Pulse Modulation

Dr Kumbirayi Nyachionjeka
• Is a process of changing a binary pulse signal to represent the
information to be transmitted.

• The primary benefits of transmitting information by binary

techniques arise from the great noise tolerance and the ability to
regenerate the degraded signal.

• Any noise that gets added to the binary signal along the way is
usually clipped off.

• Further, any distortion of the signal can be eliminated by

reshaping the signal with a Schmitt trigger, comparator, or
similar circuit.
• If information can be transmitted on a carrier consisting of
binary pulses, these aspects of binary techniques can be used
to improve the quality of communications.

• Pulse modulation techniques were developed to take

advantage of these qualities.

• The information signal, usually analog, is used to modify a

binary (on/off) or pulsed carrier in some way.
• With pulse modulation the carrier is not transmitted continuously
but in short bursts whose duration and amplitude correspond to
the modulation.

• The duty cycle of the carrier is usually made short so that the
carrier is off for a longer time than the bursts.

• This arrangement allows the average carrier power to remain

low, even when high peak powers are involved.

• For a given average power, the peak power pulses can travel a
longer distance and more effectively overcome any noise in the
system.
• There are four basic forms of pulse modulation:
• Pulse-amplitude modulation (PAM),
• pulse-width modulation (PWM),
• pulse-position modulation (PPM),
• pulse-code modulation (PCM).
 Comparing Pulse-Modulation Methods
• Fig. 8 shows an analog modulating signal and the various
waveforms produced by PAM, PWM, and PPM modulators.

• In all three cases, the analog signal is sampled, as it would be in

A/D conversion.

• The sampling points are shown on the analog waveform.

• The sampling time interval t is constant and subject to the

Nyquist conditions described earlier.
• The sampling rate of the analog signal must be at least two
times the highest frequency component of the analog wave.

Fig. 8
 Pulse Amplitude Modulation
• The PAM signal in Fig. 8 is a series of constant-width pulses
whose amplitudes vary in accordance with the analog signal.

• The pulses are usually narrow compared to the period of

sampling; this means that the duty cycle is low.
 PWM
• The PWM signal is binary in amplitude (has only two levels).
The width or duration of the pulses varies according to the
amplitude of the analog signal: At low analog voltages, the pulses
are narrow; at the higher amplitudes, the pulses get wider.
PPM
• The pulses change position according to the amplitude of the
analog signal. The pulses are very narrow.

• These pulse signals may be transmitted in a baseband form,

but in most applications they modulate a high-frequency radio
carrier.

• They turn the carrier on and off in accordance with their shape.
• Of the four types of pulse modulation, PAM is the simplest and
least expensive to implement. On the other hand, because the
pulses vary in amplitude, they are far more susceptible to noise,
and clipping techniques to eliminate noise cannot be used
because they would also remove the modulation.

• PWM and PPM are binary and therefore clipping can be used to
reduce the noise level.
• Although the techniques of pulse modulation have been known
for decades, they are no longer widely used.

• Of the three types, PWM is the most common. One example is

for remote-control purposes, e.g., in model airplanes, boats, and
cars.

• Pulse-width modulation (PWM) methods are also used in switch

mode power supplies (dc-dc convertors, regulators, etc.), motor
speed control, as well as in class D audio switching power
amplifiers.
• Today pulse-modulation techniques have been largely
superseded by more advanced digital techniques such as
pulse-code modulation (PCM), in which actual binary numbers
representing the digital data are transmitted.
 Pulse-Code Modulation
• The most widely used technique for digitizing information
signals for electronic data transmission is pulse-code
modulation (PCM).
• PCM signals are serial digital data. There are two ways to
generate them.
• The more common is to use an S/H circuit and traditional A/D
converter to sample and convert the analog signal to a
sequence of binary words, convert the parallel binary words to
serial form, and transmit the data serially, 1 bit at a time.

• The second way is to use the delta modulator described earlier.

Fig. 9
 Traditional PCM.
• In traditional PCM, the analog signal is sampled and converted
to a sequence of parallel binary words by an A/D converter.

• The parallel binary output word is converted to a serial signal by

a shift register (see Fig. 9).

• Each time a sample is taken, an 8-bit word is generated by the

A/D converter.

• This word must be transmitted serially before another sample is

taken and another binary word is generated.
• The clock and start conversion signals are synchronized so that
the resulting output signal is a continuous train of binary words.

• Fig. 10 shows the timing signals. The start conversion signal

triggers the S/H to hold the sampled value and starts the A/D
converter.

• Once the conversion is complete, the parallel word from the A/D
converter is transferred to the shift register.

• The clock pulses start shifting the data out 1 bit at a time. When
one 8-bit word has been transmitted, another conversion is
initiated and the next word is transmitted
Fig. 10
• In Fig. 10, the first word sent is 01010101; the second word is
00110011.

• At the receiving end of the system, the serial data is shifted into
a shift register (see Fig. 11).

