared by both schemes. Thus, circuitry exists that will isolate the fast serial drivers
from the slow serial bus. Setting FSDIR to input mode is sufficient to remove any
possible fast serial interactions with the slow serial bus, other than the additional device
loading, which is not a problem at slow serial bus speeds.
In order to ensure compatibility with the C64, however, the slow serial bus cannot
interfere with the fast drivers, since these drivers are shared with the User Port and a
user port device could presumably make use of them. Once C64 mode is set, the input
direction of the interface circuitry is disabled. Thus, in C64 mode, the FSDIR bit must
be set to input to remove fast to slow interference, but slow to fast interference is
automatically removed by invoking C64 mode. There is no way to disable slow to fast
interference in C l 28 mode (at least not simultaneously with the elimination of fast to
slow interference).
C l 28 HARDWARE SPECIFICATIONS 63S
THE EXPANSION BUS

The C128 Expansion Bus is compatible with the C64 Expansion Bus, while at the same
time allowing extended capabilities in C128 mode.
CARTRIDGE A D D IT IO N
The C128 can use larger and more sophisticated cartridges than the C64. One of the
main reasons for this is the new banking scheme used in the C l 28 for external
cartridges. The C64 uses two hardware control lines, /EXROM and /GAME, to control
banking out of internal facilities and banking in of cartridge facilities. The C128 uses a
software polling method, where upon power-up it polls the cartridge, according to a
defined protocol, to determine if such a cartridge exists, and if so, what its software
priority is. Since the C128 is always free to bank between cartridges and built-in ROM,
an external application can take advantage of internal routines and naturally become an
extended part o f the C128, as opposed to becoming a replacement application. See
Chapter 13 for information on the Auto Start Cartridge ROM sequence.
The elimination of /EXROM and /GAME as hardware control lines for cartridge
identification (in C128 mode) has freed up both of these lines for extended functioning.
Both of the lines appear as bits in the MMU Mode Configuration Register, and are both
input and output ports. Neither has a dedicated function other than general cartridge
function expansion, and lend themselves to act as latched banking lines or input sense
lines. O f course, neither can be asserted on C l 28 power-up or C64 mode will automati
cally be set.
DMA CAPABILITY
The C l 28 expansion bus supports DMAs in a fashion similar to that o f the C64. A C64
DMA is achieved by pulling the /DMA pin on the expansion bus low. Immediately after
this happens, the RDY and AEC lines of the processor are brought low. This can neatly
shut down the processor, but it can also cause problems, depending on what the
processor is doing at the time. The RDY input of an 8502 series processor, when brought
low, will halt the processor of the next <3>1 cycle, leaving the processor’s address lines
reflecting the current address being fetched. However, if the processor is in a write cycle
when RDY is brought low, it will ignore RDY until the next read cycle. Thus, in the
C64, a /DMA input occurring during a write cycle will tri-state the processor’s address
and data bus, but not stop it until up to three cycles later when the next read cycle
occurs. The write cycles following the /DMA input do not actually write, causing
memory corruption and often processor fatality when the /DMA line is released. Any
/DMA input during 3>2 is a potentially fatal DMA.
If a proper /DMA is asserted, the C64 tri-states and shuts down, allowing the
DMA source complete access to the processor bus. Such a DMA source must monitor
the <52 and BA outputs, as it must tri-state when the VIC is on the bus, and it must
completely DMA when a VIC DMA is called for. The VIC chip always has the highest
636 COMMODORE 128
DMA priority. When on the bus, the DMA source has access to RAM , ROM and I/O
in the C64 scheme. A proper DMA shutdown is usually achieved via some C64 software
handshaking with the DMA source.
The C128 system uses a similar DMA scheme. When the /DMA input goes low,
the RDY input to the 8502, the AEC input to the 8502, and the /BUSRQST input to the
Z80 immediately go low. Additionally, the gated AEC signal, GAEC, goes low,
causing the MMU to go immediately to its VIC CYCLE MODE, and the Z80 Data Out
buffer to tri-state. The DMA causes the Address to the Shared Address buffer to reverse
direction, and the Translated Address to the Address buffer to be enabled, giving the
external DMA source complete access to the processor address bus. The PLA is still
looking at ungated AEC and as such will allow access to I/O devices, RAM and ROM.
There can be no access to the MMU; thus for C l28 memory mapping the memory map
must be set up before being DM A’ed. For C64 mode, memory mapping is done by the
8502 processor port lines and by the external /EXROM and /GAME. Since the 8502
ports will be inaccessible by a DMA source, only the C64 map changes based upon
/EXROM and /GAME can be made during a DMA. This is the same as in a C64 unit.
All DMA sources, as with the C64, must yield to the VIC during $ 0 or BA low. The
C128 can perform a destructive DMA as easily as the C64. In order to use DMA’s, the DMA
source will most likely have to cooperate with a C128 mode program, allowing it to
handshake with a DMA source to effect D M A ’s nondestructively.
EXPANSIO N BUS P IN O U T
PIN NAME D ESCRIPTION
1 GND System Ground

2 + 5V System Vcc
3 + 5V System Vcc
4 /IRQ Interrupt Request
5 R/W System Read W rite Signal
6 DCLOCK 8.18 M Hz Video Dot Clock
7 I/O, I/O Chip Select: $DEOO-$DEFF, Active Low
8 /GAME Sensed for M emory Map Configuration
9 /EXROM Sensed for M emory Map Configuration
10 I /0 2 I/O Chip Select: $D F00-$DFFF, Active Low.
11 /ROM l External ROM Chip Select, $8000-$B FFF in C128 M ode
($8000-$9FFF in C64 mode)
12 BA Bus Available Output
13 /DMA Direct Memory Access Input (see caution on DMA capability
on p. 636)
14 d7 Data Bit 7
15 d6 Data Bit 6
16 d5 Data Bit 5
17 d4 Data Bit 4
18 d3 Data Bit 3
19 d2 Data Bit 2
20 D, Data Bit 1
21 D0 Data Bit 0
22 GND System Ground
A GND System Ground
B /ROM h External ROM Chip Select, $C 000-$FFFF in C128 M ode
($C 000-$FFFF in C64 mode)
C /RESET System Reset Signal
D /NM I Non-M askable Interrupt Request
E 1MHz System 1MHz cb0 Clock
F T A 15 Translated Address Bit 15
H TA 14 Translated Address Bit 14
J t a 13 Translated Address Bit 13
K t a 12 Translated Address Bit 12
L TA n Translated Address Bit 11
M TA,„ Translated Address Bit 10
N ta9 Translated Address Bit 9
P ta8 Translated Address Bit 8
R sa 7 Shared Address Bit 7
S sa 6 Shared Address Bit 6
T sa 5 Shared Address Bit 5
U sa 4 Shared Address Bit 4
V sa 3 Shared Address Bit 3
w sa 2 Shared Address Bit 2
X SA, Shared Address Bit 1
Y SA0 Shared Address Bit 0
Z GND System Ground
T able 16-21. Expansion Bus P inout

638 COMMODORE 128
THE VIDEO INTERFACE

The C l28 VIC video interface hardware allows the connection of a standard NTSC or
PAL commercial television and/or a color monitor. The monitor can accept either a
composite video signal or separate chroma and luminance/sync signals in addition to an
audio signal. This output is very similar to the output in later revision C64 units.
The C l 28 also provides 80-column video interfacing. The available 80-column
display is RGBI and monochrome, able to interface to most NTSC- or PAL-compatible
RGBI TYPE I monitors and most 80-column-compatible NTSC or PAL monochrome
monitors.
THE VIC VIDEO INTERFACE

The following items specify the VIC video interface for 40-column display in sixteen
colors. The VIC signal is available at analog levels at the video connector and at RF
levels at the RF output.
M O D U LA TO R SPECIFICATION
The modulator provides a broadcast-type RF signal carrying the VIC composite video
and audio signals. The NTSC modulator is switchable between channels 3 and 4 to help
minimize local broadcast interference. The signal generated by the RF modulator
complies with the FCC ruling concerning FCC Class B, TV interface devices. The RF
output is accessible via a standard RCA-type phono/video jack.
M O N IT O R O U T P U T
The VIC video output provides the signals shown in Table 16-22.
SIGNAL LEVEL IM PEDENCE DC OFFSET
Luminance/Sync 1V p-p 75 fl 0.5 V

Chroma IV p-p 75 fl 0.5 V
Composite 1V p-p 75 fl 0.5 V
Audio 1V p-p IK fl
Table 16-22. V IC Video O utput Signals
V ID E O C O N N E C TO R P IN O U T
The VIC video connector exists physically as an eight-pin DIN connector. It provides
the signals shown in Table 16-23.
PIN SIGNAL
1 Luminance/Sync
2 Ground
3 Audio Out
4 Composite
5 Audio In
6 Chroma
7 N .C.
8 N .C.
Table 16-23. Video Connector Signals
THE 8563 VIDEO INTERFACE

The following items specify the 8563 video interface for 80-column display in sixteen
colors. The 8563 signal is available at digital levels for RGBI and at a three-level
derived analog for black and white composite video.
M O N IT O R O U T P U T
Table 16-24 shows the signals provided by the 8563 output.
SIGNAL LEVEL IM PEDENCE
Red TTL TTL

Green TTL TTL
Blue TTL TTL
Intensity TTL TTL
HSync TTL TTL
VSync TTL TTL
Composite 75 n
Full Intensity 2.0 V

H alf Intensity 1.5 V
Sync 0.5 V
Table 16-24. 8563 O utput Signals
V ID E O C O N N E C TO R P IN O U T
The 8563 video connector is an IBM-style D9 connector, providing the signals shown in
Table 16-25.
640 COMMODORE 128
PIN SIGNAL
1 Ground
2 Ground
3 Red
4 Green
5 Blue
6 Intensity
7 Monochrome (non-standard)
8 Horizontal Sync
9 Vertical Sync
T ab le 16-25. 8563 Video C o n necto r Pinout
THE KEYBOARD
The C l28 Keyboard is an advance over the standard C64 keyboard, while still maintain
ing full compatibility. It has several extra keys that are used in C l 28 mode, but not in
C64 mode. It features a numeric keypad, a HELP key, extended function keys, a
true CAPS L O C K key, and a 40/80 column switch key, all of which are
strobed by the VIC chip or tied to dedicated 8502 or MMU I/O lines.
CONNECTOR PINOUT
The C l 28 keyboard is designed to be connected by one 12- and one 13-pin internal
single-in-line connector for the unit with a built-in keyboard. Table 16-26 illustrates
both connections.
D-TYPE SIGNAL
1 Ground
2 Key
3 Restore
4 + 5V
5 Row 3
6 Row 6
7 Row 5
8 Row 4
9 Row 7
10 Row 2
11 Row 1
12 Row 0
13 Column 0
14 Column 6
15 Column 5
16 Column 4
17 Column 3
18 Column 2
19 Column 1
20 Column 7
21 K„
22 Ki
23 k2
24 40/80
25 Alpha Lock
T a b le 16-26. T h e Keyboard C o nnector Pinout

642 COMMODORE 128
THE C l28 KEYBOARD TABLE
CO Cl C2 C3 C4 C5 C6 C7 K0 K1 K2 GND
PIN13 PIN19 PIN18 PIN17 PIN16 PIN15 PIN14 PIN20 PIN21 PIN22 PIN23 PIN-1
RO INS # % ) + lb i HELP ESC ALT

PIN12 DEL 3 5 7 9 1
Rl RET W R Y I P * 8 + 0
«-
PIN11
R2 t A D G J L ] CTRL 5 -
PIN10 1 »
R3 F8 $ & ( 0 - CLR TAB LINE

PIN5 F7 4 6 8 HOM 2 FEED t
R4 F2 Z C B M > RIGHT SPACE 2 ENTER

PIN8 1
FI SHIFT BAR
R5 F4 s F H K [ = 4 6
PIN7 F3 «-
R6 F6 E T U 0 @ •ir Q 7 9
PIN6 F5 -
R7 LEFT X V N < •> RUN 1 3 NO

PIN9 — > SHIFT > 1 STOP SCRL
/ (LOCKING)
NM I
PIN3
/ (LOCKING)
40/80 40/80
PIN24 DSPLY
/ (LOCKING)
P6510 CAPS
PIN25 LOCK
N O T E : Pins RO through R7 pertain to the keyboard row values for the keyboard SCAN.
These pins correspond to bits 0 through 7 of location 56321 ($DC01).
Pins CO through C7 are the keyboard’s column values, which correspond to bits 0
through 7 of location 56320 (SDC00).
Pins K0 through K2 pertain to the C128 keyboard control register bit, 0 through 2 of
location 53295 ($D02F).
APPENDIXES
A ppendix A - BA SIC Language E rro r Messages 644
Appendix B - D O S E rro r Messages 648
Appendix C - Connectors/Ports for Peripheral

Equipm ent 652
Appendix D - Screen Display Codes 658
Appendix E - A S C II and C H R $ Codes 660
Appendix F - Screen and C olor M em o ry Maps 663
Appendix G - D erived T rig o n o m etric Functions 665
Appendix H - C ontrol and Escape Codes 666
Appendix I - BA SIC 7.0 Abbreviations 670
Appendix J - Disk C om m and S um m ary 674
A ppendix K -
P a rt 1-C om m odore 128 CP/M 676
P art 11-Calling C P/M BIO S, 8502 BIO S, and CP/M

User Functions in Z80 Machine Language 702
P art lll-T h e CP/M System M em o ry M ap 709
A ppendix L - C om m odore 128 System Schematics 721
643
644 COMMODORE 128
A PPENDIX A
BASIC LANGUAGE
ERROR MESSAGES
The following error messages are displayed by BASIC. Error messages can also be
displayed with the use of the ERR$() function. The error numbers below refer only to
the number assigned to the error for use with the ERR$() function.
ERROR # ERROR NAME DESCRIPTION
1 TOO MANY FILES There is a limit of ten files OPEN at one

time.
2 FILE OPEN An attempt was made to open a file using

the number of an already open file.
3 FILE NOT OPEN The file number specified in an I/O state

ment must be opened before use.
4 FILE NOT FOUND Either no file with that name exists (disk)
or an end-of-tape marker was read (tape).
5 DEVICE NOT PRESENT The required I/O device is not available or

buffers deallocated (cassette). Check to make
sure the device is connected and turned
on.
6 NOT INPUT FILE An attempt was made to GET or INPUT

data from a file that was specified as output
only.
7 NOT OUTPUT FILE An attempt was made to send data to a file

that was specified as input only.
8 MISSING FILE NAME File name missing in command.
9 ILLEGAL DEVICE NUMBER An attempt was made to use a device im

properly (SAVE to the screen, etc.).
10 NEXT WITHOUT FOR Either loops are nested incorrectly, or there

is a variable name in a NEXT statement
that doesn’t correspond with one in FOR.
11 SYNTAX A statement not recognized by BASIC. This

could be because of a missing or extra
parenthesis, a misspelled key word, etc.
APPENDIXES 64S
12 RETURN WITHOUT GOSUB A RETURN statement was encountered

when no GOSUB statement was active.
13 OUT OF DATA A READ statement is encountered without
any data left to READ.
14 ILLEGAL QUANTITY A number used as the argument of a func

tion or statement is outside the allowable
range.
15 OVERFLOW The result of a computation is larger than
the largest number allowed (1.7014118 34E
+ 38).
16 OUT OF MEMORY Either there is no more room for program
code and/or program variables, or there are
too many nested DO, FOR or GOSUB state
ments in effect.
17 UNDEF’D STATEMENT A referenced line number doesn’t exist in
the program.
18 BAD SUBSCRIPT The program tried to reference an element
of an array out of the range specified by
the DIM statement.
19 REDIM ’D ARRAY An array can only be DIMensioned once.
20 DIVISION BY ZERO Division by zero is not allowed.
21 ILLEGAL DIRECT INPUT, GET, IN PU T #, G E T # and GET-
KEY statements are allowed only within a
program.
22 TYPE MISMATCH This error occurs when a numeric value is
assigned to a string variable or vice versa.
23 STRING TOO LONG A string can contain up to 255 characters.
24 FILE DATA Bad data read from a tape or disk file.
25 FORMULA TOO COMPLEX The computer was unable to evaluate this
expression. Simplify the expression (break
into two parts or use fewer parentheses).
26 C A N T CONTINUE The CONT command does not work if the
program was not RUN, if there was an
error, or if a line has been edited.
27 UNDEF’D FUNCTION A user-defined function that was never de
fined was referenced.
646 COMMODORE 128
28 VERIFY The program on tape or disk does not match

the program in memory.
29 LOAD There was a problem loading. Try again.
30 BREAK The STOP command was issued in a pro

gram or the S T O P key was pressed
to halt program execution.
31 CA N ’T RESUME A RESUME statement was encountered

without a TRAP statement in effect.
32 LOOP NOT FOUND The program has encountered a DO state

ment and cannot find the corresponding
LOOP.
33 LOOP WITHOUT DO LOOP was encountered without a DO state

ment active.
34 DIRECT MODE ONLY This command is allowed only in direct

mode, not from a program.
35 NO GRAPHICS AREA A command (DRAW, BOX, etc.) to create
graphics was encountered before the
GRAPHIC command was executed.
36 BAD DISK An attempt failed to HEADER a diskette,
because the quick header method (no ID)
was attempted on an unformatted diskette
or the diskette is bad.
37 BEND NOT FOUND The program encountered an “ IF . . . THEN
BEGIN” or “ IF . . . THEN . . . ELSE
BEGIN” construct, and could not find a
BEND keyword to match the BEGIN.
38 LINE NUMBER TOO LARGE An error has occurred in renumbering a
BASIC program. The given parameters re
sult in a line number greater than 63999
being generated; therefore, the renumbering
was not performed.
39 UNRESOLVED REFERENCE An error has occurred in renumbering a
BASIC program. A line number referred to
be a command (e.g., GOTO 999) does not
exist. Therefore, the renumbering was not
performed.
40 UNIMPLEMENTED A command not supported by BASIC 7.0
COMMAND was encountered.
APPENDIXES 647
41 FILE READ An error condition was encountered while

loading or reading a program or file from
the disk drive (e.g., opening the disk drive
door while a program was loading).
648 COMMODORE 128
A PPENDIX B
DOS ERROR MESSAGES

The following DOS error messages are returned through the DS and DS$ variables. The
DS variable contains just the error number, and the DS$ variable contains the error
number, the error message, and any corresponding track and sector number. NOTE:
Error message numbers less than 20 should be ignored with the exception of 01, which
gives information about the number of files scratched with the SCRATCH command.
ERROR
NUMBER ERROR MESSAGE AND DESCRIPTION
20 READ ERROR (block header not found)

The disk controller is unable to locate the header of the requested data
block. Causes: an illegal sector number, or the header has been destroyed.
21 READ ERROR (no sync character)

The disk controller is unable to detect a sync mark on the desired track.
Causes: misalignment of the read/write head, no diskette, or an unformat
ted or improperly seated diskette. Can also indicate a hardware failure.
22 READ ERROR (data block not present)

The disk controller has been requested to read or verify a data block that
was not properly written. This error occurs in conjunction with the BLOCK
commands and can indicate an illegal track and/or sector request.
23 READ ERROR (checksum error in data block)

This error message indicates there is an error in one or more of the data
bytes. The data has been read into the DOS memory, but the checksum
over the data is in error. This message may also indicate hardware ground
ing problems.
24 READ ERROR (byte decoding error)
The data or header has been read into the DOS memory but a hardware
error has been created owing to an invalid bit pattern in the data byte. This
message may also indicate hardware grounding problems.
25 WRITE ERROR (write-verify error)
This message is generated if the controller detects a mismatch between the
written data and the data in the DOS memory.
26 WRITE PROTECT ON
This message is generated when the controller has been requested to write
a data block while the write protect switch is depressed. This is caused by
using a diskette with a write protect tab over the notch or a notchless diskette.
APPENDIXES 649
ERROR
27 READ ERROR
This message is generated when a checksum error has been detected in the
header of the requested data block. The block has not been read into DOS
memory.
28 W RITE ERROR
This error message is generated when a data block is too long and
overwrites the sync mark of the next header.
29 DISK ID MISMATCH
This message is generated when the controller has been requested to access
a diskette that has not been initialized or improperly formatted. The
message can also occur if a diskette has a bad header.
30 SYNTAX ERROR (general syntax)

The DOS cannot interpret the command sent to the command channel.
Typically, this is caused by an illegal number of file names, or patterns
that are illegally used. For example, two file names appear on the left side
of the COPY command.
31 SYNTAX ERROR (invalid command)

The DOS does not recognize the command. The command must start in the
first position.

The command sent is longer than 58 characters. Use abbreviated disk
commands.
33 SYNTAX ERROR (invalid file name)

Pattern matching is invalidly used in the OPEN or SAVE command. Spell
out the file name.
34 SYNTAX ERROR (no file given)

The file name was left out of the command or the DOS does not recognize
it as such. Typically, a colon(:) has been left out of the command.

This error may result if the command sent to the command channel
(secondary address 15) is unrecognized by the DOS.
50 RECORD NOT PRESENT

Result of disk reading past the last record through IN PU T # or G E T #
commands. This message will also appear after positioning to a record
beyond end-of-file in a relative file. If the intent is to expand the file by
adding the new record (with a PR IN T# command), the error message may
be ignored. IN PU T # and G E T # should not be attempted after this error is
detected without first repositioning.
650 COMMODORE 128
ERROR
51 OVERFLOW IN RECORD
PR IN T # statement exceeds record boundary. Information is truncated.
Since the carriage return that is sent as a record terminator is counted in the
record size, this message will occur if the total characters in the record
(including the final carriage return) exceed the defined size of the record.
52 FILE TOO LARGE
The record position within a relative file indicates that disk overflow will
result.
60 W RITE FILE OPEN
This message is generated when a write file that has not been closed is
being opened for reading.
61 FILE NOT OPEN
A file that has not been opened in the DOS is being accessed. Sometimes
in this situation, a message is not generated; the request is simply ignored.
62 FILE NOT FOUND
The requested file does not exist on the indicated drive.
63 FILE EXISTS
The file name of the file being created already exists on the diskette.
64 FILE TYPE MISMATCH
The requested file access is not possible using files of the type named.
Reread the chapter covering that file type.
65 NO BLOCK
Occurs in conjunction with block allocation. The sector you tried to allocate
is already allocated. The track and sector numbers returned are the next
higher track and sector available. If the track number returned is 0, all
remaining higher sectors are full. If the diskette is not full yet, try a
lower track and sector.
66 ILLEGAL TRACK AND SECTOR
The DOS has attempted to access a track or a block that does not exist in
the format being used. This may indicate a problem reading the pointer to
the next block.
67 ILLEGAL SYSTEM T OR S
This special error message indicates an illegal system track or sector.
70 NO CHANNEL (available)
The requested channel is not available, or all channels are in use. A
maximum of five buffers are available for use. A sequential file requires
two buffers; a relative file requires three buffers; and the error/command
channel requires one buffer. You may use any combination of those as
long as the combination does not exceed five buffers.
APPENDIXES 651
ERROR
71 DIRECTORY ERROR
The BAM (Block Availability Map) on the diskette does not match the
copy on disk memory. To correct this, initialize the disk drive.
72 DISK FULL
Either the blocks on the diskette are used, or the directory is at its entry
limit. DISK FULL is sent when two blocks are still available on the
diskette, in order to allow the current file to be closed.
73 DOS VERSION NUMBER (73, CBM DOS V30 1571, 00, 00)
DOS 1 and 2 are read compatible but not write compatible. Disks may be
interchangeably read with either DOS, but a disk formatted on one version
cannot be written upon with the other version because the format is
different. This error is displayed whenever an attempt is made to write
upon a disk that has been formatted in a noncompatible format. This
message will also appear after power-up or reset and is not an error in this
case.
74 DRIVE NOT READY

An attempt has been made to access the disk drive without a diskette
inserted; or the drive lever or door is open.
652 COMMODORE 128
A PPEN DIX C
CONNECTORS/PORTS FOR
PERIPHERAL EQUIPMENT
1. Power Socket 7. Serial Port
2. Power Switch 8. Composite Video Connector
3. Reset Button 9. Channel Selector
4. Controller Ports 10. RF Connector
5. Expansion Port 11. RGBI Connector
6. Cassette Port 12. User Port
5 6 7 8 9 10 11 12
APPENDIXES 653
SIDE PANEL C O N N E C TIO N S
1. Power Socket— The free end of the cable from the power supply is attached here.
2. Power Switch— Turns on power from the transformer.
3. Reset Button— Resets computer (warm start).
4. Controller Ports— There are two Controller ports, numbered 1 and 2. Each Con
troller port can accept a joystick mouse or a game controller paddle. A light pen
can be plugged only into port 1, the port closest to the front of the computer. Use
the ports as instructed with the software.
CONTROLLER PORT 1
PIN TYPE NOTE
1 JOYAO
2 JOYA1
3 JOYA2
4 JOYA3
5 POT AY
6 BUTTON A/LP
7 + 5V MAX. 50mA
8 GND
9 POT AX
CONTROLLER PORT 2
PIN TYPE NOTE
1 JOYBO
2 JOYB1
3 JOYB2
4 JOYB3
5 POT BY
6 BUTTON B
7 + 5V M AX. 50mA
8 GND
9 POT BX
REAR C O N N E C T IO N S
5. Expansion Port— This rectangular slot is a parallel port that accepts program or
game cartridges as well as special interfaces.
654 COMMODORE 128
CARTRIDGE
EXPANSION PORT
PIN TYPE PIN TYPE
12 BA 1 GND
13 DMA 2 + 5V
14 D7 3 + 5V
15 D6 4 IRQ.
16 D5 5 R/W
17 D4 6 Dot Clock
18 D3 7 I/O 1
19 D2 8 GAME
20 Dl 9 EXROM
21 DO 10 I/O 2
22 GND 11 ROML
PIN TYPE PIN TYPE
N A9 A GND
P A8 B ROMH
R A7 C RESET
S A6 D NMI
T A5 E S 02
U A4 F A15
V A3 H A14
w A2 J A13
X A1 K A12
Y AO L A ll
Z GND M A10
22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 8 5 4 3 2
W W fW ■
.................................................................
ZY X W V U T S R P N M L K J H F E 0 C 8 A
(view of port from the back of the C128)
6. Cassette Port— A 1530 Datassette recorder can be attached here to store programs
and information.
CASSETTE PORT
PIN TYPE
1 2 3 4 5 6
A -l GND
B-2 +5V
C-3 CASSETTE MOTOR
D-4 CASSETTE READ
E-5 CASSETTE WRITE
F-6 CASSETTE SENSE
APPENDIXES 655
7. Serial Port— A Commodore serial printer or disk drive can be attached directly to
the Commodore 128 through this port.
SERIAL I/O PORT

