- 1 -

ELEC2202 Communications Engineering Laboratory 3 ---- Digital Communication

1. Objectives On completion of this assignment you will be familiar with:

Analogue-to-digital conversion, The time-division multiplexing of digital signals, Digital-to-analogue conversion, Demultiplexing.

** Make sure the PCM and Link Analysis workboard is properly connected to your PC before beginning this laboratory. 2. Theory

Time Division Multiplexing Time Division Multiplexing (TDM) is the process of switching between two, or more, signals serially in time. Consider an analogue signal and let the amplitude of this signal be sampled every 10 ms, say. This is represented in the diagram below.

It can be seen that the magnitudes of the samples follow the analogue signal waveform and that there is a significant time (almost 10 ms) between each sample. It is quite possible to sample other analogue waveforms in the periods between the samples of the first one and then to combine the two sets of samples in one waveform. This is represented in the following diagram.

Often, instead of each sample being momentary, the sampled values are held until the next sample comes along. This is called sample and hold. An example of a time-division multiplexed, sample and hold waveform is shown in the next diagram.

- 2 -

Sampling From the preceding diagrams it can be seen that the sampled waveform barely resembles the original. Eventually the original waveform must be recovered from this sampled wave, so the sampled waveform must contain all the information about the original waveform to allow this. It would seem reasonable that there would be limitations on the sampling that might stop the required recovery of the original waveform. For instance, if brief samples of a conversation were taken at 5 minute intervals, it would be impossible to recover exactly what had been said! The frequency at which samples are taken is known as the Sampling Rate (fs) and the time between successive samples is the Sampling Interval (T). thus:

fs = 1/T

It can be shown, using Fourier transform techniques, that there is indeed a lower limit to the sampling rate which is related to the bandwidth of the signal being sampled. The mathematics to prove this are not presented here, however the results are most important and must be appreciated when dealing with any sampled-data system. To be able to faithfully recover a signal from its sampled version the sampling rate must be at least equal to twice the bandwidth of the original signal. That is, fs must be greater, or equal to 2B, where B is the bandwidth of the original signal. Another way of expressing this is that the sampling rate must be at least twice the frequency of the highest frequency component of the signal being sampled. For example: to be able to recover speech on a telephone circuit that passes audio frequencies between 300 Hz and 2.7 kHz, the sampling rate must be at least 5.4 kHz.

Analogue to Digital Conversion Before an analogue signal can be transmitted down a digital link it must be converted to digital form. This may be done by an analogue-to-digital (A/D) converter. Each sampled value of the analogue waveform is applied to the input of the A/D converter in sequence and a digital value is obtained for each. This process is known as digitizing, or encoding. A simple code often used when analogue waveforms (for example: speech) are digitized is the Binary Coded Decimal (BCD) code. There are many forms of A/D converter circuits and the study of them is outside the scope of this assignment.

Digital Time Division Multiplexing In a similar way that analogue waveforms may be interleaved in time to give analogue TDM, digital words may be multiplexed to give digital TDM. For example: if two digital data streams are:

word 1 2 3 4 5 ...... stream A 1100 1010 1001 0001 0101 ...... stream B 0010 0110 1000 0000 1101 ......

these may be multiplexed in the sequence: A1, B1, A2, B2, A3, B3 ....etc. This gives: A1 B1 A2 B2 A3 B3 ……. 1100 0010 1010 0110 1001 1000

- 3 -

as the digitally multiplexed data stream.

Bandwidth Requirement Suppose we have a two-channel, multiplexed, 3-bit PCM system with the multiplexing being done at 5000 times per second. Suppose, also, that there is no dead time between the 3-bit samples. Each 'frame' will comprise 6 bits (3 bits/channel x 2 channels/frame). This gives a bit rate of 6 x 5000 = 30k bit/s. The worst-case, in terms of transitions of state, is if alternate 0's and 1's are transmitted. This would give a square wave of frequency of half the bit rate, i.e. 15k bit/s. The absolute minimum bandwidth required to transmit the basic information of such a square wave is the fundamental frequency of the wave, i.e.: BWmin = (bit frequency)/2

Duplexing The multiplexed waveform, whether it be analogue, or digitally multiplexed, has to be demultiplexed to retrieve the original constituent waveforms, or data. The demultiplexing process is the opposite of multiplexing and the switching between channels must be synchronized with the multiplexing, or corrupted data or waveforms will result. 3. Task A --- Analogue TDM

This practical introduces the concept of the time-division multiplexing of two analogue signals. Multiplexing is the term given to describe the transmission of two or more signals down a common channel. The reason for multiplexing signals is to economize on channel or link usage and to be able to convey the maximum amount of information down any given link. If two, or more, signals can use the same cable, at the same time, then the link will be running more efficiently and the cost of the service to each will be lower. 4. Task B --- Analogue to Digital Conversion

