R-F Transmission Lines - Alexander Schure

ELECTRONIC TECHNOLOGY SERIES
R-F TRANSMISSION
LINES
,, . ' ' .
4. ... • • 4J' • -:· ... , ·~·.' •

.... , ..
.
,•.,'.
,,
,·;,
'
. -:~/ \:· .. ~:
''
. ' ~.\: .
.·.~ :/·•:
·.
.,., '
l•
a publication
$1.25
R-F
TRANSMISSION
LINES
Edited by
Alexander Schure, Ph. D., Ed. D.
JOHN F. RIDER PUBLISHER, INC.

480 Canal Street • New York 13, N.Y.
Copyright 1956 by
JOHN F. RIDER PUBLISHER, INC.
All rights reserved. This book or parts thereof may

not be reproduced in any form or in any language
without permission of the publisher.
FIRST EDITION
Library of Congress Catalog Card No. 56-9744
Printed In the United States of America

PREFACE
An understanding of the fundamental concepts of transmission

lines is essential to the electronic technician or engineer. Transmis-
sion lines are links in the chain of propagation or reception of r-f
power. Regardless of the actual job that a particular kind of r-f
transmission line accomplishes, some type of line will be used to
carry radio frequency power in any complete communications system.
The book has been organized to help the student to understand
the important ideas pertaining to the basic types of r-f transmission
lines. A minimum of mathematical treatment has been employed,
but the analyses are sufficiently extensive to permit the technician,
practicing engineer, or advanced student to develop these funda-
mental concepts and basic applications to best advantage.
Specific attention has been given to the various types of line in
common use; the problem of lumped and distributed constants;
variations of constants; characteristic impedance; line termination;
standing waves; standing wave ratio; input impedance; line losses;
the half-wavelength line; the quarter-wavelength line; resonant lines;
the Lecher wire line; supporting stubs; delay lines; and artificial
transmission lines. The use of waveguides has been omitted from this
text, because it is treated separately in another electronic technology
book.
The content, then, is sufficient for an adequate understanding of
both the theory and applications of this subject. A number of sample
problems are given in sufficient detail, wherever applicable and
pertinent, to permit the reader to apply the demonstrated procedures
to situations that may confront them.
Grateful acknowledgement is made to the staff of New York Tech-
nical Institute for its assistance in the preparation of the manuscript
for this book.
New York,N. Y.
April 1956 A.S.
V
CONTENTS
Chapter Page
I Fundamental Concepts of Transmission Lines..................... I
2 Transmission Line Operation and Characteristics............ 9
3 Applications of Transmission Lines................................................ 39
vii
Chapter 1
FUNDAMENTAL CONCEPTS OF TRANSMISSION LINES
1. Introduction to Transmission Lines
A transmission line is exactly what its name implies: a line over

which something is transmitted. Water pipes, electric wires, and
telephone lines are examples of transmission lines. All types of
transmission lines have something in common: they are all con-
ductors through which some form of power is sent from the place
where it is stored or generated to a place where it is to be utilized.
If each home possessed an electric generator of its own, there
would be no reason for electric power lines. Usually, however, the
source of electric power is at a generating plant far removed from
the homes where it is actually to be used. A transmission line (in
this case, an electric power transmission line) must therefore connect
the generating plant with each home. It is through this line - as in
the case of all transmission lines - that power is transmitted from
one place to another.
There are varieties of transmission lines, from subway third rails
to transoceanic cables. We are going to deal with a type designed
chiefly to carry radio frequency power, called an r-£ transmission line.
There are many kinds of r-f transmission line, depending on the
actual job to be done, although each kind may be thought of as a
variation of either of two basic types, coaxial cable or parallel
conductor. These types, in their common practical variations, are
illustrated in Fig. I.
2 R-F TRANSMISSION LINES
,u,9 t~,
BRAIDED INSULATING
SHIELD MATERIAL
INSULATOR
TWIN LEAD ( D)
COAXIAL CABLE (A)
INSULATOR CONDUCTORS
OPEN WIRE ( E)
TWINAX ( B) SHIELDED PAIR (F)
HOLLOW TWIN LEAD (G)
AIR INSULATED (C)
TWISTED LEAD ( H)
COAXIAL CABLE PARALLEL CONDUCTOR

Fig. 1. Some common examples of rf transmission lines.
2. Types of Line
Most r-f lines consist of two conductors separated by an insulating

medium, called a dielectric. Coaxial cable is a concentric line con-
sisting of a small, round conductor situated at the center of the
other conductor, which is a surrounding tube. In solid dielectric
types, illustrated at (A) of Fig. I, the middle copper wire is held
in place by the dielectric medium, which entirely fills the space
FUNDAMENTAL CONCEPTS OF TRANSMISSION LINES 3
between the two surfaces. In some other types, air is the dielectric,
and the center conductor is supported by insulating spacers at
intervals along the line. In flexible, solid dielectric types (which
are the most common) , there is usually a plastic cover surrounding
the outer sleeve and copper braiding instead of solid pipe is used
for the outside conductor. Since the inside smaller conductor is
completely surrounded by the outer tube, the shielding effect thus
provided gives this type of line great value in the field of radio
transmission and reception. Polyethylene, a flexible plastic well
known for its insulating value at radio frequencies, is used as a
supporting and dielectric medium between the inner and outer
conductor in commonly encountered flexible types. Similar flexible
lines with two inner conductors (Fig. lB) are also available. These
are sometimes referred to as "twinax."
Only where extremely high efficiency is desired, where very high
power is involved, or in the ultra high frequency region, is the rigid
("pipe-type") coax used. The center conductor in this case is usually
supported by small polystY,rene or ceramic beads or discs, which
are separated from one another at a distance governed (except in
short sections of line) by the natural rigidity of this center wire.
(See Fig. IC.) Since there are so few of the supporting spacers, and
only air is left between the remaining surfaces, the line is said to
be air-insulated. At still higher frequencies, where the r-f insulating
properties of even the best materials begin to deteriorate, short
metallic rods, called stubs, are used for physical support; the theory
of their action will be considered later.
Parallel conductor transmission line closely resembles coaxial
cable in electrical characteristics, except that the two conductors
are equally sized and lie in a parallel plane. A popular variety
today employs two copper wires imbedded in flexible plastic (poly-
ethylene) insulation, and is termed twin lead (Fig. ID) . It enjoys
widespread use in the television industry as a cheap but reliable
method of connecting the television receiver with its antenna.
When it is wished to keep dielectric losses at a minimum, open
wire transmission line is employed (Fig. IE) . It consists of parallel
conductors separated at a constant distance from one another by
rigid ceramic or plastic spacers. Except for the minor effect of these
spacers, air is the dielectric, so losses are much lower than for the
twin lead described above. Open wire line is desired where r-f
energy must be sent over relatively great distances or at relatively
high frequencies.
In the presence of a large amount of external noise, or where

radiation is to be kept to a minimum, a shielded two-conductor
transmission line is sometimes used (Fig. IF). This is not to be
confused with coaxial line. It is similar to twinax (Fig. IB). How-
ever, instead of both conductors being imbedded in a single cylinder
of dielectric, these conductors are separately insulated and "air-
spaced" by a cord of polyethylene twisted between them. As with
twinax, the metal braid helps to prevent radiation from entering
or leaving the transmission line. It is more expensive than other
forms of parallel wire line.
Hollow dielectric parallel conductor transmission line is most
useful in seaside areas where salt spray coats the plastic dielectric.
Under these conditions most types of parallel conductor lines would
be shorted by the salt residue. The hollow dielectric line, however,
greatly increases the leakage path from one conductor to another
due to the lengthened surface of its cylindrical configuration
(Fig. IG).
Twisted lead (like lampcord) , due to its high dielectric losses,
finds little practical use at radio frequencies (Fig. IH). Its use is
confined to high signal strength areas and to where short lines are
sufficient.
As was pointed out, all of the above transmission lines are varia-
tions of two basic types - coaxial and parallel conductor. Even these
basic types, however, emanate from a fundamental theory common
to all r-f transmission lines. For the purpose of the analysis and
explanation of this theory, we will deal with the parallel conductor
variety.
3. Lumped and Distributed Constants
Since transmission lines consist of a very special arrangement of

conductors, they possess certain properties dictated by their electrical
as well as their physical makeup. For instance, all conductors have
resistance, and this property is not excluded from transmission
lines. Parallel copper wires may have resistance in the order of a
fraction of an ohm per hundred feet, but it is a factor that must
be considered, since it offers a tangible hindrance to the flow of
current. Since this resistance is distributed over the entire length
of the line, each unit length (inch, foot, yard, etc.) possesses a
certain resistance. In an attempt to visualize this distributed resis-
tance schematically, it is a practice to lump or gather it in one
place, while considering the remaining part of the unit length as

a perfect conductor (Fig. 2A) . As long as the length of line is
considered as a whole, its lumped resistance indicates its true
characteristic.
Similarly, magnetic lines of force surround all conductors that
carry an electric current. This is a property of inductance. For de
and low frequencies a pair of wires is not normally thought of as
having appreciable inductance, since its inductive effect on the
passing current is very small.
At radio frequencies this inductance is no greater, but its efject
on passing current is so magnified (because of the increased induc-
tive reactance) that it can no longer be ignored. We must, therefore,
consider each unit length of wire as possessing an inductance lumped
in one place (Fig. 2B) as we do in considering resistance.
Any two conductors separated by an insulator form a capacitor.
The actual capacitance between the two wires of a transmission
line will vary with the spacing and wire size; schematically, it can
be shown as lumped (Fig. 2C) . Any material chosen as the dielectric
between the two wires must have some leakage resistance (Fig. 2D) ;
it will vary with the material chosen, air being a near-perfect
insulator. Shunt capacitance and insulator or dielectric leakage are
thought of as being in parallel, and therefore when lumped are
represented in that manner. In addition, since resistance and
inductance exist in the same unit length of wire, they may be con-
sidered as an R-L impedance. (See Fig. 2, column 2.) Because we
are concerned with the same unit length of wire, whether referring
to series inductance and resistance or to parallel capacitance and
resistance, we may combine these constants to form a complex
impedance per unit length (Fig. 2, column 3) . Note that each wire
contains only half of the total series impedance per unit length.
4. Variations of Constants
The impedances just described are not constant but vary with
frequency. The resistance of a wire has its d-c value only when the
distribution of current is uniform throughout the cross section
area. The resistance of a wire is:
R=p~
where p = resistivity, in ohms per circular mil foot
0,.
>I
!--UNIT LENGTH-----! t---UNIT LENGTH----! UNIT LENGTH--j
I I I I I I
I I I I I I
1-------'\/v'\, I I
SERIES OHMIC I IR I I I I
(Al RESISTANCE I I
I I r (SERIES! I I I
f---'W'v-----1 ~
l I I I I I
~ I I
I
I
I
I
I ~ ~
.,,
(Bl SERIES INDUCTANCE I IL I I I r tCOMPLEXl I
~
,........rll'lOI -I
I I ~
1 I I I z
I I I I en
(Cl SHUNTCAPACITANCE
I
:
I
I
I I I I
: c > l c r u r ( S H U N T lI
I I I 5
~
en
en
z
I I I I r-
I I I I z
( l I
SHUNT O I E L E C T R I C t D
~
D LEAKAGE RESISTANCE I IR I I
I I I
I INDIVIDUAL L,C,R II II COMBINING TO FORM II I COMBINING TO FORM
j PER UNIT LENGTH
SERIES ANO SHUNT I COMPLEX IMPEDANCE I
IMPEDANCES PER I PER UNIT LENGTH I
I I I UNIT LENGTH I
I I I I I I
COLUMN 1 COLUMN 2 COLUMN 3
Fig. 2. Transmission line lumped constants.