PIN TYPE
1 SERIAL SRQIN
2 GND
3 SERIAL ATN IN/OUT
4 SERIAL CLK IN/OUT
5 SERIAL DATA IN/OUT
6 RESET
8. Composite Video Connector—This DIN connector supplies direct audio and com
posite video signals. These can be connected to the Commodore monitor or used
with separate components. This is the 40-column output connector.
(view of port while facing the rear of the C128)
COMPOSITE VIDEO CONNECTOR

PIN TYPE NOTE

2 GND
3 AUDIO OUT
5 AUDIO IN
7 NC No connection
8 NC No connection
656 COMMODORE 128
9. Channel Selector— Use this switch to select which TV channel (L = channel 3, H

= channel 4) the computer’s picture will be displayed on when using a television
instead of a monitor.
10. RF Connector— This connector supplies both picture and sound to your television
set. (A television can display only a 40-column picture.)
11. RGBI Connector— This 9-pin connector supplies direct audio and an RGBI (Red/
Green/ Blue/Intensity) signal. This is the 80-column output.
4 3 2 1
O O O 0
o o o o
9 8 7 6
RGBI CONNECTOR
PIN SIGNAL
1 Ground
2 Ground
3 Red
4 Green
5 Blue
6 Intensity
7 Monochrome
8 Horizontal Sync
9 Vertical Sync
12. User Port— Various interface devices can be attached here, including a Commo
dore modem.
APPENDIXES 657
USER I/O PO R T
PIN TYPE NOTE
1 GND
2 + 5V M AX. 100mA
3 RESET
4 CNT1
5 SP1
6 CNT2
7 SP2
8 PC2
9 SER. ATN IN
10 9 VAC MAX. 100mA
11 9 VAC MAX. 100mA
12 GND
PIN TYPE NOTE
A GND
B FLAG2
C PBO
D PB1
E PB2
F PB3
H PB4
J PB5
K PB6
L PB7
M PA2
N GND
1 2 3 4 5 6 7 8 9 10 11 12
A B C D E F H J K L M N
658 COMMODORE 128
APPENDIX D
SCREEN DISPLAY CODES

SCREEN DISPLAY CODES
40 COLUM NS
The chart below lists all the characters built into the Commodore screen character sets.
It shows which numbers should be POKEd into the VIC chip (40-column) screen
memory (location 1024 to 2023) to get a desired character on the 40-column screen.
(Remember, to set color memory, use locations 55296 to 56295.) Also shown is which
character corresponds to a number PEEKed from the screen.
Two character sets are available. Both are available simultaneously in 80-column
mode, but only one is available at a time in 40-column mode. The sets are switched by
holding down the S H I F T and C£ (Commodore) keys simultaneously. The entire
screen of characters changes to the selected character set.
From BASIC, PRINT CHR$(142) will switch to upper case/graphics mode and
PRINT CHR$(14) will switch to upper/lower case mode.
Any number on the chart may also be displayed in reverse. The reverse character
code can be obtained by adding 128 to the values shown.
SET 1 SET 2 POKE SET 1 SET 2 POKE SET 1 SET 2 POKE
@ 0 N n 14 £ 28
A a 1 0 0 15 ] 29
B b 2 P P 16 t 30
C c 3 Q q 17 «- 31
D d 4 R r 18 I SPACE p ffe 32
E e 5 S s 19 I 33
F f 6 T t 20
" 34
G 9 7 U u 21 # 35
H h 8 V V 22 $ 36
I i 9 w w 23 % 37
J j 10 X X 24 & 38
K k 11 Y y 25 - 39
L I 12 z z 26 ( 40
M m 13 [ 27 ) 41
APPENDIXES 659
SET 1 SET 2 POKE SET 1 SET 2 POKE SET 1 SET 2 POKE
* 42 □ G 71 □
100
+ 43 □ H 72 □ 101
44 1 73 m 102
□
45 J 74 □ 103
□
46 Q K 75 a 104
47 □ L 76 E 25 105
0 48 S
M 77 □ 106
1 49 0 N 78 107
2 50 n 0 79 □ 108
3 51 n P 80 □ 109
4 52 m Q 81 H 110
5 53 □ R 82 □
111
6 54 S 83 H
112
7 55 □ T 84 B 113
8 56 □ U 85 H 114
9 57 V 86 ffl 115
58 □ w 87 D 116
J 59 a X 88 E 117
< 60 □ Y 89 □ 118
= 61 a z 90 n 119
> 62 a 91 H 120
? 63 E 92 a 121
B 64 m 93 □ 0 122
A 65 0 94 B 123
B 66 a 95 a 124
m
B C 67 I SPACE
I
96 H 125
D 68 97 H 126
B 1
Q E 69 a
98 S 127
□ F 70 n
99
C o d e s from 128-255 are reversed im ages of co d e s 0-127.

660 COMMODORE 128
APPENDIX E
ASCII AND CHR$ CODES

This appendix shows you what characters will appear if you PRINT CHR$(X), for all
possible values of X. It also shows the values obtained by typing PRINT ASC (“ x ” ),
where x is any character that can be displayed. This is useful in evaluating the character
received in a GET statement, converting upper to lower case and printing character-
based commands (such as switch to upper/lower case) that could not be enclosed in
quotes.
PRINTS CHR* PRINTS CHR* PRINTS CHR* PRINTS CHR*
0 23 46 E 69
1 24 / 47 F 70
2 25 0 48 G 71
3 26 1 49 H 72
4 27 2 50 I 73
5 28 3 51 J 74
6 B53 29 4 52 K 75
7 30 5 53 L 76
oisABiEsQggB 31 6 54 M 77
ENABLES U f l j J Q 9 ■ SPACE 1 32 7 55 N 78
10 ! 33 8 56 O 79
11 " 34 9 57 P 80
12 # 35 58 Q 81
■ return^H 13 $ 36 59 R 82
r s s i
1
■ owl Rcast I 1 4 % 37 <Z 60 S 83
15 & 38 = 61 T 84
16 39 3 62 U 85
17 40 ? 63 V 86
JHB (
J E li* 18 ) 41 @ 64 w 87
19 * A 65 X 88
42
B 20 + 43
44
B 66
67
Y 89
21 . C z 90
22 - 45 D 68 [ 91
APPENDIXES 661
PRINTS CHRS PRINTS CHRS PRINTS CHRS PRINT! CHRS
£ 92 115 f4 38 B 161
] 93 □ 116 f6 39 H 162
t 94 □ 117 f8 40 n 163
- 95 118 41 □ 164
B 96 □ 119
■ S W U l ' M TO
■ u P P t H cast
V 42
I □ 165
97 120 43 m 166
CD 96 □ 121 EOL 44 □ 167
99 122 fm n 45 a 168
B
100 123 46 E 169
B ffl
□ 101 E 124 ■
l f l H clr
OME
m
MmM
47 □ 170
102 125 148 ffl 171

a m B
□ 103 0 126 Brown 149 D3
H
172
173
□ 104 a 127 Lt. Red 150 i— i
105 128 Dk. Gray 151 H 174
S I— i
106 Orange 129 Gray 152 y 175

□ i— i
q 107 130 Lt. Green 153 H 176
□ 108 131 Lt. Blue 154 B 177
K 109 132 Lt. Gray 155 B 178
0 110 f1 133 156 ffl 179
□ 111 f3 134 157 C 180
n 112 f5 135 H I* 158 c 181

n 113 f7 136 159 □ 182
□ 114 12 137 ^ .S P A C E 160 n 183
PRINTS CHRS PRINTS CHRS PRINTS CHRS PRINTS CHRS
n 184 □ 186 H 188 E 190
y 185 E 187 yj 189 i 191
CO DES 192-223 SA M E AS 96-127

CO D E S 224-254 SA M E AS 160-190
CO D E 255 SA M E AS 126
662 COMMODORE 128
N O TE: The 80-column (RGBI) output has three colors that are different
from the 40-column (composite video) color output. This means that the
character string codes that represent color codes for these three colors are
used differently depending on which video output is used. The following
character string codes represent these colors in each video output.
CHR$ 40-COLUMN (VIC COM POSITE) 8O-COLUMN (8563 RGBI)
129 Orange Dark Purple

149 Brown Dark Yellow
151 Dark Gray Dark Cyan
APPENDIXES 663
A PPENDIX F
SCREEN AND COLOR MEMORY MAPS—

C l28 MODE, 40-COLUMN
AND C64 MODE
The maps below display the memory locations used in 40-column mode (C l28 and C64)
for identifying the characters on the screen as well as their color. Each map is separately
controlled and consists of 1000 positions.
The character displayed on the maps can be controlled directly with the POKE
command.
V IC C H IP (4Q -C O L U M N ) SCREEN M E M O R Y M AP
COLUMN
20
55335
I
55296 -
55336
55376
55416
55456
55496
55536
55576
55616
55656
55696 10 o
55736
55776
55816
55856
55896
55936
55976
56016
56056
56096
56136
56176
56216
56256
The Screen Map is POKEd with a Screen Display Code value (see Appendix D).
For example:
PO K E 1024, 13 (from BANK 0 or 15)
will display the letter M in the upper-left corner of the screen.
664 COMMODORE 128
V IC C H IP (4 0 -C Q L U M N ) C O L O R M E M O R Y M AP
COLUMN
0 10 20 30 39
1063
\
1024------- -
1064
1104
1144
1184
1224
1264
1304
1344
1384
1424 10 o
1464
1504
1544
1584
1624
1664
1704
1744
1784
1824
1864
1904
1944
1984
t
2023
If the color map is POKEd with a color value, this changes the character color. For
example:
PO K E 55296,1 (from BANK 15)
will change the letter M inserted above from light green to white.
C O LO R CO DES—40 CO LUM NS
0 Black 8 Orange
1 White 9 Brown
2 Red 10 Light Red
3 Cyan 11 Dark Gray
Border Control M emory 53280

Background Control Memory 53281
APPENDIXES 665
A PPENDIX G
DERIVED TRIGONOMETRIC
FUNCTIONS
FUNCTION BASIC EQUIVALENT
SECANT SEC(X) = l/COS(X)

COSECANT CSC(X) = 1/SIN(X)
COTANGENT COT(X) = 1/TAN(X)
INVERSE SINE ARCSIN(X) = ATN(X/SQR(-X*X +1))
INVERSE COSINE ARCCOS(X) = -ATN(X/SQR
(-X*X + 1)) + tt/2
INVERSE SECANT ARCSEC(X) = ATN(X/SQR(X*X-I))
INVERSE COSECANT ARCCSC(X) = ATN(X/SQR(X*X-1))
+ (SGN(X)-1)*tt/2
INVERSE COTANGENT ARCCOT(X) = -ATN(X) + tt/2
HYPERBOLIC SINE SINH(X) = (EXP(X)-EXP(-X))/2
HYPERBOLIC COSINE COSH(X) = (EXP(X) + EXP(-X))/2
HYPERBOLIC TANGENT TANH(X) = -EXP(-X)/(EXP(X) + EXP
(-X))*2 + 1
HYPERBOLIC SECANT SECH(X) = 2/(EXP(X) + EXP(-X))
HYPERBOLIC COSECANT CSCH(X) = 2/(EXP(X)-EXP(-X))
HYPERBOLIC COTANGENT COTH(X) = EXP(-X)/(EXP(X)
-EXP(-X))*2 +1
INVERSE HYPERBOLIC SINE ARCSINH(X) = LOG(X + SQR(X*X + 1))
INVERSE HYPERBOLIC COSINE ARCCOSH(X) = LOG(X + SQR(X*X-1))
INVERSE HYPERBOLIC TANGENT ARCTANH(X) = LOG(l + X)/(l-X))/2
INVERSE HYPERBOLIC SECANT ARCSECH(X) = LOG(SQR
(-X *X+1)+1/X)
INVERSE HYPERBOLIC COSECANT ARCCSCH(X) = LOG(SGN(X)*SQR
(X*X+1)/X)
INVERSE HYPERBOLIC COTANGENT ARCCOTH(X) = LOG((X + l)/(x-l))/2
666 COMMODORE 128
APPENDIX H
CONTROL AND ESCAPE CODES

C O N TR O L CODES
The table below lists the control codes used by the Commodore 128. The print codes
in the first column are used in PRINT statements. The key codes in the second
column are the sequence of keys pressed to perform specific controls. Hold down
the C O N T R O L key (or the key specified on the left in the key code column)
and strike the key specified to the right in the key code column.
PR IN T
CO D ES K EY CODES
EFFEC TIV E
IN M ODE:
(CHR$) KEY SEQUENCE FUNCTION C64 C128
CHR$(2) CTRL B Underline (80) y

CHR$(5) CTRL 2 or CTRL E Set character color to white / y
CHR$(7) CTRL G Produce bell tone y
CHR$(8) CTRL H Disable character set change /
CHR$(9) CTRL I Enable character set change /
Move cursor to next set tab
position y
CHR$(10) CTRL J Line feed y
CHR$(11) CTRL K Enable character set change y
CHR$(12) CTRL L Disable character mode change y
CHR$(13) CTRL M Send a carriage return and line
feed to the computer and enter a
line of BASIC / y
CHR$(14) CTRL N Set character set to upper/lower
case y y
CHR$(15) CTRL 0 Turn flash on (80) y
CHR$(17) CRSR DOW N/CTRL Move the cursor down one
Q row y y
CHR$(18) CTRL 9 or CTRL R Cause characters to be printed in
reverse field y y
CHR$(19) HOME Move the cursor to the home po
sition (top left) of the display (the
current window) y y
N O TE: (40) = 40-column screen only.
(80) = 80-column screen only.
APPENDIXES 667
PR IN T
CO D ES K EY CODES
EFFEC TIV E
IN M ODE:
CHR$(20) DEL or CTRL T Delete last character typed and

move all characters to the right
o f the deleted character one space
to the left / y
CHR$(24) CTRL X, CTRL TAB
or O TAB Tab set/clear
CHR$(27) ESC or CTRL[ Send an ESC character y
CHR$(28) CTRL 3 or CTRL £ Set character color to red (40)
and (80) / y
CHR$(29) CRSR or CTRL] Move cursor one column to the
right / y
CHR$(30) CTRL 6 or CTRL Set character color to green (40)
and (80) / y
CHR$(31) CTRL 7 or CTRL = Set character color to blue (40)
and (80)
CHR$(34) Print a double quote on screen
and place editor in quote mode / y
CHR$(129) O 1 Set character color to orange (40);
dark purple (80) / y
CHR$(130) Underline off (80) y
CHR$(131) Run a program. This CHR$ code
does not work in PRINT CHR$
(131), but works from keyboard
buffer / y
CHR$(133) FI Reserved CHR$ code for FI key / y
CHR$(134) F3 Reserved CHR$ code for F3 key J y
CHR$(141) SHIFT RETURN,
CTRL ENTER, Send a carriage return and line
O ENTER or feed without entering a BASIC
O RETURN line / y
CHR$(142) Set the character set to upper
case/graphic / y
CHR$(143) Turn flash off (80) y
N O T E : (40) = 40-column screen only.

668 COMMODORE 128
PR IN T
CO D ES KEY CODES
EFFEC TIV E
IN M ODE:
CHR$(144) CTRL 1 Set character color to black (40)

and (80) y y
CHR$(145) CRSR UP M ove cursor or printing position
up one row y y
CHR$(146) CTRL 0 Terminate reverse field display y y
CHR$(I47) CLEAR HOME Clear the window screen and
move the cursor to the top-left
position y y
CHR$(148) INST Move character from cursor
position end o f line right one
column y y
CHR$(149) G 2 Set character color to brown (40);
dark yellow (80) y y
CHR$(150) G 3 Set character color to light red
(40) and (80) y y
CHR$(151) G 4 Set character color to dark gray
(40); dark cyan (80) y y
CHR$(152) G 5 Set character color to medium
gray (40) and (80) y y
CHR$(153) G 6 Set character color to light green
(40) and (80) y y
CHR$(154) G 7 Set character color to light blue
(40) and (80) y y
CHR$(155) G 8 Set character color to light gray
(40) and (80) y y
CHR$(156) CTRL 5 Set character color to purple (40)
and (80) y y
CHR$(157) CRSR LEFT Move cursor left by one column y y
CHR$(158) CTRL 8 Set character color to yellow (40)
and (80)
CHR$(159) or CTRL 4 Set character color to cyan (40);
light cyan (80) y y
N O TE: (40) = 40-column screen only.
APPENDIXES 669
ESCAPE CODES
This table lists the key sequences for the ESCape functions available on the Commodore
128. ESCape sequences are entered by pressing and releasing the ESC key, fol
lowed by pressing the key listed in the right column.
ESCAPE FUNCTION SEQUENCE KEY
Cancel quote, reverse, flash ESC o
Erase to end o f current line ESC Q

Erase to start of current line ESC P
Clear to end of screen ESC @
M ove to start of current line ESC J

M ove to end of current line ESC K
Enable auto-insert mode ESC A

Disable auto-insert mode ESC C
Delete current line ESC D

Insert line ESC I
Set default tab stop (8 spaces) ESC Y

Clear all tab stops ESC Z
Enable scrolling ESC L

Disable scrolling ESC M
Scroll up ESC V
Scroll down ESC W
Enable bell (by controI-G) ESC G

Disable bell ESC H
Set cursor to nonflashing mode ESC E

Set cursor to flashing mode ESC F
Set bottom right corner of screen window at cursor position ESC B

Set top left corner o f screen window at cursor position ESC T
Swap 40/80 column display output device ESC X
The following ESCape sequences have an effect on an 80-column screen on
ESCA PE FUNCTION SEQUENCE KEY
Change to underlined cursor (80) ESC U

Change to block cursor (80) ESC S
Set screen to reverse video (80) ESC R

Return screen to normal (non reverse video) state (80) ESC N
670 COMMODORE 128
A PPEN DIX I
BASIC 7.0 ABBREVIATIONS

N O TE: The abbreviations below operate in upper case/graphics mode. Press the letter
key(s) indicated, then hold down the S H I F T key and press the letter key follow
ing the word SHIFT.
KEYW ORD ABBREVIATION
ABS A SHIFT B
APPEND A SHIFT P
ASC A SHIFT S
ATN A SHIFT T
AUTO A SHIFT U
BACKUP BA SHIFT C
BANK B SHIFT A
BEGIN B SHIFT E
BEND BE SHIFT N
BLOAD B SHIFT L
BOOT B SHIFT O
BOX none
BSAVE B SHIFT S
BUMP B SHIFT U
CATALOG C SHIFT A
CHAR CH SHIFT A
CHR$ C SHIFT H
CIRCLE C SHIFT I
CLOSE CL SHIFT 0
CLR C SHIFT L
CMD C SHIFT M
COLLECT COLL SHIFT E
COLLISION COL SHIFT L
COLOR COL SHIFT O
CONCAT C SHIFT O
CONT none
COPY CO SHIFT P
COS none
DATA D SHIFT A
DEC none
DCLEAR DCL SHIFT E
DCLOSE D SHIFT C
DEF FN none
DELETE DE SHIFT L
DIM D SHIFT I
APPENDIXES 671
DIRECTORY D lSH IFT R

DLOAD D SHIFT L
DO none
DOPEN D SHIFT O
DRAW D SHIFT R
DS none
DS$ none
DSAVE D SHIFT S
DVERIFY D SHIFT V
EL none
END none
ENVELOPE E SHIFT N
ER none
ERRS E SHIFT R
EXIT EX SHIFT I
EXP E SHIFT X
FAST none
FETCH F SHIFT E
FILTER F SHIFT I
FOR F SHIFT O
FRE F SHIFT R
FNXX none
GET G SHIFT E
GETKEY GETK SHIFT E
GET# none
GOSUB GO SHIFT S
G064 none
GOTO G SHIFT O
GRAPHIC G SHIFT R
GSHAPE G SHIFT S
HEADER HE SHIFT A
HELP X SHIFT X
HEX$ H SHIFT E
IF . . . GOTO none
IF . . . THEN . . . ELSE none
INPUT none
INPUT# I SHIFT N
INSTR IN SHIFT S
INT none
JOY J SHIFT O
KEY K SHIFT E
LEFTS LE SHIFT F
LEN none
LET L SHIFT E
LIST L SHIFT I
LOAD L SHIFT O
LOCATE LO SHIFT C
LOG none
672 COMMODORE 128
LOOP LO SHIFT 0
MID$ M SHIFT I
MONITOR MO SHIFT N
MOVSPR M SHIFT O
NEW none
NEXT N SHIFT E
ON GOSUB ON GO SHIFT S
ON GOTO ON G SHIFT o
OPEN 0 SHIFT p
PAINT P SHIFT A
PEEK PE SHIFT E
PEN P SHIFT E
PI none
PLAY P SHIFT L
POINTER PO SHIFT I
POKE PO SHIFT K
POS none
POT P SHIFT 0
PRINT none
P R IN T # P SHIFT R
PRINT USING US SHIFT I
PUDEF P SHIFT U
RCLR R SHIFT c
RDOT R SHIFT D
READ RE SHIFT A
RECORD R SHIFT E
REM none
RENAME RE SHIFT N
RENUM BER REN SHIFT U
RESTORE RE SHIFT S
RESUME RES SHIFT u
RETURN RE SHIFT T
RGR R SHIFT G
RIGHTS R SHIFT I
RND R SHIFT N
RREG R SHIFT R
RSPCOLOR RSP SHIFT c
RSPPOS R SHIFT S
RSPR none
RSPRITE RSP SHIFT R
RUN R SHIFT U
RWINDOW R SHIFT w
SAVE S SHIFT A
SCALE SC SHIFT A
SCNCLR s SHIFT C
SCRATCH sc SHIFT R
SGN s SHIFT G
APPENDIXES 673
SIN s SHIFT I
SLEEP s SHIFT L
SLOW none
SOUND s SHIFT O
SPC none
SPRCOLOR SPR SHIFT C
SPRDEF SPR SHIFT D
SPRITE s SHIFT P
SPRSAV SPR SHIFT S
SQR S SHIFT Q
SSHAPE s SHIFT s
STASH s SHIFT T
ST none
STEP ST SHIFT E
STOP ST SHIFT O
STR$ ST SHIFT R
SWAP s SHIFT W
SYS none
TAB( T SHIFT A
TAN none
TEMPO T SHIFT E
TI none
TI$ none
TO none
TRAP T SHIFT R
TROFF TRO SHIFT F
TRON TR SHIFT O
UNTIL U SHIFT N
USR U SHIFT S
VAL none
VERIFY V SHIFT E
VOL V SHIFT O
WAIT w SHIFT A
WHILE w SHIFT H
WIDTH WI SHIFT D
WINDOW w SHIFT I
XOR X SHIFT O
674 COMMODORE 128
APPENDIX J
DISK COMMAND SUMMARY

This appendix lists the commands used for disk operation in C l 28 and C64 modes
on the Commodore 128. For detailed information on any of these commands, see
Chapter 2. Your disk drive manual also has information on disk commands.
The new BASIC 7.0 commands can be used only in C l28 mode. All BASIC 2.0
commands can be used in both C l 28 and C64 modes.
com m and use BASIC 2.0 BASIC 7.0
APPEND Append data to file J

BLOAD Load a binary file starting at the specified J
memory location
BOOT Load and execute a bootable program /
BSAVE Save a binary file from the specified mem- /
ory location
CATALOG Display directory contents of disk on screen* J
CLOSE Close logical disk file J
CMD Redirect screen output to a peripheral /
device
COLLECT Free inaccessible disk space* /
CONCAT Concatenate two data files* /
COPY Copy files between devices* /
DCLEAR Clear all open channels on disk drives J
DCLOSE Close logical disk file J
DIRECTORY Display directory of contents o f disk on J
screen*
DLOAD Load a BASIC program from disk /
DOPEN Open a disk file for a read and/or write /
operation
DSAVE Save a BASIC program to disk J
DVERIFY Verify program in memory against pro- /
gram on disk
GET# Receive input from open disk file /
HEADER Format a disk* J
LOAD Load a file from disk J
OPEN Open a file for input or output J
P R IN T # Output a data to file /
RECORD Position relative file pointers* J
RENAME Change name of a file on disk* J
APPENDIXES 675
com m and use BASIC 2.0 BASIC 7.0
RUN filename Execute BASIC program from disk J