In this practical you will investigate the analogue-to-digital conversion required to produce a digital signal from an analogue source. The number of bits in the digital output of an A/D converter is fixed by the circuit design of the converter. Typical converters have 8, 12, or 16-bit output codes. Any given A/D converter can only cope with analogue input voltages over a limited range. This means that this finite input range is converted to a digital output which has a limited number of bits. For example, an 8-bit converter can give an output of one of 256 (28) digital words. Suppose that the input range was 0 to 10 V. This means that the 10 V range will be split into 256 levels; i.e., each level covers 10 V/256 = 39 mV. This splitting the analogue range into small steps is known as Quantization. In the example given, each LSB increase in the digital output word represents a quantization step of 39 mV and the maximum error from the true value will be +/- half a LSB; i.e., 19.5 mV. This quantization error is unwanted and, as the actual instantaneous value of the error varies randomly with time, is often regarded as a noise component on the signal. It is often referred to as Quantization Noise. A simple Binary output code is often used when analogue waveforms (for example: speech) are digitized. The A/D converter used in this practical uses this code. 5. Task C --- Digital TDM

In this practical you will investigate the digital alternative method of time-division multiplexing. In a similar way that analogue waveforms may be interleaved in time to give analogue TDM, digital words may be multiplexed to give digital TDM. For example: if two digital data streams are:

word 1 2 3 4 5 ...... stream A 1100 1010 1001 0001 0101 ...... stream B 0010 0110 1000 0000 1101 ......

- 4 -

these may be multiplexed in the sequence: A1, B1, A2, B2, A3, B3 ....etc. This gives A1 B1 A2 B2 A3 B3 ……. 1100 0010 1010 0110 1001 1000 as the digitally multiplexed data stream. If the bit rate of each of the multiplexed signals is maintained, it will take a longer time to transmit the multiplexed data than it would for each individual stream alone. In the case of the example above it would take twice as long. Therefore a faster bit rate would be of advantage. However, the maximum bit rate that can be accommodated on any practical link is limited by the bandwidth of that link; i.e. maximum bit rate = 2 x (Bandwidth). If the bandwidth used is not sufficient, not only may information be lost, but data from one of the channels may interfere with that from another. This interference can be reduced by increasing the bandwidth of the channel. However this may be wasteful of bandwidth and consequently expensive. An alternate method of reducing the inter-symbol interference is by shaping the pulses so that they contain less high frequency components. 6. Task D --- D/A Conversion

In this practical you will investigate the procedure of digital-to-analogue conversion required to retrieve analogue data which has been transmitted down a digital link. Digital-to-Analogue Conversion is the reverse procedure to A/D conversion, where the digital data corresponding to an analogue sample is reconverted into analogue form. Typically, there are 8, 12 and 16-bit D/A converters, with output voltage ranges of a few volts, usually determined by a precision reference voltage source which is often integral in the IC. D/A converters produce one output analogue voltage level for each of the digital input words. Thus an 8-bit converter with an output voltage range of 0 to 2.56 V will have 256 output voltage steps of 10 mV. The output, stepped analogue waveform is normally filtered to remove the steps and produce a smooth waveform. The quantization thus produces some distortion. The higher the number of bits in the digital word, the lower will be the quantization noise and the lower will be the distortion. However, the higher the number of bits for the converters, the more expensive the system usually is. 7. Task E --- Demultiplexing

In this practical you will investigate the recovery of the two signals that have been multiplexed. This is known as Demultiplexing. The multiplexed waveform, whether it be analogue, or digitally multiplexed, has to be demultiplexed to retrieve the original constituent waveforms, or data. The demultiplexing process is the opposite of multiplexing and the switching between channels must be synchronized with the multiplexing. If there is not synchronization the wrong information may be sent to the wrong destination, or corrupted data or waveforms may result. To achieve synchronization, multiplexed data is normally grouped into 'Frames' comprising one sample of the data from each of the required number of multiplexed channels plus synchronization bit(s). The form of a typical frame for a 24-channel PCM carrier telephone system is shown in the following diagram:

- 5 -

8. Procedures 8.1 Task A --- Analogue TDM

Open the “Discovery II” by double clicking the icon on the desktop, Go to Telecommunications/ Telecommunications Principles/ Digital Communications/ Workboard: PCM & Link Analysis/ Assignment: Time Division Multiplexing/ Practical: Analogue TDM/ Perform Practical to begin this task

Two analogue signals: 1 and 2 may be switched manually, or automatically multiplexed. Set the PCM bandwidth control (5) to maximum, the Noise level control (6) to minimum and all other controls to their mid positions. Adjust Data 0 (9) and Bit clock (7) controls for synchronism. Monitor the signals using the oscilloscope and use the MUX button to change between the signals.

Answer the relevant questions. You will need to return to the practical in order to answer them.