OUTER SURFACE
oo"'~ A
ELECTRON
VELECTRON
DISTRIBUTION DISTRIBUTION
AT ZERO AT RAOIO
LINES OF FORCE FREQUENCY FREQUENCIES
INSIDE a OUTSIDE
OF CONDUCTOR
(A) (8) (C)
Fig. 3. Skin effect.
and: =
, length of conductor (in feet)
=
A cross sectional ;i.rea of conductor in circular mils
This is theoretically true only for zero frequency or de. There is a
substantial deviation from this formula as the frequency is increased,
because at high frequencies most of the current will be concentrated
at the rim (skin) of the conductor, in a phenomenon called skin
effect. The reason for this is as follows:
A wire carrying current is surrounded by concentric circular
magnetic lines of force, emanating from the center of the cross-
section of the wire. It is the presence of these lines of force that
causes the wire to have inductance. This inductance is small, but
at high frequencies it represents appreciable inductive reactance.
The inductance effect offered to current flowing near the center
of the wire is greater than the inductance effect offered to current
flowing near the edge, because only those lines of force outside the
current cause inductance. The maximum number of lines of force
surround the center of the wire; fewer and fewer lines surround
areas of greater and greater diameter, as illustrated in Fig. 3 (A) .
At high frequencies, the reactance becomes important enough
to make the decrease in inductance (and thus reactance) at the
edge of the wire significant. With less reactance at the edge, more
current flows there. At high radio frequencies, practically all the
current flows in only a thin "skin" at the edge of the conductor.
This phenomenon is known as "skin effect."
As the frequency of the current increases, less and less of the
volume of the conductor is used. Therefore, resistance increases
with frequency. The comparison between typical current distribu-
tion for direct current (B) and high frequency current (C) is shown
in Fig. 3.
As explained above, current flowing near the edge of the conductor
encounters less inductance. Thus, flow of current in the skin causes
inductance to decrease as frequency increases. However, for the
normal range of radio frequencies, the inductance change is very
small and can be neglected.
Shunt dielectric leakage increases with frequency, and must be
known when calculating line losses. Any variations in shunt capaci-
tance with frequency are usually too small to be of consequence.
5. Review Questions
(1) In one sentence define the function of a transmission line.

(2) What are the two main types of r-f transmission line?
(3) Why is the term "air-insulated" applied to some types of line?
(4) What is the main advantage of coax cable, in preference to parallel
conductor Fne?
(5) What is meant by "lumped constants"?
(6) How is the theory of lumped constants applied to the equivalent schematic
representation of transmission lines?
(7) How does "skin effect" affect the total resistance of a line at any given
frequency?
(8) Does skin effect affect the flow of direct current?
(9) Does the flux density differential (which is responsible for skin effect)
also exist at zero frequency?
(10) At high radio frequency, in what region of the conductor does the current
flow?
(II) At high radio frequency, how will the conducting qualities of a hollow
conductor compare with those of a solid conductor? Why?
(12) If the cross-sectional area of a given conductor is doubled, how will the
d-c resistance per unit length vary?
(13) If a given wire 1 inch in diameter has a d-c resistance of .01 ohm per
meter, what d-c resistance will a wire 2 inches in diameter have if both wires are
made of the same material and are the same length?
(14) In general, how do the losses on a transmission line vary as the frequency
is increased?
Chapter 2
TRANSMISSION LINE OPERATION AND CHARACTERISTICS
6. Wavelength and Velocity
Since electromagnetic waves travel in a given direction at a finite

velocity, they must cover a certain linear distance during the time
of one cycle. The distance covered in this time is called the wave-
length, and is usually expressed in the metric system; it is assigned
the symbol A. The velocity of this wave equals the distance covered
during one cycle (A) divided by the time (T) of one cycle:
Velocity (V) = TA
Remembering that l_ _ f
T-
then also (V) = A/
This fundamental equation points out an interesting fact: the
velocity of a wave is governed solely by the properties of the medium
through which it is transmitted and in no way is affected by the
frequency or the wavelength. The velocity of an electromagnetic
wave in any medium is approximately:
3 X 108 meters per second
V (meters per second) = -yk
where k = the dielectric constant of the medium; k = 1 for air.
So far, we have discussed transmission lines in a general way.
R-f lines, as opposed to those carrying lower-frequency current,
have some peculiarities all their own. They arise from the fact that
9
the r-f line is usually longer than, or is at least a large fraction of,
a wavelength at the frequency of operation. In no other application
of transmission lines is this true. At power line frequencies ,\ is
about 3000 miles long, yet few power lines exceed a hundred miles
in length. At a high radio frequency such as 300 me, ,\ equals one
meter, yet the r-f line may be many meters long.
When discussing electrical behavior, it is usually convenient to
speak in terms of electrical length, not physical length. The difference
between the electrical and physical lengths arises from the fact that
electromagnetic waves do not travel as fast in a transmission line
Transmission Line Velocity Constant
2-wire line (air dielectric) 0.98
75-ohm Twin-Lead o. 68 too. 71
150-ohm Twin-Lead o. 76 too. 77
300-ohm Twin-Lead o. 82 to 0. 84
Fig. 4. Velocity constants of
Coax (polyethylene dielectric) 0.66 various types of transmission
line.
as they do in free space. Transmission line materials have a dielectric

constant greater than I. As can be noted from the previous formula,
if the dielectric constant is greater than 1, the propagation velocity
of the wave is reduced. Only in a vacuum or in air, where k equals I,
will the wave reach its maximum velocity - the velocity of light.
The ratio of the velocity of a wave along a line to the velocity it
would possess in free space is ~ailed the velocity constant. It is a
dimensionless number (always less than 1) and is given for various
types of line in Fig. 4. The physical length of a line may be derived
by multiplying the free-space length by the velocity constant (VC).
Physical length (meters) =
free-space length
(meters) X velocity constant
Wavelength (meters) = 3~0 X vc

Wavelength (feet) = ~~4 X vc
Where f = frequency in megacycles
VC =
velocity constant
The electrical length of a line is its length expressed as velocity-
constant-corrected wavelengths. For example, "5 feet" would express
TRANSMISSION LINE OPERATION AND CHARACTERISTICS 11
a physical length, while "half-wavelength" would express electrical

length.
If a wave in a line has a finite velocity, it must require increasing
amounts of time to reach successive points at greater distances
from the source of the wave. The action is depicted in Fig. 5, where
it is seen that at each point in question (A, B, and C) there exists
a separate oscillation created by the travelling wave moving toward
the right away from the source (S). Because each of these oscillations
is actuated by a single wave front, which requires time to travel
from one point to the next, the oscillations will occur at a later
90•
/ ..... ,
I
SOURCE POINT A
I
POINT 8 POINT C POINT D
,so•~o• @s••EBo• EBO" e,;o•

gt• !. ! ! !
8 00
380" wl8o-s w w a
s
270°
.b.. _____j VECTOR
J 270°
REPRESENTATIONS
I
4
i-------4
2>. • I W • CURRENT AT POINT INDICATED
S • SOURCI! CURRENT AT SAMI! TIMI!
--------!). ------...i
---------->.-----------
Fig. 5. Relation of phases of currents at different points along a transmission line.
time at each point progressively distant from the source. Specifically,

by the time the wave has reached point A, and has started an
oscillation at that point (commencing at zero degrees), the source
vector has advanced 90 degrees. Vector A must then lag behind the
source by a phase angle of 90 degrees, or a quarter of a cycle. By
the time the wave has reached point B, and started an oscillation
at that point (commencing with zero degrees), the source vector
has advanced another 90 degrees. Point B vector now lags point A
vector by 90 degrees and the source by 180 degrees. By the time the
wave has finally reached point C, the source, point A, and point B
vectors have revolved another 90 degrees, placing C vector 270
degrees, B vector 180 degrees, and A vector 90 degrees lagging phase
angle from the source reference vector.
The phase shift introduced by travel of a wave from the source
to any physical point along the line is a function of two things:
the propagation velocity on the line, and the distance of that point
from the source. In general, however, any wave travelling down
any line will become more lagging in phase as the distance increases
from the source of that wave.
Note that if a point (D) is included in Fig. 5 and is considered
with the above discussion, the difference in phase between this
point and the source is exactly 360 degrees, one cycle, or 271' radians.
If the distance from the source to point D is the distance covered
by this wave during the time its vector has lagged or rotated one
complete cycle, or 271' radians, then this distance is also one wave-
length - since one wavelength has been previously described as the
distance covered by the wave during the time of one cycle. A wave-
length may also be defined as the distance between points whose
voltages (or currents) differ in phase by one cycle, or 271' radians.
The rate of change of phase of the wave as it travels down the
line can be expressed by the number of radians (or degrees) the
phase changes per unit length of the line. The phase change per
unit length is indicated by the symbol f3 (Greek letter beta) . It can
be expressed in any convenient units, such as radians per foot,
degrees per yard, etc., but it naturally simplifies any calculations to
express f3 in units consistent with other measurements involved.
We have already said that a wavelength is the distance in which
the phase changes by a whole cycle; a whole cycle is equivalent to
271' radians or 360 degrees. Thus wavelength multiplied by phase
change per unit length must equal the number of radians or degrees
in a full cycle.
If /3 is in radians per unit length and ,\ is in the same length
units as /3, then:
,\(3 = 271' radians
1£ f3 is in degrees per unit length and >,, is in the same length

units as {3, then:
>,,f3 =
360°
360
>,, =13
360
/3=-
A
7. Characteristic Impedance
Let us view the action of a wave as it progresses from the place

where it enters the line (called a source) to the place where it is
to be used (called a load).
Imagine a line of infinite length represented by a chain of lumped
impedances as shown in Fig 6 (A). Electromagnetic energy emitted
from the generator will be propagated down the line. Some of the
original current leaving the generator (labeled / 0 ) will be diverted
through the first shunt impedance at point a and return to the
generator. Obviously, this shunt current (/1 ) will detract from the
total current available to continue down the line; the current left
will be 10 - / 1 • Farther on at b, after shunting current / 2 has
detracted from the remaining curent, only / 0 - ( /1 +1 2) is left to
continue down the line; at c, only / 0 - ( /1 +12 +18 ) is left. It can
be shown that, as the wave progresses away from the generator, the
shunting currents increase and finally equal / 0 • At this point there
is no energy left to continue. The process is indicated graphically
in Fig. 6 (B) . Note that the decline is gradual. This is what would
happen in an actual line of infinite length, since the constants are
not lumped as shown, but are distributed throughout the entire
length of the line. The voltage decreases gradually in much the
same manner, due to the series impedance.
Since there is current flowing at the source, the generator connected
to this infinitely long line must "see" a definite impedance. The
ratio of voltage to current of the wave travelling away from the
generator establishes this impedance. It is called the characteristic
impedance (or surge impedance) of the line, and is given the
symbol Z0 •
As was stated, the relationship of E to I on an infinitely long line
is its characteristic impedance.
lo·U,+Iz)- C
(A) TO INrtNITY-
(B)
DISTANCE FROM GENERATOR -
(C)
·} Y• ~g•+l)Z • ~!b
o•fs• b•ic • <JC
(D)
SIMPLIFIED REPRESENTATION OF TRANSMISSION AT HIGH FREQUENCIES
Fig. 6. Characteristic impedance.