SAVE Store program in memory to disk /
VERIFY Verify program in memory against pro- /
gram on disk
*Although there is no single equivalent command in BASIC 2 .0 , there is an equivalent multi-comm and
instruction. See your disk drive m anual for these BASIC 2.0 conventions.
676 COMMODORE 128
A PPENDIX K
part I— COMMODORE 128 CP/M

This appendix explains each CP/M BIOS, 8502 BIOS and User Function routine and
how to call each in Z80 assembly language. This section assumes you already have
some knowledge about Z80 machine language and the basic operations of the CP/M
system. If you need more information about Z80 or CP/M, your local bookstore
probably has several good reference books about these widely covered subjects. To fully
cover these topics is beyond the scope of this reference guide. See “ Suggestions for
Reading” at the end of this guide.
Part I of this appendix first lists each CP/M BIOS, 8502 BIOS and User Function
routine by number. Part II explains how and provides examples to call these routines.
Part III lists the Z80 memory map.
The format used to describe these routines are as follows:
a) Function Name
b) Input Parameters
c) Output Parameters
d) Brief Description
e) Other required preparatory/post routines (or additional information)
The 8502 BIOS and User Function routines require certain values to be placed into
the Z80 microprocessor registers. In Chapter 5 you learned about the 8502 microproces
sor registers: A, X, Y, Status (PSW),Stack Pointer (S) and Program Counter (PC). The
Z80 also has applicable registers.The Z80 registers are named as follows:
A (Accumulator)
BC
DE
HL
PSW (Status Word)
IX (X register)
IY (Y index register)
PC (program counter)’
SP (stack pointer)
Certain registers can be used as a register pair, to represent a 16 bit address, or as

single 8-bit register. The Z80 has a duplicate set of registers for interrupt processing.
For more detailed information on the Z80 microprocessor consult the “ Sugges
tions for Further Reading.” For more detailed information on the CP/M system, refer to
the Commodore 128 System Guide to receive the full set of CP/M Plus documentation,
written by Digital Research Inc., through Commodore.
APPENDIXES 677
COMMODORE 128 CP/M BIOS ROUTINES

The Commodore 128 CP/M system has a set of routines called the CP/M BIOS, which
handle the low level input/output operations of the system. Each of these routines can
be accessed via the CP/M BIOS jump table below. The jum p vector numbers 0 through
29 are the CP/M BIOS jump vectors. The 30th jump vector is for system dependent
User Functions. These are discussed following the CP/M BIOS routines.
The 8502 BIOS routines are a subset of the system dependent User Functions
(specifically, User Function 4) contained in the Commodore 128 CP/M system. Many of
the User Functions have subfunctions which require certain parameters to be passed
through the Z80 registers. This is discussed in detail in the User Function section in this
appendix. Examples of calling the CP/M BIOS and User Function routines are provided
in Part II of this appendix.
NO. INSTRUCTION DESCRIPTION
0 JM P BOOT Perform cold start initialization

1 JM P WBOOT Perform warm start initialization
2 JM P CONST Check for console input character ready
3 JM P CONIN Read Console Character in
4 JM P CONOUT Write Console Character out
5 JM P LIST Write List Character out
6 JM P AUXOUT Write Auxiliary Output Character
7 JM P AUXIN Read Auxiliary Input Character
8 JM P HOME Move to Track 00 on Selected Disk
9 JM P SELDSK Select Disk Drive
10 JM P SETTRK Set Track Number
11 JM P SETSEC Set Sector Number
12 JMP SETDMA Set DMA Address
13 JMP READ Read Specified Sector
14 JMP WRITE Write Specified Sector
15 JMP LISTST Return List Status
16 JMP SECTRN Translate Logical to Physical Sector
17 JMP CONOST Return Output Status of Console
18 JMP AUXIST Return Input Status of Aux. Port
19 JM P AUXOST Return Output Status of Aux. Port
20 JM P DEVTBL Return Address o f Char. I/O Table
21 JM P DEVINI Initialize Char. I/O Devices
22 JM P DRVTBL Return Address o f Disk Drive Table
23 JMP MULTIO Set Number of Logically Consecutive sectors to be read or written
24 JMP FLUSH Force Physical Buffer Flushing for user-supported deblocking
25 JM P MOVE Memory to M emory Move
26 JM P TIM E Time Set/Get signal
27 JM P SELMEN Select Bank of Memory
28 JM P SETBNK Specify Bank for DMA Operation
29 JMP XMOVE Set Bank When a Buffer is in a Bank other than 0 or 1
30 JM P USERF Reserved for System Implementor
31 JM P RESERV1 Reserved for Future Use
32 JM P RESERV2 Reserved for Future Use
CP/M 3 BIOS Jump Vector Table
678 COMMODORE 128
0 BOOT
Bank: 0
Input: None
Output: None
Function: This code does all of the hardware initialization, sets up zero
page, prints any sign-on message and loads the CCP and then
transfers control to the CCP.
WBOOT
Bank: 0 or 1
Input: None
Output: None
Function: This code sets up page zero, reloads the CCP and then exe
cutes the CCP.
CONST
Bank: 0 or 1
Input: None
Output: A = 0FFH if console character
A = 00H if no console character
Function: Checks the console input status of the current console devices.
If any of the devices have a character available, FFH is
returned, otherwise 00H is returned.
CONIN
Bank: 0 or 1
Input: None
Output: A = ASCII console character
Function: Reads a character from any ONE of the assigned console input
devices. A scan of each assigned device is done until an
input character is found. The character is returned in the A
register.
CONOUT
Bank: 0 or 1
Input: C = ASCII character to display
Output: None
Function: Sends the character in C to ALL devices that are currently
assigned to the console. It waits for all assigned devices to
accept a character before exiting.
LIST
Bank: 0 or 1
Input: C = ASCII character to print
Output: None
assigned to the LIST device. It waits for all assigned devices
to accept a character before exiting.
APPENDIXES 679
AUXOUT
Bank: 0 or 1
Input: C = ASCII Character to send to AUX device
Output: None
assigned to the AUXOUT device. It waits for all assigned devices
to accept a character before exiting.
AUXIN
Bank: 0 or 1
Input: None
Output: A = ASCII character from AUX device
Function: Reads a character from any ONE of the assigned AUXIN devices.
A scan of each assigned device is done until an input character
is found. The character is returned in the A register.
HOME
Bank: 0
Input: None
Output: None
Function: Homes the head on the currently selected disk drive. This
function sets the current track to 0 and does not move the head
of the disk.
SELDSK
Bank: 0
Input: C = Disk Drive (0-15) (A = 0)
E = Initial Select Flag (LSB)
Output: HL = Address of Disk Parameter Header (DPH) if drive exists.
HL = 000H if drive does not exist.
Function: Selects the disk drive whose address is in C as the current drive
for all further disk operations. If the LSB of the E register is a
zero, then this is the first logging of this disk. The disk type
(C64 CP/M, MFM or C128 CP/M) is checked and the DPB
parameters adjusted for the diskette currently in the drive.
10 SETTRK
Bank: 0
Input: BC = Track number
Output: None
Function: Register pair BC contains the track number to be used in the
subsequent disk access. This value is saved.
11 SETSEC
Bank: 0
Input: BC = Sector number
Output: None
Function: Register pair BC contains the sector number to be used in the
subsequent disk access. This value is saved. The value in BC
is the value returned by the sector translation routine (in HL).
680 COMMODORE 128
12 SETDMA
Bank: 0
Input: BC = Direct memory access address
Output: None
Function: The value in BC is saved as the current DMA address. This is
the address where ALL disk reads or writes occur. The DMA
address that is set is used until it is changed by a future call to
this routine to change it.
13 READ
Bank: 0
Input: None
Output: A = 000H if no errors
A = 001H if nonrecoverable error
A = OFFH if media has changed
Function: Reads the sector addressed by the current disk, track and
sector to the current DMA address. If the data is read with no
errors then A = 0 on return. If an error occurs, the operation is
tried several more times, and if a successful read does not
occur then A is set to 0 0 1H. A test for media change is
performed each time this routine is called and A is set to -1 if
the media has been changed.
14 WRITE
Bank: 0
Input: C = Deblocking code (not used)
A = 002H if disk is read only
Function: Writes the sector addressed by the current disk track and
sector from the current DMA address. If thedata is written with
no errors, then A is set to 0 on return. If an error occurs, the
operation is tried several more times, and if a successful write
does not occur, then A is set to 001H. A test for media change
is performed each time this routine is called and A is set to
-1 if the media has been changed. Also, if an attempt is made
to write to a read-only disk, then the A register is set to 002H.
15 LISTST
Bank: 0 or 1
Input: None
Output: A = 00H if list device is not ready to accept a character.
A = OFFH if list device is ready to accept a character.
Function: This routine scans the currently assigned list devices and returns
with A set to OFFH if ALL assigned devices are ready to accept a
character. If any assigned device is not ready then A is set
to 00H.
APPENDIXES 681
16 SECTRN
Bank: 0
Input: BC = Logical sector number (0-n)
DE = Translation table address (from DPB)
Output: HL = Physical sector number
Function: This routine converts the physical sector number to a logical
sector number. If no translation is needed then it moves the
BC register to HL and returns.
17 CONOST
Bank: 0 or 1
Input: None
Output: A = OFFH if Ready
A = 000H if not Ready
Function: This routine scans the currently assigned console devices and
returns with A set to OFFH if ALL assigned devices are ready
to accept a character. If any assigned device is not ready then
A is set to 000H.
18 AUXIST
Bank: 0 or 1
Input: None
Output: A = OFFH if console character present
A = 000H if no console character
Function: Checks the status of the current AUXIN device. If any of the
devices have a character available, OFFH is returned, otherwise
OOOH is returned.
19 AUXOST
Bank: 0 or 1
Input: None
Output: A = OFFH if ready
A = OOOH if not ready
Function: This routine scans the currently assigned AUXOUT devices and
returns with A set to OFFH if ALL devices are ready to accept a
character. If any assigned device is not ready then A is set
to 00H.
20 DEVTBL
Bank: 0 or 1
Input: None
Output: HL = address of character I/O table
Function: This routine returns the address of the Character I/O table.
This table is used to name each of the driver modules and
set/control the baud rate and XON/XOFF logic for each driver.
Note: the device drive mechanism is used to replace the IOBYTE
used with CP/M 2.2.
682 COMMODORE 128
21 DEVINI
Bank: 0 or 1
Input: C = device number
Output: None
Function: Initializes the physical character device specified in the C
register to the BAUD rate in the DEVTBL.
22 DRVTBL
Bank: 0
Input: None
Output: HL = address of the drive table
Function: Returns the address of the drive table in HL (NOTE: first
instruction MUST be LXI H, DRVTBL). The drive table is a
list of 16 word pointers that point to the DPH for that drive. If
a drive is not present in the system, then the pointer for that
drive is set to zero.
23 MULTIO
Bank: 0
Input: C = multisector count
Output: None
Function: The multisector count is set before the track, sector, and DMA
address and the read/write of the sectors occur. A maximum
of 16K can be transferred by each multisector count.
24 FLUSH
Bank: 0
Input: None
A = 002H if disk is read-only
Function: This routine is used only if blocking/deblocking is done in the
BIOS. This code ALWAYS returns with A = 000H .
25 MOVE
Bank: 0 or 1
Input: HL = destination address
DE = source address
BC = count
Output: HL = HL(in) + BC(in)
DE = DE(in) + BC(in)
Function: Moves a block of data. Data to be moved is to/from the current
memory bank (or common) unless the XMOVE routine is
called first, then the move is an interbank data movement.
26 TIME
Bank: 0 or 1
Input: C = 000H (Time Get) / OFFH (Set Time)
Output: None
APPENDIXES 683
Function: This function is called with C = 000H if the system time in the
SCB needs to be updated by the clock. If C = 0FFH, then the
time in the SCB has just been updated and the clock is
set to the SCB time. NOTE: HL and DE MUST be preserved.
27 SELMEM
Bank: 0 or 1
Input: A = memory bank
Output: None
Function: Changes the current memory bank. This code MUST be in
common memory.
N O TE: ONLY A can be changed.
28 SETBNK
Bank: 0
Input: A = DMA memory bank
Output: None
Function: Sets the DMA bank for the next READ/WRITE operation.
29 XMOVE
Bank: 0
Input: B = destination bank
C = source bank
Output: None
Function: Provides the system with the ability to perform memory-to-
memory DMA through the entire system space.
30 USERF
Banks 0 or 1
Input: A = function number, L = subfunction number
Output: Outputs depend on the called function or subfunction.
Function: This calls the user functions and 8502 BIOS routines, which
are defined in the Commodore 128 System Dependent User
Function section.
31 RESERV1
Bank: N/A
Input: N/A
Output: N/A
Function: Not available to the user.
32 RESERV2
Bank: N/A
Input: N/A
Output: N/A
Function: Not available to the user.
684 COMMODORE 128
DATA STRUCTURES
SYSTEM CONTROL BLOCK—(SCB)

The System Control Block is a 100-byte data structure. The data structure is
used as the basic communication between the various modules that make up the
CP/M Plus system. The contents of the data structure are system parameters and
variables.
DRIVE TABLE (DRVTBL)
PHO through D P H I5
A list of 16 word pointers (reverse-byte format). The first pointer (DPHO) is
drive A and the last pointer (DPH15) is drive P. The pointers point to the XDPH
for that disk drive. Any drive that is not in the system has its pointers set to
zero.
E XTEN D ED DISK PARAMETER HEADER (X D P H )

(NO R M A L DPH W IT H A HEADER)
W RT READ LOGIN INIT TYPE UNIT XLT -0- M F DPB CSV ALV DIRBCB DTABCB HASH HBANK
16b 16b 16b 16b 8b 8b 16b 72b 8b 16b 16b 16b 16b 16b 16b 8b
-1 0 - 8 -6 -4 -2 -1 0 2 11 12 14 16 18 20 22 24
WRT Contains the address of the sector write routine for this drive.
READ Contains the address of the sector read routine for this drive.
LOGIN Contains the address of the login routine for this drive.
INIT Contains the address of the first time initialization routine for this drive.
TYPE This byte is used by the BIOS to keep track of density and media type.
UNIT Contains the drive number relative to the disk controller.
XLT Contains the address of the sector translation table or zero if none.
-0- BDOS scratch area (9 bytes).
MF Media flag cleared to zero if disk logged in. BIOS sets to OFFH if media
has changed.
DPB Contains a pointer to the current DPB that describes the current media type.
CSV Contains a pointer to directory checksum area (one per disk drive).
ALV Contains a pointer to allocation vector area (one per disk drive).
DIRBCB Contains a pointer to a single directory Buffer Control Block (BCB).
DTSBCB Contains a pointer to a single data Buffer Control Block (BCB).
HASH Contains a pointer to an optional directory hashing table (FFFFH is
not used).
HBANK Contains a bank number of the directory hashing table.
APPENDIXES 685
DISK PARAMETER BLOCK— DPB
SPT BSH BLM EXM DSM DRM ALO AL1 CKS O FF PSH PHM
16b 8b 8b 8b 16b 16b 8b 8b 16b 16b 8b 8b
SPT Number of 128 records per track

BSH Data allocation block shift factor
BLM Block mask
EXM Extent mask
DSM Number of allocation block on disk minus one
DRM Number of directory entries minus one
ALO First byte: directory block allocation vector. Filled from MSB to LSB
AL1 Second byte: (Up to 16 allocation blocks can be used for the directory)
CKS Size of the directory check vector, (D R M + l)/4
OFF Number of reserved tracks at the beginning of the disk
PSH Physical record shift factor
PHM Physical record mask
BUFFER C O N TR O L BLOCK (LRU C O N TR O L BLOCK— BCB)
DRV REC# W FLG 00 TRACK SECTO R BUFFAD BANK LINK
8b 24b 8b 8b 16b 16b 16b 8b 16b
DRV Drive associated with this record. Set to OFFH when not used.
REC# Contains the absolute sector number of the buffer.
WFLG Set to OFFH when buffer contains data that must be written to disk.
00 Scratch byte used by BDOS
TRACK Physical track address of buffer.
SECTOR Physical sector address of buffer.
BUFFAD Address of the buffer associated with this BCB.
BANK Bank number of buffer associated with this BCB.
LINK Contains the address of the next BCB in this linked list. Set to zero if last.
COMMODORE 128 SYSTEM DEPENDENT

USER FUNCTIONS
The Z80 machine language routines discussed in this section pertain explicitly to the
Commodore 128 Personal Computer. These routines are not part of the standard CP/M
system that is transportable from one manufacturer’s computer to another. They are
written to run only on the Commodore 128.
Most of these customized C l 28 Z80 routines require certain parameters to be
passed from the user into the appropriate Z80 registers. All these routines, listed by
686 COMMODORE 128
function number, require that the function number be placed in the Z80 A register
(accumulator). Many of these routines have subfunctions which require the user to place
a value in the eight bit L register, which is used to call the appropriate subfunction. All
other additional required parameters are noted if they are necessary. This section follows
the same format as the preceding CP/M BIOS section:
a) Function Name
b) Input Parameters
c) Output Parameters
d) B rief Description
e) Additional Information
0 a) User Function 0: Read any Byte in RAM Bank 00
b) A = 0 (function number)
DE = 16 bit address to be read
c) C = Value read from address in RAM Bank 0
A = 0 on RETurn from routine
d) This reads a value from RAM BANK 0 ($0-$0FFF, $1000-$FFFF RAM)
and places it in register C.
I. a) User Function 1: Write Function in RAM Bank 0
DE = 16-bit address where write operation occurs
C = Value to be written to the address in Bank 0
c) A = 0 on RETurn from routine if write issuccessful
A = -1 ($FF) on RETurn from routine if write is unsuccessful
d) This function writes the value in register C to RAM in bank 0 specified
by the address in register pair DE. An error is flagged if a write to ROM
($0-$DFFFH) or the top page of memory ($FF0Q-$FFFF) is attempted.
II. a) User Function 2: Keyscan Function
c) B = -1 ($FF) if no key is pressed
B = Matrix position in the keyboard matrix table (if pressed)
A = Contents (character value) of matrix position whether A is lower- or
uppercase or containing the C O N T R O L key.
HL = Address (pointer) where A (contents of matrix position) is found
in memory
Address = Tablestart + (4 * B) 4- C (bits 1 and 0)
C = Returns control code information, where each bit has a specific
meaning as follows:
Bits 1 & 0
0 0 = lower case
0 1 = upper case
1 0 = shifted
1 1 = control key
APPENDIXES 687
C = Bit 2 1 = control key down, 0 = up

Bit 3 Not implemented
Bit 4 1 = Right shift key down, 0 = up
Bit 5 1 = C* key is active (not necessarily down)
Bit 6 Not implemented
Bit 7 1 = left shift key down, 0 = up
d) The keyscan function allows the user to bypass the normal I/O BIOS
keyboard processing and check for a particular key orkey sequence
being pressed.
e) Additional Information— Important Addresses:
$FD09 = Pointer to Tablestart (low byte first)
Each ASCII character has four coded definitions. Each key has a defined code for
the following:
a) lower case
b) upper case
c) shifted key and character
d) control key and character
These four definitions are labeled in columns in the ASCII table.
ASCIISTBL: A B C D
1292 7F7F7F16 DB 7FH ,7FH ,7FH ,16H INS DEL matrix 0 position
1296 0D0D0D0D DB 0DH ,0D H ,0D H ,0D H RETURN
129A 06060101 DB 06H ,06H ,01H ,01H LF RT
129E 86868787 DB 86H ,86H ,87H ,87H F7 F8
12A2 80808181 DB 80H ,80H ,81H ,81H FI F2
12A6 82828383 DB 82H ,82H ,83H ,83H F3 F4
12AA 84848585 DB 84H ,84H ,85H ,85H F5 F6
12AE 1717171A DB 17H,17H ,17H,1AH UP DOW N
12B2 333323A2 DB 33H,33H ,23H,0A2H 3 #
12B6 77575717 DB 77H ,57H ,57H ,17H W
12BA 61414101 DB 61H ,41H ,41H ,01H A
12BE 343424A3 DB 34H ,34H ,24H ,0A 3H 4$
12C2 7A5A5A1A DB 7AH,5AH,5AH,1AH Z
12C6 73535313 DB 73H ,53H ,53H ,13H s
12CA 65454505 DB 65H ,45H ,45H ,05H E
12CE 00000000 DB 00H ,00H ,00H ,00H (LF SHIFT)
12D2 353525A4 DB 35H,35H ,25H,0A4H 5 %
12D6 72525212 DB 72H ,52H ,52H ,12H R
12DA 64444404 DB 64H ,44H ,44H ,04H D
12DE 363626A5 DB 36H,36H ,26H,0A5H 6 &
12E2 63434303 DB 63H ,43H ,43H ,03H C
12E6 66464606 DB 66H ,46H ,46H ,06H F
688 COMMODORE 128
ASCIISTBL: A B C D
12EA 74545414 DB 74H ,5 4 H ,5 4 H ,1 4 H ;T

12EE 78585818 DB 78H ,5 8 H ,5 8 H ,1 8 H ; x
12F2 373727A 6 DB 37H ,3 7 H ,2 7 H ,0 A 6 H ;7 ’

12F6 79595919 DB 79H ,5 9 H ,5 9 H ,1 9 H ;y
12FA 67474707 DB 67H ,4 7 H ,4 7 H ,0 7 H ;G
12FE 383828A 7 DB 38H ,3 8 H ,2 8 H ,0 A 7 H ;8(
1302 62424202 DB 62H ,4 2 H ,4 2 H ,0 2 H ;b
1306 68484808 DB 68H ,4 8 H ,4 8 H ,0 8 H ;h
130A 75555515 DB 75H ,5 5 H ,5 5 H ,1 5 H ;u
130E 76565616 DB 76H ,5 6 H ,5 6 H ,1 6 H ;v
1312 39392900 DB 39H ,3 9 H ,2 9 H ,0 0 H ; 9)
1316 69494909 DB 69H ,4 9 H ,4 9 H ,0 9 H ;I
131A 6A 4A 4A 0A DB 6 A H ,4A H ,4A H ,0A H ;J
131E 30303000 DB 3 0 H ,3 0 H ,3 0 H ,0 0 H ; o
1322 6D 4D 4D 0D DB 6 D H ,4 D H ,4 D H ,0 D H ;M
1326 6B 4B 4B 0B DB 6B H ,4 B H ,4 B H ,0 B H ;K
132A 6F 4F 4F 0F DB 6F H ,4 F H ,4 F H ,0 F H ; o
132E 6E 4E 4E 0E DB 6E H ,4 E H ,4 E H ,0 E H ;N
1332 2B 2B 2B 00 DB 2B H ,2 B H ,2 B H ,0 0 H ; +
1336 70505010 DB 70 H ,5 0 H ,5 0 H ,1 0 H ;p
133A 6C 4C 4C 0C DB 6C H ,4 C H ,4 C H ,0 C H ;l
133E 2D 2D 2D 00 DB 2 D H ,2 D H ,2 D H ,0 0 H
1342 2E 2E 3E 00 DB 2E H ,2 E H ,3 E H ,0 0 H ;, <
1346 3A 3A 5B 7B DB 3A H ,3 A H ,5 B H ,7 B H ;:[{
134A 40404000 DB 40 H ,4 0 H ,4 0 H ,0 0 H ;@
134E 2C 2C 3C 00 DB 2C H ,2 C H ,3 C H ,0 0 H ;,<
1352 23232360 DB 23 H ,2 3 H ,2 3 H ,6 0 H ; POUND
1356 2A 2A 2A 00 DB 2 A H ,2 A H ,2 A H ,0 0 H . *
135A 3B 3B 5070 DB 3B H ,3 B H ,5 D H ,7 D H ;;]}
135E 000000F 5 DB 00 H ,0 0 H ,0 0 H ,0 F 5 H ; C L E A R /H O M E
1362 00000000 DB 00 H ,0 0 H ,0 0 H ,0 0 H ; (R T SH IF T )
1366 3D 3D 3D 7E DB 3D H ,3 D H ,3 D H ,7 E H
136A 5E 5E 7C 7C DB 5E H ,5 E H ,7 C H ,7 C H ; PI
136E 2F 2F3F 5C DB 2 F H ,2 F H ,3 F H ,5 C H ;/?\
1372 313121A 0 DB 31H ,3 1 H ,2 1 H ,0 A 0 H ;1
1376 5F 5F 5F 7F DB 5F H ,5 F H ,5 F H ,7 F H ; <-
137A 09153000 DB 09 H ,1 5 H ,3 0 H ,0 0 H ; (C O N T R O L ) SO U N D 1
SO U N D 2
137E 3 2 3222A 1 DB 32H ,3 2 H ,2 2 H ,0 A 1 H ; 2"
1382 20202000 DB 20H ,2 0 H ,2 0 H ,0 0 H ; SP A C E
1386 21200000 DB 2 1 H ,2 0 H ,0 0 H ,0 0 H ; (C O M M O D O R E ) SO U N D 3
138A 71515111 DB 71 H ,5 1 H ,5 1 H ,1 1 H ;Q
138E 000000F 0 DB 00 H ,0 0 H ,0 0 H ,0 F 0 H ; R U N ST O P
APPENDIXES 689
ASCIISTBL: A B C I)
1392 9F9F9F9F DB 9FH ,9FH ,9FH ,9FH /HELP/

