Beruflich Dokumente
Kultur Dokumente
Introduction to
Introduction to
Communications
Lecture 15: Pulse Modulation
This lecture:
Pulse‐Amplitude Modulation
Pulse‐Time Modulation
Pulse‐Duration and Pulse‐Position Modulation
Ref: Carlson, Chapter 6.2‐6.3; Haykin, Chapter 3.3‐3.4
Pulse Modulation
Pulse modulation (PM) offers two potential advantages over CW 2
modulation
The transmitted power can be concentrated into short bursts
Th t itt d b t t di t h tb t
instead of being generated continuously. This enables the use of
devices which operate better in pulse regime than in CW regime,
mostly due to thermal reasons (e.g. semiconductor lasers,
microwave active devices)
Time between pulses can be filled by sampled signal form other
Time between pulses can be filled by sampled signal form other
sources – a scheme termed time‐division multiplexing.
Analogue PM has a disadvantage of requiring large bandwidth
compared to message bandwidth
dt b d idth
COMS3100 Lecture 15
Pulse Amplitude Modulation (PAM)
If a message waveform is adequately described by periodic 3
sample values, it can be transmitted using analogue pulse
mod lation herein the sample al es mod late the
modulation wherein the sample values modulate the
amplitude of pulse train.
Therefore, the amplitudes of regularly spaced pulses are
Therefore, the amplitudes of regularly spaced pulses are
varied in proportion to the corresponding sample values
of a continuous message signal x(t).
This technique is termed Pulse Amplitude Modulation
COMS3100 Lecture 15
Generation of the PAM signal
There are two operations involved in the generation of 4
the PAM signal:
1. Instantaneous sampling of the message signal x(t) every
Ts seconds, where the sampling rate fs = 1/Ts is chosen in
accordance with the sampling theorem
d h h l h
2. Lengthening the duration of each sample so obtained to
t t l τ (sample‐and‐hold)
some constant value ( l d h ld)
s (t )
x(t )
τ
Ts
COMS3100 Lecture 15
Flat‐top Sampling and PAM
Practical method for obtaining PAM (or implementing the 5
steps 1. and 2. is the sample‐and‐hold (S/H) technique.
This method produces flat‐top pulses
sample‐and‐hold circuit
waveform obtained
f
COMS3100 Lecture 15
Generation of PAM – holding network impulse response
The finite width pulse obtained by the sample‐and‐hold 6
circuit can be interpreted mathematically using the
following holding network. p(t )
Impulse response of
holding network
p (t
() ⎡t −τ / 2⎤
p(t) = Π ⎢
P( f ) ⎣ τ ⎥⎦
arg[ P( f )]
Amplitude response of
holding network
Phase response of
Phase response of
holding network
P( f ) = τ sinc fτ
COMS3100 Lecture 15
Flat‐top Sampling and PAM – sampled wave
Periodic gating of the S/H circuit generates the sampled wave 7
Sampling pulse train, p(t)
Signal value at kTS
So far we have not yet specified
th “ h
the “shape” of the p(t) or P(f)
” f th (t) P(f)
COMS3100 Lecture 15
PAM
Flat‐top sampling is equivalent to passing an ideal sampled wave 8
through a network having transfer function P( f ) = [p(t)]
Loss of high frequency content is called aperture effect
Loss of high frequency content is called aperture effect
The larger the pulse duration or aperture τ, the larger the effect
Can be corrected using equalizer
No equalization is needed if t/Ts<<1
Spectrum for ideal
sampling when
X( f ) = Π( f/2W )
−W W
aperture effect in
P( f ) = τ sinc fτ
flat‐top sampling
⎡t −τ / 2⎤
p(t) = Π ⎢ ⎥
⎣ τ ⎦ −W W
COMS3100 Lecture 15
PAM
There are many similarities between PAM and AM CW modulation
10
Modulation index
Spectral impulses
Spectral impulses
DC block
PAM spectrum extends from DC up through several harmonics of fs
Required transmission bandwidth can be estimated based on time‐
domain considerations
Assuming small pulse duration compared to time between pulses
Assuming small pulse duration compared to time between pulses
Adequate pulse resolution then requires
COMS3100 Lecture 15
Pulse‐Time Modulation
The sample values of a message can also modulate the time 11
parameters of a pulse train:
1. Pulse width – pulse‐duration modulation (PDM)
2. Pulse position – pulse‐position modulation (PPM)
The pulse width or pulse position varies in direct proportion to
p
the sample values of x(t)
()
COMS3100 Lecture 15
Pulse‐Duration and Pulse‐Position Modulation
In both cases a time parameter of the pulse is being modulated 12
In both cases amplitude remains constant
Methods for producing PDM and PPM are similar
COMS3100 Lecture 15
Generation of PDM and PPM
When x(t) exceeds the sawtooth wave ‐ comparator output is a 13
positive constant A
Otherwise comparator output is zero
This is an example of PDM
with trailing edge
modulation of the pulse
modulation of the pulse
duration
For PPM signal, the PDM
signal triggers a monostable
p
pulse generator (triggers on
g ( gg
trailing edge and produces
short pulse of fixed duration
COMS3100 Lecture 15
Generation of PDM and PPM
Sample values are nonuniformly spaced 14
This can be tolerated if tk- kTs<<Ts.
For nearly uniform sampling the duration of the
k‐th pulse in the PDM signal is
where the unmodulated duration τ0 0 represents x(kT
p ( s) = 0
PPM pulses have fixed duration and amplitude ‐> there can be
no potential missing pulses. k‐th pulse begins at time
where the unmodulated position kT
where the unmodulated position kTs+td represents x(kT
represents x(kTs) = 0 and
and
the constant t0 controls the displacement of the modulated pulse
COMS3100 Lecture 15
Generation of PDM
An approximation for the PDM can be formulated if we assume 15
rectangular pulses centred at t = kTs and assuming tk varies
slowly from pulse to pulse
slowly from pulse to pulse
where
PDM signal contains the message signal plus a dc component
and phase modulated signal at the harmonics of fs
COMS3100 Lecture 15
PTM to PAM conversion
Message can be reconstructed by first converting PDD/PPM to PAM. 16
Middle waveform is produced by a ramp generator that starts at time
kTs and stops at t
d t t tk
Demodulation requires received pulses with short risetime to
preserve accurate message information
preserve accurate message information
COMS3100 Lecture 15