Z _ E, where E, = voltage across infinite line

0
- I, I, = current into infinite line
The characteristic impedance of a transmission line can also be
shown to be as indicated by the following formula:
Z0 =v/f
where Z 0 is characteristic impedance in ohms
Za is series impedance per unit length (ohms)
Y is shunt admittance per unit length (mhos)
The factors in Z 0 and Y are illustrated in Fig. 6 (C). Z 0 is developed
from R and L in the well-known relation:
Za = ✓ R2+w2L2
The shunt leakage resistance and capacitance are effectively in
parallel. For this reason, it is simplest to combine them as reciprocals.
The reciprocal of the leakage resistance is conductance (g) and the
reciprocal of capacitive reactance is susceptance (b). These two
combine to form the total admittance, Y, which is used in the
formula above, as follows:
Y=-Jg2+b2
But b is the reciprocal of X 0 , and is thus equal to wC
thus Y=y g2
,---::::;..:;.===w:;;,21'="~'!c"2
and
g +w2 2
R and g both increase with frequency. However, w2 increases so
much faster that, for practical r-f lines, at all but the lowest radio
frequencies R and g become negligible, with the line assumed to
be as in Fig. 6 (D) . The expression for characteristic impedance is
then simplified as follows:
L
Zo=c
It should be noted that, since L and C are substantially constant
(for all but very low radio frequencies), Z 0 must also be constant.
Thus the conclusion: for a given geometric cross-section and insu-
lating material, characteristic impedance is constant for all radio
_frequencies. Measuring the inductance and capacitance of a line
is not necessary for determining Z 0 • By substituting formulas for
inductance and for capacitance (in terms of the line's actual con-
struction) and substituting for L and C in the above formula, we
may deduce for parallel conductor line:
800
750
SIZE
OF
700
WIRE
USED 150
DIAM.
'"zu 120
OF
USING "'
C
USED lf 110
~ vi
u ~ 100
~9
ir
300 ::;'"
250 "'"'
"'
:i:
u
200
15 0
1.5 2 2.5 3 4 5 6 7 8 9 IO .15 .2 .25 .3 .4 .5 .6 .7 .8 .9 1.0

SPACING BETWEEN CONDUCTORS INNER OIAM. OF OUTER CONDUCTOR
(INCHES) (INCHES)
Fig. 7. Characteristic impedance of two types of transmission line with air dielectric.
D
.921 log10 d X IO - 6
12.08 k X IO - 12
D
log107
. l"f
an d s1mp .
1 ymg: 276 log10 7
Z = -:;::Tr D
0
where: D center to center spacing between wires

d = radius of each conductor (same units as D)
k = dielectric constant (air= 1)
and in the same manner for coaxial line as:
138 D
Zo=;::rtlog107
where: D = inner surface radius of outside conductor
d = radius of inner conductor (same units as D)
k = dielectric constant
From the ratio of ~ it is seen that using a larger wire or a closer

spacing of parallel conductors will decrease the impedance. Increas-
ing the inner conductor radius or decreasing the outer conductor
radius will have the same effect upon coaxial cable. Figure 7 shows
how wire size and spacing affect the impedance of parallel conductor
and coaxial line respectively.
The dielectric constant (k) of most materials varies with fre-
quency. However, for materials used for insulation in practical
transmission lines (notably polyethylene), k is substantially constant
over the entire r-f range. Thus Z 0 for a transmission line of given
cross-section is constant.
At this point the reader may question the practicality of discussing
an infinitely long line, which obviously does not exist in a physical
sense. Under certain circumstances, however, the wave reaching the
end of a line is unable to distinguish between this end and an
infinite extension of the line; the line as a whole has the charac•
teristics of an infinite line. Let us examine the reasons for this.
8. Line Termination
Assume the generator in Fig. 8 (A) is looking into an infinitely

long line of 300 ohms characteristic impedance (for simplicity
lumped constants have not been shown). If we break the connection
at a and b and insert a section of line (say 100 feet) between a and
b and the generator, it will not make any difference to the generator,
since a finite distance added to an infinite distance is still an
infinite distance. The new l 00-foot insertion is shown in Fig. 8 (B) .
It is agreed that the generator will still "see" a 300-ohm impedance
- even with the 100-foot piece of line inserted. In addition, however,
when the electrons bounding down the line reach a and b, they
will also see 300 ohms, the Z 0 of an infinite line. This is logical
because it corresponds to the Z 0 the generator saw when it was
located at a and b. Now if we were to again break the connection
at a and b, but this time insert a 300-ohm resistor (called the load
resistance) in place of the infinitely long line (Fig SC) , the electrons
reaching this resistor would not find conditions altered in the least;
in fact, they would interpret this 300-ohm resistor as being the
same infinitely long 300-ohm line - even though it is physically
quite different. Under these conditions it is said that there is an
impedance match betwen the line and load.
9. Reflections
It would be well for the reader to reread the previous paragraph,

since it contains one of the most important concepts in this book;
a
~o •300.n TO INFINITY-
(A)
100 FT
·I
:,
1· a
$ ~o • 300.n ~o • 300.n TO INFINITY-
( B)
I. IOOFT
•1A
$ ~o • 300.n ~-,· ,.,,,

B
(C)
Fig. 8. Impedance matching.
namely, if a wave reaches the end of a transmission line and encoun•

ters a resistive load equal to the characteristic impedance of that
line, it is unable to differentiate between this resistor and an infinite
extension of the line. Only when the load matches the line's
impedance does this phenomenon take place. If the two are not
equal, some of the energy reaching the end of the line will not be
absorbed by the load resistor; the part that is not absorbed is
reflected back along the line toward the source. The wave travelling
from the generator toward the load is called the incident wave,
and the one (if any) traveling from the load back to the generator
is called the refiected wave.
In Fig. 9 (A) there is a water trough in which a large stone has
been dropped. The stone evidently must be the generator of a
disturbance, and the wave thus created will travel down the length
of the trough. Upon reaching the end, the incident wave finds the
wall unable to absorb its energy, and hence reflection takes place.
Almost everyone has observed how the sun is reflected from the
windshield of a passing automobile. The person inside the car is
WATER -+ DIRECTION OF PROPAGATION
TROUGH (A)
,,, __ ------------- ---·--
( B)
R1::FLECTED WAVE\
(C)
Fig. 9. Reflections in water trough analogous to electrical reflection in transmission
line.
not aware of this because he sees only the effects of the light coming
through the glass. It is evident to those on the outside that some
of the total light energy hitting the glass is not passed but is reflected.
Both of the above examples are analogous to reflections on a trans-
mission line.
The rate at which the load can absorb the incident wave is deter-
mined by its resistance, or impedance. The rate at which the line
can supply the load with energy is determined by its impedance -
the characteristic impedance (Z0 ) of the line. If the Z 0 and the
load resistance (RL) are equal, the load will absorb all of the energy
supplied to it by the line as soon as it becomes available.
A transmission line and its load are analogous to a simple voltage
source and its load, as illustrated in Fig. 10. The most power available
from an electrical source of given source resistance is that power
delivered when the load resistance equals the source impedance.
Figure 10 illustrates this for a source voltage of 100 volts with a
source (internal) resistance of 50 ohms. The maximum power is
delivered to the load when the load resistance is 50 ohms; if the
load resistance is changed to a value higher or lower than the 50-ohm
source resistance, the power output drops off. Note that higher
currents mean higher output power only for currents up to the
50.0.
50
,, -
MATCHING LOAD
--
~- -./ I
I
I
I
I
LOAD
R 45
I-
::> I
I
I
I
I
-
I ~L Q.
I I I-
::>
I I 0 40
L .J Ill
I-
I-
;
20 30 40 50 60 70 80 90 IOO
LOAD RESISTANCE RL, OHMS
SOURCE RESISTANCE Rs • 50 A (CORRESPONDING TO ~o OF LINE)
!!b. 2_ Po
!Al (AMPS) (WATTS)
20 1.43 40.8 SAMPLE CALCULATION
30 1.25 46.8 RL • 40A

40 1.11 49.2
TOTAL RESISTANCE•Rs + RL • 50 + 40 • 90A
50 1.0 50
60 .909 49.5
I• .!Q.QY. I.II AMPS
90A
70 .833 48.5 2 2
Po• I RL • ll.11) X40
80 .769 47.2 • I.II X I.II X 40 • 49.2 WATTS
90 .714 45 .9
100 .667 44.5
Fig. 10. Relations between load resistance and power output for voltage source
of given fixed resistance. Relations are the some for transmission line in which
Zo corresponds to source resistance.
maximum power value; although lower load resistances increase

current, they do not absorb as much power.
The same principle applies in a transmission line. The charac-
teristic impedance of the line corresponds to the source resistance,
and the resistance of the circuit connected to the end of the line
corresponds to the load resistance, RL. The first wave of electrical
energy that travels down the line is called the incident wave. If the
load resistance equals Z 0 , all of the incident energy is absorbed.
If not, some energy "bounces back" in what is called a refiected wave.
It may then be said that if there are no reflections on a line the
incident wave is the only wave on the line, and the ratio of I to E
of that wave is the same at every point on the line. Z 0 is now
defined as:
E,
Za=-r;
Where: Ei = the voltage of the incident wave
Ii = the current of the incident wave
Z0 =
the characteristic impedance of the line
This is true for every point on the line, including the load end.
It means that the ratio of the amount of current flowing through
the line to the value of voltage existing across the line is determined
by the Z 0 or E/1, when there are no reflections.
The ratio of the amount of current that will flow through the
load to the amount of voltage across the load is determined by the
load resistance (RL), or ELf IL.
Since the form in which the line can supply energy to the load
is governed by f' and the form in which the load can accept this
power is governed by ;: , it should be clear that if the load is to

absorb all of the power supplied to it by the line, the two ratios
must be equal. Therefore, if
E, EL
-r;=
IL
where EL= voltage across the load
IL= current through the load
then Z 0 = RL
If, however, RL does not equal Z 0 , the ratios oft to ;: will not
be equal and there will be reflection. The magnitude of the reflected
wave will be determined by the amount of impedance mismatch, or
the amount by which the ratios of7t and EILdiffer. The magnitude
i L
of the reflected wave will always be just large enough to make the
ratio of 7; of the part not reflected just equal to the ratio of 7:
of the load resistance. This means that the load will absorb all the
incident wave power only when the ratio of voltage to current of
that power is numerically equal to the load resistance. If incident
wave power possessing any other ratio of E, to I, reaches the load,
the correct amounts of I, and E, will be rejected by this load to
make the amount remaining (in the form of-}) just equal to the
1
load resistance.
It may be easier to understand the effects in transm1S1on lines

if the concept of a "wave" is considered further. A high-frequency
pulse, or wave, takes a finite time to travel along the line. Unlike
a steady-state voltage or current, it is not necessarily governed by
the whole circuit. Rather, for the moment during which the incident
wave is first sent down. the line, its energy is "released" from the
remainder of the circuit. Its voltage and current are related according
to the Z 0 of the line, and are not affected by the source and the load.
However, when the termination (load) of a line that is not matched
is encountered, the incident wave voltage-current relationship is
upset, and reflected energy must be released to keep the load voltage
and current in proper Ohm's law relation to load resistance. At the
same time the reflected wave, as it leaves the load and moves toward
the source, must (like the incident wave) maintain a voltage-current
ratio equal to Z 0 •
The currents and voltages in the line and at the load are governed
by basic electrical laws as follows:
a. The vectorial sum of the voltage of the incident wave and the
voltage of the reflected wave must equal the load voltage.
b. The ratio of incident voltage to incident current, and the ratio
of reflected voltage to reflected current, must both equal Z 0 of the line.
c. The ratio of load voltage to load current must equal load
resistance.
If all the above conditions are combined and solved mathemat-
ically, an expression for the ratio of the reflected wave voltage to
the incident wave voltage is obtained. This ratio is known as the
reflection coefficient (k). Its value can be calculated in any given
case from:
ZL - I RL - I
20
k - .,.- - - - ; for resistance load: k = R_Z..; .o___
-~+I _.!:_+I
Zo Zo
RL-Zo
or: k =R +Z
I, 0
In the simple (and fortunately most common) case in which the

load is (or can be considered to be) a pure resistance, k is not a
complex quantity, but may be plus or minus. Examination of the
above equation shows that when load impedance (resistance) is
greater than Z 0 , k is positive. When load resistance is less than Z 0 ,
k is negative.
To clarify the above ideas, consider some examples. First, suppose

a 300-ohm (Z0) transmission line is terminated in a 300-ohm resis-
tance (load) . Substituting in the equation we obtain:
300 - I
300 0
k= 300 =y=O
300 + I
The zero value of k indicates that there is no reflected wave. Thus
the line is matched, because the load resistance equals Z 0 • Next,
let us assume that the load resistance for the same line is reduced
to 150 ohms:
150 - 1
300 0.5 - 1 -0.5 I
k = ..,.15=0,---- - 0.5+ I 1.5 T
300 + I
Thus the reflected wave is of opposite polarity to the incident wave,
and has one-third of its amplitude.
It is this unique control the load has over the action of a line that
leads us to state that all calculations involving transmission lines
must start at the load end of the line.
In summary, the identifying characteristics of a transmission line
operating into a matched resistive load are as follows:
a. At every point on the line current and voltage maintain the
same phase relationship with one another.
b. At every point on the line, the ratio of voltage to current is
always the same and is numerically equal to the characteristic
impedance of the line.
c. Maximum amount of power is transferred from the line to
the load.
d. No power is reflected back to the generator.
10. Standing Waves
1£ the line we are discussing is being used to feed r-f power from
one place to another, it is natural that we do not want mismatches,
since they represent a loss of power in the form of reflections. Let
us look deeper into what happens when a wave travels back toward
its source.
1£ an incident and a reflected wave both exist on a line, the voltage
at any given point along the line is the vector sum of the incident
wave and reflected wave voltages at that point. Current at any point
I
IMPEDANCE
SHORTED
LOAD
END
CURVES ARE SHOWN WITHOUT

