Pulse Modulation

Pulse Modulation: The Bridge Between Analog and Digital

Pulse modulation represents a fascinating intermediate step between purely analog and purely digital communication. Instead of modulating a continuous carrier wave, pulse modulation encodes information onto a train of discrete pulses. These techniques form the conceptual bridge that helps us understand how continuous analog signals get transformed into digital representations suitable for modern communication systems.

What Is Pulse Modulation?

Think of it this way: instead of continuously varying a carrier (like AM or FM does), pulse modulation takes periodic "snapshots" of the analog signal and represents each snapshot as a characteristic of a pulse — its height, width, or position. This is fundamentally different from continuous-wave modulation because information exists only at discrete time instants.

Pulse modulation comes in two families:

• Analog pulse modulation — The pulse parameter varies continuously (PAM, PWM, PPM)
• Digital pulse modulation — The pulse parameter is quantized to discrete levels (PCM, DM, DPCM)

Let us explore the analog pulse modulation techniques first, as they provide the intuition for understanding digital encoding.

Pulse Amplitude Modulation (PAM)

PAM is the simplest pulse modulation technique — the amplitude of each pulse is proportional to the instantaneous value of the analog signal at the sampling instant.

Mathematical expression: xPAM(t) = Σ x(nTs) × p(t - nTs)

Where x(nTs) is the sample value at time nTs, and p(t) is the pulse shape (usually rectangular with width τ < Ts).

Natural PAM vs. Flat-Top PAM:

• Natural PAM: Pulse amplitude follows the signal shape during the pulse width (like cutting out slices of the original waveform)
• Flat-top PAM: Pulse amplitude is held constant at the sample value (sample-and-hold) — this is what practical circuits produce

Flat-top PAM introduces a slight high-frequency attenuation described by the sinc function: H(f) = τ × sinc(fτ). This "aperture effect" can be compensated by an equalizer at the receiver.

Key Parameters:

• Sampling rate: fs = 1/Ts ≥ 2fm (Nyquist criterion)
• Duty cycle: τ/Ts (typically 10-50%)
• Bandwidth: Approximately 1/(2τ) for the pulse train

Pulse Width Modulation (PWM)

In PWM (also called Pulse Duration Modulation or PDM), the width (duration) of each pulse is proportional to the signal amplitude at the sampling instant. The amplitude and position of all pulses remain fixed.

How it works:

• Larger signal value → wider pulse
• Smaller signal value → narrower pulse
• Zero signal → pulse width equals nominal value (50% duty cycle typical)

Generation method:

1. Sample the analog signal at regular intervals
2. Compare each sample with a sawtooth (ramp) waveform
3. Output is HIGH when signal exceeds ramp, LOW otherwise

This produces pulses whose width linearly tracks the input signal.

Advantages of PWM:

• Inherently noise-resistant (information is in timing, not amplitude)
• Easy to generate with simple comparator circuits
• Pulses can be transmitted at full power (constant amplitude)
• No amplitude-dependent noise effects

Disadvantage: Bandwidth varies with signal amplitude — wide pulses occupy more time, potentially causing inter-pulse interference. Maximum bandwidth efficiency is lower than PAM.

Applications: PWM is widely used for power control rather than communication: LED dimming, motor speed control, audio Class D amplifiers, and switching power supplies.

Pulse Position Modulation (PPM)

PPM encodes the signal amplitude as the time displacement of a pulse from its nominal position. All pulses have the same amplitude and width — only their timing varies.

How it works:

• Larger signal value → pulse shifted later from reference position
• Smaller signal value → pulse shifted earlier
• Zero signal → pulse at nominal position

PPM can be derived from PWM: differentiate the PWM signal to get narrow pulses at the rising and falling edges, then retain only the trailing edges (which carry the timing information).

Key properties:

• Constant amplitude — Can use efficient non-linear amplifiers
• Constant pulse width — Fixed bandwidth per pulse
• Better noise performance than PAM and PWM — Timing information is more noise-resistant
• Requires synchronization — Receiver must know the reference position

Bandwidth: PPM has the widest bandwidth requirement among analog pulse modulation schemes because the receiver must handle the full range of possible pulse positions.

Comparison of Analog Pulse Modulation Techniques

From Analog to Digital: The PAM-PCM Connection

PAM is the critical first step in the analog-to-digital conversion process. The PCM encoder works as follows:

1. Anti-aliasing filter — Limits input bandwidth to fm (prevents aliasing)
2. Sample-and-hold — Creates flat-top PAM signal
3. Quantizer — Rounds each PAM sample to nearest discrete level
4. Encoder — Converts quantized level to binary code

The transition from PAM to PCM is the quantization step — replacing continuously-valued pulse amplitudes with discrete levels. This irreversible step introduces quantization noise but enables all the advantages of digital transmission.

Signal Recovery and Reconstruction

To recover the original analog signal from pulse-modulated waveforms:

PAM recovery:

1. Low-pass filter with cutoff at fm (removes pulse spectrum, retains baseband)
2. Aperture equalization (compensates for sinc roll-off)
3. Output: reconstructed analog signal

This works because of the sampling theorem — the PAM signal contains all the information needed to reconstruct the original waveform perfectly, provided sampling was at or above Nyquist rate.

PWM recovery:

1. Low-pass filter — averages the variable-width pulses
2. Alternatively: integrate-and-dump followed by sample

PPM recovery:

1. Convert PPM back to PWM (using flip-flop triggered by reference and PPM pulses)
2. Low-pass filter the resulting PWM signal

Multiplexing with Pulse Modulation

Pulse modulation naturally enables Time-Division Multiplexing (TDM). Since each channel occupies only a brief time slot (pulse duration << sampling period), multiple channels can share the same medium:

Channel capacity: N channels possible if N × τ ≤ Ts

Example: With fs = 8 kHz and τ = 5 μs per channel:

• Available time per frame: 1/8000 = 125 μs
• Maximum channels: 125/5 = 25 channels

This principle underlies all digital telephone multiplexing (E1: 32 channels in 125 μs frame).

Modern Applications

While analog pulse modulation is rarely used for long-distance communication today, the concepts remain vital:

1. PAM-4 — Used in 400 Gigabit Ethernet (four amplitude levels = 2 bits/symbol)
2. Class D audio amplifiers — PWM drives speakers with 90%+ efficiency
3. Optical PPM — Deep-space optical communication uses PPM for power efficiency
4. Delta-sigma ADCs — Modern audio converters use oversampled 1-bit PWM principles
5. Servo control — PWM drives motors, servos, and actuators in robotics

Key Takeaways

1. Pulse modulation encodes analog information onto discrete pulses via amplitude (PAM), width (PWM), or position (PPM).
2. PAM is the first step in digitization — adding quantization converts PAM to PCM.
3. PWM and PPM offer better noise immunity than PAM because timing information is more robust than amplitude.
4. The sampling theorem guarantees perfect reconstruction from PAM samples taken at or above the Nyquist rate.
5. Pulse modulation naturally enables TDM — multiple channels interleaved in time on a shared medium.
6. These foundational concepts persist in modern systems: PAM-4 in high-speed networking, PWM in power electronics, PPM in optical space communication.