Digital and Computer Systems

Course Sample

This module is offered as part of the Certificate II in Electronics, providing knowledge and skills to enable you to identify a digital system or a computer system and to describe the basic principles of operation of common systems.

Here is a sample of the material covered in this module.

The simulation below shows a TDM system with 4 input sources and 4 destinations. To set up the data to be sent, click the mouse in the source box and type in some characters. It is best to use not more than 6 characters or letters and the letters will be easier to read if you use upper case. Once you have set up the sources, click on the 'Run' button to start transmission - leave the 'Slow' selection for the moment. Data is taken a character at a time from each source in turn and transmitted via the switches to the corresponding destination. In this simulation the same data is sent over and over again until you click on 'Stop'.

You need at least IE3 or NetScape 3 to see the applet

Watch the simulation on 'Slow' for a few times round until you can see how it works - then click on 'Fast'. Try stopping and changing the data. This time leave one of the sources empty.

What do you think the '*'s are there for?

The sources are often called channels and the set of data comprising one sample from each channel is called a frame. The amount of time in any one frame that the data for any one channel is being transmitted is called a channel time slot or just a time slot.

The diagram contained in the link below simulates an ADC set up to digitise an analog signal. The maximum value of the signal is 15V. You can see the conversion process by clicking on the 'Run' button below. The table that appears lists the samples S0 - S9, the associated voltage and the digital value assigned to that value.

Try it now.

If a 4 bit ADC is used then there are 16 (ie. 2 to the power 4) possible digital values 0 - 15. If we have a signal ranging from 0V to 15V and 16 digital values then each digital step will be 1V. Note that for this combination (15V signal and 4 bit converter) the digital value is simply the nearest whole number to the input voltage. The green dots are a plot of the digital values. This allows you to see the difference between the analog signal and its digital representation. Not too bad in this case.

You can change the ADC to be 8 or 4 bits by clicking on the '4 bit' button. Running the simulation again shows a closer fit of the digital points. This is because there are now 256 possible digital values so each step in only 15/256 = 0.059V and thus the error introduced by the ADC process is reduced.

The ADC being simulated has been set up to deal with a maximum of 15 volts. You can change the size of the Vin input signal by using the 'Inc Vin' and 'Dec Vin' to increase or decrease Vin in the range 1 - 15V. You will find that the errors introduced at low input voltages become more significant. For instance, with a 3V input signal and a 4 bit ADC, the errors are quite noticeable. If you change to 8 bits the fit is much improved.

Can you work out how the digital values come about for the case of the 8 bit ADC? If not, ask your tutor.

The diagram above shows a possible use of an ADC and a DAC to control the temperature of some equipment by measuring its temperature and controlling the speed of a fan to keep the temperature within given limits. Note that the input is temperature which has to be converted to an electrical signal before it can be used as input to the ADC. The device which does this is called a temperature transducer. A wide range of transducers is available to convert most physical quantities to electrical signals. Some examples are:

• pressure transducers to measure pressure
• photocells to measure light intensity
• flow meters to measure flow of liquids or gases in pipes
• thermocouples to measure temperature
• tachometers to measure speed of rotation of a shaft.

Just about anything that can be measured will have its associated transducer.

A computer involved in the system allows a lot of flexibility. For example, by reading its ADC input every second and saving the last 10 values the computer not only knows what the current temperature is, but whether it is increasing or decreasing and how rapidly it is doing so. So if the temperature is rising very rapidly the fan speed can be increased to compensate even though the temperature may still be within the allowable limits. This could prevent an over temperature condition occurring.

The notes do not mention the problem of synchronising the data sent by the transmitter. It is no use the transmitter squirting out a stream of bits if the receiver does not know how to interpret what it receives.

Consider a serial connection between two digital systems. As well as the data, the transmitter sends a clock. Remember the clock in the counter of unit 1. The counter counted each time there was a '0' to '1' transition in the clock input. In the same way, the receiver uses one of the clock edges to catch the incoming data.

Digital System Receiving a Serial Signal

Notice how the data is only received when a clock pulse occurs. The level of the data input at other times does not matter. This means that the clock pulse can be timed such that it does not occur while the data line is changing.

It is not always necessary to actually have a separate clock line, although that is the simplest way. More sophisticated techniques are possible whereby the data is encoded into special forms which include the clock - something to look forward to finding out about later in your studies.

The diagram shows 5 lots of 8 bits or 5 bytes being transferred from the transmitter to the receiver. The transmitter first selects the data to be transferred and then outputs the data onto the bus. It then generates a clock pulse. When the receiver detects the clock pulse it reads in the data present on the bus. A short while after generating the clock pulse the transmitter removes the data from the bus and restarts the cycle to transfer the next byte.

Since 8 bits are transferred with each clock pulse the data transfer rate is 8 times that of the serial transfer. You can see why the buses tend to get wider and wider - 8 bits, 16 bits, 32 bits, 64 bits..... Each time the width of the bus doubles, twice as much data is transferred without having to speed up the clock.

The data transfer rate is calculated by multiplying the clock rate (or frequency) by the width of the bus (often measured in bytes). For example, if the clock rate is 1 Mhz (ie. 1,000,000 cycles/sec) and the bus width is 16 bits or 2 bytes then:

Data Xfer Rate = 1,000,000 x 16 = 16 Megabits/sec or 2 MegaBytes/sec

There are, of course, disadvantages to making the bus wider and wider - the main one being the large number of connections required between the components.