REGARD TO POLARITY
Fig. 11. Conditions olong o line with o short circuit load.
is similarly derived. The resulting voltage and current distribution

along the line depends upon the magnitude and phase relationship
between the incident and reflected waves. If the incident and reflected
waves are equal in magnitude and in phase, the resultant wave will
be double either individual value. At other points where they are
exactly 180 degrees out of phase, they will cancel each other. Points
of intermediate values of phase and amplitude will produce a result
somewhere between these two values. Since the incident and reflected
waves cannot exist independently and their voltage and currents
always combine vectorially, we will discuss only resultant voltage
or current, which is what is actually present and measurable.
The phase relation between the incident and reflected waves is
such that their vectorial sum voltage (and current) varies along the
line. If a measuring device is moved along the line, it will show
that the voltage rises to a peak value, then falls to a minimum, then
rises to a peak again. Maxima and minima of voltage alternate with
equal spacings. Current also varies between minimum and maximum
values: where the voltage is maximum the current is minimum, and
where voltage is minimum the current is maximum. These variations
of voltage and current are present in any line in which there is a
reflected wave [that is, any line in which the load resistance is not
matched (equal to) Z0 ] .
A minimum point of voltage or current is called a node. A maxi-
mum point of voltage or current is called a loop. Thus a current
node is also a voltage loop, and a voltage node is a current loop.
These variations of voltage and current with physical position along

a transmission line are referred to as standing waves. Each successive
node is spaced a half wavelength from its predecessor and each
successive loop is spaced a half wavelength from its predecessor
(allowing for velocity factor). Their actual position in relation to
the load end of the line depends on whether the incident wave finds
a short or open circuit, or reactive impedance at the load.
Where it is desired to transmit r-f power over a distance with
greatest efficiency and economy, standing waves have the following
disadvantages:
a. "Copper" losses are very high. Power loss in the conductors
is the summation of the I 2 R losses at all points along the line. With
appreciable standing waves, the high values of current at and near
current loops result in a large increase in copper losses over those
of a matched line. The increase in power loss at current loops is
much greater than the decrease in power loss at current nodes, so
the overall copper losses are increased.
b. Dielectric losses are also increased in the same manner.
c. Peak voltage at voltage loops is often so high (in transmitter
applications) that special attention must be given to the provision
of insulation. In other words, a heavier duty, more expensive,
physically heavier, and more bulky transmission line may have to
be used, as compared to matched line requirements.
d. As will be further explained presently, a line with standing
waves has an input impedance that is very critical with respect to
line length. If the length of the line is not just right at a given
frequency, it may be difficult to couple power into it from the source.
When reflections occur, it is sometimes helpful to think of the
load as the generator of the reflected wave. If the incident wave
strikes a short circuit, no power is absorbed by the load and complete
reflection of the incident wave takes place, producing standing waves.
Because there is zero resistance at the load point, the reflected current
will be maximum and will just equal the incident current. The two
currents will add vectorially and produce a resultant current double
the individual value. It is therefore said that the two currents are
in phase at the shorted end of a line. At a distance one-half wave-
length from the load (Fig. 11) the incident and reflected currents
will again be in phase. The condition is repeated at half-wavelength
multiples from the load.
It might be well to mention that in Fig. 11 it is assumed that the
voltage and current both vary in sine wave fashion. Since in actual
practice we are mostly concerned with magnitude and not polarity,

the graph has been simplified by flipping the negative part up so
that both parts of the cycle are above the zero axis.
If the load is shorted, not only will there be maximum current
but, according to Ohm's Law, the voltage must be essentially zero
at this point. Its value may also be explained by noting that the
incident wave and reflected wave voltages are of equal value but
opposite polarity; therefore, they cancel. This is verified by substi-
tuting zero for ZL in the previous formula; k is shown to be - I,
thus the reflected wave is of magnitude equal to the incident wave
but of opposite polarity. Because this also is repeated at half-wave-
length multiples from the load, any point of maximum current is
a point of minimum voltage.
As noted in Fig. 11, a point one-quarter wavelength from the load
is a point of minimum current and maximum voltage. It is at this
point that the voltages are in phase and the currents out of phase;
load conditions have been exactly reversed. The same thing is dupli-
cated at odd multiples of a quarter-wavelength from the shorted
load. It can then be said that any point of maximum voltage is a
point of minimum current.
Note the peculiar shape of the impedance curve. It is at a zero
value whenever the voltage is zero and the current maximum; it
rises nearly to infinity when the voltage is maximum and the current
zero.
The impedance at any point on the line is equal to the ratio of
voltage divided by the current. At the shorted load voltage is zero,
and therefore so is the ratio E/1. This accounts for the zero impedance
across the shorted load. At a point a quarter-wavelength away from
the short, the voltage is maximum and the current is zero; the ratio
of E/1 is now infinite, and so is the impedance. The impedance thus
varies in much the same way as the voltage; that is, their maximum
and minimum points are th~ same points on the line, and conditions
of impedance and voltage are repeated every half-wavelength and
reversed every quarter-wavelength.
It should be remembered that we are expressing line lengths only
electrically (half-wavelength, quarter-wavelength, etc.) . If the fre-
quency of operation is changed and the physical line length is not,
the whole voltage and current distribution changes accordingly.
If the incident wave strikes an open circuit at the end of the line,
as shown in Fig. 12, there can be no current through this near-infinite
resistance; only a large voltage may exist. Again, no power will be
IMPEDANCE
i!o
OPEN
CIRCUIT
LOAD
1..). §.). ~). ~). ..! ). _g_).

4 4 4 4 4 4
CURVES ARE SHOWN WITHOUT
REGARD TO POLARITY
Fig. 12. Conditions along a line with an open circuit load.
absorbed. It will be reflected and will cause standing waves as

before, but this time the voltage and current relationships are exactly
opposite to those of the short-circuited line of Fig. 12. The currents
will cancel and the voltages add, producing a voltage loop and
current node. At this point the ratio of E/l approaches infinity,
and so does the impedance. Again, these conditions are repeated at
half-wavelength intervals from the load, and are reversed at odd
multiples of a quarter wavelength.
It is interesting to note that at points that are odd multiples of
Ys wavelength from the shorted and open-circuited load, the im-
pedance curve equals the characteristic impedance of the line. (If
the reader has some knowledge of trigonometry, he will recognize
the impedance curve as the familiar tangent graph without regard
to polarity.) The exact relationship expressing the value of
impedance at a given angular distance from the short-circuit load is:
Z,=Z 0 tan 8
from an open circuited load:
z,, =Z0 cot 8
where z,, = the impedance at any point on the line
Z0 = the characteristic impedance of the line
8 = the angular distance from the load.
At Ys wavelength from the load the impedance is always the Z0
because Ys wavelength equals 45 degrees and the tangent of 45
degrees is 1. The previous formulas now reduce to:
z, = Z 0 tan (45°)
=Z 0 (1)
=Zo
and:
z, = Z 0 cot (45°)
=Z 0 (1)
=Zo
A significant comparison may be made here. Impedance, voltage,
and current on an open-ended line are everywhere opposite to the
same quantities for a shorted line; that is, where shorted line current
is high, open line current is low, and vice versa.
11. Standing Wave Ratio
If the line is matched to the load, no reflections occur. If the line

is terminated in either a short or an open, complete reflection occurs,
with the incident and reflected waves equal in amplitude. If the
line is terminated in some value of resistance other than its charac-
teristic impedance, but not an open or short, only some of the power
is.reflected- the remaining power being absorbed by the load. Under
these circumstances the power in the reflected wave never equals,
and is always less than, the incident wave. Maximum and minimum
points of voltage and current are still found, but the maximum
points are no longer double the value of the reflected or incident
waves, and the minimum points are never zero. The ratio of maxi-
mum to minimum voltage or current along any line is called the
standing wave ratio. It is also a measure of the ratio of mismatch
between line and load, or load and line. A mathematical statement
of this is:
SWR = Imai»
[min
Em=
= Emtn
if RL <Z 0, then it can be shown that
SWR=~L
if Z 0 < RL, then it can be shown that:
SWR=:L
0
where
Imaa, = current at current loop
Im,n = current at current node

Em= =
voltage at voltage loop
Emtn = voltage at voltage node
RL =
load resistance in ohms
Z0 =
characteristic impedance of line in ohms
SWR =
the standing wave ratio of the line
The standing wave ratio (SWR) is a measure of the mismatch
of the line termination. Accordingly, it is tied into the coefficient
of reflection, k. The two are related as follows:
SWR=: ±:
But the relation between k, Z 0 , and RL (ZL) was given previously.
Substituting the equation for coefficient of reflection in the above
equation gives:
1 - RL-Zo
RL Z0+
Examination of this equation shows how it simplifies for the con-
ditions given, as follows:
Rr, > Z 0 :
SWR _ RL + ,; + RL - Zv _ 2 RL _ RL
- R..t, + Z 14, + Z
0 - 0 - 2 Z0 - Z0
R 1, <Z 0:
l Z 0 -RL +
SWR- RL+Zo - ~ + z0 +z0 -1\
- l _ Zo - RL - RL + z; - z; + RL
RL Z0 +
If a certain line possesses a SWR of 5, it means that maximum current
or voltage is five times minimum current or voltage. It also means
that the Z0 is five times RL or ¼ RL. If Z0 equals RL, the ratio is l: 1,
and there are no standing waves. The SWR is always l for a matched
line. If the load resistance is less than the Z0 , there is a current loop
and voltage node at the load; if the load is greater in resistance than
the Z 0 , a voltage loop and current node exist at the load.
Impedance in general is composed of a resistance and a capacitive
or inductive reactance. Up to this point the load of a transmission
line has been considered as a pure resistance. Many times the line
is terminated in an impedance that, although predominantly resis-
P • PARALLEL RESONANT CIRCUIT POINTS

S • SERIES RESONANT CIRCUIT POINTS
Fig. 13. Impedance curves for short circuited line.
tive, has a reactive component. When a line terminates in a reactive

impedance, the reactance does not absorb energy, therefore power
is reflected. The amount of reflection (and resulting SWR) is pro-
portional to the amount of reactance contained in the impedance.
Because of the phase shifting abilities of capacitive and inductive
reactance, the standing waves thus formed are shifted along the line
with respect to their position if the line were equally mismatched
to a resistive load. The effect is the same as if all the curves in Figs.
11 and 12 were shifted a few degrees to the right or left. If the line
is terminated in a purely reactive load, the pattern is similar to an
open or shorted line except for this shift.
The graphs of Figs. 11 and 12 do not show the nature of the
impedance existing along a shorted or open line. Figure 13 denotes
the same short-circuited line impedance characteristics as does Fig.
11, except that the reactive component is now indicated, and positive
and negative reactances are separated in their appropriate positions
above and below the horizontal axis. Following this curve from the
right, the reader will note that the impedance rises gradually and
finally reaches a near-infinite value at a distance a quarter wavelength
back from the short-circuit termination. During this rise it is above
J
>
;::
~~
cz
<>:!
:rf~
z"'
c"'
0
"'II.
!
I
.!. ).
4
P • PARALLEL RESONANT CIRCUIT POINTS
S • SERIES RESONANT CIRCUIT POINTS
Fig. 14. Impedance curves for open circuited load.
the zero axis and thus is inductive reactance. At a quarter-wave

distance, this near-infinite impedance is resistive in nature; i.e.,
neither capacitive nor inductive. As distance from the load becomes
greater than a quarter wavelength and approaches a half wavelength,
the impedance gradually decreases. During the swing, however, it is
below the zero line and acts like a capacitor. When it reaches the
half-wavelength mark, it is again a resistive impedance, but this time
of virtually zero value. The entire process is duplicated during the
next half wavelength, varying from zero to maximum inductive
reactance, and then a near infinite resistance; from maximum to
minimum capacitive reactance, and to zero resistance. The distance
along the line back from the load will determine whether it acts
like a large or small resistance, or a large or small inductive or
capacitive reactance. At multiples of odd eighth wavelengths, it
acts as an inductive or capacitive reactance equal to the characteristic
impedance. The only limiting factors imposed on the value of
impedance at different points are the losses of the line. In the lossless
line, the impedance would actually vary from zero to infinity.
Figure 14 shows the impedance curve for a line terminated in an
open circuit. The impedance is almost infinitely negative, or capaci-
tive, at this point. It decreases to zero a quarter wavelength back

