CCIoT

Lecture 4 – DAC and ADC

Dr. Aftab M. Hussain,
Associate Professor, PATRIoT Lab, CVEST

18-03-2024 Lecture 4 1
Sensor systems

Read-out
Interface
• Converts physical circuit • Converts the
signals into analog voltage into
• Converts electrical • Send the digital
electrical digital at some
properties into information to the
properties sampling frequency
voltage processor

Digitizing
Transducer
circuit

Reverse process for actuators

2
Need of ADCs and DACs
• We know that sensor outputs (e.g., a voltage measured with a
thermocouple or a speech signal recorded with a microphone) are analog
quantities, varying continuously with time
• However most of the processing, storage and happens in digital format
• An ADC (Analog-to-Digital Converter) is used to convert an analog signal to
the digital format
• The reverse conversion (from digital to analog) is also required (mostly for
actuator operation)
• For example, music stored in a DVD in digital format must be converted to
an analog voltage for playing out on a speaker
• A DAC (Digital-to-Analog Converter) is used to convert a digital signal to the
analog format

18-03-2024 Lecture 3 3
DAC
• A DAC is a circuit that takes digital
signal and provides a corresponding
analog output VR

• For a four bit DAC with input

A3A2A1A0, the output voltage is:
𝑉𝑅
𝑉𝐴 = 8 𝐴3 + 4 𝐴2 + 2 𝐴1 + 𝐴0
15
• In general, the output is:
𝑛−1
𝑉𝑅
𝑉𝐴 = 𝑛 ෍ 𝐴 𝑘 2𝑘
2 −1
𝑘=0

18-03-2024 Lecture 3 4
DAC – Kelvin divider
• The most intuitive way of
implementing a DAC is using the
Kelvin divider
• In this case, we use 2n resistors of
equal values in series with the
reference voltage
• The voltage is thus divided into 2n
equal intervals
• The output is then tapped from an
appropriate position using a decoder
circuit in conjunction with 2n
switches
18-03-2024 Lecture 3 5
DAC – Kelvin- divider
• Advantages:
• Simple design
• It is configurable to non-linear DACs if
specific step size is required for
particular applications

• Disadvantages
• 2n switches and resistors can explode
exponentially
• Large power consumption
• Decoding circuit can become
complicated for large values of n

18-03-2024 Lecture 3 6
DAC – Weighted resistors
• We can connect the input signals to
switches such that the VR appears at
a terminal if the input is high
• Such a connection can be used in
conjunction with resistors of specific
value to obtain the analog voltage
• If the input bit Ak is 1, the terminal
gets connected to VR ; else, it gets
connected to ground
• Thus,
𝐴𝑘 𝑉𝑅 − 0 𝐴𝑘 𝑉𝑅
𝐼𝑘 = =
𝑅𝑘 𝑅𝑘

18-03-2024 Lecture 3 7
DAC – Weighted resistors
• From the non-inverting opamp, we
have:
𝑛−1
𝐴𝑘 𝑉𝑅
𝑉𝐴 = −𝑅𝑓 𝐼 = −𝑅𝑓 ෍
𝑅𝑘
𝑘=0
2𝑛−1 𝑅
• If we use 𝑅𝑘 = :
2𝑘 𝑛−1
𝑅𝑓 𝑉𝑅 𝑘
𝑉𝐴 = −𝑅𝑓 𝐼 = − ෍ 𝐴 𝑘 2
𝑅 2𝑛−1
𝑘=0
• Thus, with only n resistors and n
switches, the circuit can be made
• Rf can be used to introduce
additional gain

18-03-2024 Lecture 3 8
DAC – Weighted resistors
• Advantages:
• n resistors and 2n switches
• Current is only drawn when input is
applied, no static power consumption
• The Rf value can be varied to obtain a
gain along with D to A conversion

• Problems:
• Hard to find resistors of exact values
such as R, 2R… 16R, 32R etc.

18-03-2024 Lecture 3 9
DAC – R-2R ladder
• The R-2R ladder circuit consists of
the input voltage applied to a
resistance network made of resistors
of values R and 2R as shown
• The output can be thought of as an A0VR A1VR A2VR A3VR
Thevenin equivalent circuit with,
𝑅𝑡ℎ = 𝑅
𝑉𝑅
𝑉𝑡ℎ = 𝐴3 23 + 𝐴2 22 + 𝐴1 21 + 𝐴0
16
• Thus,
𝑅𝑓 𝑉𝑅
𝑉0 = − 𝐴3 23 + 𝐴2 22 + 𝐴1 21 + 𝐴0
𝑅 16

18-03-2024 Lecture 3 10
DAC – R-2R ladder
• Advantages:
• Does not require resistors of specific
values
• Can be made using only 2n resistors
and 2n switches
A0VR A1VR A2VR A3VR
• No static power consumption