1396 383838B7 DB 38H ,38H ,38H ,0B7H /8/
139A 353535B4 DB 35H ,35H ,35H ,0B4H 151
139E 09090900 DB 09H ,09H ,09H ,00H /TAB/
13A2 323232B1 DB 32H,32H ,32H ,0B1H 121
13A6 343434B3 DB 34H,34H ,34H ,0B3H 141
13AA 373737B6 DB 37H ,37H ,37H ,0B6H 111
13AE 313131B0 DB 31H ,31H ,31H ,0B0H III
13B2 1B1B1B00 DB 1BH ,1BH,1BH ,00H /ESC/

13B6 2B2B2B00 DB 2BH,2BH,2BH ,00H /+/
13BA 2D2D2D00 DB 2D H ,2D H ,2D H ,00H /-/
13BE 0A0A0A0A DB 0AH,0A H ,0A H ,0A H /LINE FEED/
13C2 0D0D0DFF DB 0D H,0DH,0DH,0FFH /ENTR/
13C6 363636B5 DB 36H ,36H ,36H ,0B5H 161
OCA 39393900 DB 39H ,39H ,39H ,00H 191
BCE 333333B2 DB 33H ,33H ,33H ,0B2H 131
13D2 00000000 DB 00H,00H ,00H ,00H /ALT/
13D6 30303000 DB 30H ,30H ,30H ,00H 101
13DA 2E2E2E00 DB 2EH ,2EH ,2EH ,00H l.l
13DE 05050512 DB 05H,05H ,05H ,12H /UP/
13E2 18181803 DB 18H,18H ,18H,03H /DN/
13E6 1313138D DB 13H,13H ,13H,08DH /LF/
13EA 0404048E DB 04H ,04H ,04H ,08EH IR T I
13EE F1F1F1F2 DB 0F1H,0F1H,0F1H ,0F2H /NO SCROLL/
5
» LOGICAL COLOR TABLE (USED WITH ESC ESC ES
J (W HERE CHAR IS 50H
>
13F2 00112233 DB 000H,011H ,022H,033H
13F6 44556677 DB 044H,055H ,066H,077H
13FA 8899AABB DB 088H,099H ,0AAH,0BBH
13FE CCDDEEFF DB 0CCH ,0DDH,0EEH,0FFH
III. a) User Function 3: Execute a Z80 ROM function

b) A = 3
L = subfunction number
Additional input parameters may be necessary
c) Most of the ROM functions do not output values; they instead perform
an action, which in this user function is assumed to be the output.
d) This function executes a Z80 ROM function. These ROM routines
primarily perform screen manipulation routines.
All subfunction numbers are even numbers. The first 80 subfunctions are
screen manipulating routines for the 40- and 80-column display. Only the
80-column subfunction is listed here. Add 2 to the 80-column subfunction
number to get the corresponding 40-column subfunction number.
690 COMMODORE 128
III. 0a) Subfunction 0: Write Character

b) wr$char
c) Register D = character to write auto advance cursor to next position
d) —
III. 4a) Subfunction 4: Cursor Position
b) cursor$pos
c) Register D = row value
Register E = column value
d) This subfunction sets the current position of the cursor on the 80-column
screen.
III. 8a) Subfunction 8: Cursor Up one position
b) cursor$up
c) no values returned
d) This subfunction moves the cursor up one row on the current screen (40
or 80), but not past the top of the screen.
III. 12a) Subfunction 12: Cursor down one row
b) cursor$down
d) This subfunction moves the cursor down one row on the current screen
(40 or 80). The screen scrolls down if on bottom line.
III. 14a) Subfunction 14: Not Implemented
b)
c) ------
III. 16a) Subfunction 16: Cursor left one column
b) cursor$left
d) This subfunction moves the cursor left one column, but not past the
left margin.
III. 20a) Subfunction 20: Cursor right one column
b) cursor$rt
c) no values returned.
d) This subfunction moves the cursor right one column, but not past the
right edge of the screen.
III. 24a) Subfunction 24: Execute a carriage return
b) do$cr
c) No values returned.
d) This subfunction executes a carriage return and places the cursor at the
left margin.
III. 28a) Subfunction 28: Clear to end of line
b) CEL
d) This subfunction clears the cursor row starting where the cursor is
currently located and ending at the end of the line.
APPENDIXES 691
III. 32a) Subfunction 32: Clear to end of screen

b) CES
d) This subfunction clears the screen starting where the cursor is currently
located through the end of screen.
III. 36a) Subfunction 36: Character insert
b) char$ins
d) This subfunction inserts a character at the current cursor position, the
last character on the line is lost.
III. 40a) Subfunction 40: character delete
b) char$del
d) This subfunction deletes a character at the current cursor position, and
places a space in the last position in the line.
III. 44a) Subfunction 44: Line insert
b) line$ins
d) This subfunction inserts a line of spaces on the current cursor row, the
current cursor row is moved down one and the last line (24) is lost.
III. 48a) Subfunction 48: Line delete
b) line$del
d) This subfunction deletes a line at the character row marked by the
current cursor position. The bottom line (24) is filled with spaces.
III. 50a) Subfunction 50: Not Implemented
b) ------
c) ------
d) ------
III. 52a) Subfunction 52: Set Cursor Color
b) color
B register returns color code value
d) This function sets the current cursor color code with the value in color.
III. 56a) Subfunction 56: Set 80-column Attributes
b) attr
B = bit to set/clear
C = bit value
d) This function enables/disables the 8563 attributes for the 80-column screen.
Attributes are extra screen features such as foreground R,G,B and I,
character blinking, underlining, reverse video and the alternate character
set. The 40-column function (subfunction 58) controls only reverse video.
692 COMMODORE 128
III. 60a) Subfunction 60: Read Character Attribute (80)

b) rd$chr$atr
D = 8563 character row
E = 8563 character column
c) D = 8563 character row
L = 8563 character column
B = character value
C = attribute bit pattern of selected character location
d) This function reads the attribute byte of the selected character row and
column on the 8563 screen, and returns the true RGBI color. The
corresponding 40-column subfunction (62) controls only reverse video.
III. 64a) Subfunction 64: Write Character Attribute (80)
b) wr$chr$atr
B = character value
C = attribute
c) No values returned
d) This function writes an attribute byte of the selected character value
on the 8563 screen.
III. 68a) Subfunction 68: Read Color
b) rd$color
c) A = character color
B = background color
C = border color (40-column only)
D = 8563 attribute
d) This function reads the current color of the 8563 color sources. Subfunction
70 returns the VIC color sources.
III. 80a) Subfunction 80: Convert record (GCR only)
b) @trk
c) VICStrk, VICSSEC
d) This function returns the physical track and sector from the disk drive,
including skew, given the logical track.
III. 82a) Subfunction 82: Check CBM code from disk (GCR only)
b) check$CBM
c) zero flag = 1 if CBM (C l28 mode disk) is present in drive
zero flag = 0 if CP/M 2.2 (C64 CP/M disk) is present in drive
A = 0 if single sided
A = $FF if double sided
d) This function detects which type of GCR disk is in the drive. Reads
data from t = 1, s = 0 into buffer at $FE00.
III. 84a) Subfunction 84: Bell Function
b) Sound 1, Sound 2, Sound 3 ($FD10, $FD12, $FD14, respectively)
c) Outputs a bell sound
d) This function outputs a bell sound from the SID chip.
Subfunctions 86-95 are not defined
APPENDIXES 693
III. 96a) Subfunction 96: Track 40

b) trk$40
c) @ off 40
d) This function computes the logical (offset) position of the cursor on the
physical 80-column screen. The variable @ off40 is computed on
8-character boundaries, and is used in the next subfunction.
III. 98a) Subfunction 98: Set cursor position
b) set$cur$40
c) No value returned.
d) This function calls Track 40 to check the offset between the logical
(window) and physical (40-column) screen. It keeps the cursor within
the 40-column logical window on the scrolled 80-column virtual screen
for VIC screen output.
III. 100a) Subfunction 100: Line Paint
b) line$paint
d) This function updates the current character line only if @ off 40 and
old $offset are the same. If old $offset = - 1 , TRACK 40 is called.
If @off40 and old $offset are not equal, Screen Paint is called to
scroll the 40-column logical screen (window) over the virtual 80-
column screen.
III. 102a) Subfunction 102: Screen paint
b) screen$paint
d) This function updates (scrolls) the 40-column logical (window) screen
across the VIC 80-column virtual screen based on the value of @off40.
This and the three previous subfunctions (96, 98, 100) are intertwined
to make the scrolled 40-column logical screen display as fast as
possible. Normally you would only call Set Cursor Position or Line
Paint, since they call the other related routines, if necessary. See file
CXROM.ASM listing on the DRI disks that come with the documenta
tion for the source listings.
III. 104a) Subfunction 104: Print Message (BOTH)
b) prt$msg$both
d) This function prints simultaneous output to both screens, displaying
the string pointed to by the top value on the stack. Place the address of
the string on the stack, and terminate the string with a zero. Execution
resumes with the byte following the zero terminator. This works like
an “ in-line Print”
III. 106a) Subfunction 106: (Print Message (BOTH)
b) prt$de$both
694 COMMODORE 128
d) This function works like the previous one, except the start address of
the string is taken from DE, and execution resumes with the return
address from the stack.
III. 108a) Subfunction 108: Incripted Messages
b) update$it
c) No value returned
d) This function displays incripted messages.
Subfunction 110 is not implemented.
III. 112a) Subfunction 112: ASCII to PET ASCII Conversion
b) ASCII$pet, B register = ASCII character ($20—$7F)
c) A register = converted PET ASCII character.
d) This function performs a standard ASCII to Pet ASCII conversion on
the characters printed to the screen (from any input device)
Control codes are not translated.
III. 114a) Subfunction 114: Place 40-column cursor at specified address
b) cur$adr$40
c) HL = address of cursor on screen
DE = cursor line (row) start address
BC = # of characters to end of line (< 8 0 , not counting cursor)
d) This function places the cursor at the address specified in HL (in RAM
bank 0). This address is of the logical screen, not the virtual one.
III. 116a) Subfunction 116: Place 80-column cursor at specified address
b) cur$adr$80
c) HL = address of cursor on screen
DE = cursor line (row) start address
BC = # of characters to end of line (< 8 0 , not counting cursor)
d) This function places the cursor at the address specified in HL (in
8563 RAM).
III. 118a) Subfunction 118: Look up color
b) look$color, B = color code ($30-$3F), C = max. Value of color code.
c) HL = pointer to logical color table (lower nybble = 80, high nybble = 40-
B = $0 (Character color)
$10 (Background color)
$20 (Border color)
d) This function sets the 8563 screen colors to the VIC screen colors.
Subfunction 120 is not defined.
III. 122a) Subfunction 122: Block Fill
b) blk$fill, put start address on stack (8563)
BC = # of bytes to fill
D = fill character
E = attribute
d) This function fills a 256-byte block with data specified in the D register.
APPENDIXES 695
III. 124a) Subfunction 124: Block move

b) blk$move, place source address (in 8563 RAM) on the stack
DE = destination address (in 8563 RAM)
BC = count
d) This function moves 256-byte blocks from one memory block to another
in 8563 RAM.
N O TE: Subfunctions 122 and 124 and 126 must be direct Z80 calls. The
example routines provided in this section for subfunctions 122, 124 and 126
will not work. They are usually not user-accessible, and must run in BANK 0.
III. 126a) Subfunction 126: Character Install

b) chr$inst, stack = 8563 RAM address to install character definition
DE = address of system memory (C l28 RAM) bank 0) of new
character definition (8 bytes per character).
B = number of consecutive characters
c) No value
d) This function installs a user-defined character in 8563 RAM.
IV. a) User Function 4: 8502 BIOS Functions
L = Subfunction number (-1 through 11)
Additional input parameters may be required
c) Outputs depend on each subfunction
d) User function 4 allows you to call 8502 input/output functions that are
performed by the 8502 processor. These functions are not part of the
standard CP/M system and are completely hardware-dependent. This
function enables you to go back and forth between the Z80 and 8502
processors.
e) Each subfunction is discussed in detail using the defined conventions.
I V .- 1 . a) Subfunction -1 : Reboot C128 hardware
b) A = 4
L = -1
c) none
d) This subfunction reboots the Commodore 128 hardware when the value
in the L register is equal to -1 ($FF). This sub function is not normally
used. It performs the same actions as pressing reset.
IV. O.a) Subfunction 0: Initialize 8502 BIOS
b) A = 4
L = 0
c l) Sets up interrupt vector $0314—$0319 to vector to the 8502 interrupt
handler
c2) Copies ROM interrupt vector to RAM
696 COMMODORE 128
c3) Sets up PAL and NTSC variables (SYSFREQ)

c4) Closes all open channels
d) This subfunction intializes 8502 system variables, interrupt processing
and system frequencies so that the Z80 and 8502 may communicate
back and forth. Processor control is given to either the Z80 or the 8502
at one particular time. The processors cannot run simultaneously.
IV. l.a ) Subfunction 1: 1541 Read (GCR format only)

b) A = 4
L = 1
VICTRACK = (1—35)— Variable for track number on disk
VICSECTOR = (0-21)— Variable for sector number on disk
VICDRV = — Variable for disk drive device number where:
LO W ER NYBBLE EQUIVALENT OPEN

O F VICTORY
BIT VALUE DEVICE NUMBER DRIVE STATEM ENT IN BASIC
0001 8 0 OPEN 8,11,15

0010 9 0 OPEN 9,12,16
0100 10 0 OPEN 10,13,17
1000 11 0 OPEN 11,14,18
Values o f the Lower Nybble of VICDRV
c) VICDATA = 11 ($0B) if disk in drive has been changed

VICDATA = 13 ($0D) if read/write or channel error occurs
VICDATA = 0 if read is successful
VICDATA = 15 ($0F) if device is not present
d) This function reads a particular track and sector on the disk in the drive
as specified by VICTRACK and VICSECTOR and VICDRV respec
tively. Data is read into the buffer at $FE00H. The value returned in the
variable VICDATA depends on the conditions described in c.
e) Additional Information: This subfunction assumes that both the data and
command channels have been opened previously. An error will occur if
this routine is called to read from a 1571.
IV. 2 .a) Subfunction 2: 1541 Write (GCR format only)
b) A = 4
L = 2
VICTRACK = Same as for 1541 Read
VICSECTOR = Same as 1541 Read
VICDRV = Same as 1541 Read
c) VICDATA = Same as 1541 Read
d) This subfunction writes data to the specified track and sector on the disk
in the drive as specified by VICTRACK, VICSECTOR and VICDRV
respectively. The value returned in VICDATA depends on the condi
tions described for the outputs in c. See subfunction 1 for details.
APPENDIXES 697
e) Additional Information: This subfunction assumes that both the data and
command channels have been opened previously. Data is written from
the buffer at $FE00H.
IV. 3 .a) Subfunction 3: 1571 Read Set Up (MFM or GCR formats)

b) A = 4
L = 3
* VICTRACK = (1-35)— Variable for track number on disk
* VICSECTOR = (0-21)— Variable for sector number on disk
* VICDRV = — Variable for disk drive device number (See VICDRV
table above.)
* = Ranges apply to GCR format only. The ranges are different for
MFM disks depending on the manufacturer.
VIC$COUNT = Number of sectors to read (on the track)
VICDATA = 12 ($0C) if drive isnot a fast (1571) disk drive
VICDATA = 13 ($0D) if channel error occurs
If FAST ANDed with VICDRV = 0 meaning drive isa 1541
If FAST ANDed with VICDRV = 1 meaning drive isa 1571
d) This subfunction sets up the 1571 disk drive for a read operation.
However, the data transfer is not performed by the 8502 BIOS. The
data is transferred by the Z80.
e) Additional Information: To access the back side of an MFM disk set bit
7 ($80) of VICSECTOR. For MFM formats, a dash between the track and
sector on the display window means that the drive accesses the back
side of the disk. This is usually performed by the BIOS.
IV. 4 .a) Subfunction 4: 1571 Write Set Up (MFM or GCR formats)
b) A = 4
L = 4
* VICTRACK = (1-35)— Variable for track number on disk
* VICSECTOR = (0-21)— Variable for sector number on disk
VICDRV = — Variable for disk drive device number. See VICDRV
table on previous page.
VIC$COUNT = Number of sectors to read
* = Ranges apply to GCR format only. The ranges are different for
MFM disks depending on the manufacturer.
VICDATA = 12 ($0C) if drive is not a fast (1571) disk drive
VICDATA = 13 ($0D) if channel error occurs
If FAST ANDed with VICDRV = 0 meaning drive is a 1541
If FAST ANDed with VICDRV = 1 meaning drive is a 1571
d) This subfunction sets up the 1571 disk drive for a write operation.
However, the data is not performed by the 8502 BIOS. The data is
transferred by the Z80.
698 COMMODORE 128
e) Additional Information. This is how the user should select between a

1541 and 1571 Drive. To access the back side of an MFM disk set bit 7
of VICSECTOR. To perform a write operation, the user will have to do
so in their application.
IV. 5 .a) Subfunction 5: Interrogate 1541 or 1571 Disk Drive

b) A = 4
L = 5
VICDRV = (8-11)— Disk drive device number variable
c) VICDATA = lower four bits returns status for FAST read/write
(same as previous VICDATA Disk error codes, which
are listed in the Disk Error Status table on the next page).
= upper four bits return sector size for MFM disks
d) This subfunction interrogates the disk drive for the disk sector size
(MFM or GCR) and the drive status. In addition, this subfunction
initializes the FAST variable, closes and reopens the channel for the
corresponding drive, and clears the drive status.
IV. 6 .a) Subfunction 6: Query to Disk
b) A = 4
L = 6
VICTRACK = (1-35)— Variable for track number on disk
VICSECTOR = (0-21)— Variable for sector number on disk
VICDRV = — Variable for disk drive device number
c) If error-free
VICDATA = lower four bits returns status for FAST read/write
= upper four bits return sector size for MFM disks and the
subfunction inputs 6 bytes into a memory buffer starting at
location $FE00 (and ending at $FEFF). These 6 bytes are
defined as :
1) $FE00-TRACK STATUS (on track below)

2) $FE01-Number of Sectors
3) $FE02-Logical Track
4) $FE03-Minimum Sector Number (on this track)
5) $FE04-Maximum Sector Number (on this track)
6) $FE05-Physical Interleave
If an error occurs:
VICDATA = 11 (SOB) if disk in drive has been changed

VICDATA = 12 ($0C) if drive is not a fast (1571)disk drive
VICDATA = 13 (SOD) if channel error occurs
d) This subfunction queries the disk, and returns the disk status and sector
size (if MFM format). In addition, the buffer located between $FE00
and $FEFF receives 6 data items as described above.
APPENDIXES 699
Disk E rro r Status T ab le
MD DN II 12 Dl D2 D3 D4
MD ................ Mode: 1 = MFM, 0 = GCR.

DN .............................................. Drive Number.
II, 12 ........................................... Sector Size.
a) 00 = 128 bytes
b) 01 = 256 bytes
c) 10 = 512 bytes
d) 11 = 1024 bytes
D 1-D 4 .................................... Controller Status
GCR
OOOx = Ok.
0010 = Sector not found.
0011 = No Sync.
0100 = Data block not found.
0101 = Data block checksum.
0110 = Format error.
0111 = Verify error.
1000 = Write protect error.
1001 = Header block checksum.
1010 = Data extends into next block.
1011 = Disk ID mismatch/Disk change.
1100 = Drive is not fast (1571).
1101 = Channel Error.
1110 = Syntax.
1111 = No Drive present.
MFM
OOOx = Ok.
0010 = Sector not found.
0011 = No address mark.
0100 = Unused.
0101 = Data CRC error.
0110 = Format error.
0111 = Verify error.
1000 = Write protect error.
1001 = Header CRC error.
1010 = Unused.
1011 = Disk change.
1100 = Drive is not fast (1571).
1101 = Channel Error.
1110 = Syntax.
1111 = No Drive present.
Disk E rro r Status T ab le

700 COMMODORE 128
IV. 7 .a) Subfunction 7: Print characters to a serial bus printer

b) A = 4
L = 7
VICDRV = Printer number (either 4 or 5)
VICTRACK = Secondary address in which device is opened as
VICDATA = Character to be printed to the serial bus printer (if
VICCOUNT = 0)
c) VICDATA = -1 ($FF) if device is not present
d) This subfunction outputs characters to the previously opened serial bus
printer.
e) Additional Information: If the secondary address (which is normally 7)
is changed, the device is closed and reopened with the new secondary
address. If a serial bus error besides device not present occurs, the
channel is closed, reopened and the original operation is executed
again.
If VICCOUNT is not equal to zero, the data is printed from the buffer
pointed to by $FE00. The number of bytes printed is supplied in VICCOUNT.
IV. 8. a) Sub function 8: Format a 1541 or 1571 Diskette

b) A = 4
L = 8
D R IV E#
FAST
c) VICDATA
d) This subfunction formats a 1541 or 1571 diskette in the appropriate
drive. If FAST is enabled, (FAST ANDed with D R IV E # (not equal to
0)), the length of a disk command is fetched from memory buffer
location $FE00. The command starting at location $FE01 and ending at
the location specified by the length of the command in $FE00 is sent to
the drive. For example, if $FE00 = $06, the command in the memory
buffer between $FE01 and $FE06 is sent to the drive. All commands
have a “ U 0” preceding them, so only the command from the memory
buffer must be supplied.
IV. 9 .a) Subfunction 9: User Call to 8502 Code Routine
b) A = 4
L = 9
VICOUNT = ($FD05) low byte address of 8502 routine (pointer to the
start of execution of the user routine)
VICDATA = ($FD06) high-byte address of 8502 routine
c) U ser defined outputs only
d) This is the routine that allows you to call an 8502 machine language
subroutine from Z80 mode. The 8502 coded routine is usually user
defined. It must execute in RAM bank 0 with the input/output registers
enabled. The MMU value = $3E. Control is transferred to the 8502
processor with the KERNAL disabled. If you want to call a C l 28
APPENDIXES 701
KERNAL routine, you must enable the KERNAL after you have trans
ferred control to the 8502. Before you return to the Z80, you must
disable the KERNAL again.
When control passes from the Z80 to the 8502 processor, the Z80
is idle. To return control to the Z80 processor, place the customary
8502 RTS instruction at the end of your 8502 coded routine and control
is passed back to the Z80.
e) Additional Information: Once control is passed from the Z80 to the
8502, the 8502 is running at the speed of 1 Mhz. You can increase the
speed to 2 Mhz to speed up processing on the 8502 side of the
computer. However, YOU MUST RETURN TO 1 Mhz SPEED BEFORE
RETURNING TO THE Z80 OR A SYSTEM CRASH WILL OCCUR.
The nature of the timing of the two processors dictates this. If you
don’t return to 1 Mhz, the clock cycle timing is thrown off and the
system crashes.
IV. 10.a) Subfunction 10: RAM Disk Read

b) A = 4
L = 10
c) Data is transferred to expansion RAM from RAM (BANK 0)
d) All expansion RAM registers must be set up prior to calling this routine.
IV. 11.a) Subfunction 11: RAM Disk Write

b) A = 4
L = 11
c) Data is transferred to expansion RAM from RAM (BANK 0)
d) All expansion RAM registers must be set up prior to calling this routine.
V. a) User Function 5: Read 40/80 Column Key

c) A = Value stored in $D505 (C128 Mode Configuration register)
If bit 7 is high (1), 40/80 key is up, otherwise 40/80 isdown.
d) This function returns the value of location $D505, the mode configura
tion register. Only bit 7 is significant as noted above. The 40/80 key is
not in the keyboard matrix table, so this function is dedicated to reading
its position.
VI. a) Functions 6 through 254 are not implemented.
b) none
c) HL = number of days (in binary) since 1/1/78
VII. a) User Function 255: System Date
b) A = -1 ($FF)
c) HL = number of days (in binary) since 1/1/78
d) By specifying A = -1 , the system date is returned.
702 COMMODORE 128
A PPENDIX K
PART II
CALLING CP/M BIOS, 8502 BIOS A N D
CP/M USER FUNCTIONS IN
Z80 MACHINE LANGUAGE
The Commodore 128 CP/M system allows you to call the CP/M BIOS, 8502 BIOS and
CP/M user functions in your own Z80 assembly language programs. However, in order
to program in Z80 assembly language, you need either an assembler or machine
language monitor. Many Z80 assemblers and monitors are available on the market;
however, the full featured Digital Research CP/M Plus (3.0) system now available on
the Commodore 128 comes with two Z80 assemblers, MAC and RMAC and CP/M plus
documentation. These programs or documentation are not included in the Commodore
128 Personal Computer package; refer to the Commodore 128 System Guide for infor
mation on obtaining them.
Assuming you have the MAC and RMAC assemblers, you can now enter and
assemble Z80 assembly language programs. At this point, this reference guide must
make a substantially large assumption about its readers and their knowledge of Z80
assembly language programming. As you probably agree, this reference guide could not
possibly introduce Z80 machine language and thoroughly cover it. That is simply
beyond the scope of this book, considering how voluminous it is already. Your trusty
local bookstore undoubtedly offers several excellent books on Z80 programming. See
“ Suggestions for Further Reading” for a few Z80 book titles.
Now that all the assumptions are out of the way, here’s how to call a Z80 user or
BIOS function. First, the user function call.
CALLING A CP/M SYSTEM USER FUNCTION

