Encoding and modulation of digital signals

We know that digital signals are stored as binary numbers in memory cells. Let’s see how these digital signals can be transferred in other words – how they can be encoded in to electrical signal shape. When signal has to be sent it has to be modulated.

Pulse Code Modulation (PCM)

Probably simplest and most common modulation method is pulse code modulation (PCM).

PCM_pulse_code_modulation.JPG

In CPM coding information is modulated as voltage levels. If supply voltage is 5V, then binary”1″ is represented as +5V while 0 as 0V level. Only two voltage levels are used to modulate signal. Signal can be sent in parallel (bus mode) or serial mode. In parallel mode whole (let’s say 8 bit) signal can be sent at once via 8 wires, while in serial mode data has to be sampled in time. This transmission method is most popular in embedded systems, for instance, ADC sends converted data to DSP using CPM. CPM method is OK but it has several drawbacks like number of wires required for parallel buses, but probably most dangerous problem is that transmission may experience skew problems when each data wire may have different time delays because of geometrical differences. Especially when data rates is high – all bits may not arrive at same time to receiver what may lead to errors. So CPM encoding method is mainly used for short distances. Of course there can be serial method used instead of parallel, but then we are loosing speed. In other hand we are saving space in PCB because only one wire is needed.

Serial transfer mode has more possible modulation methods like PAM, PPM, PNM, PWM, and PDM. Each of them has a different schema.

Pulse Amplitude Modulation (PAM)

During PAM modulation the actual amplitude of the pulse represents the number being transmitted. This is similar to analog signal transmission but it is discrete in time.

PAM_pulse_amplitude_modulation.JPG

Pulse Position Modulation (PPM)

During PPM transmission, signal pulses are fixed width and amplitude, but the actual number is represented by pulse position in time.

PPM_pulse_position_modulation.JPG

Pulse Number Modulation (PNM)

In this case digital number is represented by number of pulses in a given time slot. Signal width and amplitude is fixed.

PNM_pulse_number_modulation.JPG

Pulse Width Modulation (PWM)

This is more common encoding method in power and motor control. In this case pulse width in a given time period represents the number being transmitted.

PWM_pulse_width_modulation.JPG

Usually in PWM mode signal amplitude isn’t interesting as it is fixed value. But in some cases second information value can be carried as amplitude. In this case we get combination of PAM and PWM – simplest way to achieve this is to multiply PWM and PPM signals.

Pulse Density Modulation (PDM)

In PDM case information value is represented as signal period (frequency). In other words digital value is represented by pulse density.

PDM_pulse_density_modulation.JPG

In order to achieve one or another modulation there has to be some standardization needed. To transmit and receive signals properly there are CODEC (coder-decoder) chips used. CODECs may be separate chips or integrated in to microprocessors and microcontrollers. Usually they operate in hardware level and you don’t need to take care of signal formation. Probably most common is PCM modulation interface. All processors and other digitals devices have PCM interfaces that can handle this type of signal.

Leave a Reply