• Due to its many advantages, R-2R

ladder is often used in fabricating
commercial DACs

18-03-2024 Lecture 3 11
ADCs
• To convert from analog to digital, we
8(𝑉𝑅 /8)
may think of dividing the reference 111
voltage by 2n and consider each voltage 7(𝑉𝑅 /8)
interval (corresponding to 000, 001, 110
etc.) as a bin 6(𝑉𝑅 /8)
101
• If the input voltage VA falls in the 100 5(𝑉𝑅 /8)
bin; therefore, the output of the ADC 100 VA
would be 100 4(𝑉𝑅 /8)
011
• Thus, the basic idea behind an ADC is 3(𝑉𝑅 /8)
simple: 010
• Generate reference voltages V1, V2, etc. 2(𝑉𝑅 /8)
001
• Compare the input VA with each of Vi to 1(𝑉𝑅 /8)
figure out which bin it belongs to 000
• If VA belongs to bin k, convert k to the 0(𝑉𝑅 /8)
binary format

18-03-2024 Lecture 4 12
ADC – parallel/flash
• In case of the parallel ADC, the input
voltage is compared with the Vref
divided into bins using a voltage
divider

• The output of the comparators

depends on the level of the input
voltage with respect to these bins

• This output is decoded into the

digital output

18-03-2024 Lecture 4 13
ADC – parallel/flash
• Advantages:
• Speed – the ADC is not called flash for
nothing! Flash ADCs handling 10+ Gsps
are commercially available
• Disadvantages:
• Number of comparators and resistors is
2n
• Static power is consumed because Vref
is continuously subjected to voltage
divider
• The comparators may not settle to the
correct output value together for a
changing input – leading to error in
output

18-03-2024 Lecture 4 14
ADC – Ramp type (or counting)
• An interesting way of making an ADC
is using an internal DAC to compare
the output with the input signal
• We start a digital counter at the start
of every conversion
• The digital counter output is
converted into analog and compared
with the input signal
• When the comparison just becomes
high, the counter is stopped
• The output of the counter is the
digital equivalent of the analog input

18-03-2024 Lecture 4 15
ADC – Ramp type (or counting)
• Advantages:
• Simpler circuit compared to the flash
ADC for large value of n

• Disadvantages:
• Requires a DAC
• Is very slow – in worst case, requires 2n
clock cycles to complete. On average
2n-1. This reduces the sampling
frequency

18-03-2024 Lecture 4 16
ADC – Successive approximation
• We can use a DAC and adjust one bit
at a time to obtain the correct digital
output
• Suppose we have a 4-bit DAC
• Start with D3D2D1D0 = 0000
• Set MSB to 1 (D3 = 1) keeping other bits
unchanged
• If VDAC > VA, set D3 back to 0, else keep
D3 at 1
• Repeat these steps for successively
lower bits
• At the end of four steps, the digital
output is given by D3D2D1D0

18-03-2024 Lecture 4 17
ADC – Successive approximation

18-03-2024 Lecture 4 18
CCIoT

Sensors outputs - digital

Dr. Aftab M. Hussain,
Associate Professor, PATRIoT Lab, CVEST

18-03-2024 Lecture 4 19
Sensor outputs – Digital
• Digital communication is preferred over analog because it is less susceptible
to noise
• There are multiple ways in which you can obtain digital
• Parallel – with each bit on a separate wire
• Serial – with bit transmitted one after the other
• In serial communication we can different protocols:
• UART (asynchronous)
• SPI (synchronous)
• I2C (synchronous)

18-03-2024 Lecture 4 20
Sensor outputs – Digital – Parallel
• Parallel interfaces transfer
multiple bits at the same time
• They usually require buses of
data - transmitting across eight,
sixteen, or more wires
• Advantages:
• Very high data rates (single clock
transfer)
• Easy to implement
• Disadvantages:
• Large number of data lines
required, specially if number of
peripherals are large

18-03-2024 Lecture 4 21
Sensor outputs – Digital – Serial
• Serial interfaces stream their data,
one single bit at a time
• These interfaces can operate on as
little as one wire
• Serial interfaces can be
synchronous and asynchronous
• A synchronous serial interface
always pairs its data line(s) with a
clock signal, so all devices on a
synchronous serial bus share a
common clock
• Asynchronous means that data is
transferred without support from
an external clock signal

18-03-2024 Lecture 4 22
Sensor outputs – Digital – UART
• A universal asynchronous
receiver/transmitter (UART) is a
serial communication protocol that
employs two lines Tx and Rx for
communication
• UART support is commonly found
inside microcontrollers
• For example, the Arduino Uno -
based on the "old faithful"
ATmega328 - has just a single UART,
while the Arduino Mega - built on
an ATmega2560 - has a whopping
four UARTs
• NodeMCU has two UARTs

18-03-2024 Lecture 4 23
Sensor outputs – Digital – UART
• Advantages:
• Two line communication
• Simple to implement in software
• Legacy protocol
• Disadvantages:
• No synchronization means we have to
make “baud rates” equal manually
before communication
• Low data rate – general baud rate is
9600 bits per second
• Hardware implementation is complex
• Needs start and stop bits to sync –
which can be wasteful
• Rx and Tx pins can be very confusing!

18-03-2024 Lecture 4 24