from the open end, where it represents a resistance. The curves in
Figs. 13 and 14 are exactly the same except that one is displaced
90 degrees (quarter wavelength) from the other. This only corrobo-
rates the idea that impedance conditions are reversed every quarter
wavelength; a distance a quarter wavelength from an open circuit
looks like a short circuit and vice versa. Conditions are duplicated
every half wavelength.
12. Input Impedance
So far we have not been overly concerned with the impedance

the generator "sees" as it looks into the input end of the line. When
the line is matched to its load, the generator sees the Z 0 of the line,
regardless of its length. Such a line is called a fiat or nonresonant
line. If the line is not matched, and standing waves are produced,
the impedance the generator "sees" will vary between a maximum
and minimum value as the line is made longer or shorter. This
impedance is called the input impedance, and is, as always, depen-
dent upon the ratio of E/1 at that point. The amount of possible
deviation of input impedance above and below the Z 0 depends on
the SWR. Because the SWR is infinite when the line is terminated
in a short or open circuit, the impedance will vary from near zero
to a near-infinite value, depending on the exact line length (in
terms of half wavelengths) from the load. All this is to say that if
the point we have been discussing in connection with standing wave
curves is located at the input of the line, the generator will "see"
all the phenomena previously attributed to that point (as shown
in Figs. II, 12, 13, and 14). Not only will the input impedance vary
with a change in line length, but also with a change in frequency,
since in either case the physical length in relation to the electrical
length is changed. This means that if the line length is held constant
and the frequency is varied, or if the frequency remains the same
while the line length is changed, the effect is the same.
Since the input impedance is a function of frequency (and line
length) the line is said to be tuned or resonant. Only flat lines
with a SWR of I are nonresonant. The effect can be further appre-
ciated by again referring to Fig. 13. At even multiples of a quarter
wavelength from the shorted load, the line has a minimum im-
pedance that is purely resistive, like a series resonant circuit. At odd
multiples of a quarter wavelength, the impedance is infinite and
resistive in nature, corresponding to a parallel resonant circuit. The

points at which the line exhibits properties of parallel and series
resonance are indicated by the letters P and S respectively.
The same sort of conditions exist on an open circuited line (Fig.
14) , but this time the parallel resonant points are even multiples
of a quarter wavelength, and series resonant points are odd multiples,
away from the load. A composite graph showing variations of input
impedance for a short circuited line is contained in Fig. 15, and for
open circuited lines in Fig. 16. Perhaps the reader sees the obvious
similarity between the two graphs. Every half wavelength the entire
cycle of input impedance characteristics repeats itself. The only
difference between the shorted and open circuited lines is the point
at which the cycle commences. For example, an open circuited line
a quarter (2/8) wavelength long has the same input impedance as
a shorted line a half wavelength long. Because the reverse also
applies, we may say that a shorted line will possess the same input
impedance as an open line of any length if the shorted line is longer
or shorter than the open line by even multiples of a quarter wave-
length. Although losses are greater on a resonant line than on a
nonresonant line, it is sometimes more convenient to use a resonant
line in practice. This is discussed further in Chap. 3.
13. Line losses
Standing waves can represent a power loss because some of the

incident wave power is reflected. Assuming there is no power loss
on a mismatched line, the percentage of power absorbed by the
load for the same line and source conditions, with a given mismatch
ratio, compared to perfect match power is:
percentage of matched-line power absorbed by the load =
2 (SWR)
I+
(SWR)2 X 100
Applying this formula, it can be seen that when the line is matched,
the SWR is I, and the percentage of power absorbed by the load is:
1+2 (I)
(I) 2 X 100 = 100 percent
Every practical transmission line has losses. Although for good high-
frequency commercial lines, these losses do not materially affect the
characteristic impedance, they nevertheless place a practical limit
on the efficiency of any line.
~{
0
,-····\
•{,_/
2 I- •
,< ,<
-I m o~. ix -Jm
~{ •
-0 ~{
,< ,<
OI )m °'Im
• I
,<
Or • Ox ,.,Im
-1r-"ffl'-
i{ • I
,<
,..Im
,<
on)m
~{
Or • 1X
• ,<
101m
.,,::c
I-
Cl)
:,
:!
.
ii
.
~{
z
Ill u
..J C
• I Ill ,,.
0
;>
-0 ~{ ,.
Cl.
,<
'°Im :!:
-~'5
I
Ill
()
z
-~
Cl.
01
,<
.,,~ -~
C
or .ox ""Im ci .c .
,<
mlm
~{
1r-"ffl'-
• •
,<
m)m
Ill
z
:i
.c
Cl.
e01
-~
,<
a,Jm
~{
0~ .,x
• ' u
8.E
0
!{ • I
-0 ~{ •
S33S 'N39 lltHM
0
0 ,,-,
""A
()
.~~,_..,
2 } '
>C ;(
oil!.ox Im
0
,~ ""V
()
>C
' ' ... ,..,
;(
ot• ix
~{
1110
' ' ,.,,..,
;(
-ci
1
.
.
C
a.
0
J!
A<
.J ' ' .,.., ;
-0
>C ;( UI
C
0
.
..!!
Ill
A()
1/1 .
:§
>C
' ' ;(
::c
I-
Cl)
..,z
:,
.
:!
ot.ox
0
inlm
..,
.J .
>
u
C
> a
""t C
>C
' I
..<
:ii:
l
!: .!
1~ 0
colm
::c
I-
;;
a.
.5
""~ Cl)
..,
>C
' ' ,..,..,
..<
z
.J
UI
C
1-5
oil!. ,x ..,
0 z z.
""A<
.J ' I
.J a.
e
UI
-0
>C ..<
.!
mlm .;
0 8.E
""A
() ' ' u
0
>C
-
..< ,0
oil!. ox OIICD
~ff-
§{ ' ' ;(
2lm
QI
iL:
II
'
S33S 'N39 .Ll1HM
Perhaps the most serious of these losses is radiation. All conductors

carrying r-f power tend to act as antennas. The amount of radiation
is largely dependent upon the length of the conductor in relation
to the wavelength of operation. As mentioned earlier, transmission
lines are usually a number of wavelengths long; on this basis they
would seem to be good radiators.
We speak of the term "radiation," which applies primarily to a
line being used to connect a radio transmitter to its antenna. The
object of the antenna system 1s to radiate r-f energy as efficiently as
possible. However, it is not desirable that the transmission line itself
radiate, because (a) it is normally not as high and "in the clear"
as the antenna, and (b) the field from the transmission line interacts
with the field from the antenna to produce undesirable directivity
effects and losses.
Although we use the term radiation for transmitting applications,
the same conditions that produced it also adversely affect reception
with the same antenna system. For receiving, the situation is reversed,
and what was radiation becomes pickup of unwanted signals, such
as man-made noise and other undesired signals the antenna is de-
signed to reject. Therefore, in all discussions, it should be remem-
bered that both radiation and induction losses affect receiving as
well as transmitting applications.
Parallel conductor lines contain two wires. If the line has been
correctly constructed and installed, the conductors are balanced;
i.e., hold the same relationship to ground, are the same size, etc.
In balanced line, the currents in the two conductors are equal but
flow in opposite directions. Opposing magnetic fields are generated.
Theoretically, these opposing fields cancel the total magnetic field
surrounding the line, and radiation is eliminated. Unfortunately,
this cannot happen completely unless the conductors occupy the
same space - an obvious impossibility. Practical lines, however, have
little radiation when conductor spacing is kept to .01 wavelength
or less. Radiation loss from any two-conductor line varies directly
as the square of the frequency and the square of the spacing.
Radiation is proportional to the electromagnetic field intensity;
this field is always proportional to the square of the RMS current
in the conductors. In the presence of standing waves, the current
will be much greater in certain places than would exist on a matched
or flat line. Radiation power losses can therefore be expected to
increase as the SWR increases.
Even if most of the overall magnetic field surrounding a line is

canceled by proximity of the conductors, the constituent lines of
force extend for a significant distance from each conductor. If these
lines happen to cut a nearby conducting object, and power is dissi-
pated in that object, it must be supplied by the line. This form
of power loss is known as induction loss. To keep induction losses
to a minimum, parallel conductor lines are always kept as far away
from conducting objetts as possible.
The construction of coaxial lines is such that, when these lines
are properly used, losses through radiation and induction can be
greatly reduced, compared to parallel conductor types. This is
possible because:
a. The conductors are concentric. The fields developed by the
two conductors occupy the same position in space, and thus nearly
perfect cancellation takes place.
b. The receiver, or other load device, to which power is delivered,
is connected to the line in such a way that currents in the shield
cannot enter the load. Ordinarily, this is accomplished by connecting
the shield to a ground potential point oii the load, which is also
connected to earth ground. Thus, although radiation and pickup
currents are still present in the shield conductor, they cannot couple
to the load and, having a direct ground return, dissipate little power.
(In radio systems the source always works against ground.)
Conductor resistance is an important source of power loss in a
transmission line. It is often called l 1 R or copper loss. The lost
power is always dissipated in the form of heat. Since skin effect will
only add to the natural resistance of the conductor, a rise in frequency
will increase conductor losses. The impedance characteristic of the
line will also affect the conductor loss. Specifically, conductor power
loss will vary inversely as the square of Z0 , because as Z0 decreases
the current for a given power transfer must increase; and, for any
conductor with a given resistance, the power dissipation rises with
a rise in current.
Dielectic leakage is always responsible for some power loss. It
also manifests itself in the form of heat, but rises with an increase
in frequency. Of course, the quality of dielectric has a considerable
bearing on the loss at any frequency. The heat loss in the dielectric
varies directly as does Z 0 ; it is the voltage, which va!ies directly as
does Z 01 that is the determining factor in shunt dielectric conduction
for a given line.
14. Review Questions
(1) What is the relationship between the frequency and the period of a given
oscillation?
(2) Why is the electrical and not the physical length of a transmission line
most important in line calculations?
(3) How does the phase angle compare on a line at different distances from
the generator?
(4) At 2-½ wavelengths from the generator, what is the phase angle of a
traveling wave at a given instant with reference to the generator?
(5) In actual transmission lines, are constants lumped or distributed?
(6) What is the physical length in feet of one wavelength on a line possessing
a velocity constant of .75 and operating at a frequency of 300 me?
(7) What is the physical length in meters of the line in the above problem?
(8) What is the characteristic impedance of a line at a given point where the
voltage is 58 volts and the current is 0.29 amperes?
(9) What is the characteristic impedance of a line whose shunt admittance
per unit length is 0.04 mhos and whose series impedance per unit length is 400
ohms?
(10) Why is it that at high frequencies we need consider only the inductance
and capacitance per unit length of a line when calculating its characteristic
impedance?
(11) What is the characteristic impedance of a parallel conductor transmission
line whose center-to-center spacing between wires is one inch, the radius of each
conductor 40.3 thousandths of an inch, and the dielectric constant is 0.9?
(12) What is the characteristic impedance of a coaxial cable when the inner
surface radius of the outside conductor is 0.125 inches, the radius of the inner
conductor is .015 inch, and the dielectric constant is 0.66?
Chapter 3
APPLICATIONS OF TRANSMISSION LINES
Transmission lines are most noted for their use as carriers of r-f
power. Typically, they are employed as a connection link between
the transmitter and its antenna, or antenna and receiver; as an
interstage coupling device; or for connecting studio and transmitter.
15. The Half-Wavelength Line
Let us re-examine the voltage, current, and impedance relationship

on a shorted line a half-wavelength long, as shown in Fig. 17 (A) .
The shorted end of the line has minimum voltage and impedance,
and maximum current. Conditions are reversed a quarter wavelength
back, but at the generator end a half wavelength away, conditions
are duplicated; the generator also "sees" a short. In Fig. 17 (B)
the end of the line is an open circuit and the generator also sees
an open circuit. In either case the half-wavelength line has caused
the source to see whatever load conditions might exist. The idea
may be carried further, as in Fig. 18, where it is apparent that the
generator will see a resistance if the load is a resistance. Even if the
load is a capacitor, the generator, due to a capacitor's current-leading
characteristics, will see a capacitive reactance equal to the reactance
of the capacitive load; if the load is a coil, the generator will see
an inductive reactance equal to the reactance of the coil.
The statement may be broadened to say that anything looking
into one end of the half-wavelength line - or any multiple of a
39
half-wave line - will see whatever is connected to the other end.