As you saw in Part I of this appendix, certain user functions had subfunctions, others
did not. User function 4 is the 8502 BIOS call function. The 8502 BIOS functions have
13 (—1—11) subfunctions which perform the machine level 8502 input and output
routines. Do not confuse the 8502 BIOS with the CP/M BIOS. The CP/M BIOS comes
standard with every CP/M Plus system regardless of the hardware running it. Remember
that CP/M was designed to be transportable from one microcomputer to another. Each
different microcomputer has its own machine level input and output which it must
perform. The 8502 BIOS is completely hardware-dependent and will only run on the
Commodore 128’s 8502 microprocessor.
Another way of understanding the difference between the two BIOS types is
APPENDIXES 703
recalling the CP/M jum p vector on page 677 of Appendix K. The first 30 jumps (0-29)
are direct calls to CP/M BIOS routines. Jump number 30 is the call to the user functions, of
which user function 4 is the 8502 BIOS. The 8502 BIOS is a subset of the CP/M user
functions. The user function call is 1 of 31 system routine calls within the CP/M BIOS
jum p table.
The following example calls user function 2 the keyscan function. The calling
routine starts at “ waitkey” and the subroutine starts at “ user$fun” .
MVI useroffset,30
waitkey:
MVI A ,2 ; load A reg. with function no. 2
CALL user$fun ;calls subroutine user$fun
INR B ;increment B— test for -1 if no key is pressed
JZ waitkey ;jump to wait key if none
DCR B ;decrement B if B not equal to 0
;this restores original matrix value
Rest of Program
user$fun
PUSH H ;place HL reg. pair on stack
LHLD 1 ;load address of jump vector 1 (WBOOT) into HL
MVI L, useroffset * 3 ;get offset of 90 (30*3) to L
XTHL ;exchange HL with top of stack
RET ;jump to new HL pointer location of routine
First, the main program loads the user function number (2) into the A register. To
call a user function, place the required input parameter, the user function number, in the
A register. If a subfunction is going to be called, like in user function 4 (8502 BIOS), the
input parameter for the subfunction number must be placed in L.
The second instruction in the main routine is a CALL to the subroutine user$fun.
However, in this example of the keyscan user function, the returned value B is -1 if
no key is pressed. If this is the case, B is incremented to zero and the main program
jumps to waitkey and scans the keyboard again. Otherwise, the rest of the program
continues processing.
When the subroutine is called, the first instruction saves the HL register pair, the
address pointer, on top of the stack. The next instruction loads memory location 1 (low)
and 2 (high) into register pair HL. The high byte points to the page number that the
jum p vectors are on and the low byte is always 3 (Jump #1 the warm boot vector). Next
the jum p number times 3 is loaded into the low byte (L) of register pair HL. This adds the
offset of 90 memory locations to base address of the CP/M BIOS jump table, which now
points to jump number 30, USERF.
704 COMMODORE 128
m
The XTHL instruction exchanges the HL register pair with the top of the stack.
This places the computed address on the top of the stack and the entry values of HL
back in HL. When the RETum instruction is reached by the Z80, program control is
therefore passed to the USERF vector, entry 30 in the CP/M BIOS jump table. When
the function has completed, control returns to the instruction immediately following the
CALL User$Fun instruction (INR B).
In order for the routines to be called successfully, the proper required input
parameters must be placed in the appropriate registers. The user function number must
be placed in the A register, the subfunction number is placed in L, if any. Additional
inputs must be placed in the correct register or variable prior to calling the user function.
The above example calls user function 2, the keyscan function. To call any other
user function, load the subfunction number into the A register. To call a sub function,
load the L register with a subfunction number on the beginning of the main (calling)
program as follows:
MVI L, subfun
The way this program is written, the value in L, which you will load at the beginning
of the program with the above instruction, is placed back in L from the stack when the
XTHL is reached at the end of the subroutine. Use this example as a template when
calling other subfunctions. Make sure the proper inputs are present in the correct
locations prior to calling the user function.
CALLING A CP/M BIOS ROUTINE

Making a direct BIOS call is different from calling a user function. User function calls
always enter jum p number 30 in the CP/M BIOS jump table. A direct BIOS call enters
any of the first 30 (0-29) jum p vectors. These are the input and output routines that are
a part of any CP/M system on any microcomputer. The standard method of calling a
CP/M BIOS routine is via BIOS function 50. This function handles the banking of the
two 64K RAM banks in the C l 28.
You could call a CP/M BIOS routine similar to the first example, but there is a
limitation. With the user function call method, you are only able to call BIOS routines
that reside in RAM bank 1. The TPA resides in bank 1, while the Z80 function code is
stored in RAM bank 0. If you try to call a direct BIOS routine in bank 0, the system will
crash because bank 1 is in context. This is why it is important to make direct
CP/M BIOS calls through user function 50. Refer to the Digital Research CP/M Plus
documentation for more details on user function 50.
H ere’s a program example that illustrates how to call a CP/M BIOS function.
main
MVI C,50 ;store function no. 50 in reg. C
LXI D,bios$pb ;load immediate a 16 bit pointer in DE
CALL 5 standard BIOS call
APPENDIXES 705
Rest of Program
RET
bios$pb: ;BIOS variable table
db FUNNUM ;BIOS function no. variable
db AREG tem porary A reg. storage
dw BCREG ;temp BC reg. pair storage
dw DEREG ;temp DE reg. pair storage
dw HLREG ;temp HL reg. pair storage
The first instruction in the main routine stores the function number 50 in the C
register. The system expects the input parameter C to be the BIOS function number. The
next instruction loads the address of the BIOS variable table bios$pb into register pair
DE. The third instruction calls BDOS function 50 through the call BDOS vector, the
standard BIOS vector through which all direct BIOS functions are called. BDOS
function 50 manipulates the banking in and out of RAM banks 0 and 1. This is the
recommended way of directly calling a CP/M BIOS function over that of the first
example which is designed primarily for calling user functions.
The variable table “ bios$pb” contains the necessary input parameters required for
BDOS function 50. The “ db” ’s stand for a byte of storage (like .byte in 8502) while the
“ dw ” ’s stand for a 16 bit word (like .word in 8502).
This appendix is the only section of the book that crosses the Z80 programming
barrier. It at least points you in the right direction and gives you the “ hooks” into the
machine level routines of the CP/M system. For more detailed Z80 programming
information see the “ Suggestions for Further Reading” at the back of the book. For
more detailed CP/M information, refer to the Digital Research CP/M Plus documentation.
MFM DISK FORMATS

The abbreviation MFM stands for Modified Frequency Modulation. This type of disk
format is variable and programmable according to the specifications of the particular
computer manufacturer.
The Commodore 128 CP/M system supports four standard MFM disk formats. These
four formats constitute the majority of CP/M software formats available on themarket.
This does not mean that the Commodore 128 CP/M system can read every single CP/M
disk format in the universe; however, the majority of available CP/M software can run
on the Commodore 128, if the particular application is not hardware dependant.
The four major formats that are supported are as follows:
a) Epson QX10 (double sided)

a l) Epson QX10 (double sided)
b) IBM-8 (single sided)
b l) IBM-8 (double sided)
c) KayPro IV (double sided)
d) KayPro II (single sided)
e) Osborne (single sided)
706 COMMODORE 128
Each of these formats is compatible with the Commodore 128 CP/M system. At
the present time, the system cannot format these disk types and successfully use them on
the host system, but they can be used on the C l28 system. This portion of the system
is still in development.
The following table lists the parameters that these disk formats are looking for
when reading third party CP/M software.
M anufacturers
a b c d e al bl
Disk Type DSDD SSDD DSDD SSDD SSDD DSDD DSDD
Starting pos. (t/s) 2/1 1/1 0/10 1/0 3/1 2/1 1/1
Sector size (bytes) 256 512 512 512 1024 512 512
Number of sec/trk 32 8 10 10 5 20 8
Number of tracks 40 40 80 40 40 40 80
Block alloc size 2048 1024 2048 1024 1024 2048 2048
# of dir entries 128 64 128 64 64 128 64
# of resvd tracks 2 1 1 1 3 2 1
SS = single sided
DS = double sided
DD = double density
M F M Disk F o rm a t T ab le
N O TE: Epson (a) labels sector numbers 1 through 16. The other Epson
QX10 format (al) labels sectors from 1 to 10. Both the top and bottom of
the disk are labeled the same way.
IBM (b) labels sector numbers 1 through 8. Both the top and
bottom of the disk are labeled the same way.
KayPro IV and KayPro II (c and d) label sector numbers 0 through
9 on top and 10 through 19 on the bottom.
These values are taken from the MFM table. The vector at $FD46 holds the
pointer to the start of the table labeled MFM$table. In the current system, these are the
formats that are read and write compatible on the Commodore 128 CP/M Plus system.
The following is a listing of the MFM format table:
APPENDIXES 707
db S256*2 + (16*2-8) + 1 256 byte sect, 16 sect/trk

db MFM + S256 + TypeO + CO + SI DSDD
dw 0 start on track 2 sect 1 (2 ale)
dpb 256,32,40,2048,128,2 se c t# 1 to 16
db 16 (top and bottom numbered the same)

db ‘Epson QX10’ 1 Epson QX10
db 80h + S512*2 + ( 10*2-8) + 1 ,512 byte sect, 10 sect/trk

db S256*2 track 0 is 256 bytes/sector
db MFM + S512 + TypeO + CO + SI DSDD
dw 0 start on track 2 sect 1 (2 ale)
dpb 512,20,40,2048,128,2 se c t# 1 to 10

db ‘Epson QX10’ 2
db S512*2 + (8*2-8) + 1 512 byte sect 8 sect/trk

db MFM + S512 + Type2 + CO + SI SSDD
dw 0 start on track 1 sector 1 (2 ale)
dpb 512,8,40,1024,64,1 sec t# 1 to 8
db 8
db ’ IBM-8 SS’
db S512*2 + (8*2-8) + 1 512 byte sect 8 sect/trk

db MFM + S5 12 + Type2 + CO + SI DSDD
dpb 512,8,80,2048,64,1 sec t# 1 to 8

db ’ IBM-8 D S’ 4
db S512*2 + (1 0 * 2 -8 )+ 0 512 byte sector, 10 sect/trk

db MFM + S512 + T yp el + C l + SO DSDD
dpb 512,10,80,2048,128,1 se ct# 0 to 9 on top (even tracks)
db 10 se c t# 10 to 19 on bottom (odd tracks)

db ‘KayPro IV ’ 5
db S512*2 + (10*2-8)4-0 512 byte sect, 10 sect/trk

db MFM + S512 + TypeO + C l + SO SSDD
dpb 512,10,40,1024,64,1 sec t# 0 to 9
db 10
db ‘KayPro IP 6
MFM Format Table
708 COMMODORE 128
db S1024*2 + (5*2-8) + 1 ; 1024 byte sect, 5 sect/trk

db MFM + S1024 + TypeO + CO + SI ; SSDD
dw 0 ; start on track 3 sector 1 (2 ale)
dpb 1024,5,40,1024,64,3 ; se c t# 1 to 5
db 5 ;
db ‘Osborne D D ’ ;7
M FM F o rm a t T a b le (continued)
APPENDIXES 709
APPENDIX K
PART III
THE CP/M SYSTEM MEMORY MAP
The following pages contain the Z80 CP/M memory map for the Commodore 128.
The memory map includes all the key CP/M locations, vectors and variable tables.
Use it as a guide through the Z80 CP/M system within the Commodore 128.
The CP/M memory map is available as a disk file “ CXEQU.LIB” on the disk
that comes with the computer.
710 COMMODORE 128
$*MACRO
fa ls e equ 0
tru e equ n o t fa ls e
banked equ tru e
EXTSYS equ fa ls e ; u se e x te rn a l s y s te m a s d isk and ch a r
p r e $ r e le a s e equ fa ls e
; s ta r t a t Jan 1 ,1 9 7 8 78 79 80 81 82 83 84
d t$ h x $ y r equ 365+365+366+365+365+365+366
1985 1 2 4 5 6 7 8 9 10 11 12
d a te $ h e x equ d t$ h x $ y r+ 3 1 + 2 8 + 3 1 + 3 0 + 3 1 + 3 0 + 3 1 + 3 1 + 3 0 + 3 1+30+ 6
d a te m ac ro
db '6 D ec 8 5 '
endm
; b o o t m e m o ry m a p (b an k 0 o n ly )
b io s 02 equ 3000h
b lo c k $ b u ffe r equ 3 4 0 Oh ; u s e s 2K
b o o t$ p a rm equ 3C00h ; u s e s a b o u t 256 b y te s
; bank 0 lo w m e irp ry rre p
ROM equ OOOOh

V IC $ c o lo r equ lO O O h ; I /O p a g e o n ly (I0 $ 0 s e le c te d )
S Y S $ k ey $ area equ lO O O h ; 3 256 b y te b lo c k s (a llo w 4)
sc re en $ 4 0 equ 1400h ; 2 X 80 X 25 = 4 000
B A N K $ p a rm $ b lk equ 240G h ; a l l o w 0 .5 K o f p a r a m e t e r s
B IO S 8 5 0 2 equ 2600h ; 1 . 5K
V IC $ sc re en equ 2C 00h ; IK
c c p $ b u ffe r equ 3000h ; 0 c 8 0 h ( a l l o w 4K )
0
bank $ fre e equ 4 0 0 Oh ; s t a r t o f fre e a re a in bank 0
; m apped I /O lo c a tio n s
V IC equ ODOOOh
S ID equ 0D 400h
MMU equ 0D 500h
DS8563 equ 0D 600h 8563
V IC $C $H equ 0D 800h ( m e m o ry m a p p e d o n l y i n I $ 0 0)
V IC $ C $ L equ OlOOOh (m e m o ry a n d i / o m a p p e d i n 10$
C IA l equ ODCOOh 6526
C IA 2 equ ODDOOh 6526
USART equ ODEOOh 6551 ( e x tr n c a rd )
R A M $ d sk $ b ase equ ODFOOh 8726
; C o m r o n m e m o ry a l l o c a t i o n
in t$ b lo c k equ OFCOOh ; m ode 2 i n t e r r u p t p o i n t e r s ( to FD FD h)

p a rm $ b lo c k equ OFDOOh ; s y s te m p a r a m e te r s
@b u f f e r equ OFEOOh ; d i s k b u f f e r (2 5 6 b y t e s )
OFFOOh ; t o O FFFFh u s e d b y 8 5 0 2
th e f o l l o w i n g a r e C l 28 s y s te m e q u a te s
e n a b le $ z 8 0 equ OFFDOh ; 8502 co d e

re tu rn $ z 8 0 equ OFFDCh
e n a b le $ 6 5 0 2 equ OFFEOh ; Z80 c o d e
re tu rn $ 6 5 0 2 equ O FFEEh
; 1 s t b y te u se d a s I n tt e r r u p t p o in te r
v ic $ a n d equ p a rm $ b lo c k + l ; ; b i o s 8 5 0 2 ccm m and b y t e
v ic $ d rv equ v ic $ c m d + l ; b io s8 5 0 2 d r iv e ( b i t 0 s e t , d r v 0)
v ic $ trk equ v ic $ d rv + l ; ; b i o s 8 '5 0 2 t r a c k #
APPENDIXES 711
v ic $ s e c t equ v ic $ trk + l b io s8 5 0 2 s e c t o r #
v ic $ c o u n t equ v ic $ s e c t+ l b io s8 5 0 2 s e c t o r c o u n t
v ic $ d a ta equ v ic $ c o u n t+ l b i o s 8502 d a ta b y te to /f r o m
c u r $d r v equ v ic $ d a ta + l c u rr e n t d is k in s ta lle d to V ir. d r iv e
fast equ c u r$ d rv + l b it 0 s e t, d rv 0 is f a s t. e c t.
k e y $ tb l equ fa st +1 p o in te r to k e y b o a rd t a b l e
f u n $ tb l equ k e y $ tb l +2 p o in te r to fu n c tio n ta b le
c o lo r$ tb l$ p tr equ fu n $ tb l +2 p o in te r to lo g ic a l c o lo r ta b le
fu n $ o ffse t equ c o lo r $ tb l$ p tr +2 fu n c tio n # t o b e p rep o rm ed
sou n d $ l equ fu n $ o ffs e t+ l unused
sound $2 equ so u n d $ l+2
so u n d ?3 equ sound$ 2+2
@t r k equ s o u n d ? 3+2 c u r r e n t t r a c k num ber
@ d im equ @ trk +2 c u r r e n t DMA a d d r e s s
; b e l o w h e r e n o t u s e d b y ROM
@ sect equ @ dm a +2 c u r r e n t s e c to r num ber

@ cnt equ @s e c t +2 re c o rd c o u n t f o r m u ltis e c to r tr a n s f e r
(3 c b n k equ @cn t +1 bank f o r p ro c e s s o r o p e ra tio n s
@ dbnk equ @cbnk +1 b a n k f o r DMA o p e r a t i o n s
(a a d rv equ +1
(? d b n k c u rr e n tly s e le c te d d is k d riv e
@ rd rv equ @ a d rv + l c o n tr o lle r r e l a ti v e d is k d riv e
ccp$count equ (a rd rv +1 n u m b e r o f r e c o r d s i n t h e CCP
s ta t$ e n a b le equ c c p $ c o u n t+ l s t a t u s l i n e e n a b le
7 k y b r d , k e y c o d e s ( 1 ) , f u n c t i o n s (0 )
6 4 0 c o lu m n t r a c k i n g o n ( 0 ) , o f f ( l )
0 d is k s t a t , e n a b le ( l) , d isa b le (O )
e m u la tio n ? a d r equ s ta t$ e n a b le + l a d d r e s s o f c u r r e n t e m u la tio n
u s a rt$ a d r equ a n u la tio n $ a d r +2 p o i n t e r t o USART ( 6 5 5 1 ) r e g i s t e r
; C X IO e q u a t e s
in t$ h l equ u s a rt$ a d r +2 i n t e r r u p t HL h o l d l o c a t i o n
in t$ s ta c k equ in t$ h l+ 2+20 c u r r e n tly o n ly 5*2 u se d
u s e r$ h l$ te m p equ in t$ s ta c k u s e r f u n c t i o n HL h o l d l o c a t i o n
h l$ te m p equ 1
u s e r $ h $ te m p +2 m i s c t e m p s t o r a g e ( u s e d b y VECTOR)
d e $ te m p equ h l$ te m p +2 m i s c t e m p s t o r a g e ( u s e d b y VECTOR)
a $ te m p equ d e $ to n p+2 m i s c t e m p s t o r a g e ( u s e d b y VECTOR)
so u rc e $ b n k equ a $ te m p + l i n t e r b a n k m ove s o u r c e b a n k #
d e s t$ b n k equ s o u rc e $ b n k + l i n t e r b a n k m ove d e s t b a n k #
M F M $ tb l$ p tr equ d e s t$ b n k + l p o i n t e r t o MFM t a b l e
; 1 s t r e le a s e end
p rt$ c o n v $ l equ M F M $ tb l$ p tr+ 2
p rt$ c o n v $2 equ p rt$ c o n v $ l +2
k e y $ F X $ fu n c tio n e q u p rt$ c o n v $ 2+2
X x D $ c o n fig equ k e y $ F X $ fu n c tio n + 2
b it 7 0 = no p a rity 1= p a rity
b it 60= m a rk /sp a c e 1= o d d /e v e n
b it 5 0 = s p a c e /e v e n 1= m a rk /o d d
b it 10=1 s to p b i t 1=2 s to p b i t s
b it 00= 7 d a ta b i t s 1=8 d a ta b i t s
R S 2 3 2 $ s ta tu s equ X x D $ c o n fig + l b it 7, l= se n d d a ta , 0 =no d a ta

b it 6 , l= s e n d in g d a ta
b it 5, l= re c v q u e a c tiv e
b it 4, l= p a rity e rro r
b it 3, l= fra m in g e r r o r
b it 2 , l= re c v o v e r ru n (n o u se d )
b it 1, l= re c e iv in g d a ta
b it 0, l= D a ta b y t e r e a d y
x m it$ d a ta equ R S 2 3 2 $ s ta tu s + l d a ta b y te t o sen d

re c v $ d a ta equ x m it$ d a ta + l r e c e iv e d d a ta b y te
T he f o llo w in g e q u a te s a r e u sed by th e i n te r r u p t d riv e n k e y b o a rd h a n d le r

712 COMMODORE 128
in t$ ra te equ re c v $ d a ta + l
1 s t b y te i s a p o in te r in to ta b le , nd to2 12
t h b y te r e p r e s e n t
t h e k e y b o a r d s c u r r e n t s t a t e ( a c t i v e l o w ) , N O TE: o n l y
c u rre n t i f k e y $ b u ffe r i s n o t f u ll
k e y $ s c a n $ tb l equ in t$ r a te + l
k e y b o a rd r o l l over b u ffe r
k e y $ b u f$ s iz e equ 8*2 m u st b e an
k e y $ g e t$ p tr equ k e y $ s c a n $ tb l+ l 2
k e y $ p u t$ p tr equ k e y $ g e t$ p tr +2
k e y $ b u ffe r equ k e y $ p u t$ p tr +2
; s o f t w a r e UART r e c v b u f f e r
R x D $ b u f$ s iz e equ 64
RxD $buf $ c o u n t equ k ey $ b u f fer+ k e y $ b u f $s iz e
R x D $ b u f$ p u t equ R x D $ b u f$ c o u n t+ l
R x D $ b u f$ g e t equ R x D $ b u f$ p u t+ l
R x D $ b u ffe r equ R x D $ b u f$ g e t+ 1
tic k $ v o l equ R x D $ b u ff e r + R x D $ b u f $ s iz e
equ tic k $ v o l+ l
I N T $ v e c to r equ OFDFDh c o n ta in s
; ( i n com m on
;= > 4 0 c o lu m n m is c p a rm
te m p $ l equ B A N K $ p a rm $ b lk ;
@ o f f 40 equ te n p $ l+2 ;
c u r$ o ffse t equ @ o f f 40
o ld $ o ffs e t equ @ o ff4 0 + 2 ;
p rt$ flg equ o ld $ o ffs e t+ l :
f la s h $ p o s equ p rt$ flg + l ;
; = = > 40 c o lu m n p o s i t i o n a n d c o l o r s t o r a ge
p a in t$ s iz e equ fla s h $ p o s +2 ;
c h a r$ a d r$ 4 0 equ p a in t$ s iz e + l ;
c h a r$ c o l$ 4 0 equ c h a r$ a d r$ 4 0 + 2 ;
c h a r$ ro w $ 4 0 equ c h a r$ c o l$ 4 0 + l ;
a ttr$ 4 0 equ c h a r$ ro w $ 4 0 + l ;
b g $ c o lo r $ 4 0 equ a ttr$ 4 0 + l
b d $ c o lo r $ 4 0 e q u b g $ c o lo r $ 4 0 + l ;
rev $ 4 0 equ b d $ c o lo r $ 4 0 + l ;
? = > 8 0 c o lu m n p o s i t i o n a n d c o l o r s t o r a g e
c h a r$ a d r equ re v $ 4 0 + l ;
c h a r$ c o l equ c h a r$ a d r +2 ;
c h a r$ ro w equ c h a r$ c o l+ l ;
c u rre n t$ a tr equ c h a r$ ro w ^ l ;
b g $ c o lo r $ 8 0 equ c u rre n t$ a tr+ l ;
c h a r$ c o lo r$ 8 0 equ b g $ c o lo r $ 8 0 + l ;
; ROM u s e s l o c a l t i o n s a b o v e t h i s p o i n t
; = > E m u la tio n p a r a m e te r s
p a rm $ b a se equ c h a r$ c o lo r$ 8 0 + l
p a rm $ a re a $ 8 0 equ p a rm $ b a se +2
ds 2 ; 8 0 c o lu m n
• ds 1 80 c o lu m n
p a rm $ a re a $ 4 0 equ p a rm $ a re a $80+3
ds 2 40 c o lu m n
ds 1 40 c o lu m n
b u ffe r$ 8 0 $ c o l equ p a rm $ a re a $ 40+3
; = > C X IO p a r a m e t e r s
; i n t $ c o u n t n o t u s e d b y r e l e a s e s p a s t 10 O c t 85
in t$ c o u n t qu b u f fe r $ 8 0 $ c o l+ 4 0 * 2 ; o ne ad ded e v e ry l/6 0 th sec
key$buf ev. i in t$ c o u n t+ l
APPENDIXES 713
;==> CXKEYS parameters

key$down$tbl equ key$buf+l ; not used any more (int code)
free space above, new interrupt driven code does not require this space
; c o n tro l$ k e y s equ 1
k e y $ d o w n $ t b l + l *^ ; b y te , n o t u s e d a n y m o re ( in t code)
c o m m o d o re$ m o d e equ k e y $ d o w n $ tb l+ l 1*2

