Unit-IV-Pulse modulation & Digital
Modulation
Staff: Dr. R.Kalidoss
Objective
1. To introduce the concepts of pulse modulation
2. To introduce and understand sampling theorems
3. To discuss about different types of pulse modulation
techniques
Pulse Modulation
• The process of changing the characteristics of
pulse carrier in accordance with the modulating
signal is called pulse modulation.
• The major points of difference between analog
modulation and pulse modulation are; in analog
modulation technique simple or complex
sinusoidal signal is considered as a carrier
signal, whereas in pulse modulation, carrier
signals are periodic rectangular trains of pulse
signals.
• Fig Carrier signal format in pulse modulation
Types of Pulse modulation
• The various types of analog pulse modulation techniques
based on these characteristics are:
1. Pulse amplitude modulation (PAM)
2. Pulse width modulation (PWM)
3. Pulse position modulation (PPM)
• Types of Digital Pulse modulation:
1. Pulse code modulation (PCM)
2. Delta modulation
3. Adaptive delta modulation
Pulse amplitude modulation (PAM)
• The process of changing the amplitude of the pulse
carrier signal in accordance with the modulating signal is
called “Pulse Amplitude Modulation” (PAM) which is also
referred to as "sampling process".
Fig: Model of a sampler
Generation of PAM signals / Sampling
operation
• PAM signals can be generated using electronic switches.
The inputs for the electronic switches are continuous
time modulating signal and train of periodic pulses.
When the switch is closed, the corresponding instant
message signals are arrived at the output side. The
output is zero when the switch is open. Continuous time
signals are converted into discrete-time signal due to this
process. The output signals are sampled signals.
Sampling operation
Input or
x (t)
message signal = x (t)
t
g(t)
Carrier signal = g(t)
t
0 T 2T 3T 4T 5T 6T 7T 8T 9T
xs (t)
Sampled signal = xs (t)
t
Mathematical representation
Let x(t) be the modulating signal (continuous)
g(t) be the carrier (pulse) signal and xs (t ) sampled signal.
Now g(t) are periodic train of pulses. Any periodic signal can represent
by Fourier series. Hence Fourier series representation of carrier
signal g(t) is given by
∞
()
g t = ∑ Cn e j 2πnf s t
(1)
n = −∞
where Cn is a n th Fourier coefficients and it can be expressed as,
fs
2
Cn =
1
∫ g (t )e − j 2πnf s t
(2)
T fs
−
2
The output of a electronic switch or sampler circuit is given as
x s (t ) = x(t )g (t ) (3)
This expression describes the sampling operation in time domain.
• The sampled signal in frequency domain can also be
expressed by finding the Fourier transform of a signal xs (t )
x(t ) →
FT
x( f )
∞
( )
x f– = ∫x t e ()
− j 2πft
dt (4)
−∞
The frequency domain signal of sampled signal can be
obtained in a similar manner.
x s (t ) →
FT
xs ( f )
∞
( )
xs f = ∫ xs t e ()
− j 2πf t
dt (5)
Sub (3) in−(5),
∞ we get
∞
xs ( f ) = (t )g (t )e − j 2πf t
∫x dt (6)
−∞
Sub value of g(t) from eq. (1) to eq. (6)
∞ ∞
= ∫ x (t ) ∑ Cn e
j 2πnf s t − j 2πf t
e dt
−∞ n = −∞
∞ ∞
= ∑ C ∫ x(t )e
n = −∞
n
− j 2π ( f − nf s )
dt
−∞
From the definition of Fourier transform,
∞
∫ x (t )e − j 2π ( f − nf s )
dt = X ( f − nf s )
−∞
∞
xs ( f ) = ∑ C X ( f − nf )
n s
n = −∞ (7)
Eq. (7) can be represented pictorially, as shown in figure. It shows
sampling in time domain introducing periodicity in the frequency
domain. i.e. the same band limited spectrum repeated for every
sample interval.
• The conclusion drawn from the figure are:
1. Band limited signal x(t ) can be perfectly recovered at the
receiver side only when the sampling frequency is
fs ≥ 2 fh .
2. When sampling frequency does not satisfy the above
condition, recovery of the original signal at the receiver
side is not possible. Also the signals are also affected by
adjacent samples. This effect is referred to as "Aliasing
effect". Hence the condition for Aliasing : f s ≤ 2 f h .
In other words, for perfect reconstruction 1
of sampling
interval (Ts ), is always higher than 2T . i.e. rate of closer of
electronic switch at the transmitter side must satisfy the
condition of Nyquist interval.
Sampling Theorem or Nyquist Criterion for
sampling operation
A band limited signal x(t ) which has no frequency
components above f h , can be completely specified by
samples at a rate greater than or equal to 2 f h .
i.e. fs ≥ 2 fh (8)
Perfect recovery of the signal is not possible at the receiver
side, when the sampling interval at the transmitter side is
not followed.
Detection of PAM signal
• Just passing the spectrum of x s ( f ) to low pass filter may be done
for reconstructing the signal. Low pass filter allows the spectrum
from the band − f h to + f h and suppresses all other side bands.
Fig. illustrates the detection circuit of PAM signal.
Fig: Detection of PAM signal