• The clock signal is derived from the data to ensure exact

synchronization with the transmitted data. (The process of clock
recovery is discussed in Chap. 11.)

• Once one 8-bit word is in the register, the D/A converter

converts it to a proportional analog output.
• Thus the analog signal is reconstructed one sample at a time as
each binary word representing a sample is converted to the
corresponding analog value.

• The D/A converter output is a stepped approximation of the

original signal. This signal may be passed through a low-pass
filter to smooth out the steps.
 Companding.
• Companding is a process of signal compression and expansion
that is used to overcome problems of distortion and noise in the
transmission of audio signals.

• The range of voice amplitude levels in the telephone system is

approximately 1000 : 1.
• In other words, the largest-amplitude voice peak is
approximately 1000 times the smallest voice signal or 1000 : 1,
representing a 60-dB range.

• If a quantizer with 1000 increments were used, very high-quality

analog signal representation would be achieved.

• For example, an A/D converter with a 10-bit word can represent

1024 individual levels

• If the maximum peak audio voltage were 1 V, the smallest

voltage increment would be 1/1023 of this, or 0.9775 mV
• As it turns out, it is not necessary to use that many quantizing
levels for voice, and in most practical PCM systems, a 7- or 8-
bit A/D converter is used for quantizing.

• One popular format is to use an 8-bit code, where 7 bits

represents 128 amplitude levels and the eighth bit designates
polarity (0 = + , 1 = -). Overall, this provides 255 levels; one-half
are positive, and one-half negative.

• Although the analog voltage range of the typical voice signal is

approximately 1000 : 1, lower-level signals predominate.
• Most conversations take place at a low level, and the human
ear is most sensitive in the low-amplitude range. Thus the upper
end of the quantizing scale is not often used.

• Since most signals are low-level, quantizing error is relatively

large.

• That is, small increments of quantization become a large

percentage of the lower-level signal.

• This is a small amount of the peak amplitude value, of course,

but this fact is irrelevant when the signals are low in amplitude.
The increased quantizing error can produce garbled or distorted
sound.
• In addition to their potential for increasing quantizing error, low-
level signals are susceptible to noise.

• Noise represents random spikes or voltage impulses added to

the signal. The result is static that interferes with the low-level
signals and makes intelligibility difficult.

• Companding is the most common means of overcoming the

problems of quantizing error and noise.

• At the transmitting end of the system, the voice signal to be

transmitted is compressed; i.e., its dynamic range is decreased.

•
• The lower-level signals are emphasized, and the higher-level
signals are deemphasized.

• At the receiving end, the recovered signal is fed to an expander

circuit that does the opposite, deemphasizing the lower-level
signals and emphasizing the higher-level signals, thereby
returning the transmitted signal to its original condition.

• Companding greatly improves the quality of the transmitted

signal.
• Originally, companding circuits were analog, and the concept is
most easily understood when described in analog terms.

• One type of analog compression circuit is a nonlinear amplifier

that amplifies lower-level signals more than it does upper-level
signals. Fig. 12 illustrates the companding process.

• The curve shows the relationship between the input and output
of the compander. At the lower input voltages, the gain of the
amplifier is high and produces high output voltages.
• As the input voltage increases, the curve begins to flatten,
producing proportionately lower gain.

• The nonlinear curve compresses the upper-level signals while

bringing the lower-level signals to a higher amplitude.

• Such compression greatly reduces the dynamic range of the

audio signal. Compression reduces the customary ratio of
1000 : 1 to approximately 60 : 1.

• The degree of compression can be controlled by careful design

of the gain characteristics of the compression amplifier, in which
case the 60-dB voice range can be reduced to more like 36 dB.
• In addition to minimizing quantizing error and the effects of
noise, compression lowers the dynamic range so that fewer
binary bits are required to digitize the audio signal.

• A 64 : 1 voltage ratio could be easily implemented with a 6-bit

A/D converter, but in practice, a 7-bit A/D converter is used.

• Two basic types of companding are used in telephone systems:

the μ-law (pronounced “mu law”) compander and the A-law
compander.

• The two companders differ slightly in their compression and

expansion curves.
• The μ-law compander is used in telephone systems in the
United States and Japan, and the A-law compander is used in
European telephone networks.

• The two are incompatible, but conversion circuits have been

developed to convert μ-law to A-law and vice versa.

• According to international telecommunication regulations, users

of μ-law companders are responsible for the conversions.

• The voltage formulas for both are as follows:

where Vout = output voltage
Vm = maximum possible input voltage
Vin = instantaneous value of input voltage
The value of μ is usually 255; A is usually 87.6.
 Example 3

The input voltage of a compander with a maximum voltage range

of 1 V and a μ of 255 is 0.25. What are the output voltage and
gain?
 Example 4

The input to the compander of Example 4 is 0.8 V. What are the

output voltage and gain?