S5-1.3 Pulse Analog & Pulse Digital Modulation
You can compare PAM, PWM and PPM, generate and reconstruct PCM, estimate quantization noise, apply companding, and analyse DPCM, DM and adaptive DM.
This unit is the crossing from analog to digital, and PCM is the crossing point - sample, quantize, encode, and the signal becomes numbers. Quantization noise is the price of that conversion and the 6 dB per bit rule of thumb is worth carrying. Companding exists because speech spends most of its time at low amplitudes, so uniform quantization wastes levels where the signal never goes; delta modulation is the minimal case of the same idea and its slope overload is the classic failure to be able to describe.
Work through these
Types of pulse analog modulation: PAM, PWM, PPM (qualitative)
Three ways to carry a message on a train of pulses, taken qualitatively. They are the bridge between the continuous schemes above and the digital ones below, and are worth being able to sketch rather than analyse.
NPTEL: Analog Communication · CoursePCM generation and reconstruction
The crossing point of the whole subject: sample, quantize, encode, and the signal becomes numbers. Everything in modern communication sits downstream of this item.
Estimation of quantization noise
The price of turning a continuous value into a number. Carry away the rule of thumb that each extra bit buys about six decibels, because it is the fastest way to size a converter in your head.
Non-uniform quantization and companding
Speech spends most of its time at low amplitudes, so equal sized steps waste levels where the signal never goes. Companding fixes that by making the steps unequal, and it is why telephone audio sounds as good as it does at the bit rate it uses.
Differential PCM (DPCM)
Encoding the change instead of the value, because consecutive samples are usually similar. It is the first appearance of prediction as a compression idea, which is the same insight the source coding topic formalises.
Delta modulation and adaptive delta modulation
The minimal case of the idea above, with a single bit per sample. Slope overload is its classic failure and the one to be able to describe: the staircase cannot climb fast enough to follow the signal.
Noise in PCM and DM
What actually limits these schemes in practice, and why the two failure modes are different. Quantization noise you designed for; slope overload means the design does not fit the signal.
Lab: sampling and reconstruction (PAM); PWM and PPM modulation and demodulation
The bench version of the pulse schemes. Reconstruction is the part to watch, because seeing the original come back out of a train of pulses is what makes the sampling theorem concrete.
Lab: PCM generation and detection; linear and adaptive delta modulation
The bench version of the digital half. Comparing linear against adaptive delta modulation on the same input shows you the slope overload problem and its fix in one sitting.
Sign in to keep your progress.
Free resources
We haven't checked most of these for screen reader use yet.
Links last checked 30 Aug 2026.
Stuck here?
Ask a mentor. A real person answers, and they can see exactly which topic you're on. Usually within a couple of working days.
Checking your session…
Topics shown in module order.