m sg p tr equ c o m m o d o re$ m o d e + l
o ffse t equ m sg p tr +2
c u r$ p o s equ o ffse t +1
s tr in g $ in d e x equ c u r$ p o s+ l
; 1 st re le a se e n d (3 J u n e 8 5 )
sy s$ fre q equ s tr in g $ in d e x + l ; - l= 5 0 H z , 0 = 6 0H z
; 2nd r e le a s e e n d (1 A ug 8 5 )
equ sy s$ fre q + l
; = = > t e m p ROM b o o t d a t a s t o r a g e
b lk $ p tr$ c n t equ 32
lo a d $ c o u n t equ b o o t$ p a rm num ber o f 128 b y te b lo c k s t o lo a d
ld $ b lk $ p tr equ lo a d $ c o u n t +2 c u r r e n t s e c t o r dm a p o i n t e r
b lk $ u n ld $ p tr equ ld $ b lk $ p tr +2 r e a d m e m o ry b l o c k ( l k , 2 K ) p o i n t e r
b lo c k $ s iz e equ b lk $ u n ld $ p tr +2b l o c k s i z e (1 K = 3 2 o r 2 K = 6 4 )
b lo c k $ e n d equ b lo c k $ s iz e + l a l l o w 48K c p m . s y s t o l o a d
b lo c k $ p trs equ b lo c k $ e n d +2 e n d o f b l o c k l o a d b u f f e r (+ 1 K o r +2K
in fo $ b u ffe r equ b lo c k $ p trs + b lk $ p tr$ c n t
; C f^ l3 .s y s l o a d a d r ' s a n d c o u n t s
e x t$ n u m equ in fo $ b u ffe r +12
C P M 3 .S Y S c u r r e n t e x t #
re try equ e x t$ n u m + l
b o o t$ s ta c k equ re try + 1 + 6 4 a llo w 64 b y te s o f s ta c k
equ b o o t$ sta c k
•
;= = = > s p e c i a l e q u a t e s u s e d b y CXKEY
s p e c ia l equ 00010111b
S F $ e x it equ 00lh RETURN KEY
S F $ in s e rt equ 028h PLU S KEY
S F $ d e le te equ 02B h M INUS KEY
a lp h a $ to g g le equ 03D h c c m m o d o re k e y
a lt$ k e y equ 0 5 Oh a lte r a n t key
S F $ le ft equ 055h l e f t a rro w k ey
If$ a rro w equ 05 5h l e f t a rro w k ey
S F $ rig h t equ 056h r i g h t a rro w k ey
rt$ a rro w equ 05 h 6 r i g h t a rro w k ey
b u ff$ la rg e equ 25
b u ff$ s m a ll equ 7
b u ff$ p o s equ 7
;= = > E x te rn a l R S232 i n te r fa c e c o n tro ls

rx d $ 6 5 5 1 equ USART+0 rea d
tx d $ 6 5 5 1 equ USART+0 w rite
s ta tu s $ 6 5 5 1 equ USART+1 rea d
re s e t$ 6 5 5 1 equ USART+1 w rite
canm and$6551 equ USART+2 re a d /w rite
c o n tr o l$ 6 5 5 1 equ USART+3 re a d /w rite
tx rd y equ lO h
rx rd y equ 08h
c rrd $ in it equ O bh ; n o p a r i t y , e n a b le tx d + r x d , i n t e r r u p t s o ff
c n t r $ i n i t $ 19200 equ lF h ; 1 s to p , 8 b i t s , 19200 baud
c n tr $ in it$ 9 6 0 0 equ lE h ; 1 s to p , 8 b i t s , 9600 b au d ( in te r n a l)
c n tr $ in it$ 6 0 0 equ 017h ; 600 b au d
;= > m e m o ry m a n a g e m e n t l o a c t i o n s
m m u $ s ta r t equ MMU
c o n f$ re g equ MMU 3eh
c o n f$ re g $ l equ MMU+1 3 fh
c o n f$ re g $2 equ MMU+2 7 fh
c o n f$ re g $ 3 equ MMU+3 3eh
c o n f$ re g $ 4 equ m u+4 7eh
rro d e $ re g equ MMU+5 b lh
714 COMMODORE 128
ram ? r e g equ MMU +6 O bh 16K t o p C annon

page$ $ l 0 equ MMU+7 OOh
p ag e$ $h 0 equ MVIU +8 O lh
p a g e $ l$ l equ MMU+9 O lh
p a g e $ l$ h equ MWJ +10 O lh
r rm u $ v e rs io n equ MMU +11 ??h
e n a b le $ C 6 4 equ 11110001b FS=0

z 8 0 $ o ff equ 10110001b v a lu e to b e w rite to
z80$on equ 10110000b
f a s t$ rd $ e n equ Z80$on+0 f a s t s e r ia l rea d
fa s t$ w r$ e n equ Z80$on+8 f a s t s e r i a l w rite
com m on$4K equ 09h to p 4K cc m m o n
c o rrm o n $ K8 equ O ah to p 8
K cc m m o n
co m m o n $ 1 6 K equ O bh t o p 16K c cm m o n
; = > p re c o n fig u r a ti o n m aps

fo rc e $ m a p equ O ffO O h
bank $0 equ O ffO lh 3 fh
bank$l equ 0 02
ff h 7 fh
io equ O ff0 3 h 3eh
io $0 equ O ff0 3 h 3eh
io $ l equ O ff0 4 h 7eh
;= > 8 0 c o lu m n d i s p l a y e q u a t e s
D S $ in d e x $ r e g equ D S8563
D S $ s ta tu s $ r e g equ D S8563
D S $ d a ta $ re g equ D S8563+1
; = > r e g is te r p o in te rs
D S $ c u rs o r$ h ig h equ 14
D S $ c u rs o r$ lc w equ 15
D S $ rw $ p tr$ h ig h equ 18
D S $ rw $ p tr $ lc w equ 19
D S $ rw $ d a ta equ 31
D S $ c o lo r equ 26
;= > s ta tu s b its
D S $ re a d y equ Oh 8
D S $ lt$ p e n equ 40h
; = > d i s p l a y m e m o ry l a y o u t (1 6K ) 0 -3 fffh
D S $ screen equ OOOOh
D S $ a ttrib u te equ 0 8 OOh
D S $ ch a r$ d ef equ 2000
h
; = > V IC e q u a t e s
; v ie c o lo rs
b la c k equ 0
w h ite equ 1
red equ 2
cyan equ 3
p u rp le equ 4
g ree n equ 5
b lu e equ 6
y e llo w equ 7
o ran g e equ 8
b ro w n equ 9
lt$ re d equ 10
d a rk $ g re y equ 11
m ed $ g ray equ 12
lt$ g r e e n equ 13
lt$ b lu e equ 14
lt$ g re y equ 15
R M $ s ta tu s equ R A M $ d sk $ b ase ;re a d

; b it 7 I n t e r r u p t p e n d in g i f 1
6 T r a n s f e r c o m p le te i f 1
/ 5 B lo c k v e r i f y e r r o r i f 1
; n o te : b its 5 -7 a r e c le a r e d w hen r e a d
4 128K i f 0 , 512K i f 1
3 -0 V e r s io n #
APPENDIXES 715
RM$ co m m a n d equ R A M $ d sk $ b ase+ l ;r/w

b it 7 e x e c u te p e r c u r r e n t c o n fig . i f s e t
6 reserv ed
5 e n a b le a u to r e lo a d i f s e t ( r e s to r e s a l l r e g i s t e r t o
v a l u e b e f o r e com m and w a s d o n e , e l s e p o i n t t o
n e x t b y te to r e a d /w r ite .)
4 d i s a b l e F F 0 0 d e c o d e i f s e t ( d o o p e r a t i o n a f t e r com m and w r i t e n )
3 ,2 rese rv e d
1,0 00 = t r a n s f e r C l 2 8 — > R am D i s k
01 = T r a n s f e r C l 2 8 <— R am D i s k
10 = sw ap C l 2 8 < - > Ram D i s k
11 = V e r i f y C l 28 = Ram D i s k
R M $ 1 2 8 $ lo w equ R A M $ d sk $ b ase+ 2 ;r/w

; b its 0 t o 7 o f C l28 a d d r e s s
R M $ 1 2 8 $ m id e q u R A M $ d sk $ b ase+ 3 ;r/w
; b its 8 to 15 o f th e C128 a d d r e s s
R M $ e x t$ lo w e q u R A M $ d sk $ b ase+ 4 ;r/w
; b its 0 t o 7 o f R am D i s k a d d r e s s
R M $ e x t$ m id equ R A M $ d sk $ b ase+ 5 ;r/w

; b its t o 1 5 o f Ram D i s k a d d r e s s
R M $ e x t$ h i equ R A M $ d sk $ b ase +6 ;r/w

b it 16 o f Ram D i s k a d d r e s s i f 1 2 8 K v e r s i o n
b its 16 to 1 8 o f R am D i s k a d d r e s s i f 5 1 2 K v e r s i o n
R M $ c o u n t $ lo w equ R A M $ d sk $ b ase+ 7 ;r/w

; lo w b y t e t r a n s f e r c o u n t ( b i t s 0 - 7 )
R M $ c o u n t$ h i equ R A M $ d sk $ b ase +8 ;r/w

; h i b y te t r a n s f e r c o u n t (b its 8 -1 5 )
R M $ i n tr $ m a s k equ R A M $ d sk $ b ase+ 9 ;r/w

b it 7 l= e n a b le c h ip i n t e r r u p t s
6 l= e n a b le e n d o f b lo c k i n t e r r u p t s
5 l= e n a b le v e r i f y e r r o r i n t e r r u p t s
R M $ c o n tro l equ R A M $ d sk $ b ase+ 1 0 ; r / w

; b it 7 ,6 00 I n c r o r ie n t b o th a d d r e s s e s (d e fa u lt)
01 F ix e x p a n sio n a d d re s s
10 F ix C l28 a d d re s s
11 F ix b o th a d d re s s e s
;= = > C IA e q u a t e s
D a ta $ a equ OOh
D a ta $ b equ O lh
D a ta $ d ir$ a equ 02 h
D a ta $ d ir $ b equ 03h
tim e r$ a $ lo w equ 04h
tim e r$ a $ h ig h equ 05h
tim e r$ b $ lc w equ 06h
tim e r$ b $ h ig h equ 07h
to d ? s e c $ 6 0 equ 08h
to d $ se c equ 09h
to d $ m in equ O ah
to d $ h rs equ Obh
s y n c $ d a ta equ O ch
in t$ c trl equ O dh
c ia $ c trl$ a equ O eh
c ia $ c trl$ b equ O fh
C IA $ h o u rs equ C IA l+ to d $ h rs
k e y ? ro w equ C I A 1 + D a ta $ a ; o u tp u t
716 COMMODORE 128
key$col equ C I A 1 + D a ta $ b ; in p u t
V IC $ k ey $ ro w equ 0 02
d fh
; o u tp u t
d a ta $ h i equ 4 R S232 d a ta lin e HI

d a t a $ lo w equ 0 R S232 d a t a l i n e LOW
lf$ s h ift$ k e y equ 80h

rt$ s h ift$ k e y equ lO h
co m m o d o re$ k ey equ 20 h
c o n tro l$ k e y equ 04h
ty p e $ lo w e r equ 0
ty p e $ u p p e r equ 1
ty p e $ s h ift equ 2
ty p e $ c n trl equ 3
ty p e $f i e I d equ 00000011b
bnkl equ 1
pageO equ 0
pagel equ 1
M M U $tbl$M m acro
db 3 f h , 3 f h , 7 f h , 3 e h , 7eh c o n f ig r e g 's
db 8 8
z 0 $ o n , c a rm o n $ K m o d e & mem
db p a g e O ,b n k l,p a g e lfb n k l p a g e r e g 's
endm
; ROM f u n c t i o n s
TJM P m acro X
rs t 2 ! db x
endm
TCALL rac ro X
rn v i l , x ! rs t 4
endm
RJM P m ac ro X
rs t 3 ! db x
endm
RCALL ira c ro X
rn v i l , x ! rs t 5
endm
FR $40 equ 2 o ffs e t to 4 0 c o l u m n ROM f u n c t i o n s
E R $ w r$ ch ar equ OOh D =char a u to ad v an ce

F R $ c u rso r$ p o s equ 04h B = r o w , C = c o lu m n
F R $ c u rs o r$ u p equ 08h
F R $ c u rso r$ d c w n equ OCh
F R $ c u rs o r$ le ft equ lO h
F R $ c u rso r$ rt equ 14h
FR $do$cr equ 18h
FR $CEL equ IC h
FR $CES equ 20 h
F R $ c h a r$ in s equ 24h
F R $ ch ar$ d el equ 28h
F R $ lin e $ in s equ 2C h
F R $ lin e $ d e l equ 3 Oh
F R $ c o lo r equ 34h B = c o lo r
F R $ a ttr equ 38h B = b it t o s e t / c l e a r , O b i t v a lu e
F R $ rd $ c h r$ a tr equ 3C h in D = ro w , E = c o l
o u t H = ro w , L = c o l , B = c h a r , C = a t t r ( r e a l )
F R $ w r $ c h r $ a tr equ 40h ; in D = ro w , E = c o l , B = c h a r , C = a t t r ( r e a l )
o u t H = ro w , L = c o l
F R $ rd $ c o lo r equ 44h
;F R $ w r$ c o lo r equ 48h
; equ 4C h
APPENDIXES 717
F R $ tr k $ s e c t equ 5 Oh
F R $ ch e c k $ C B M equ 52h
F R $ b e ll equ 54h
equ 56h
equ 58h
equ 5Ah
equ 5C h
equ 5E h
F R $ tr k $ 4 0 equ 6Oh
F R $ se t$ c u r$ 4 0 equ 62h
F R $ lin e $ p a in t equ 64h
F R $ sc re e n $ p a in t equ 66h
F R $ p rt$ m s g $ b o th equ 68h
F R $ p rt$ d e $ b o th equ 6Ah
F R $ u p d a te $ it equ 6Ch
'*
equ 6Eh
F R $ A S C II $ to $ p e t equ 7 Oh
F R $ c u r$ a d r$ 4 0 equ 72h
F R $ c u r$ a d r$ 8 0 equ 74h
F R $ lo o k $ c o lo r equ 76h
; equ 78h
F R $ b lk $ f ill equ 7Ah HL p a s s e d o n t h e s ta c k
F R $ b lk $ m o v e equ 7C h
F R $ c h a r $ in s t equ 7Eh
fix e d ROM l o c a t i o n s
R $ c m p $ h l$ d e equ 10Oh-6
R $ w rite $ m e m o ry equ 180h+0
R $ re a d $ m e m o ry equ 180h+3
R $ s e t$ u p d a te $ a d r equ 180h+6
R $ w a it equ 180h+9
R $ s ta tu s $ c o lo r$ tb l equ 1 0 0 0 h -2 4 6 -1 6
R $ c o lo r$ c o n v e rt$ tb l equ 1 0 0 0 h -2 3 0 -1 6
D is k t y p e b y te d e f i n i t i o n
b it 7 0= G C R , 1=MFM
if b it 7 is 1 (MFM)
6 C0=0, C l= l ( s i d e 2 #, 0 t o ( n / 2 ) - l o r n/'.
5 ,4 00=128, 01=256, 10= 512, 11=1024 b y te /s e c to r
3 ,2 ,1 d i s k t y p e (MFM)
0 s ta r tin g s e c to r # ( 0 o r 1)
if b it 7 is 0 (GCR)
6 unused ( s e t t o 0
)
5 ,4 a lw a y s 01 (2 5 6 b y te s e c to r s )
3 ,2 ,1 d i s k t y p e (GCR)
T ypeO = n o n e , s e t tra c k and s e c to r a s p assed
T y p e l = C64 ty p e d is k
T y p e 2 = C l 28 t y p e d i s k
0 u nused ( s e t to 0
)
MFM equ 1*128

CO equ 0*64 ; 2n d s id e s t a r t a t b e g in in g
Cl equ 1*64 ; 2n d s id e c o n tin u e s fro m f i r s t
C l$ b it equ 6
T ypeO equ 0*2 ; (MFM) t o p , b o t t c m t h e n n e x t t r a c k
(TRK # 0 t o ( 3 4 o r 3 9 ) )
T ypel equ 1*2 ; (MFM) t o p ( t r k 0 e v e n ) , b o t t c m ( t r k
(TRK # 0 t o ( 6 9 o r 7 9 ) )
Type 2 equ 2*2 ; (MFM) t o p TRK# 0 t o 3 9 , b o t t o n TRK#
(TRK # o n b a c k s t a r t a t 3 9 a n d g o t o 0)
718 COMMODORE 128
Type 7 equ 7*2 ; (MFM) p a s s t h e b y t e v a l u e s s u p p l i e d

; and ' s e c t
T ypeX equ 7*2
SO equ 0*1 ; s ta r t a t s e c to r 0
Si equ 1*1 ; s t a r t a t s e c to r 1
S i 28 equ 0*16
S256 equ 1*16
S512 equ 2*16
S1024 equ 3*16
d sk $ n o n e equ T ype0+S256 ; access to any s e c to r on th e

dsk$c64 equ T y p e l+ S 2 5 6
d s k $ c l2 8 equ Type2+S256
d ir$ tra c k equ 18 ; C64 d i s k d ir tra c k
\ 6 5 1 0 com m ands
v ic $ re s e t equ -1 ; r e b o o t C l 28
v ic $ in it equ 0 ; i n i t i l i z e t h e b io s8 5 0 2
v ic $ rd equ 1 ; r e a d o n e s e c t o r o f d a t a (2 5 6 b y t e s )
v ic $ w r equ 2 ; w r ite one s e c to r o f d a ta
v ic $ rd F equ 3 ; s e t - u p f o r f a s t r e a d (m any s e c t o r s )
v ic $ w rF equ 4 ; s e t- u p f o r f a s t w rite
v ic $ te s t equ 5 ; t e s t c u rr e n t d is k in d riv e
v ic $ q u e ry equ 6 ; g e t s t a r t s e c to rs and # s e c to r /trk
v ic $ p rt equ 7 ; p r in t d a ta c h a ra c te r
v ic $ frm t equ 8 ; fo rm a t a d i s k (1 5 4 1 )
v ic $ u s e r$ fu n equ 9
v ic $ R M $ rd equ 10 ; RAM d i s k r e a d
v ic $ R M $ w r equ 11 ; RAM d i s k w r i t e
; c o n tro l c h a ra c te rs
eo m equ OOh
b e ll equ 07h
bs equ 08h
If equ O ah
cr equ O dh
xon equ llh
x o ff equ 13h
esc equ lb h
APPENDIXES 719
is , ©
ATTR40 fa
BG CO LO R40
BLK PTR CN T
H
BLO CK PTRS
& a* O O —, A Tt
00 f! & fa J
p o
<
H C
as
oog zo, S 5 os as 72 ®
N
fa
2j
~
S
a “
BQ
jM < ( OS
U
S
^ fa, D
z
H
y Sos os 2 3 ss oES^SS; J u W J<(a! P S * !« 0

^ P£ h os
h nSywuP o CS Hi j
SZSlC iiH yU Q jw Cj«CQUU Q HZ H
0 3 " 3 S 2 z < o os os ea as OS 3 Z
CQHUUUO o u u a o a a o a a fa fa fa fa fa fa fa fa fa- K
240D
240E
o © 2 Tt oo fS oo «
© ®
© ©
H © ® i/) c* t- »-h T( X h-
3C09
0020
r. -t o « o n © P Tt
U s o t x fa © fa © © © ©
© fa c-j
rtO O N N U fa © fa © © © ©
H
fa
fa fa
2 ® * ^^
a -T fa P es
O
H
Tf sa g VO 0j 05 PH fa «
BLOCKEND
fisS z o u U O C4 o o o «inU
< o w s ^ 05 z § w us o» ps O C/3 O Tt
BANKED
On “ « g s O S -3 o C/3 fa 05 £5 yo ps o Sj
z
ATEMP
H fa O P P
BLACK
1 . .“z.
S
O •< H H u U U U U PS H Z 5
BELL
o -B 5SZ 00«> -C H »l tc PC PC 05 C5 m ps a g
aa P u u u p U U U U O O Q Q fa fa fa fa fa fa fa &
0\ ff)f i ff) P<) - . © ^ ®
FFFF
FD43
2 © os rf fa t*- © P ? n © s © o o © T t© o o
0007
0000
3 S ° °
3C07
P, Tt © © »-< © © © © © ^ [^{^©©r^rjrî-i©
sj
p^j
© rt r f © ©
c r) m o o
©
©
»T)© © © P ©
© © © © fa c<
©
©
O S © © T t © © © © ©
©©©<S©©©©©
S
o oo fa
OP fa
a
OS fa ? aa £4
O ^0= O
S £4 w [— o o CQ
H f - ^ 0
" 1 2S g
^ K/ j J
© z g Z O ^ ® <5 fa
*
p
e> <c PS S3 U ps^sa^sa
30
PS —i gflZ O >JH K ^ o PS o ^ w S p ffl fa
PS Q PS H q o W O Z B ? e fa§« u
< PSssasaSHQa 55 a P S p
fa « H o |s f a H a < ( £
PS vo g hJ ^ ^
w ffl
U h-3fa P<c/3- I
5 fa P P U S
hi fa - j S S z z k h K V bi PS u u
z 3 o * ^ ooooopcw t» C/3 C/3 C< PH & & cm p£ ca PC g
sa ea p o u U U U U U U flfl QQQ w w ts< ta fa fafa fa fa fa £
ALTKEY
BDCOLOR4G
BIOS8502
BANKl
*-H ©
© © 1H
CJ © 1-H f©a § o S a < M '9 -i« rj!§
S < s o o o - i o g © i-H »N
R
W 1/1 «5 U <s
IT)
Tt Tt
H so
 © T t © p © © © feUo © © © © ©© © © u
© m © n © fa©©©©<s©ta © © © fa © © © © © © © © © f'i
z
fa H ^
4000 BANKOFREE FF02
o
BLOCKBUFFER
H
0050
Jj 05
2600
Oh p < PS £
H H t» Z O ^ z M <
^ Z a o x £2 U H a - O pHfa
fa
240F
oo <( g
1 p p s OS
as
OCQXh
W ct O 53 S SS ^
a o « < «
fa
fa ft. < W H c/s fa ffi P g P5 a
3000 BIOS02
z s h h 3 ^ ^ UteCBUO 3
o D <! -< 03 C/DCC c/3 sc •< OS PS PS PS as PS P
CSCOCQUUUU <J u a a a QQQ fa fa aa
so oo r - eo r - © fa P -H p O CQ © © u fa u u
© © OS H Q H oo © Z2 © r<5
3400
CQ t^> ©
© © ©P Tt 5T ©
© © © fa fS Q ©
T t © CQ oo
c* © © ©
© © © © © ©
© © © © © ©
© ®
o p
fa
fa
Q
O PC H
aa gH Z
o Q C/3
BANKPARMBLK
irj td
ALPHATOGGLE
< © o
s § fa< H
*S u
BLKUNLDPTR
fa OS 2 ^ So h a Z
BGCOLOR30
O o Z H
o ^ a > os S ^ Z O fa fa
Z D
-<! O a o *fa co o5o S
n ^ fa c
d | Q ° §
fa 05 < c/3 S « ^ ~
PS PS M ^ < ps os S a o z
us5 2 p y z 5 S S
fa H o - c J<« «C 3< oS w p o „ H o f a 3 = P P & 2o H
BANKO
u^faoa^^KPS « * fa fa
S S Z Z « ^ ^ oo p 5 Cfi5 X K | « ° B U P U J p fa m P W
S S p u B ^ S z o o O O P < < C/3 C/3 C/3 <; O- <PSPSCS0SC<0! at os OS PS ps
W facq up p op p o P P U Q Q f l Q Q Q Q fa fafafafafafafafa fa fa fa fa o
1
9CSJ)^
© r©- ©©<©33 n©©Q ^5 M
W x© 5o So ©
§ ©o f©S 5© ©© S® xW© §S©©©^S©® tS
©^©r^ ^t T
t-r< ' *,r-©
'^‘r2ro
) ®r•>r-so
'' 0' 00 0so
«^ So^ ©
sso !^
©
FFOl
3C04
M Oi*
003D
2416
2400
U © © © T t V © © > S T t J £ ? ^ ^ © © ^ © ^) £ ©w© P ^ ©W | »t
Q es
H©“»i
©©es©
i**
©“©© ©>>—
©<
©i-©© ©© © ©© ©^ © ©ifTi
^ © 0 n N H © © Q M H Q N 0 0 Q © Q © © f a 0 f a <*© © © © © 0 © 0 0 © <■**>
720 COMMODORE 128
o
a! « n z >
g i o < !>
| s o h ! 3 £*i
jc oa |ow^ gS Bu 3S <w W
S
td pa § o X
c u w o ^ w g w w o s u q o '*;
^ ^ R ^ o a c s
w o S a * < ; o S s s
s a ii J ^ « 5 0 s . f t . D . a£ as a!
p C£o
o rj „ WH Tt
B*. 'S© O
___„ . 'M
— $“ ' o ! S 3 u o o
q a Q O o « o S 5 o
U- ta tiim o Q o Q c J i© fa O Q
ta
23
H
H
Z C*
O ta
ta ta . o ^
O 05 > 00 Z H 3
O5 ta
Z E- Q < O ^ H QS ^
a z * W U Q H O 53 ,_ H «< 9 H
y c*htah *< h ^ 55 o £ 5 £ t£4
o
z® £h j| *C 2 Z h u g a ^ >
g S e g l i l S s i l S g i g - S5
W=r*g^
g
w^2« g s 9 g g u Q f t , K H ^ Q ^
Q os a 3
£ g g g £ £ i S S 3 S 8 £ s io X _ „ ta
_, O
« § § “ &£ a a a a a 3 3 ®
ta^^ :*ltata S O ta c i< a < Q !;o !!a !!tfc * C£ CO U
GO CO CO
° - ^ E o < § S s < g < ^ S § § if'- ’T* * : 5 w O^ Ho O« W