For example, if we ran short of 300-ohm line when connecting a
transmitter to its antenna, we could employ another piece of line
of any impedance to make up the difference, providing the added
piece is exactly a half-wavelength or multiple of a half-wavelength
long. This is illustrated in Fig. 19. It should be clear, in view of
(A)
Fig. 17. Characteristics af

(A) shorted-end and (BJ open-
end half-wavelength lines.
(B)
GEN.SEES LOAD IS GEN.SEES LOAD IS
CAMC"A'CEUCAMOTA"°'
1--½--j
""". Li """. 1-½---l

Fig. 18. Characteristics of half-wavelength line In which impedance is neither a
short nor an open circuit.
APPLICATIONS OF TRANSMISSION LINES 41
ANTENNA
300.'1. 300.'1.LINE
TRANSMITTER - - - - - . ANY ~o
' - - ) . OR ANY I
2 MULTIPLE---!
(A) ANTENNA
72.n
TRANSMITTER
( B)
Fig. 19. Use of half-wave line to extend insufficiently long line af different z•.
the previous discussion, that conditions are repeated every half

wavelength; i.e., the impedance presented at one end is merely
"repeated" at the other end. The disadvantage of the above is that
it will work only on one frequency, or harmonics of one frequency.
If the frequency changes, the electrical length of the line changes
and is no longer a half wavelength. On the other hand, if the fre-
quency is exactly doubled, or trebled, or multiplied by any integer,
the length of the line becomes an integral number of half-wave-
lengths, and can still function as described above. As indicated in
Fig. 19, either parallel conductor or coaxial types may be used.
However, use of both in one installation is not desirable because
of the different distribution of current.
The half-wavelength line has a polarity-inverting ability similar
to a 1: 1 isolating transformer. This is illustrated in Fig. 20 (A) ,
where polarities of voltage and current are shown. The same reversal
of polarity would occur every odd half wavelength (Fig. 20B) ;
similar phase conditions are met every even multiple of a half
wavelength (Fig. 20C) . A half-wavelength line is most important at
high frequencies, at which it possesses moderate physical dimensions.
Other uses will be discussed when we cover radar applications.
+E~OI
~-E PHASE
REVERSAL
~ (A)
-E~--
1• t .,
+Er----. I
OI~-~....,_-~"'1,;.:----::;.,,,C.-----::..,j"!!=--~.:::---31
Fig. 20. Polarities on half-wavelength lines and multiples af half-wave lines.
16. The Quarter-Wavelength Line
In many ways the quarter-wavelength line is the reverse of the

half-wavelength line. In Fig. 21 (A) the quarter-wavelength line
segment has a short circuit at one end, but this time the generator
"sees" an open circuit, not a short. In Fig. 21 (B) the quarter-wave-
length line has an open circuit at the load end, and the generator
sees a short. Whereas a half-wavelength line "repeated" the load
impedance to the source, a quarter-wavelength line has the power
to invert the load impedance. Thus a low impedance at one end
of the quarter-wavelength line looks like a high impedance at the
other end. The same thing happens if the line is an odd multiple
of a quarter wavelength in length. Because of the impedance-in-
verting property, we call a quarter-wavelength line an impedance
trans/ormer_: it may be used, among other things, as an impedance
I I
'<i!
\
~/
I
\. I
"" / E
GENERATO._R_S_E_ES_ _ _ _ _O_P~
SHORT CIRCUIT CIRUIT
SHORTED OUARTER OPEN OUARTER

WAVELENGTH LINE WAVELENGTH LINE
(A) (B)
Pig. 21. Conditions on a quarter-wave line section.
matching device. Below the vhf region, a quarter wavelength is a

number of feet (at 14 me it is approximately 16 feet) and is only
useful where there is sufficient room for its application.
A dipole antenna (at resonance) has a center impedance of 72
ohms, but this is not necessarily the most desirable impedance for
the connecting transmission line. The line, for example, might well
be of the 400-ohm open wire variety. Now if we connect a quarter
wavelength piece of transmission line between the 72-ohm antenna
and the 400-ohm line, and if we make this quarter wevelength
segment a particular characteristic impedance, the 72-ohm antenna
will "see" 72 ohms when looking into the segment, while the 400-ohm
line will "see" 400 ohms when looking into the other end. The
arrangement is shown in Fig. 22. The value of Z 0 for this matching
segment may be calculated from:
z. = V 400 X 72 = V 28,800
Z8 = 170ohms
Where Z, Required characteristic impedance of quarter-wave-
length matching section; it is always equal to the
geometric mean of the two impedances to be matched.
In practice one would choose the commercially available type of
line with the nearest Z 0 , which in this case is 150-ohm twin lead.
Otherwise, sufficient open wire line of proper dimensions for the
desired Z 0 must be constructed. The physical length of this segment

is shorter than a quarter wavelength in free space because of the
velocity constant.
The same line segment may be used to obtain an impedance
match between two transmission lines with dissimilar Z0 's. Figure
23 shows how the connection is to be made between two lines of
Fig. 22. Open quarter-wave-

length line as matching trans-
former between antenna and
MATCHING
TRANSFORM ER
line.
~o• 110.n.
400A OPEN WIRE
/<AIR INSULATED)
52 ohms and 600 ohms Z0 respectively. The value of the Z0 for the
line segment may be derived in the same manner as follows:
Z8 = V 600 X 52 =V 31,200
z. = 170ohms
17. Resonant Lines
As has been explained, the most efficient method of transmission

of r-f energy is over nonresonant lines. However, it is not always
sufficiently economical or physically practical to use a matched line.
This is especially true of cases in which an antenna is to be used at
a number of different frequencies at different times. A good example
is the operation of a single antenna in a number of harmonically
related frequency bands by radio amateurs. In such cases, a resonant
transmission line is frequently used.
A resonant line has the disadvantage that its input impedance
varies with both line length and frequency. To overcome this disad-
vantage, arrangements must be made to use a line length which
places a current loop or node at the transmitting end, because at
loops or nodes the impedance is purely resistive. If the line length

cannot be made such an exact length, the reactance in the input
impedance must be "tuned out" by some matching device. The
matching device must also match the high or low input impedance
of the line to the impedance of the plate tank of the transmitter
from which it receives energy.
Figure 24 illustrates the use of the popular "zepp" antenna used
in amateur and other communications. The illustration shows how
it may be used on two harmonically related frequencies and the
type of matching circuit required. By choice of either high or low
impedance points along an antenna, and use of a transmission line
an integral number of quarter wavelengths along, the principles of
line sections previously discussed can be used to determine whether
the input impedance is high or low.
It should be emphasized that standing waves always go along
with resonant lines, and that the proper line length and matching
arrangement simply make it easier to couple power into and out of
the line. The resonant line still has the disadvantage of high losses,
radiation and induction effects, and high peak voltages. For this
reason, it is considered desirable to use only air-insulated open-wire
type lines in this kind of application.
18. Line Matching with Stubs
Another method of removing standing waves from a line whose

load does not match Z 0 is by use of stubs. Stubs are short lengths
of transmission line so cut that they exhibit a desired impedance.
i!O • 177 ll.

,..,
Fig. 23. Quarter-wavelength
line as matching transformer
52A
r' 600A
between two transmission ~

lines.
I. 4
,..
:I
A stub connected across the unmatched line at the proper point
(as illustrated in Fig. 25) cancels the reactance due to mismatch
and makes the line between the stub and the source nonresonant.
Standing waves then exist only on the stub and on the line between
the stub and the load, normally a negligible part of the system.
>-12ANTENNA >, ANTENNA
TRANSMISSION LINE _ TRANSMISSION LINE

- lENGTH • >-14 LENGTH• >-12
SERIES TUNING FOR

-LOW IMPEDANCE
~-TRANSMITTER ~-TRANSMITTER
I I
PLATE TANK I I
PLATE TANK
(A) AT f•fL (B) AT f•2fL
Fig. 24. Example of use af resonant line to couple rf to a single antenna at either
of two harmonically related frequencies.
A short-circuited or open-circuited stub exhibits a pure reactance,

except at exactly the quarter wavelength and multiples thereof. The
stub therefore does not influence the resistive component of the
line; it only equalizes the reactance. The stub must be connected
at a point on the line where the input resistance component of the
parallel combination of the stub and line to the load equals Z,,, so
that the line from there to the source will be matched. When such
a point on the unmatched line is found, the reactance at that point
is the reactance that must be canceled by the stub. The stub must
therefore be of such a length that its reactance is equal in magnitude
to the reactance of the line at that point, but of opposite polarity.
The stub and the line from it to the load then form a parallel-
resonant circuit, of which the effective resistance must be equal to Z 0 •
The parallel connection complicates the problem of determining
proper stub connecting point. The unmatched line input resistance
we ordinarily speak of is the resistance effectively connected in series
with input reactance of the unmatched line. However, this is not
the input resistance which remains after the stub has canceled the
input reactance. To analyze the line in such a way as to find the
point at which the input resistance is Z 0 after the stub has canceled
the reactance, we must deal in admittance (Y) , which is the recip-
rocal of (1 divided by) the impedance. Y is made up of two com-
ponents: (1) a resistance-like component, called conductance (g)
and (2) a reactance-like component, called susceptance (b). Com-
MATCHED
LOAD
"~
Fig. 25. How a stub may be conneded to a line to provide matching.
ponents g and b combine to form Y in the same way that R and X

combine to form Z. The stub is chosen so that the b of the stub is
equal and opposite in polarity to the b of the line. Then the con-
ductance remaining is the true total conductance, and its reciprocal
is the resistance component, which must be equal to Z0 at the stub
connecting point.
The variations of conductance, g, and susceptance, b, for a typical
case of a line with a SWR of 2 are shown in Fig. 26. These variables
are expressed in relation to g0 , the conductance represented by Z 0 •
The stub connecting point is chosen at the place where line g = g0 •
By following the vertical line down to the b curve, we see that b
at that point is approximately 0. 72 g0 • Since b is negative (inductive) ,
a stub with a positive b (capacitive) must be used. The length of
the stub can be determined by the formulas for z, given earlier,
remembering that b = - I /x. In this case, an open-ended stub
should be about 0.1,\ long and is connected to point 0, and a
shorted stub should be about 0.15,\ long and connected to point S.
Notice in Figs. 26 and 27 that there are two possible stub con-
necting points near each voltage maximum. The connecting point
nearer the load is always best matched by an open-ended stub
(because for the same g it is much shorter than a shorted stub).
The connecting point nearer the source from the voltage maximum
is always best matched by a shorted stub. If the SWR of the line is
measured, and the locations of the voltage maxima (or minima)
are noted, the stub location can be determined with reference to
the voltage maximum nearest to the load, without knowledge of
the characteristics of the load. (See Fig. 27.)
The calculation of stub locations and lengths is a rather complex
process, and the foregoing description has been included to explain
G•Qo•.L
---
i!o
1.0
-I~
"0
0.5 ~ "'
l&I
>
5
--------------,---+---~o ~
II:
.Cl
0
z
C
0.5 .,.
.15ll.t-----l.15ll.
._____.__ _.__..____._ _.__..____._ _.__..____._ _. 1.5

