Communication
Systems
Pulse Code
Modulation
By
Engr. Jawwad Ahmad (Ph.D.)
1
Today’s Goal
PCM
Linear PCM
Non Linear PCM
Engr. Dr. Jawwad Ahmad 2
Pulse Code Modulation
PCM is a digital transmission system with an analog-to-digital
converter (ADC) at the input and a digital-to-analog converter
(DAC) at the output.
PCM convert each sample of digital time samples (DTS) into
digital code word.
A digital code word is generated corresponding to each
sample.
The steps are Sampling, Quantization and Encoding
(converted to serial).
Engr. Dr. Jawwad Ahmad 3
Pulse Code Modulation
Engr. Dr. Jawwad Ahmad 4
Pulse Code
Sampling generates Discrete
Modulation
Time Continuous Value
(DTCV) or Discrete Time
Continuous Amplitude
(DTCA) signal.
Continuous amplitude is
adjusted to fit into digital code
that is, to develop Discrete Time
Discrete Amplitude (DTDA)
signal.
Engr. Dr. Jawwad Ahmad 5
Quantization
Two methods are there to convert DTCA to DTDA.
Approximation.
Truncation
The number of quantized level (L) are 2n, where n is an integer
and it is actually number of encoding bits.
Engr. Dr. Jawwad Ahmad 6
Encoding
This uses logic circuitry.
A two bit PCM generator is to be considered.
Engr. Dr. Jawwad Ahmad 7
V Sampling,
o Quantization and
l Coding
t
a
g
e
Time
7 111
L 6 110 B
e 101 i C
5
v 4 100 n o
e 3 011 a d
l 2 010 e
r
s 1 001 s
0 000 y
Time
V
o 010101110111110101010
l
t
a
g
e
Engr. Dr. Jawwad Ahmad 8
Time
x(t)
0
t
Consider the analog Signal x(t).
Engr. Dr. Jawwad Ahmad 9
x[n]
0
n
The signal is first sampled
Engr. Dr. Jawwad Ahmad 10
3
0
n
Engr. Dr. Jawwad Ahmad 11
3
0
n
Engr. Dr. Jawwad Ahmad 12
3
Sample n
Engr. Dr. Jawwad Ahmad 13
3
And Hold n
Engr. Dr. Jawwad Ahmad 14
3
0
n
Engr. Dr. Jawwad Ahmad 15
Assign Closest
Level
3
0
n
Engr. Dr. Jawwad Ahmad 16
3
0
n
Engr. Dr. Jawwad Ahmad 17
3
0
n
Engr. Dr. Jawwad Ahmad 18
3
0
n
Engr. Dr. Jawwad Ahmad 19
3
0
n
Engr. Dr. Jawwad Ahmad 20
Each quantization level corresponds to a unique combination of
bits. The analog signal is transmitted/ stored as a stream of bits
and reconstructed when required.
3
0
n
Engr. Dr. Jawwad Ahmad 21
Each quantization level corresponds to a unique combination of
bits. The analog signal is transmitted/ stored as a stream of bits
and reconstructed when required.
3
0
n
00 01 10 11 10 01 00
Engr. Dr. Jawwad Ahmad 22
x(t) Original Signal
0
t
Engr. Dr. Jawwad Ahmad 23
x~(t)
Quantized Signal
0
t
It is quite apparent that the quantized signal is not exactly the same as the original
analog signal. There is a fair degree of quantization error here. However; as the
number of quantization levels is increased the quantization error is reduced and
the quantized signal gets closer and closer to the original signal
Engr. Dr. Jawwad Ahmad 24
x~(t)
Quantized Signal
0
t
It is quite apparent that the quantized signal is not exactly the same as the original
analog signal. There is a fair degree of quantization error here. However; as the
number of quantization levels is increased the quantization error is reduced and the
quantized signal gets closer and closer to the original signal
Engr. Dr. Jawwad Ahmad 25
Quantization and Encoding of a Sampled signal
Engr. Dr. Jawwad Ahmad 26
Flash Type Linear PCM (ADC)
Flash Type ADC is based on the principle of comparing analog
input voltage with a set of reference voltages.
Since all the quantization level are equal linear PCM
generated, hence known as Linear PCM.
To convert the analog input voltage into a digital signal of n-bit
output, (2n – 1) comparators are required.
The three op-amps are used as comparators.
The non-inverting inputs of all the three comparators are
connected to the analog input voltage.
Engr. Dr. Jawwad Ahmad 27
Flash Type Linear PCM (ADC)
The inverting terminals are connected to a set of reference
voltages (V/4), (2V/4) and (3V/4) respectively which are obtained
using a resistive divider network and power supply +V.