O O ®g ^§ Ot -O CS 2f ®
t f &t Q2
P ’t P p o o ^ T f O ’t r r f a U ^ ^ ©
fct<MCnfa®©Qc*®<sc^©f<5QQ© U. © —<© U- © ©
© © ©
© ©
o ©
o s
Q ^b o© ©o ^b ^b ^
b ^
b
X
ta ?
O
Z ta HO u
5 ^ cfl
ta t/3
h
H g O H « 5 s fc n L^!
owawSzz><
.B
« lS£h?S?S£SsW« W
P UUiOSH
Ut*!0SH ™ O SQ «
I/} ^ O ta ta W rN^ ^ »•■*^ M t-M ^ 00 /r-}. /rî
S | N c m h g t q g Q ^ o o o o i l b e S e o l , 1 - Q ® n o a - iw$^fff eSet taNM ^ S i0a taso

p | ^ w S § l ? l 5 | f e « l l l g w 2 g i 2 S g ^ ? § g § §«O| O ^| | £r..§© g g QSQS
b b N Q o Q Q N H M b o Q Q C i Q b o o E i i o o o t a M N o Q o o Q b Q b o o o U- U-
APPENDIXES 721
APPENDIX L
COMMODORE 128 SYSTEM SCHEMATICS

The following eight pages contain the full system schematics for the Commodore 128.
Each two-page spread represents one full-size engineering schematic sheet. For easier
readability, the right edge of the left-hand page and the left edge of the right-hand page
have portions of the schematic that are duplicated. This overlap is provided so you can
read the circuit diagram from either half of the two-page spread, then move to the
adjacent page and pick up where you left off from the point where the opposite page
ends. The arrow at the top of each page provides a frame of reference to mark the
portion of the diagram that is overlapped.
722 COMMODORE 128
7 6 5
APPENDIXES 723
4 3 2 1
724 COMMODORE 128
A U D IO O U T r
ID A C C
7 A 5 -T A II
CLR BN K
L
I t -
kA 7 2 18 SA7y
\A 4
lc
10>
14
SA<0/
SA5,
6 l£ SA4,
vA;- <? U55
sA2 7 13 SA2,
15 SAI/
kA0 -i 17 5A0/
D/ft OE
VAI4
VA15
VMA® -VMA7
7 6 5
APPENDIXES 725
726 COMMODORE 128
APPENDIXES 727
CN 13
LED
R42
l8o -n_
R32
I/O CONNECTOR
m te ;:
/\ L O W E N D ONLY
fe \ H I G H O J D ONLY
I/O — C o n tro l P o rts , K e y b o a rd ,
U s e r P o rt, S e ria l P o rt, C a s s e tte P o rt
S C S C H E M A T IC 3 1 0 3 7 8
728 COMMODORE 128
T,
OPEN- 1 2 8 k ROM
CLOSE - 25G>K ROM
- 0 1 ONLY
- 0 1 ONLY | R A S
~dwT
G C A 50
MAffl - M A T
D 0- Dl
8 7 6 5
APPENDIXES 729
C /PS . o t p f
RAM , ROM
S C H E M A T IC 3 1 0 3 7 8
3 2 1
r
GLOSSARY
This glossary provides brief definitions of frequently used computing terms.
Acoustic C oupler o r Acoustic M odem : A device that converts digital signals to

audible tones for transmission over telephone lines. Speed is limited to about 1200
baud, or bits per second (bps). Compare Direct Connect Modem.
Address: The label or number identifying the register or memory location where a unit
of information is stored.
Alphanum eric: Letters, numbers and special symbols found on the keyboard, exclud
ing graphic characters.
A L U : Arithmetic Logic Unit. The part of a Central Processing Unit (CPU) where
mathematical operations are performed.
A n im ation: The use of computer instructions to simulate motion of an object on the
screen through gradual, progressive movements.
A rra y: A data storage structure in which a series of related constants or variables is stored
in consecutive memory locations. Each constant or variable contained in an array
is referred to as an element. An element is accessed using a subscript. See
Subscript.
A S C II: Acronym for American Standard Code for Information Interchange, which is a
seven-bit code used to represent alphanumeric characters. It is a useful communi
cation code for such things as sending information from a keyboard to the
computer, and from one computer to another. See Character String Code.
Assembler: A program that translates assembly language instructions into machine
language instructions.
Assembly Language: A machine-oriented language in which mnemonics are used to
represent each machine language instruction. Each CPU has its own specific
assembly language. See CPU and Machine Language.
Assignment S tatem ent: A BASIC statement that sets a variable, constant or array
element to a specific numeric or string value.
Asynchronous Transmission: A communication scheme in which data characters are
sent at time intervals, independent of the system clock. Limits phone line trans
mission to about 2400 baud (bps). See Synchronous Transmission.
A ttack: The rate at which the volume of a musical note rises from zero to peak volume.
Background C olor: The color of the portion of the screen that the characters are
placed upon.
731
732 COMMODORE 128
BASIC: Acronym for Beginner’s All-purpose Symbolic Instruction Code.

Baud: A unit of serial-data transmission speed. The term was originally used for
measuring telegraph transmission speed. Three hundred baud is approximately
equal to a transmission speed of 30 bytes, or characters, per second.
Binary: A base-2 number system. All numbers are represented as a sequence of Os and
Is.
Bit: The abbreviation for Binary diglT. A bit is the smallest unit in a computer’s
memory. Each binary digit can have one of two values, 0 or 1. A bit is referred to
as set or “ on” if it equals 1. A bit is clear or “ off” if it equals 0.
B it C ontrol: A means of transmitting serial data in which each bit has a significant
meaning and a single character is surrounded with start and stop bits.
B it Map Mode: An advanced graphic mode in the Commodore 128 in which you can
control every pixel on the screen.
B order Color: The color of the edges around the screen.
Branch: To jum p to a section of a program and execute it. GOTO and GOSUB are
examples of BASIC branch instructions.
Bubble M em ory: A relatively new type of computer memory, it uses tiny magnetic
“ pockets” or “ bubbles” to store data.
Burst Mode: A special high-speed mode of communication between a 1571 disk drive
and a C l 28 computer, in which information is transmitted at many times the speed
of the 1541 Disk Drive.
Bus: Parallel or serial lines used to transfer signals between devices. Computers are
often described by their bus structure (i.e., S-100-bus computers, etc.).
Bus N e tw o rk : A system in which all stations or computer devices communicate by
using a common distribution channel or bus.
Byte: A group of 8 bits that make up the smallest unit of addressable storage in a
computer. Each memory location in the Commodore 128 contains 1 byte of
information. One byte is the unit of storage needed to represent one character in
memory. See Bit.
C a rrie r Frequency: A constant signal transmitted between communicating devices
that is modulated to encode binary information.
C haracter: Any symbol on the computer keyboard that is printed on the screen.
Characters include numbers, letters, punctuation and graphic symbols.
C h aracter M em ory: The area in Commodore 128’s memory that stores the encoded
character patterns that are displayed on the screen.
C h aracter Set: A group of related characters. The Commodore 128 character sets
consist of upper case letters, lower case letters and graphic characters.
C h aracter String Code: The numeric value assigned to represent a Commodore 128
character in the computer’s memory.
Chip: A miniature electronic circuit that performs a computer operation such as
graphics, sound and input/output.
Clock: The timing circuit for a microprocessor.
Clocking: A technique used to synchronize the sending and receiving of data that is
modulated to encode binary information.
C oaxial Cable: A transmission medium, usually employed in local networks.
GLOSSARY 733
Collision D etection: Determination of collision between two or more sprites or

between sprites and data.
C o lo r M em ory: The area in the Commodore 128’s memory that controls the color of
each location in screen memory.
C om m and: A BASIC instruction used in direct mode to perform an action. See Direct
Mode.
C om piler: A program that translates a high-level language, such as BASIC, into
machine language.
C om posite M onitor: A device used to provide the C128 40-column video display.
C om puter: An electronic, digital device that inputs, processes, and outputs information.
Condition: Expression(s) between the words IF and THEN, in an IF . . . THEN
statement, evaluated as either true or false. The conditional IF . . . THEN
statement gives the computer the ability to make decisions.
C oordinate: A single point on a grid having vertical (Y) and horizontal (X) values.
C ounter: A variable used to keep track of the number of times an event has occurred in
a program.
C P U : Acronym for Central Processing Unit, the part of the computer containing the
circuits that control and perform the execution of computer instructions.
Crunch: To minimize the amount of computer memory used to store a program.
Cursor: The flashing square that marks the current location on the screen.
D ata: Numbers, letters or symbols that are input into the computer and are to be
processed.
Database: A large amount of related data stored in a well-organized manner. A
database management system is a program that allows access to the information.
D a ta Link Layer: A logical portion of data communications control that mainly
ensures that communication between adjacent devices is error-free.
D a ta Packet: A means of transmitting serial data in an efficient package that includes
an error-checking sequence.
D a ta Rate or D a ta Transfer Rate: The speed at which data issent toa receiving
computer— given in baud, or bits per second (bps).
Datassette: A device used to store programs and data files sequentially on tape.
Debug: To correct errors in a program.
Decay: The rate at which the volume of a musical note decreases from its peak value to
a midrange volume called the sustain level. See Sustain.
D ecrem ent: To decrease an index variable or counter by a specific value.
Dedicated Line o r Leased Line: A special telephone line arrangement supplied by
the telephone company, and required by certain computers or terminals, whereby
the connection is always established.
D e la y Loop: An empty FOR . . . NEXT loop that slows the execution o f a
program.
D ia l-U p Line: The normal switched telephone line that can be used as a transmission
medium for data communications.
D igital: Of or relating to the technology of computers and data communications
where all information is encoded as bits of Is or Os that represent on or off
states.
734 COMMODORE 128
Dimension: The property of an array that specifies the size and direction along an axis
in which the array elements are stored. For example, a two-dimensional array has
an X-axis for rows and a Y-axis for columns. See Array.
D ire ct Connect M odem : A device that converts digital signals from a computer into
electronic impulses for transmission over telephone lines. Contrast with Acoustic
Coupler.
D irect Mode: The mode of operation that executes BASIC commands immediately after
the R E T U R N key is pressed. Also called Immediate Mode. See Command.
Disable: To turn off a bit, byte or specific operation of the computer.
Disk D rive: A random access, mass-storage device that saves and loads files to and
from a floppy diskette.
Disk O p eratin g System: A program used to transfer information to and from a disk.
Often referred to as a DOS.
D uration: The length of time a musical note is played.
Electronic M ail, o r E-Mail: A communications service for computer users in which
textual messages are sent to a central computer, or electronic “ m ailbox,” and
later retrieved by the addressee.
Enable: To turn on a bit, byte or specific operation of the computer.
Envelope G enerator: Portion of the Commodore 128 that produces specific wave
forms (sawtooth, triangle, pulse width and noise) for musical notes. See Waveform.
EPROM: A PROM that can be erased by the user, usually by exposing it to ultraviolet
light. Abbreviation for Erasable Programmable Read Only Memory. See PR O M .
E rro r Checking or E rro r Detection: Software routines that identify, and often
correct, erroneous data.
Execute: To perform the specified instructions in a command or program statement.
Expression: A combination of constants, variables or array elements acted upon by
logical, mathematical or relational operators that return a numeric value.
File: A program or collection of data treated as a unit and stored on disk or
tape.
F irm w are: Computer instructions stored in ROM, as in a game cartridge.
Frequency: The number of sound waves per second of a tone. The frequency corre
sponds to the pitch of the audible tone.
Full-D uplex Mode: In this mode, two computers can transmit and receive data at the
same time.
Function: A predefined operation that returns a single value.
Function Keys: The four keys on the far right of the Commodore 128 keyboard. Each
key can be programmed to execute a series of instructions. Since the keys can be
SHIFTed, you can create eight different sets of instructions.
GCR: The abbreviation for Group Code Recording, a method of storing information on
a disk. The 1541 and 1571 disk drives can read, write and format GCR disks.
G raphic Characters: Nonalphanumeric characters on the computer’s keyboard.
Graphics: Visual screen images representing computer data in memory (i.e., charac
ters, symbols and pictures).
Grid: A two-dimensional matrix divided into rows and columns. Grids are used to
design sprites and programmable characters.
GLOSSARY 735
H alf-D u p lex Mode: In this mode, data can be transmitted in only one direction at a
time; if one device is sending, the other must simply receive data until it’s time for
it to transmit.
H ard w are : Physical components in a computer system, such as the keyboard, disk
drives and printer.
H exadecim al: Refers to the base-16 number system. Machine language programs are
often written in hexadecimal notation.
H o m e: The upper-left corner of the screen.
1C: The abbreviation for Integrated Circuit. A silicon chip containing an electrical
circuit made up of components such as transistors, diodes, resistors and capaci
tors. Integrated circuits are smaller, faster and more efficient than the individual
circuits used in older computers.
Increm ent: To increase an index variable or counter with a specified value.
Index: The variable counter within a programming loop.
Input: Data fed into the computer to be processed. Input sources include the keyboard,
disk drive, Datassette or modem.
Integer: A whole number (i.e., a number containing no fractional part), such as 0, 1,
2, etc.
Interface: The point of meeting between a computer and an external entity, whether an
operator, a peripheral device or a communications medium. An interface may be
physical, involving a connector, or logical, involving software.
I/O : The abbreviation for Input/Output. Refers to the process of entering data into the
computer, or transferring data from the computer to a disk drive, printer or storage
medium.
Keyboard: Input component of a computer system.
Kilobyte (K): 1024 bytes.
Local N e tw o rk : One of several short-distance data communications schemes typified
by common use of a transmission medium by many devices at high-data speeds.
Also called a Local Area Network, or LAN.
Loop: A program segment executed repetitively a specified number of times.
Machine Language: The lowest-level language the computer understands. The com
puter converts all high-level languages, such as BASIC, into machine language
before executing any statements. Machine language is written in binary form, which
a computer can execute directly. Also called machine code or object code.
M a trix : A two-dimensional rectangle with row and column values.
M em ory: Storage locations inside the computer. ROM and RAM are two different
types of memory.
M em o ry Location: A specific storage address in the computer. There are 131,072
memory locations (0-131,071) in the Commodore 128.
M FM : The abbreviation for Modified Frequency Modulation, a method o f storing
information on disks. The 1571 disk drives can read and write to MFM disks.
Microprocessor: A CPU that is contained on a single integrated circuit (IC). M icropro
cessors used in Commodore personal computers include the 6510, the 8502 and
the Z80.
Mode: A state of operation.
736 COMMODORE 128
M odem : The acronym for MOdulator/DEModulator. A device that transforms digital

signals from the computer into analog electrical impulses for transmission over
telephone lines, and does the reverse for reception.
M onitor: A display device resembling a television set but with a higher-resolution
(sharper) image on the video screen.
M otherboard: In a bus-oriented system, the board that contains the bus lines and edge
connectors to accommodate the other boards in the system.
M u lti-C o lo r B it Map Mode: A graphic mode that allows you to display one of four
colors for each pixel within an 8 by 8 character grid. See Pixel.
M u lti-C o lo r C h aracter Mode: A graphic mode that allows you to display four
different colors within an 8 by 8 character grid.
Multiple-Access N e tw o rk : A flexible system by which every station can have access
to the network at all times; provisions are made for times when two computers
decide to transmit at the same time.
N ull String: An empty character (“ ” )• A character that is not yet assigned a character
string code. Produces an illegal quantity error if used in a GET statement.
O ctave: One full series of eight notes on the musical scale.
O p eratin g System: A built-in program that controls everything a computer does.
O p e rato r: A symbol that tells the computer to perform a mathematical, logical or
relational operation on the specified variables, constants or array elements in the
expression. The mathematical operators are + , *, / and f . The relational
operators are < , = , > , < = , > = and < > . The logical operators are AND,
OR NOT and XOR.
O rd e r of Operations: Sequence in which computations are performed in a mathemat
ical expression. Also called Hierarchy of Operations.
Parallel Port: A port used for transmission of data 1 byte at a time using 8 data lines,
one for each bit.
P a rity Bit: A 1 or 0 added to a group of bits that identifies the sum of the bits as odd or
even, for error checking purposes.
Peripheral: Any accessory device attached to the computer such as a disk drive, printer,
modem or joystick.
Pitch: The highness or lowness of a tone that is determined by the frequency of the
sound wave. See Frequency.
Pixel: Computer term for picture element. Each dot that makes up an image on
the screen is called a pixel. Each character on the screen is displaced within
an 8 by 8 grid of pixels. The entire screen is composed of a 320 by 200 pixel
grid. In bit-map mode, each pixel corresponds to a bit in the com puter’s
m em ory.
Pointer: A register used to indicate the address of a location in memory.
Polling: A communications control method used by some computer/terminal systems
whereby a “ master” station asks many devices attached to a common transmis
sion medium, in turn, whether they have information to send.
Port: A channel through which data is transferred to and from the CPU.
P rinter: Peripheral device that outputs the contents of the computer’s memory onto a
sheet of paper. This paper is referred to as a hard copy.
GLOSSARY 737
Program : A series of instructions that direct the computer to perform a specific task.
Programs can be stored on diskette or cassette, reside in the computer’s memory,
or be listed on a printer.
P rogram Line: A statement or series of statements preceded by a line number in a
program. The maximum length of a program line on the Commodore 128 is 160
characters.
Program m able: Capable of being processed with computer instructions.
PROM : The acronym for Programmable Read Only Memory. A semiconductor memory
chip whose contents can be changed.
Protocol: The rules under which computers exchange information, including the
organization of the units of data to be transferred.
Random Access M em o ry (R A M ): The programmable area of the computer’s memory
that can be read from and written to (changed). All RAM locations are equally
accessible at any time in any order. The contents of RAM are erased when the
computer is turned off.
Random N um ber: A nine-digit decimal number from 0.000000001 to 0.999999999
generated by the RaNDom (RND) function.
Read O nly M em o ry (R O M ): The permanent portion of the computer’s memory. The
contents of ROM locations can be read, but not changed. The ROM in the
Commodore 128 contains the BASIC language interpreter, character-image pat
terns and the operating system.
Register: Internal storage compartments with the microprocessor that communicate
between system ROM, RAM, and themselves.
Release: The rate at which the volume of a musical note decreases from the sustain
level to 0.
Rem ark: Comments used to document a program. Remarks are not executed by the
computer, but are displayed in the program listing.
Resolution: The fineness of detail of a displayed image, determined by the density of
pixels on the screen.
RGBI M onitor: A high-resolution display device necessary to produce the C128 80-
column screen format. RGBI stands for Red/Green/Blue/Intensity.
Ribbon Cable: A group of attached parallel wires, usually made up of 25 lines for
RS-232 communication.
Ring N e tw o rk : A system in which all stations are linked to form a continuous loop or
circle.
RS-232: A recommended standard for electronic and electromechanical specifications for
serial communication. The Commodore 128 parallel user port can be treated as a
serial port if accessed through software, sometimes with the addition of an
interface device.
Screen: A video display unit, which can be either a television or a video monitor.
Screen Code: The number assigned to represent a character in screen memory. When
you type a key on the keyboard, the screen code for that character is entered into
screen memory automatically. You can also display a character by storing its
screen code directly into screen memory with the POKE command.
Screen M em ory: The area of the Commodore 128’s memory that contains the infor
mation displayed on the video screen.
738 COMMODORE 128
Serial Port: A port used for serial transmission of data; bits are transmitted one bit
after the other over a single wire.
Serial Transmission: The sending of sequentially ordered data bits.
Softw are: Computer programs (set of instructions) stored on disk, tape or cartridge that
can be loaded into random access memory. Software, in essence, tells the com
puter what to do.
Sound Interface Device (S ID ): The MOS 6581 sound synthesizer chip responsible
for all the audio features of the Commodore 128.
Source Code: A nonexecutable program written in a higher-level language than
machine code. A compiler or an assembler must translate the source code into an
object code (machine language) that the computer can understand.
Sprite: A programmable, movable, high-resolution graphic image. Also called a M ov
able Object Block (MOB).
Standard C h aracter Mode: The mode the Commodore 128 operates in when you
turn it on and when you write programs.
S ta rt Bit: A bit or group of bits that identifies the beginning of a data word.
S tatem ent: A BASIC instruction contained in a program line.
Stop Bit: A bit or group of bits that identifies the end of a data word and defines the
space between data words.
String: An alphanumeric character or series of characters surrounded by quotation
marks.
Subroutine: An independent program segment separate from the main program that
performs a specific task. Subroutines are called from the main program with the
GOSUB statement and must end with a RETURN statement.
Subscript: A variable or constant that refers to a specific element in an array by its
position within the array.
Sustain: The midranged volume of a musical note.
Synchronous Transmission: Data communications using a synchronizing, or clock
ing, signal between sending and receiving devices.
Syntax: The grammatical rules of a programming language.
Tone: An audible sound of specific pitch and waveform.
Transparent: Describes a computer operation that does not require user intervention.
V ariable: A unit of storage representing a changing string or numeric value. Variable
names can be any length, but only the first two characters are stored by the
Commodore 128. The first character must be a letter.
Video Interface C ontroller (V IC ): The MOS chip (8564) responsible for the 40-column
graphics features of the Commodore 128.
Voice: A sound-producing component inside the SID chip. There are three voices
within the SID chip so the Commodore 128 can produce three different sounds
simultaneously. Each voice consists of a tone oscillator/waveform generator, an
envelope generator and an amplitude modulator.
W aveform : A graphic representation of the shape of a sound wave. The waveform
determines some of the physical characteristics of the sound.
W ord: Number of bits treated as a single unit by the CPU. In an 8-bit machine, the
word length is 8 bits; in a 16-bit machine, the word length is 16 bits.
INDEX
A BASIN, 423-433 C l28 BASIC, 219-220
Abbreviations, 670-673 BCC, 163 C64 BASIC, 222
ABS, 73 BCS, 163 multi-color, 233-237
Accumulator, 127-129 BEGIN/BEND, 28-29 programmable characters, 230-233
addressing, 138 BEQ, 164 screen location, 224, 234
loading, 147-148 BIOS (Basic Input Output System), screen memory data, 224—225
ACPTR, 422-423 486-489, 500, 677-683, 704- standard, 223-233
ADC, 162 705 CHKIN, 429-430
Addition, 19 BIT, 153-154, 164 CHR$, 74, 660-662
Addressing Bit map mode 112, 221, 222 CIA (6526) chip, 611-623
absolute, 138-139, 143 data, 241-243 control registers 622-623
accumulator, 138 80-column (8563) chip, 314-320 description, 611, 618-622
immediate, 138 multi-color, 243-245 electrical characteristics, 613-615
implied, 139 standard, 239-243 interface signals, 615-616
indexed, 141-143, 144-145 standard sprites, 283-284 interrupt control, 621-622
indirect, 143-145 video matrix, 240-241, 244 serial port, 620-621
modes, 137, 141-142 Bits timing, 616-617, 618-620
relative, 140-141 masking, 97-98 CINT, 410, 414
16-bit, 133-135 16-bit addressing, 133-135 CIOUT, 423
table, 161-179 values in a byte, 98-99 CIRCLE, 32-33, 115-116
zero-age, 139, 142 BLOAD, 29 CKOUT 430-431
ALT Mode, 497 BMI, 164 CLALL, 439
AND, 152, 162 BNE, 164 CLC, 166
APPEND, 27 BOOT, 29-30, 446-447 CLD, 166
Arithmetic BOX, 30, 113-114 CLI, 166
instructions, 151-152 BPL, 165 CLOSE, 33, 428-429
operations, 18-20 BRK, 165 CLOSE ALL, 443
Arrays, 13, 16-18 BSAVE, 31 CLR, 33
ASC, 73, 660-662 BSOUT, 433 CLRCH, 431-432
ASL, 163 Buffer CLV, 166
Assembler, 126-127 control block, 685 CMD, 33
ATN, 73 routine, 93 CMP, 167
AUTO, 27 BUMP, 73-74 CMPSTA, 456
Bus COLLECT, 34
B architecture, 560-562 COLLISION, 34, 267
BACKUP 27-28 color data, 562 COLOR, 34-35, 116-117
BANK, 28 display, 562 Color mode
BASIC expansion, 635-637 extended background, 237-239
advanced programming techniques, loading, 567-568 memory map, 664
103-107 multiplexed address, 561 sprites, 283-285
color RAM in C128, 218 processor, 560 See also Memory, color RAM
color RAM in C64, 218-219 serial, 633-634 Commands, 12
C l28 bit map mode, 221 shared address, 561-562 basic, 27-72
crunching of programs, 95 translated address, 560-561 CP/M, 481-482, 483
C64 bit map mode, 222 BVC, 165 format, 25-27
C64 character modes, 222 BVS, 165 graphics, 113-122
entering machine language subrou machine language monitor, 186-194
tines through, 198-202 C sprites, 267-270
error messages, 644-647 Cassette connector, 398 summary, 674—675
intelligent use of, 97 CATALOG, 31 See also specific commands
mixed with machine language, CHAR, 31-32, 115 Commodore 128. See C l28 Mode
198-205 Character mode Commodore 64. See C64 Mode
placement of machine language accessing character ROM, 229 Complex Interface Adapter. See CIA
routines with, 203-205 character memory, 226-229, (6526) chip
relocating, 106 234-235 CONCAT, 35-36
screen memory in C l28, 215-217 color data, 225-226, 235-237 C128 Mode, 2-3, 5
screen memory in C64, 217 color memory, 226 BASIC bit map mode, 221
739
740 IN D EX
C l28 Mode (continued) DEY, 168 FETCH, 43