Questions Q1. Use the MUX button to switch to Signal 2. What is the form of Signal 2? Q2. Adjust the dc Channel 0 control (1). What happens to Signal 2? Q3. Use the MUX button to switch to Signal 1. What is the form of Signal 1? Q4. Click on Change MUX. Can you describe the shape of the output signal? Q5. Adjust the dc Channel 0 control (1). Does the output waveform change as you would expect it to?

8.2 Task B --- Analogue to Digital Conversion Go to Telecommunications/ Telecommunications Principles/ Digital Communications/ Workboard: PCM & Link Analysis/ Assignment: Time Division Multiplexing/ Practical: A/D Conversion/ Perform Practical to begin this task In this practical a variable dc voltage is applied to the input of an analogue-to-Digital converter. Set the PCM bandwidth control (5) to maximum, the noise level control (6) to minimum and all other controls to their mid positions.

- 6 -

Answer the relevant questions. You will need to return to the practical and make some measurements in order to answer them.

Questions Q1. Set the dc Channel 1 control (2) to minimum and monitor the A/D output with the Oscilloscope. What is the waveform you see? Q2. Turn up the dc Channel 1 control (2). Does the A/D output waveform change? Q3. Reset the dc Channel 1 control (2) to minimum and then very slowly turn it up until there is a change in state visible. Continue to turn it up extremely slowly and observe carefully the next change Repeat this for the next change. Can you see any pattern in the output changes? Q4. Continue to turn up the dc control and verify the answer to the last question. Turn up the dc control further and observe carefully the digital output. Does the digital output go through a complete set of counts?

8.3 Task C --- Digital TDM Go to Telecommunications/ Telecommunications Principles/ Digital Communications/ Workboard: PCM & Link Analysis/ Assignment: Time Division Multiplexing/ Practical: Digital TDM/ Perform Practical to begin this task

Signals 1 and 2 are dc levels. Set the PCM bandwidth control (5) to maximum, the dc Channel 0 (1) and Channel 1 (2) controls to minimum and all other controls to their mid positions. Adjust Data 0 (9) and Bit clock (7) controls for synchronism.

- 7 -

Questions Q1. What is the yellow trace monitoring? Q2. Click on monitor point 23 and increase the dc Channel 0 control 1. The dc Channel 0 control 1 adjusts Signal 1. Which state of the multiplexing waveform corresponds to Signal 1 being passed? Q3. Turn the dc Channel 0 control 1 back to minimum. Now increase the dc Channel 1 control 2. The dc Channel 1 control 2 adjusts Signal 2.

a. Which state of the multiplexing waveform corresponds to Signal 2 being passed? b. Is this the same state as for Question 2?

Q4. Turn the dc Channel 1 control 2 back to minimum. Change the monitor point. Increase the dc Channel 0 control 1 and observe the oscilloscope. Does the digitized value of the dc level appear in the correct time-slot of the A/D output waveform? Vary both the dc level controls and observe the results on the oscilloscope.

8.4 Task D --- D/A Conversion Go to Telecommunications/ Telecommunications Principles/ Digital Communications/ Workboard: PCM & Link Analysis/ Assignment: Time Division Multiplexing/ Practical: D/A Conversion / Perform Practical to begin this task

The Digital Signal is a multiplexed waveform produced from a triangle wave and a dc level. Set the PCM bandwidth control (5) to maximum, the dc Channel 0 (1) and Channel 1 (2) controls to minimum and all other controls to their mid positions. Adjust Data 0 (9) and Bit clock (7) controls for synchronism.

Answer the relevant questions. You will need to return to the practical in order to answer them.

Questions Q1. Monitor point <8> and slowly adjust the DC Channel 0 control <1>. Does the digital waveform vary? Q2. Change monitor point to point <24> and slowly adjust the DC Channel 0 control <1>. Does the analogue waveform vary? Q3. What form of waveform is this? Q4. What now has to be done to retrieve the constituent analogue signals from this waveform?

8.5 Task E --- Demultiplexing Go to Telecommunications/ Telecommunications Principles/ Digital Communications/ Workboard: PCM & Link Analysis/ Assignment: Time Division Multiplexing/ Practical: Demultiplexing/ Perform Practical

- 8 -

to begin this task

The TDM is an analogue, sample-an-hold, multiplexed waveform produced from a triangle wave and a dc level. Set the dc Channel 0 (1) control to minimum and all other controls to their mid positions. Adjust Data 0 (9), Data 1 (8) and Bit clock (7) controls for synchronism.

Answer the relevant questions. You will need to return to the practical in order to answer them.

Questions Q1. Monitor point 24. Does the waveform look familiar? Q2. Monitor point 27 and adjust the dc Channel 0 control (1). Does the demultiplexed wave at this point correspond with the originating dc level? Q3. Monitor point 28. Does the demultiplexed wave at this point correspond with the originating triangle wave? Q4. Why do you think that the triangle has been rounded off and the A/D, multiplexing, D/A and demultiplexing processes?