TO SOURCE LINE ,..____,
-------------------11 LOAD
Fig. 26. Variation of conductance Cg) and susceptance Cb) of a line with SWII = 2.
Either point where g = 9• is appropriate for stub connection.
the principle of this matching. All the necessary data have been
derived and plotted. Such a plot is shown in Fig. 28.
19. Quarter-wave Line as Impedance Matching Transformer
If the frequency does not change, a shorted quarter-wavelength

line may serve simultaneously as a resonant circuit and an impedance
matching transformer. In Fig. 29 (A) a shorted quarter-wavelength
line is acting as a resonant tank circuit and at the same time as a
matching transformer to a lower impedance transmission line. By
varying the point of connection with the transmission line, a match
may always be obtained, since the impedance on the quarter-wave-
length section varies from zero at the shorted end to near-infinite at
the open end. Parallel conductor line is usually balanced to ground
EMAX
i----~-----l
4
SHORTED S T U B r - - - - '
THIS SIDE OF
II
EMAX I ,___ OPEN STUB THIS
.-----. I SIDE OF EMAX
I
I
I
LOAD
Fig. 27. Locations of shortest open ond shorted stubs relative to E... point
on unmotched line•
. 22
LOCATION OF STUB
- --- -
FROM EMAX POINT ,_
in .20
::c I i..- 1- i-
t; t- - i..,,, l
z .18
l&I
.J
\
\ 1.....
~" V ~"'"
I
I
~ .16 STUB OPEN
cri ;
'I/ ,,"
I.I~ ~
~ ~ .14
c:> )(
z C
'I(
~
l&I
i :l w
2 .12
! J ,...
~
~~ .10 .... STUB SHORTED

31= 0:
zi..
;;; ~ .08
I ' ~
""i....
~I;;
11> 1&. .06 '"~ - ....
.._o 1-- ......
01&1
::c () .04
I- z
c:>~
~ !!?-02
.J 0
0
2 34 56 7 8 9 10
STANDING WAVE RATIO• EEMAX
MIN
Fig. 28. Graph showing proper length and connecting point for open and shorted
stubs for SWII volues up to 10.
u,
0
UHF EQUIVALENT CONVENTIONAL UHF EQUIVALENT CONVENTIONAL
f
,, /
TUNED I
CIRCUIT //
//
I
/ I
/
I f
I I
I I
/
/-AOJUSTAIU
CONTACTS TRANSMISSION ........... , :
.>\ '\
I LINE
I
I
'ADJUSTABLE
\ :,a
CONTACTS .;..
-I
J: l ~
z
I?' (A) BALANCED VERSION I
?' (B) UNBALANCED VERSION
en
3';
c;;
UHF EQUIVALENT CONVENTIONAL UHF EQUIVALENT CONVENTIONAL en
0
TUNED
CIRCUIT z
C:
z
C:
--,)\ TRANSMISSION
LINE
. '\\
'ADJUSTABLE.
CONTACTS
q q (0) BALANCED VERSION

(C) UNBALANCED VERSION
Fig. 29. Use of quarter-wove line sectloM 01 motchlng transfonter.

r 2" -ANTENNA
L7
HIGH~/
~o---.....
"
2 a b
ANTENNAS
½ '
4" -t TRANSFORMER
J
MATCHING
TRANSFORMER
(A) J ANTENNA ( 8) HALF-WAVE ANTENNAS IN PHASE
Fig. 30. Use of quarter-wave transformer with antennas.
and thus requires a balanced matching device; hence both sides of the
line are connected to the same impedance point on either conductor.
Coaxial cable is usually unbalanced, with its outer conductor
grounded. Figure 29 (B) shows an unbalanced impedance trans-
former. In both examples, the line segment is used as a step-down
transformer from the high impedance plate of an amplifier or
oscillator to a lower impedance transmission line.
In a receiver, a problem arises where the grid of the first tube
has a high impedance. It is necessary to employ a similar arrange-
ment, in which the low impedance line is matched to the high
impedance grid. In Fig. 29, C and D show graphically how balanced
and unbalanced transmission line segments may be employed to
perform this duty. The same principle may be used in matching
antennas, as illustrated in Fig. 30. At A is the popular "J" antenna,
which is simply a half-wave whose high-impedance end runs right
into a quarter wave whose other end is shorted and grounded. The
transmission line is then moved down the quarter-wave section until
its impedance is matched. At (B) is the connection for two half-wave
antennas in phase. Again, the high-impedance ends of the antennas
connect to the top of the quarter-wave transformer and the line is
"tapped down" on it. In this case the transformer also keeps voltage
at point a 180 degrees out of phase with voltage at point b, so that
the antennas operate in phase.
20. Lecher Wire Line
Often the operational frequency of a transmitter must be measured.

A popular device for this purpose is a Lecher wire frequency meter.
The principle of operation directly follows the previously discussed
theory; by creating standing waves on a transmission line segment
(usually greater than a half wavelength) and measuring the linear
distance between maximum or minimum points of current (or
voltage) , we may determine the wavelength of operation, and con-
sequently the frequency.
R-f energy is usually coupled into the system by use of an inductive
loop (Fig. 31A), one end of which is connected to the Lecher wires,
with the other end within inductive proximity of the oscillator tank
coil. Energy induced in this manner flows down the Lecher wires
to a shorting bar whose position along the line is adjustable. At this
point reflection takes place and energy bounces back to the tank
circuit. What the tank circuit sees when looking into the Lecher
wires depends upon the position of the shorting bar. If the shorting
bar is located a half wavelength from the tank, a short circuit is
seen; if it is located at odd multiples of a quarter wavelength, an
open circuit is seen. At points where the tank sees a short circuit,
most of its energy will be absorbed by the Lecher wires.
Of course, no energy is absorbed when the tank sees an open
circuit. To sense the change of energy in the tank circuit at different
positions of the shorting bar, a light bulb is also inductively coupled
to the tank circuit. Except when the tank sees a short circuit in the
Lecher wires, its energy is sufficient to light the bulb.
Operation is very simple. The light bulb is first made to light by
placing it close enough to the tank, and the shorting bar is moved
to either end of the Lecher wire system. As the shorting bar is
gradually moved to the other end, a point of series resonance (mul-
tiple of half wavelength distance from tank) is reached. This point
(where the light dies out) is carefully marked. The shorting bar is
now moved in the same direction until the next point is reached
where the bulb is extinguished. The distance between these two
points is a half wavelength. From this measurement the frequency
may be determined from:
OOBUL.B
MOVABL.E
~!~K SHORTING
: ] ] COIL. / BAR
E' " " ~ ' - -

L.OOPS 2
(A)
SHORTED
END\
___...\~
TRANSMISSION~ BUL.B
L.INE SEGMENT
TANK Cl RCUIT
OF OSC
Fig. 31. Lecher wire frequency meter.
. 5905
Frequency m megacycles = d.1stance between • m
pomts • me
• h es
Instead of a bulb, the plate current of the oscillator, indicated by
a meter, may be observed. At half-wave intervals, the plate current
rises sharply.
Because of the size, this device is of little use below 100 megacycles.
For example, at broadcast frequencies it would be hundreds of feet
long.
21. Supporting Stubs
If the frequency is increased to 1000 me and over, the dielectric

losses of ordinary transmission line are prohibitively high. A rigid
SHORTED QUARTER- PARALLEL CONDUCTOR

WAVELENGTH SUP· TRANSMISSION LINE
PORTING STUBS
(A)
GROUND
COAXIAL CABLE
SUPPORTING QUARTER
WAVELENGTH STUB ( B)
CLOSED END
Fig. 32. Use of quarter-wave stubs as physical supports for line.
quarter wavelength transmission line stub is often used as a support

at these frequencies.
Figure 32 (A) shows just such an arrangement for parallel con-
ductor line. Note that the shorting bar is grounded and that the
open end is connected to the transmission line. Now any r-f energy
in the line that attempts to enter the stub will find an extremely
high impedance at the input of the stub. The impedance would be
resistive, because the stub is actually a parallel resonant circuit. An
additional advantage of the quarter-wavelength supporting stub
is that it provides a low resistance path for de to fl.ow to ground.
As an example, suppose a line is carrying energy at 400 me and
has a velocity constant of 0.9. The length of a quarter-wave stub
can be obtained as follows:
984 984 .
A./4 = ¼ x-,-X VC = ¼ X 400 X 0.9 = 0.55 ft.= 6% mch
As the frequency is increased, stubs are used almost exclusively as
the insulating and supporting medium. Figure 32 (B) illustrates the
coaxial version of this stub. The disadvantage of the system is that
the stubs are effective insulators at only one frequency. By proper
Q-lowering design of stubs, a limited frequency deviation is
permissible.
Since, when measuring the SWR, we must not disturb the condi-
tions already existing on the line, it is usually easier to use a high
resistance voltmeter for the purpose. Figure 33 shows a high resis-
tance r-f voltmeter being used. It consists of a quarter-wavelength
transmission line segment with an r-f ammeter connected to one
end, while the other end is provided with sliding contacts to move
back and forth over the transmission line on which we wish to
measure the SWR. The r-f ammeter has a very low internal resistance,
and is effectively a short circuit. The same shorted quarter wave-
length line segment is employed here. The only modification is the
insertion of an ammeter for measuring the current through the
MOVABLE
CONTACTS TRANSMISSION LINEi
Fig. 33. SWR meter employ-

1n g a quarter-wavelength
shorted stub.
THEMOCOUPLE
AMMETER
shorting bar. Of course the transmission line is the generator for

energy entering the quarter-wavelength stub, and the line always
sees a maximum impedance.
The reader is probably wondering how an ammeter can measure
voltage. The enigma may be solved by remembering one of the
first transmission line equations, which stated that Z0 =
E/1. If we
measure the current at the shorted end, and if the Z 0 of the stub is
known, the voltage at the other end may be calculated from:
E =Z0 XI
Even this calculation is not necessary since we are only interested
in the ratio of maximum to minimum voltage or current. As the
sliding contacts are moved along the line, maximum and minimum
readings on the ammeter may be used directly in determining the
SWR.
22. Delay Lines
A composite color television signal is composed of brightness and

chrominance signals. In all color receivers, the chrominance portion
is filtered out of the total signal - leaving just the brightness signal.
Due to the reactive elements in these filters, an inherent delay (time
lag) is introduced, which causes the chrominance signals to lag the

brightness signal in some receivers by about 1.2 microseconds.
To eliminate color misregistration, it is essential that something
be done to the brightness signal so that it does not reach the picture
tube before the lagging chrominance signals. The time delay of the
brightness signal is accomplished by using a delay line.
There are two major varieties of delay lines: the distributed
constant line and the lumped constant line. (The latter is often
called an "artificial transmission line.") The principle of operation
of a distributed constant delay line follows closely the previous
development of transmission line theory. In fact, the only theoretical
difference between the two is the amount of inductance and capaci-
tance per unit length.
The velocity of an electromagnetic wave in a transmission line
at radio frequencies is given by ~ - From this relationship it
can be seen that an increase of L and C of a transmission line will
lower the velocity of any wave travelling through the line. As the
velocity decreases, so does the wavelength; hence a lower velocity
will mean a greater number of wavelengths for a given physical
distance of transmission line. Since the wave's lagging phase angle
with respect to the source increase~ at the rate of 271' radians (360°)
for each wavelength covered, the total phase angle of displacement
for a given distance from the source will be greater if the velocity
decreases.
Physically, a distributed constant delay line is slightly different
from ordinary coaxial cable. (See Fig. 34.) The delay line consists
of a center helical or spiral wire conductor wound around a circular
insulating form. Over this wire is a layer of insulation. The next
layer is a braided conducting shield that is grounded, and surround-
ing the entire cable is a protecting and insulating plastic sheath.
Any transmission line could be used for a delay line, but since the
propagation velocity on most lines is well over 50 percent of the
speed of light, a delay of 0.5 microsecond would necessitate a line
more than 500 feet long. Some commercial delay lines decrease the
wave velocity to such an extent that a delay time of up to 0.5 micro-
second is possible in one foot of length.
Many delay lines have a characteristic impedance ranging around
1000 ohms. All the rules about terminating the line in its Z 0 must
be followed. Reflections will cause a distortion (of the wave) that
is dependent on frequency.
SCHEMATIC
OUTER PLASTIC REPRESENTATION
COATING
INSULATION
Fig. :U. Distributed constant delay line.
23. Artificial Transmission Lines
An artificial or lumped constant transmission line is a network

designed to simulate a real transmission line. Its advantage over an
actual line lies in its compactness, higher characteristic impedance,
and ease of construction. An artificial line a few inches long can
T T-f:T T
SCHEMATIC REPRESENTATION
Fig. 35. Lumped-constant

artiflcial transmission line.
simulate an actual line thousands of feet in length. For this reason