character memory, 219-220 DIM, 38 Files
character set availability, 222 DIRECTORY, 38-39 CP/M, 479-481
color RAM in BASIC, 218 Directory, 375 creating and storing, 376-378
CP/M disk format, 493-494 Disk drive disk drive, 374-376
memory map, 502-540 copies, 482, 485 merging, 106-107
ROM cartridge, 471-472 device number, 378 FILTER, 43, 337-338, 348-351
screen memory in BASIC, 215-217 directory, 375 FNxx, 74
switching from mode to mode, 6 formatting, 372-373 FOR/TO/STEP/NEXT, 44
using C64 function key values, 95 replacing files or programs, 374 FRE, 75
Configuration Register. See Memory retrieving files or programs, 375-376 Function keys
Connectors, 652-657 saving programs, 373-374 changing, 95
Constants, 12-15 verifying files or programs, programming, 94
floating-point, 13 374-375 using C64 values, 95
integer, 13 Disk Parameter Block, 685 Functions, 72-86
string, 14-15 Division, 19-20 errors, 101
CONT, 36 DLCHR, 450 user, 685-704
Control codes, 666-668 DLOAD, 39 See also specific functions
COPY, 36 DMA CALL, 444-445
COS, 74 DO/LOOP/WHILE/UNTIL/EXIT, 39-40 G
CP/M, 676 DOPEN, 40 GET, 44—45
BIOS routines, 677-683, 704-705 DOS errors, 101, 648-651 G E T #, 45
calling user function, 702-704 Dot matrix printer, 378-379 GETCFG, 452
commands, 481-482, 483 DRAW, 41, 117-118 GETIN, 438-439
control characters for line editing, Drive Table, 684-685 GETKEY, 45
482, 484 DSAVE, 41 G 064, 45
copies of disks and files, 482, 485 DVERIFY, 42 GOSUB, 45-46
disk organization, 491-495 GOTO/GO TO, 46
enhancements, 479 E GRAPHIC, 46, 119
files, 479-481, 482, 485 Editor. See Screen editor Graphics
keyboard scanning, 496-497 80-column (8563) chip, 292-334 commands, 113-122
memory map, 709-720 bit map mode, 314-320 power behind, 208-263
mode, 4 Block Write and Block Copy, programming, 110-122
requirements for system, 478 312-313, 333 system, 215-223
switching from mode to mode, 7 characters, 296-297, 301-304, 325, GSHAPE, 46, 119-120
system layout, 486-487 328, 333
system memory organization, cursor, 313-314, 326 H
489-491 display, 299-301, 326, 327-334 Hardware
system operations, 500 frames, 297-299 components, 4-5
system release, 8 RAM, 304-305, 309-313, 327-334 specifications, 556-641
CPX, 167 registers, 304-309, 324-334 system architecture, 557-558
CPY, 167 scrolling of screen, 320-323, See also specific components
C64 Mode, 3, 5, 444 328-331 HEADER, 47
BASIC bit map mode, 222 8502 microprocessor, 569-574 HELP, 47
BASIC character modes, 222 description, 569 HEXS, 75
color RAM in BASIC, 218-219 electrical specification, 569-571 Hexadecimal notation, 136-137
CP/M disk format, 492-493 processor timing, 571-574
input/output assignments, 546-554 END, 42 I
memory map, 540-554 ENVELOPE, 42, 336-337, 347-348 IF/THEN/ELSE, 47-48
ROM cartridge, 472 Environmental specifications, 568 INC, 169
screen memory in BASIC, 217 EOR, 152, 168 INDFET, 454-455
switching from mode to mode, 7 Errors INDSTA, 455
using function key values, 95 BASIC messages, 644-647 INPUT, 48
Cursor, 313-314, 326 DOS, 101, 648-651 IN PU T#, 49
functions, 101 Input/output, 5, 372-400, 727
D logic, 99 BIOS (Basic Input Output System),
Daisy wheel printer, 378 syntax, 99 486-489, 500, 667-683, 704-
DATA, 36 tracing of, 101 705
Datassette, 389-390 trapping of, 100 controller ports input, 390-393
Data structures, 684-685 See also Programming, debugging C64 assignments, 546-554
DCLEAR, 37 Escape codes, 669 Datassette output, 389-390
DCLOSE, 37 Exponentiation, 20 disk drive, 372-376
Debugging. See Programming Expressions, 18 files, 376-378
DEC, 168 arithmetic, 18 modem output, 381
DEF FN, 37 string, 24 output control, 393-394
DELETE, 38 pinouts, 394-400
Device numbers, 457 F printer output, 378-381
DEX, 168 FAST, 43 RS-232 channel, 382-388
IN D E X 741
Input/output (continued) routine, 93-94 screen (RAM), 215-218

screen outpuut, 388-389 LOCATE, 51, 120 16K video banks, 210-212
INSTR, 75 LOG, 77 64K RAM banks, 208-210
Instructions Logic split-screen mode, 246-247
arithmetic, 151-152 errors, 99 storage, 12-13, 148
branching, 154-156 instructions, 151, 152-153 switching banks, 459-460
compare, 150-151 Logical operators, 21-22 system organization, 562-567
counter, 148-149 LSR, 171 See also Character mode
entering machine language in Memory Management Unit (MMU).
monitor, 183-184 See Memory, management
jum p, 159-160 Machine language, 124-179 MEMTOP, 419-420
logical, 151, 152-153 character memory, 223 Menu, 92
machine language, 145-179, color RAM, 219 MFM. See Modified Frequency
183-184 definition, 124 Modulation
multiple, 96 entering programs, 182-195 MID$, 77
register to memory, 147-148 entering subroutines through Modem, 381
register transfer, 156 BASIC, 198-202 Modified Frequency Modulation
return, 160 executing programs, 184-186 (MFM), 705-708
rotate, 156, 157 instructions, 145-179, 183-184 Module communication, 486-487
set and clear, 158-159 mixed with BASIC, 198-205 MONITOR, 51
shift, 156-157 monitor, 127 Monitor, 388-389
stack, 160 monitor commands, 186-194 entering machine language pro
table, 161-179 operand field, 126 grams in, 183-184
See also specific instructions operation code field, 125 field descriptors, 187-188
INT, 75-76 placement of programs in memory, machine language commands,
Interrupt service routine, 258-263 202-203 186-194
INX, 169 placement of routines with BASIC, manipulating text within, 194—195
INY, 169 203-205 Mouse, 392
IOBASE, 442 programming of SID chip, 352-358 Movable object blocks. See Sprites
IOINIT, 409-410, 415-416 screen memory, 217-218 MOVSPR, 51-52, 267-268
IRQ pin, 411-414 Z80, 702-708 Multiplication, 19
MEMBOT, 420-421 Music, 336-369
J Memory, 4 coding a song from sheet music,
JMP, 169 banked, 208-213, 218, 490 341-344
JMPFAR, 453-454 character (ROM), 219-222, 226-229 equal-tempered scale values, 366,
JOY, 76 color RAM, 218-219, 225-226, 607-608
Joysticks, 390-392 236, 238, 243, 245 instruments, 336-337
JSR, 170 Configuration Register, 460-463 notes, 338-340, 341-343, 366-369
JSRFAR, 453-454 CP/M system memory organization, statements, 336-341
489-491
K crunching, 95 N
Kemal calls, 414-457 dynamic RAM, 624-626 NEW, 52
Kemal/Editor flags, 539-540 80-column (8563) chip, 299-301, Non-Maskable Interrupt (NMI) vector,
Kemal jump table, 537-539 304-305, 309-313, 327-334 407-408
Kemal routines, 403-406 management, 5, 458-471, 583-587 NOP, 160, 171
KEY, 49, 421-422 maps, 502-554, 663-664, 709-720
Keyboard, 640-642 , 727 Mode Configuration Register, O
connector pinout, 640-641 465-466 ON, 53
scanning, 496-497, 588-589 placement of machine language OPEN, 53-54, 427-428
Keywords. See Reserved system programs, 202-203 Operating system, 402-475
words preconfiguration, 462-465 CP/M components, 486
RAM Configuration Register, Kemal calls, 414—457
L 467-469 Kemal routines for programs,
LDA, 147-148, 170 RAM and 80-column (8563) chip, 403-406
LDX, 170 304-305, 309-313 327-334 vectors, 407-414
LDY, 171 RAM organization, 566-567 Operations
LEFTS, 76 RAM and system architecture, arithmetic, 18-20
LEN, 77 557-558, 729 hierarchy of, 22-23
LET, 49-50 ROM, 627-632 string, 24
Light pen, 393, 593 ROM banking, 627 ORA, 152, 172
LIST, 50 ROM cartridge startup, 471-472 Output. See Input/output
LISTN, 425 ROM chip, 630, 632
LKULPA, 448-449 ROM organization, 564 P
LKUPSA, 448-449 ROM pinout, 629, 631 Paddles, 392
LOAD, 50, 434-435 ROM and system architecture, Page pointers, 470-471
Loading 557-558, 729 PAINT, 54-55, 120-121
accumulator, 147-148 ROM timing, 628 P E E K ,77
bus, 567-568 RS-232 channel, 387-388 PEN, 78
742 IN D EX
Performance specifications, 567-568 80-column (8563) chip, 304-309, SETLFS, 426-427

PFKEY, 450-451 324-334 SETMSG, 418
PHA, 172 8502 microprocessor, 127 SETNAM, 427
PHOENIX, 448 8563 video controller, 595 SETTIM, 436-437
PHP, 172 interrupt, 591-592 SETTMO, 422
Pinouts, 394-400 raster, 591 SGN, 84
PLA, 172 shadow, 213-214, 539-540 SID (Sound Interface Device) chip,
See also Programmed Logic Array SID chip, 359-365, 528-530 336, 723
PLAY, 55-56, 338-339 status, 130-132 audio input, 351-352
PLOT, 441 System Version, 471 electrical characteristics, 605-606
PLP, 173 VIC chip, 524-527, 591-592 envelope generators, 608-610
PO IN T E R ,78 X and Y index, 129-130, 141-143 filter, 337-338, 348-351, 363
POKE, 56 Relational operators, 20-21 pins, 600, 602-604
Ports, 372, 727 REM, 60, 96 programming in machine language,
controller, 390-393, 398-399 RENAME, 60 352-358
expansion, 399-400 RENUMBER, 60-61 registers, 359-365, 528-530
for peripheral equipment, 652-657 Reserved system symbols, 88-89 specifications, 599-604
serial, 394 Reserved system words, 86-88 synchronization and ring modula
user (RS-232 channel), 394-396 RESET, 408 tion, 358-359
See also specific controller devices RESTOR, 416-417 and system architecture, 558
POS, 79 RESTORE, 61, 407-408 timing, 606-607
POT, 79 RESUME, 61-62 SIN, 84
PRIMM, 456-457 RETURN, 62 SLEEP, 64, 99
PRINT, 56, 100 RGBI video connector, 397-398 SLOW, 64
PR IN T#, 57, 380 RGR, 80-81 SOUND, 65, 339-340
Printer, 378-380 RIGHTS, 81 Sound, 5, 336-369
control, 380-381 RND, 81 characteristics, 345-348
daisy wheel, 378 ROL, 173 statements, 336-341
dot matrix, 378-379 ROM. See Memory volume, 347
PRINT USING, 57-58 ROR, 173 Sound Interface Device. See SID chip
Processors, 4, 560 RSPCOLOR, 82 Space elimination, 95
Program counter, 132 RSPPOS, 82 SPC, 84
Programmed Logic Array, 581-582, RSPRITE, 83 SPINP, 442-443
723 RS-232 channel, 382-388, 394-396 Split-screen mode, 245-248
Programming closing, 385-387 organization in memory, 246-247
advanced BASIC techniques, data, 385 raster interrupt program, 248-258
103-107 memory locations, 387-388 SPOUT, 442-443
debugging, 99-101 opening, 382-385 SPRCOLOR, 65-66, 268-269
o f 80-column (8563) chip, 292-334 sample program, 387 SPRDEF, 66, 269-270, 276, 279
escape, 106 RTI. 174 SPRITE, 66-67, 272-273
of function keys, 94 RTS, 174 Sprites, 266-290
graphics, 110-122 RUN, 62 adjoining, 274-276
of SID chip in machine language, RWINDOW, 83 collision priorities, 289-290
352-358 color, 283-285
Programs S commands, 267-270
definition, 376 SAVE, 63, 435-436 creation of image, 279-281
Kemal routines, 403-406 SBC, 174 creation procedure in definition
printing program listing, 379 SCALE, 63, 121-122 mode, 270-274
printing through a program, 380 Schematics, 721-729 display priorities, 288-289
replacing, 374 SCNCLR, 64 enablement, 282
retrieving from disks, 375-376 SCRATCH, 64 expansion of size, 287-288
saving, 373-374 Screen editor inner workings, 279-290
verifying, 374-375 control codes, 474 pointers, 281-282
PUDEF, 58-59 escape codes, 473 positioning on screen, 285-287
intermediate storage, 213-214 program examples, 276-278
R interrupt-driven, 214-215, 247-248 SPRSAV, 67, 273
RAM. See Memory jump table, 474-475 SQR, 84
RAMTAS, 416 Screen output, 388-389 SSHAPE/GSHAPE, 68-69, 122,
Raster interrupt split screen program, See also Graphics; Video 273-274
248-258 Scrolling STA, 175
RCLR, 79-80 of 8563 screen, 320-323, 328-331 Stack pointer, 132-133
RDOT, 80 o f 8564 VIC chip, 593 STASH, 69
RDTIM, 437 SCRORG, 440 Statements, 12
READ, 59 S E C ,175 basic, 27-72
READSS, 426 SECND, 418-419 format, 25-27
RECORD, 59-60 SED, 175 See also specific statements
Registers, 126 SEI, 175 STOP, 69, 100, 437-438
CIA (6526) chip, 622-623 SETBNK, 451 Storage. See Files; Memory, storage
IN D E X 743
STORE, 148 Text composite connector, 396-397

STR$, 84-85 display, 111-112 display codes, 658-659
String, 497 manipulation within machine 8563 controller, 595-598
STX, 176 monitor, 194-195 features, 110-112
STY, 176 TKSA, 419 40/80 column display, 498-499
Subtraction, 19 TouchTone frequencies, 103-104 interface chip, 524-527, 557-558,
SWAP, 69 TRAP, 70 588-594, 638-639
SWAPPER, 449 Trigonometric functions, 665 RGBI connector, 397-398
Symbols. See Reserved system TROFF, 70 16K banks, 210-212
symbols TRON, 71 types of screen display, 111-112
Syntax, 96, 99 TSX, 177 VOL, 71,341
Synthesizers, 336 TXA, 177
SYS, 69-70 TXS, 177 W
System Control Block, 684 TYA, 177 WAIT, 71-72
System schematics, 721-729 WIDTH, 72, 122
System Version Register, 471 U Wildcard characters, 480
UDTIM, 439-440 WINDOW, 72, 102
T UNLSN, 424 Windowing, 102
TAB, 85 UNTLK, 424
TALK, 425 User functions, 685-704 X
TAN, 85 User number, 480 XOR, 86
TAX, 176 USR, 85-86
TAY, 176
Telephones V Z80 microprocessor, 574-580, 685
detecting carrier, 105 VAL, 86 bus interface, 575
programming to be on or off hook, Variables, 12-13, 15-16, 96 control interface, 575
104-105 VECTOR, 417 electrical specifications, 579-580
ringing, 104 VERIFY, 71 machine language, 702-708
rotary (pulse) dialing, 105-106 VIC chip. See Video, interface processor switching, 576-577
TouchTone frequencies, 103-104 chip signal description, 577-578
TEMPO, 70, 340-341 Video, 4, 725 system description, 575-577
A B O U T THE AUTHORS
LARRY GREENLEY, the principal author of the Commodore 128 Programmer's

Reference Guide, is a Senior Technical Writer in the Commodore Documentation
Group. In addition to writing text for the Guide, Larry created most of the examples and
programs shown in the book. His previous publications include the C128 System Guide,
the new (1984) C64 User’s Manual, and the Modem/300 User’s Manual, as well as a
number of articles in computer magazines.
FRED BOW EN, the Senior Systems Engineer at Commodore, originated the overall
C l 28 design concept, which involves incorporating three computers in a single unit. He
designed the C l 28 software system architecture and the software operating system,
including the C l 28 Kernal, the Machine Language Monitor and the C l28 Editor. He
also contributed to the development of the BASIC 7.0 language.
BILL HERD, the Senior Hardware Design Engineer at Commodore, was Project Leader
for the C128 hardware design effort. He originated the C128 hardware design and
architecture and was responsible for the integration of the Memory Management Unit.
He is co-author of the C l 28 hardware specifications.
DAVE HAYNIE, a Hardware Design Engineer at Commodore, was a principal contrib

utor to the C l 28 hardware design effort. He was responsible for the intricate timing
analysis and for integrating the Programmed Logic Array. He is co-author of the C l 28
hardware specifications.
TERRY RYAN, a Software Engineer at Commodore, was the major contributor to the
development and implementation of the BASIC 7.0 language and cocreator of the
software architecture.
VON ERTW INE, an independent software engineer, customized the Digital Research
CP/M 3.0 operating system for use with the C128. This included creating the 8502
BIOS routines, adapting several standard disk formats, and providing several enhance
ments to CP/M 3.0.
KIM ECKERT, an Integrated Circuit Design Engineer at Commodore, was responsible

for the design and development of the 8563 chip, which controls the C128’s 80-column
display.
M ARIO EISENBACHER, an independent technical writer, is the author of a number

of books and magazine articles in the computer field.
NORMAN M CVEY is independent technical writer/editor with a particular interest in

information technologies. He is the co-author of the C128 Introductory Guide, the C128
System Guide and the new (1984) C64 User’s Manual.
GET FREE SOFTWARE
WHEN YOU SUBSCRIBE TO
COMMODORE MAGAZINE
Y ou’ll find u n b iased in -d ep th rev iew s o f d ie b e s t n e w softw are and h a r d w a r e . ..
d isco v er h o w to m ake y o u r o w n p ro g ra m s . . . learn p ractical h o m e an d business
applications. A nd m u ch , m u c h m ore.
Plus, in ev ery issue, y o u ’ll re c e iv e free p ro g ram s— b o th gam es and p ractical
applications— y o u can ty p e in and use rig h t away!
And, if y o u a c t n o w w e ’ll se n d you a FREE "B est o f L oadstar" disk. . full o f great
games, p ractical p ro g ram s, p lu s utilities, graphics, m usic, tu to rials an d m u c h m ore!
Subscribe o r re n e w y o u r su b sc rip tio n n o w at th e lo w ra te o f S26.95 an d w e ’ll send
you a full y ear o f C o m m o d o r e M a g a z in e (12 issues, to tal ) PLUS y o u r FREE “Best of
L oadstar” disk (S6.95 valu e).
TO ORDER T OLL
FREE 8 0 0 - 3 4 5 - 8 1 1 2
CALL i n PA: 8 0 0 -6 6 2 -2 4 4 4
I C S , I w a n t to s a v e 10% o ff th e
b a s ic s u b s c r ip tio n r a te o f $30.00f o r
C om m odore M agazine a n d r e c e iv e th e
“B est o f L o a d s ta r ” D isk , FREE.
N a m e _____________________________________________________
A ddress___________________________________________________
C ity ______________________State___________Z ip__________
S ignature__________________________________________________
□ E nclosed is m y ch eck o r m o n ey o rd e r for $ _______ (M a ke
check o r m o n e y order, f o r S2 6 .9 5 p e r subscription,
p a y a b le to COMMODORE PUBLICATIONS)
□ Bill m e 128PRG
□ C harge m y VISA o r M asterC ard ( circle o n e ) Card n u m b er
E xpiration D ate____________________ a l l p r i c e s in us c u r r e n c y .
o r call 800-345-8112 to o rd e r (in Penna. 8 0 0 -6 6 2 -2 4 4 4 )
Canadian add 110.00 to each subscription to cover postage. Overseas: $50.00/12 issues
NO POSTAGE
NECESSARY
IF MAILED
IN THE
UNITED STATES
Box 651
Holmes, PA 19043
Commodore 128
Programmer’s Reference Guide
This book is an indispensable reference guide and sourcebook for anyone using
the new and powerful Commodore 128 computer. This machine has many new
and exciting built-in features, such as the advanced BASIC programming
language Version 7.0, superior graphics, and excellent sound and music
capabilities. All information on these and other technical details, such as
machine language programming, memory maps, input/output guide, pinout
diagrams of primary chips, and schematics of the computer, are here in this, the
only official Commodore 128 Programmer’s Reference Guide.
Whether you are a new user or an advanced programmer, you’ll benefit from all
of the material in this book. Find out more about:
■ The New BASIC 7.0—Explaining new BASIC with advanced features
■ Graphics—Utilizing the Commodore 128’s graphics programming
■ Sound and Music—Getting the notes out of the C128
■ Machine Language—Programming in machine language and combining it
with BASIC
1 Operating System — Understanding the C128 operating system, the kernal,
and memory management
■ Screen Editor and Memory M aps—Deciphering the C128, C64, and CP/M
modes
■ Input/Output G uide—Controlling peripherals through software
■ Chips—Specifications and pinouts of all important chips
■ All this and much, much more
N D-SS3-34ETE-4>EnS 34292-4 ■ IN U.S. $21.95 (IN CANADA $24.95) ■ BANTAM COMPUTER BOOKS

Commodore 128 Programmers Reference Guide

Uploaded by

Copyright:

Available Formats

Commodore 128 Programmers Reference Guide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Commodore 128 Programmers Reference Guide

Uploaded by

Copyright:

Available Formats

T .M .

THE AMIGADOS MANUAL THE MACINTOSH

THE APPLE //c BOOK THE NEW jr. A GUIDE TO IB M ’S PCjr

THE COM MODORE 64 SURVIVAL ORCHESTRATING SYMPHONY

EXPLORING ARTIFICIAL INTELLIGENCE POWER PAINTING: COMPUTER GRAPHICS

COMM ODORE BUSINESS MACHINES, INC.

Commodore 64 and Commodore 128 are registered trademarks o f Commodore

Grateful acknowledgment is made fo r permission to reprint two bars o f Invention

Book design by Ann Gold.

Cover design by Jo Ellen Temple.

All rights reserved.

Published simultaneously in the United States and Canada

W ritten by Larry Greenley

We also want to recognize the contributions of Frank Palaia of Commodore Hardware

The Commodore 128 Personal Computer is a versatile, multimode computer. The

■ 128K bytes of RAM, optionally expandable to 256K or 640K

■ 8502 processor running at 1.02 or 2.04 MHz

■ Sound (three voices) using SID chip

■ 8502 processor running at 1.02 MHz

■ Integral Z80 processor running at 2.04 MHz

6 581: SID Chip

COM PATIBILITY W ITH

SW ITCHING FROM MODE TO MODE

■ C l 28 Mode with 80-column display

O FF C128 C128 C64 CP/M CP/M

C l 28 Check that 1. Press ESC I . Check that 1. Check that

Figure I —I . C o m m o d o re 128 M ode Sw itching C h a rt

OFF C128 C128 C64 CP/M CP/M

C64 1. Hold 1. Type GO 1. Type GO 1. Turn com ­ 1. Turn com ­

CP/M 1. Turn disk 1. Turn disk 1. Turn disk 1. Press 1. Insert

Figure I —I . C o m m o d o re 128 Mode Sw itching C h a rt (continued)

NOTE: If you are using a Commodore 1902 dual monitor, remember to

CP/M 3.0 SYSTEM RELEASES

H O W TO USE THIS GUIDE

C H A P T E R 2. BASIC BUILDING BLOCKS AND BASIC 7.0 ENCYCLOPEDIA—

C H A P T E R 5. MACHINE LANGUAGE ON THE COMMODORE 128— Defines,

C H A P T E R 15. COMMODORE 128 AND COMMODORE 64 MEMORY MAPS—

The BASIC language is composed of commands, operators, constants, variables, arrays

NUMERIC MEMORY STORAGE:

C O N STAN TS: INTEGER,

1. Simple Number Notation

In simple number notation, the floating-point number is calculated to ten digits of

1. The mantissa is the leftmost number separated by a decimal point.

SIM PLE NUMBER SCIENTIFIC

NOTE: The values in either column are not equivalent.

Subscripts are a convenient way to access elements within an array. If subscripts

20 DIM B(9, 9) (100 elements)

5 bytes for the array name

Here are some example arrays:

SUM(CNT%( 1)) = FP*K% (floating-point array)

EXPRESSIONS AND OPERATORS

ARITHM ETIC EXPRESSIONS

ARITHM ETIC OPERATIONS

SU BTR A C TIO N (-)

M U LTIP LIC A TIO N (*)

< LESS THAN

5 -4 = 1 result true (-1)

Relational operators may be used to compare strings. For comparison purposes,

“ A” < “ B” result true (-1)

C64 1. Hold 1. Type GO 1. Type GO 1. Turn com 1. Turn com

parentheses ( ), creating a “ subexpression.” The parts of an expression enclosed in pa

The BASIC Interpreter performs operations on expressions by performing arithme

AUTO 10 Automatically numbers program lines in incre

BOOT BOOT a bootable disk (CP/M Plus for ex