Data Converters

Courtesy: Doug Hinckley Lee Huynh

Dooroo Kim
Byron Johns
Danny Carpenter
Stephanie Pohl
Harry “Bo” Marr
2
 Methods of Sampling

Ideal sampling
- an impulse at
each sampling
instant

Fig. Ideal Sampling

 Methods of Sampling
Natural
sampling - a
pulse of short
width with
varying
amplitude
with natural
tops

Fig. Natural Sampling

 Methods of Sampling
Flat-top
sampling - a
pulse of short
width with
varying
amplitude
with flat tops

Fig. Flat-top Sampling

Sampling Process: Introductory Note
Sampling of the signal
Periodic signal
spectrum in the
in the time
frequency domain
domain

By Duality

Sampling of the Making the spectrum of

signal in the time the signal periodic in
domain the frequency domain

6
Motivation
Filtering & Sampling (1)

Technical Presentation
 Page 7
Motivation
Filtering & Sampling (2)

Technical Presentation
 Page 8
Motivation
Filtering & Sampling (3)

Technical Presentation
 Page 9
 Sampling Theorem
• Sampling Theorem for Band-limited Signals:
– A band-limited signal of finite energy which has no frequency
components higher than W Hz is completely described by specifying
the values of the signal at instants of time separated by 1/2W seconds
– A band-limited signal of finite energy which has no frequency
components higher than W Hz may be completely recovered from
knowledge of its samples taken at the rate of 2W samples per second

• fS=2W is called the Nyquist Rate

• tS=1/2W is called the Nyquist interval

10
Motivation

Filtering & Sampling (4)

➢ Most of the speech contents lies in between 300 – 3400 Hz
➢ According to Nyquist theorem Fs >= 2 fm (to avoid aliasing)
➢ A value of 8kHz is selected (8 >= 2*3.4).
➢ For good quality16 bits are used to represent each sample.
➢ Bit-rate = 8kHz *16 bits = 128 kbps Input Rate

The Input rate could even be more, for example in Skype: 16 kHz sampling
frequency is used in skype and so resulting to an input rate of 192 kBit/s.
But, this is a waste of bandwidth that could rather be used by other services
and applications.

Source Coding (Speech Coding in this Context)

Technical Presentation
 Page 11
 Quantization
Understanding Quantization
To understand quantization a bit more let’s have a look at the following Example:

Technical Presentation
 Page 12
13
 Uniform Quantization

• A quantizer with equal quantization level is a Uniform Quantizer

• Each sample is approximated within a quantile interval
• Uniform quantizers are optimal when the input distribution is uniform
– i.e. when all values within the range are equally likely

• Most ADC’s are implemented using uniform quantizers

• Error of a uniform quantizer is bounded by q q
− e
2 2
14
 Signal to Quantization Noise Ratio
◼ If q is the step size, then the maximum quantization error
that can occur in the sampled output of an A/D converter is q
V pp
q=
L
where L = 2n is the number of quantization levels for the converter. (n is
the number of bits).

• The mean-squared value (noise variance) of the quantization error is given

by:

2 1 
q/2 q/2
1 q/2 2
 2 =  e p(e)de =  e   de =  e de
2

−q / 2 −q / 2 q q −q / 2
1 e q/2
q 2
=q =
3

3 −q / 2 12
16
17
 Data Conversion
➢ Basically we have two types of data in nature: Analog and Digital.

➢ Data transmission – Electrical wires (for analog) and databus (for digital).

➢ Most of the real-world physical quantities are in analog form.

➢ Data storage and processing – Analog versus Digital. (That’s the idea for
data conversion!!!)

CD players
 Motivation for ADC
• Analog signals are continuous, with infinite values in a given
range.
• Digital signals have discrete values such as on/off or 0/1.
• Limitations of analog signals
– Analog signals pick up noise as they are being amplified.
– Analog signals are difficult to store.
– Analog systems are more expensive in relation to digital systems.
 Comparators
• Determines which
input is larger
• A small difference
between inputs results
maximum output
voltage (high gain)
• Zero-level detection
• Non-zero-level
detection
Digital to Analog (D/A, DAC, or
D-to-A) Conversion
 Types of DACs
• Many types of DACs available.
• Usually switches, resistors, and op-amps used
to implement conversion
• Two Types:
– Binary Weighted Resistor
– R-2R Ladder
 Binary Weighted Resistor
Voltages V1 through Vn are either Vref
Vref if corresponding bit is high or
ground if corresponding bit is low V1
R

V1 is most significant bit V2 2R I Rf

V3 4R
Vn is least significant bit
- Vout
+
Vn 2n-1R

MSB
LSB
 V1 V2 V3 Vn 
Vout = − IRf = − Rf  + + +  n -1 
 R 2R 4R 2 R
 Binary Weighted Resistor

If Rf=R/2
 V1 V2 V3 Vn 
Vout = − IRf = − + + +  n 
2 4 8 2 
For example, a 4-Bit converter yields

 1 1 1 1
Vout = −Vref  b3 + b2 + b1 + b0 
 2 4 8 16 
Where b3 corresponds to Bit-3, b2 to Bit-2, etc.
 Binary Weighted Resistor
• Advantages
– Simple Construction/Analysis
– Fast Conversion