Engr. Dr. Jawwad Ahmad 28
Flash Type Linear PCM (ADC)
The output of the comparator is at logic 1, when voltage at
non-inverting terminal is greater than voltage at inverting
terminal and is in negative saturation otherwise, as shown.
Engr. Dr. Jawwad Ahmad 29
Flash Type Linear PCM (ADC)
Consider first condition, where analog input voltage V is less
in
than (V/4).
In this case, the voltage at the non-inverting terminals of all the
three comparators is less than the respective voltages at
inverting terminals and hence the comparator outputs are
C1C2C3 = 000.
This comparator outputs are applied to the further coding
circuit to get the digital outputs as B1B0 = 00.
Similarly the digital outputs are calculated for other three
conditions also.
Engr. Dr. Jawwad Ahmad 30
Flash Type Linear PCM (ADC)
To obtain serial PCM, use parallel to serial convertor.
It is the fastest type of ADC because the conversion is
performed simultaneously through a set of comparators, hence
referred as flash type ADC.
The construction is simple and easier to design.
It is not suitable for higher number of bits.
To convert the analog input voltage into a digital signal of n-bit
output, (2n – 1) comparators are required.
Engr. Dr. Jawwad Ahmad 31
Non-Linear PCM
The magnitude change between two quantize levels is not
uniform rather the step size increases with the amplitude of the
input signal.
Non-Linear PCM is very complicated but it is widely in use.
PCM was first used in Telephony by AT&T (American
Telephone & Telegraph, USA) under the supervision of board
named as CCITT (Consultative Committee for International
Telegraphy and Telephony).
The committee concluded that PCM did not work properly.
Engr. Dr. Jawwad Ahmad 32
Non-Linear PCM
Problems were observed regarding SNR (Signal to Noise
Ratio) variation.
SNR varied over 40dB for
Close and Far Users
Faint and Loud Talkers
Problem was investigated and it was found that the problem
was due to uniform quantization levels for each case, i.e.,
linear PCM.
Solution was suggested to be Nonlinear or Companded PCM.
Engr. Dr. Jawwad Ahmad 33
Non-Linear PCM
Engr. Dr. Jawwad Ahmad 34
Non-Linear PCM
In actual the solution was to use compress quantized level at
the start of the scale and then expanded quantized levels for
higher values.
It was proposed to employ logarithmic scale for Nonlinear
PCM thus Companded PCM generation uses logarithmic
amplifiers.
Two standards were described
by CCITT
μ-Law PCM
A-Law PCM
Engr. Dr. Jawwad Ahmad 35
Non-Linear PCM
Engr. Dr. Jawwad Ahmad 36
μ-Law PCM
The input and output of μ-Law was standardized by CCITT as
Vi μ-Law Vo
Companded
where μ is called Companded parameter (the range varies
from 100 to 500, but CCITT recommended to use μ = 255) and
the ratio of Vi to Vp gives the normalized value.
μ should always be greater than 100.
Adopted in United States, Japan, India and Pakistan.
Engr. Dr. Jawwad Ahmad 37
SNR and Transmission Bandwidth width of μ-Law
PCM
For μ-Law, the SNR is given by the following equation
where L is number of quantized level (L = 2n).
While the transmission bandwidth is described as
where n is the encoding bits and B is the signal bandwidth.
For example, voice band is 32KHz at 8 bits.
Engr. Dr. Jawwad Ahmad 38
Effect of One Bit on SNR
Substitute L = 2n in SNR Equation of μ-Law as
Now let then
Convert it into dB
Engr. Dr. Jawwad Ahmad 39
Effect of One Bit on SNR
With an increment of a bit, the equation becomes
Subtract above two results
An increment of
Engr. Dr. Jawwad Ahmad 40
A-Law PCM
In this also algorithmic amplifiers are used but with different
topologies. Vi A-Law Vo
Companded
The input and output of A-Law was standardized piece wise
by CCITT as
where A is called Companded parameter (for optimal
operation CCITT recommended to use A = 87.6).
Adopted in Europe.
Engr. Dr. Jawwad Ahmad 41
PCM Decoder
To recover an analog signal from a digitized signal the
following steps should be taken:
Use a hold circuit that holds the amplitude value of a pulse till the next
pulse arrives.
Pass this signal through a low pass filter with a cutoff frequency that is
equal to the highest frequency in the pre-sampled signal.
The higher the value of L, the less distorted a signal is
recovered.
Engr. Dr. Jawwad Ahmad 42
PCM Decoder
Engr. Dr. Jawwad Ahmad 43
Thank you
Engr. Dr. Jawwad Ahmad 44
