Lab three questions and answers

Remember that the list of control signals in each microcode step is a list of which ones are supposed to be 1 for that clock cycle, and all other control signals are supposed to be 0. (This is crucial to understanding what these microcode instructions mean!)

Q: On the handout, it says that the IR will determine which function is to be performed by the ALU. So does that mean that we just arbitrarily assign a code for each function?

A: Yes. Of course there are only two possibilities. The IR contains the instruction being executed, and since it's only 1 bit and there is an opcode field, the opcode is that entire one bit. So either 1 means AND and 0 means XOR, or vice versa.

Q: What does the "decoding" circuitry connected to the IR do?

A: In the case of this simple architecture, nothing. The IR contents is just one bit and there's nothing interesting to decode there. It's all opcode; that one bit goes directly into the ALU to select which operation is performed.

Q: Must we use a multiplexer to implement the ALU?

A: No, although you can if you like. But I suggest doing it more simply. Just write out the complete formula for what the ALU calculates, with IR being one of the variables in your algebraic expression.

Actually, what you'll get is the same as what you'd get if you expanded all the circuitry for a two-input multiplexer. But I think you'll find it easier to understand and work with if you do it all within boolean algebra, by just making formulas involving A, B, and IR.

Q: On the D flip-flop chips 74LS174 and 74LS175, there is only one Clock line. So I need a separate chip for each register, right?

A: You would indeed, if you were to use those chips for your registers. I suggest instead using the 74LS74 chip (note number 74 instead of 174!) which only has two D flip-flops but they are separately-clocked.

Q: Do we use a separate tri-state device for every 'out' control signal? And if so, does this mean that we can't use the 74LS367A chip?

A: You do indeed use a separate tri-state device for every 'out' control signal. That means that you only need one data bit for that tri-state device.

The 74LS125 chip is optimal for this purpose. However, there might not be enough to go around, so it would be best if you were prepared to be able to use the 74LS367A chip instead. Each 74LS367A chip gives you only two usable tri-state drivers because you only want one bit of each of the identically-controlled group of them. That is, you'll just want to ground or power three of the four data bits corresponding to one of the control signals, and one of the two data bits corresponding to the other control signal.

Q: How do I make a "bus"?

A: Just connect all the wires together. All the tristated data outputs, and all the D inputs, all get connected together in a big bundle. That's the bus.

In practice in working with the breadboard you probably want to chain them; i.e. if A, B, C, D, E, and F are all connected to the bus, you might connect A to B, and B to C, and C to D, and D to E, and E to F; rather than attempting to connect all of A, B, C, D, and E to F (for which there aren't enough holes in that column in the breadboard).

As stated on the lab 3 handout, "the control input of the tri-state bus drivers used in the lab means 'output disable' instead of 'output enable'"; so the input should be 0 to turn them on and 1 to turn them off. But this doesn't mean you necessarily need any extra circuitry. Because the tri-state device control lines will be going directly from input switches. So just consider the input switches to be upside-down, and then you have achieved the necessary inversion.