• Disadvantages
– Requires large range of resistors (2000:1 for 12-bit DAC)
with necessary high precision for low resistors.
– Can be expensive. Therefore, usually limited to 8-bit
resolution.
 R-2R Ladder

• One form of ladder network is commonly used to scale down voltages to

certain weighted values for digital-to-analog Conversion called R/2R Ladder
Network
• Each bit corresponds to a switch: If the bit is high, the corresponding switch is
connected to the inverting input of the op-amp.
• If the bit is low, the corresponding switch is connected to ground.
Lets assume all the binary inputs are grounded at 0 volts, that is: VA = VB = VC = VD =
0V (LOW). The binary code corresponding to these four inputs will therefore be: 0000.
Therefore with a binary code of “0000” applied as inputs, our basic 4-bit R-2R digital-to-
analogue converter circuit would look something like this:
 R-2R Ladder
• Advantages
– Only two resistor values (R and 2R)
– Does not require high precision resistors
• Disadvantage
– Lower conversion speed than binary weighted
DAC
 Resolution
• Smallest analog increment corresponding to 1 LSB change
• An N-bit resolution can resolve 2N distinct analog levels
• Common DAC has a 8-16 bit resolution
 Applications
• Digital Motor Control
• Computer Printers
• Sound Equipment (e.g. CD/MP3 Players, etc.)
• Electronic Cruise Control
Back to A/D Converters….
Flash type A/D

Courtesy: Prof. Indranil Sengupta

 Flash
Advantages Disadvantages
• Simplest in terms of
operational theory • Lower resolution
• Expensive
• Most efficient in terms of • For each additional
speed, very fast output bit, the number of
• limited only in terms of comparators is doubled
comparator and gate
• i.e. for 8 bits, 256
propagation delays
comparators needed
 Successive Approximation A/D Converter
Circuit

Output of the
successive
approximation
register

Courtesy: Prof. Indranil Sengupta

 Successive Approximation A/D Converter
Circuit
• The SAR (successive approximation
register) begins by turning on the
MSB Bit7.
• Vo of the D/A converter is compared
with the analog input voltage Vin in
the comparator.
• If analog voltage is less than the
digital voltage, Bit7 is turned off and
Bit6 is turned on.
• If analog voltage is greater than the Display
digital voltage, Bit7 is kept on and Bit6
Bit 7 is set: b7=1
is turned on. If Va < Vd → b7=0; b6=1
• The process of turning bit on/off is If Va > Vd → b7=1; b6=1
continued until Bit0. …..
• Now the 8-bit input to the D/A If Va < Vd → b7=0; …b0=1
converter represents the digital If Va > Vd → b7=1; … b0=1
equivalent of the analog signal Vin. Done
 Successive Approximation ADC

Output of DAC

Courtesy: Prof. Indranil Sengupta

 Successive Approximation
Advantages Disadvantages

• Capable of high speed and • Higher resolution successive

reliable approximation ADC’s will be
• Medium accuracy compared slower
to other ADC types • Speed limited to ~5Msps
• Good tradeoff between speed
and cost

• Capable of outputting the

binary number in serial (one
bit at a time) format.
ADC Types Comparison
ADC Resolution Comparison
Dual Slope
Flash
Successive Approx
Sigma-Delta

0 5 10 15 20 25
Resolution (Bits)

Type Speed (relative) Cost (relative)

Dual Slope Slow Med
Flash Very Fast High
Successive Appox Medium – Fast Low
Sigma-Delta Slow Low
 TUTORIAL QUESTIONS
• Question 1
– Assumes the input analog voltage is changing between 0-5 V.
– Using a 3-bit A/D converter draw the output as the input signal ramps
from 0 to 5V.
– Calculate the resolution.
– What is the maximum possible voltage out? (this is called the full-scale
output)
– If the output is 1000 0000, what is the input?
• Question 2
– Assumes the input analog voltage is changing between -5 to 5 V; using
a 10-bit A/D converter.
– Calculate the number of quantization levels.
– Calculate the voltage resolution.
 References

• Callis, J. B. “The Digital to Analog Converter.” 2002.

http://courses.washington.edu/jbcallis/lectures/C464_Lec5_Sp-02.pdf. 14 March
2006
• “DAC.” 2006. http://en.wikipedia.org/wiki/Digital-to-analog_converter#DAC_types.
14 March 2006.
• Johns, David and Ken Martin. “Data Converter Fundamentals.” © 1997.
http://www.eecg.toronto.edu/~kphang/ece1371/chap11_slides.pdf. 14 March
2006
• Goericke, Fabian, Keunhan Park and Geoffrey Williams. “Digital to Analog
Converter.” © 2005.
http://www.me.gatech.edu/mechatronics_course/DAC_F05.ppt. 14 March 2006
• http://www.engin.brown.edu/courses/en123/Lectures/DAconv.htm
• http://www.seas.upenn.edu/~ese206/labs/adc206/adc206.html
• Interesting project ideas: http://www.byonics.com/
• More interesting project ideas… http://www.ke4nyv.com/picprojects.htm