(d) 3000H.
5 A UVEPROM:
(a) is programmed by ultraviolet light.
(b) loses the data if the power supplies are disconnected.
(c) is a form of non-volatile memory.
(d) is used in a SIMM.
7. A microprocessor-based system
How simple can a microprocessor-based system actually be? It must obviously contain a microprocessor otherwise it is simply another electronic circuit. A microprocessor must be programmed. This means that it must be provided with a series of instructions to be followed. However we program the microprocessor, the result is a series of binary numbers that represent the simple step by step instructions to be followed. These instructions must be stored in some memory. But do the instructions have to be stored in RAM or ROM? It must be in ROM. Remember that RAM will hold random data when first switched on and if our microprocessor was controlling the operation of a dynamite factory, the last thing that we would want is for it to start following random instructions at the rate of a million a second!
What determines how fast the microprocessor carries out the instructions? For the moment we will say that regular pulses of voltage applied to the microprocessor determine its speed. This voltage pulse is called a clock pulse.
A clock circuit controls the operation of the microprocessor. This produces a series of voltage pulses like a ticking clock. The whole system runs sequentially, doing the required jobs one after the other. One step completed for each tick of the clock system.
The clock circuit can be mostly internal to the microprocessor or it can be entirely external. It is unlike a normal watch or clock in that the exact speed is not important. It may be running at a nominal rate of 200 MHz but if it actually ran at 199 MHz or 201 MHz there would be no great panic. With a modern watch, the equivalent error of seven minutes a day would mean a trip back to the supplier. In Figure 7.1 a typical clock pulse is shown.
Figure 7.1 A 200 MHz clock pulse
The shape of the clock pulse is stipulated in detail in the microprocessor specification. If it does not stay within the limits stated, serious problems can occur. Let’s assume that two circuits have to switch at the same time but they operate at slightly different voltages. In Figure 7.2 the two operating voltages occur at virtually the same moment due to the very fast rise time. In Figure 7.3 the clock pulse has a very slow rise and fall time and Circuit 1 will switch before Circuit 2. In this example, the difference is about 2 ns. Whether or not a 2 ns difference is significant will depend on the circuit being considered.
Figure 7.2 An almost vertical leading edge means zero rise time
Figure 7.3 The rise time is about 2 ns
All microprocessor technical data will stipulate the maximum rise time and fall time and between which voltages it is being measured. For example, the Intel 80386 has a supply voltage of 0–5 V and must have a rise time (and fall time) of less than 2 ns when measured between 0.8 V and 4.2 V.
If the rise time (written as tr or trise) is not shown, it is normally measured between 10 and 90% values of the supply voltage. Each microprocessor has a maximum frequency for the clock pulses and a minimum value.
Why is there a minimum speed?
Microprocessors have internal registers and storage areas called capacitors that are like dynamic RAM and need refreshing at intervals. As the clock speed is reduced, the interval between refreshing gets longer until the register or capacitor can no longer hang on to the information and the whole operation collapses. Typically, the minimum clock speed is about a quarter of the maximum speed. The Intel 80386 will run down to 800 kHz.
The first microprocessor, in 1971, was the Intel 4004. It ran at a clock speed of 0.108 MHz and handled 4 bits at a time. In 1998, the 64-bit Alpha 21164 runs at 600 MHz, king for a day but will, no doubt, soon take its place on the museum shelf. (It is 700 MHz now and the ink is barely dry!)
The most basic system is therefore going to include a microprocessor, clock and a ROM chip to provide the built-in sequence of instructions. The only other essential is to have wires or other conductors to connect the circuit. Figure 7.4 shows the most basic microprocessor based system with just the bare essentials. As usual, the power supplies are not shown.
Figure 7.4 The most basic microprocessor-based system
All microprocessors are connected internally and to the rest of the system by collections of conductors called buses.
These conductors are in groups, since many will be going to the same place for much the same purpose. For example, an 8-bit microprocessor normally uses 8 connectors to carry the data between the microprocessor and the memory. It would make diagrams very complicated if each wire were to be shown separately so we group them together and refer to them as the data bus. A bus is therefore a collection of conductors providing a similar function.
In a microprocessor-based system we have three main buses: the data bus, the address bus and a control bus. The data bus is a two-way bus carrying data around the system. Information going into the microprocessor and results coming out. The address bus carries addresses and is a one-way bus from the microprocessor to the memory or other devices. The control bus is rather different from the other two. It is a somewhat looser collection of conductors. If we look at a microprocessor-based system we can easily see the data and address buses since they consist of many parallel connections. However, the control bus is just an association of all the other necessary connections such as those to the chip select and read/write pins.
Address information, data and control signals have to be carried around inside the microprocessor as well as in the external system. We will therefore meet internal as well as external buses.
The signal on the data bus has only a very low power level and to be of use it must be amplified.
The first response was to put a series of amplifiers; one on the end of each of the connections on the data bus but this was soon superseded by a more sophisticated chip with more facilities. For example, if we wanted to send a signal to a printer at 1 ms intervals it would not be sensible to tie up the main processor with counting out these time intervals. It would be much better to tell the output chip to do its own timing and thus release the microprocessor for more important jobs.
The output devices have become quite complex and now go under a variety of names:
I/O controller – input/output controller
PIA – programmable interface adapter
VIA – versatile interface adapter
PIO – programmable input/output
PPI – programmable peripheral interface
…and many others.
Whatever name is given to the device, they are basically the same and are essential to any microprocessor-based system. An improved basic system is shown in Figure 7.5.
Figure 7.5 An input/output chip has been added
Apart from in the most minimal of circuits, some RAM is needed. Even if the microprocessor-based system is controlling an oven, we still need the facility to vary the instructions to change the temperature, the time cycle, the fan speed etc., so some RAM must be added. Some microprocessors have a small amount of RAM included internally, enough for this sort of system to work but still quite limited. If we add some external RAM, the microprocessor is controlling the operation of three chips: ROM, RAM and I/O. To control the flow of information it needs to send chip select and read/write information along the control bus.