it is used in compact equipment too small to accept an actual line,
and is called upon to perform a variety of tasks, from a resonant
circuit to a delay line.
An artificial transmission line is constructed by lumping together
all the constants of a long real line (Fig. 35) ; actual coils and
capacitors replace the distributed constants. If the wavelength of
the signal applied to such a line is long, compared to the distance
between each of the lumped constants, the network "looks" to the
travelling wave as an actual transmission line. The upper frequency

limit at which the line ceases to look "real" is called the critical
frequency.
24. Transmission Line Oscillator Tank
At low frequencies, coils and capacitors are relatively efficient,

while a quarter wavelength line segment is large and unwieldy. 'As
the frequency is increased above 300 or 400 megacycles, the amount
of inductance and capacitance needed is so small and the losses in
coils and capacitors become so great that they are of little use.
Now the size of a quarter-wavelength line section becomes con~
veniently small enough so that it may be used for every one of the
applications discussed in this chapter.
Perhaps the most common use of a quarter-wavelength line is to
act as a resonant circuit, such as might be used as an oscillator tank
circuit. Let us again consider what happened at the input terminals
of the shorted quarter wave line in Fig. 36 (A) . The generator sees
a maximum impedance which is purely resistive at one frequency.
If the frequency is raised, the impedance drops rapidly and is
capacitative; if the frequency is lowered, the impedance drops and
is inductive. The line has zero impedance for direct current. The
description above fits exactly a parallel resonant circuit. If, on the
other hand, the line is open-circuited (Fig. 36B) , the generator sees
a minimum impedance only at the resonant frequency. The
impedance rises on either side of resonance - inductive above and
capacitive below - and will not pass de. An open-circuited quarter-
wavelength line thus looks like a series resonant circuit. An example
of how a transmission line can be used as a tuned circuit in a simple
uh£ oscillator is given in Fig. 37.
25. Transmission Lines in UHF and Microwave Region
The problems confronting conventional transmission line

operation in the high uh£, or sh£ (above 3000 me) region become
enormous. Radiation, dielectric, and conductor losses, in many
cases, consume most of the power intended to be propagated down
the line.
Although the losses in both parallel conductor line and coaxial
line increase with frequency, they do not increase at the same rate.
Due to high radiation losses, parallel conductor line is not usually
0=~ ~ _£7
~
1-~-1
4
(A)
1-~-J - ~
4
(B)
Fig. 36. Quarter-wavelength line as resonant circuit.
e+
ULTRAUDION
OSCILLATOR
(LOW FREQ.)
- (A)
RFC
B+
ULTRAUDION
OSCILLATOR
(UHF)
(B)
-
Fig. 37. Use of transmission line section as tuned circuit in a uhf oscillator.
used much above l 000 me. Coax, as previously explained, is not

susceptible to radiation loss at any frequency, when properly used.
The stipulation was made for parallel wire line that radiation
would be low only when the separation between conductors was
less than 0.01 wavelength at the frequency of operation. To maintain
a conductor spacing of less than this amount, and at the same time
increase the frequency (decreasing the wavelength), the conductor
spacing must be reduced. There is, however, a lower limit on the
distance, governed (if not by constructional difficulties) by the
. E2
amount of power (P) transmitted on the line. Smee P Z' and = 0
E =y P X Z0 , for a given Z0, E varies as the square root of P.

Thus, for a given power transfer, there must be a certain potential
difference between the conductors. If the distance between these
wires is too small for a given potential, arcing will occur. Considering
only the constructional problem, 0.1..\. at 2000 me is less than 0.06
inch.
Because parallel conductor losses are intolerable at very high
frequencies, coaxial cable is the only type of transmission line used
in that region. Coax does not, however, provide lossless transmission
line operation. Even with the most efficient stub-insulated lines, over
50 percent of the power may be dissipated through skin effect -
copper (l'R) losses. If the frequency of operation is so high that
the distance from the inner surface of the outer conductor to the
outer surface of the inner conductor is greater than a half wave-
length, the previously described method of propagation is altered,
and for all normal purposes, the line becomes useless as a conductor
of r-f power. Naturally, the distance between conductors may be
reduced, but if the spacing becomes inordinately small, voltage
breakdown again becomes a limiting factor. Coaxial lines are used
quite extensively in the radar, tv, and £-m fields - up to a frequency
of about 3000 megacycles (IO cm). Above approximately this fre-
quency, transmission lines, as described herein, are not frequently
used. Instead, another type of transmission device, called a wave-
guide, is most commonly employed.
26. Review Questions
(l) How do the magnitude and the phase angle of the output of a half-wave-
length line compare with its input?
(2) If there is not sufficient transmission line of the same z. available to
complete a run, how may the line be extended without interfering with the
impedance match?
(3) Explain the relations between the input and output impedances of a
quarter-wavelength line section.
(4) A 400-ohm transmission line is to be matched to a 72-ohm transmission
line, using a quarter-wave matching section. The frequency of operation of the
system is 50 me. How long must the matching section be physically, and what
must be its characteristic impedance?
(5) Explain a situation in which it might be desirable to use a resonant
transmission line for a use other than matching.
(6) How and why may one vary the length of a resonant transmission line
to effect an impedance match to the generator?
(7) What kind of impedance is offered by a matching stub?
(8) What is the function of a matching stub in terms of its impedance?

(9) Explain how and why stub matching is possible.
(10) If a quarter-wavelength impedance matching transformer such as that
illustrated in Fig. 29-B is adjusted to correctly match the output of a transmitter
to its transmission line, what will be the quantitative effect on the input and
output connections of this transformer with a decrease in frequency?
(II) When is it practical to use a shorted quarter-wavelength transmission line
as a resonant circuit? Why?
(12) If a Lecher wire frequency meter is used to measure an unknown frequency
from a transmitter, and it is noted that a dip in the indicating bulb's brightness
occurs at two points when the shorting bar is moved from one end of the Lecher
wire system to the other, what is the frequency of operation if the distance between
these two points is 3.2 feet?
(l!l) Why are Lecher wire frequency meters not used extensively below about
100 me?
(14) What is the main disadvantage of transmission line stub insulators?
(15) How can a low-resistance r-f ammeter be used to measure transmission
line voltage without disturbing the line?
(16) Why are delay lines necessary in color tv receivers?
(17) How do the total inductance and total capacitance of a delay line affect
its time delay?
(18) How does the length of a delay line affect its delay?
INDEX
Admittance, 15, 46 Parallel wire line, 4

Artificial trans111ission lines, 56, 57, 58 Phase shift, 12, 13
Physical length of line, formula for, 10
Characteristic impedance, 13, 15, 16, 17, Pipe-type coax, 3
20, 21, 23 Polyethylene, 3
Coaxial cable, 1, 2, 37, 51, 58, 59, 60
Conductance, 15, 46 Quarter-wavelength
Copper loss, 25, 37 line, 26, 33, 42, 43, 44, 48, 51
Critical frequency, 58
Radiation loss, 36
Delay lines, 55, 56
Reflected wave, 20, 21, 22, 23, 24, 26
Dielectric leakage, 37
Reflections, 17, 18, 19, 20, 21
Distributed constant line, 56
Reflection coefficient, 22
Resonant lines, 32, 33, 44, 45
Half-wavelength line, 26, 39, 40, 41
Hollow dielectric parallel conductor, 4
Shielded two-conductor transmission line, 4
Impedance curve, 26 Shunt capacitance, 5
1.mpedance match, 17 Shunt dielectric leakage, 5, 8, 37
Impedance mismatch, 21 Skin effect, 7, 8
Impedance per unit length, 5 Solid dielectric type, 2, 3
Impedance value, open circuit load, 27, 28 Supporting stubs, 53, 54, 55
Impedance value, short circuit load, 27, 28 Susceptance, 15, 46
Incident wave, 20, 21, 22, 24 Standing wave ratio, 28, 29, 30, 31, 32,
Induction loss, 37 36, 55
Input impedance, 32, 33 Standing waves, 23, 24, 25, 26, 27
Stubs, 3
"J" antenna, 51
Transmission line oscillator tank, 58
Lecher wire line, 52 Transmission lines, introduction, 1
Line losses, 33, 36, 37 Transmission lines, uhf, microwave region,
Line matching with stubs, 45, 46, 47 58, 59, 60
Line termination, 17, 22 Tuned line, 32
Loop, 24 Twinax, 3, 4
Lumped constants distributed, 4, 5 Twin lead, 3
Lumped inductance, 5 Twisted lead, 4, 5
Types of lines, 2, 4
Maximum power transfer, 23
Variations of constants, 5, 7, 8
Node, 24
Open wire, 3 Wavelength and velocity, 9, 10, 11, 12

Wire resistance, 5, 6
Parallel conductor transmission
line, 3, 24, 36, 37, 51, 58, 59, 60 "Zepp" antenna, 45
63

R-F Transmission Lines - Alexander Schure

Uploaded by

Copyright:

Available Formats

R-F Transmission Lines - Alexander Schure

Uploaded by

Document Information

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

R-F Transmission Lines - Alexander Schure

Uploaded by

Copyright:

Available Formats

ELECTRONIC TECHNOLOGY SERIES

4. ... • • 4J' • -:· ... , ·~·.' •

JOHN F. RIDER PUBLISHER, INC.

JOHN F. RIDER PUBLISHER, INC.

All rights reserved. This book or parts thereof may

Library of Congress Catalog Card No. 56-9744

Printed In the United States of America

An understanding of the fundamental concepts of transmission

FUNDAMENTAL CONCEPTS OF TRANSMISSION LINES

1. Introduction to Transmission Lines

A transmission line is exactly what its name implies: a line over

COAXIAL CABLE (A)

TWINAX ( B) SHIELDED PAIR (F)

HOLLOW TWIN LEAD (G)

AIR INSULATED (C)

COAXIAL CABLE PARALLEL CONDUCTOR

Most r-f lines consist of two conductors separated by an insulating

In the presence of a large amount of external noise, or where

3. Lumped and Distributed Constants

Since transmission lines consist of a very special arrangement of

place, while considering the remaining part of the unit length as

Fig. 2. Transmission line lumped constants.

Fig. 3. Skin effect.

(1) In one sentence define the function of a transmission line.

TRANSMISSION LINE OPERATION AND CHARACTERISTICS

6. Wavelength and Velocity

Since electromagnetic waves travel in a given direction at a finite

Transmission Line Velocity Constant

2-wire line (air dielectric) 0.98

75-ohm Twin-Lead o. 68 too. 71

150-ohm Twin-Lead o. 76 too. 77

as they do in free space. Transmission line materials have a dielectric

Wavelength (meters) = 3~0 X vc

a physical length, while "half-wavelength" would express electrical

,so•~o• @s••EBo• EBO" e,;o•

Fig. 5. Relation of phases of currents at different points along a transmission line.

time at each point progressively distant from the source. Specifically,

1£ f3 is in degrees per unit length and >,, is in the same length

Let us view the action of a wave as it progresses from the place

DISTANCE FROM GENERATOR -

SIMPLIFIED REPRESENTATION OF TRANSMISSION AT HIGH FREQUENCIES

Fig. 6. Characteristic impedance.

Z _ E, where E, = voltage across infinite line

1.5 2 2.5 3 4 5 6 7 8 9 IO .15 .2 .25 .3 .4 .5 .6 .7 .8 .9 1.0

where: D center to center spacing between wires

From the ratio of ~ it is seen that using a larger wire or a closer

Assume the generator in Fig. 8 (A) is looking into an infinitely

It would be well for the reader to reread the previous paragraph,

$ ~o • 300.n ~o • 300.n TO INFINITY-

$ ~o • 300.n ~-,· ,.,,,

namely, if a wave reaches the end of a transmission line and encoun•

WATER -+ DIRECTION OF PROPAGATION

,,, __ ------------- ---·--

SOURCE RESISTANCE Rs • 50 A (CORRESPONDING TO ~o OF LINE)

30 1.25 46.8 RL • 40A

maximum power value; although lower load resistances increase

power is governed by ;: , it should be clear that if the load is to

It may be easier to understand the effects in transm1S1on lines

In the simple (and fortunately most common) case in which the

To clarify the above ideas, consider some examples. First, suppose

.___. _...__._ _...__._ _...__._ _. 1.5