In the AC characteristics for the WD65C02, it lists the voltages matched with the frequency.

Does this mean that if I want to run 5V, I should be running 14Mhz, or 1.8V for 2Mhz???

If I can run 5V at 2Mhz, how does this affect the timings?


------------ ADDED LATER -------------

I was looking at the 74LS30 and 74LS20 (8-1 NAND and 4-1 NAND) and the datasheets for both specify a typical high output of 3.4v - is this 'standard' for logic?? I thought the high state was supposed to be 5v.

At the bottom of the DC Charactaristics page, they list maximum frequency vs voltage. If you give it 2V, it won't run above 6MHz.The timing will depend on the voltage, not the frequency. At 5V, it needs 10ns read data setup. At 1.8V, it needs 60ns. It's those sorts of limits (along with undocumented internal timing) that determine the maximum frequency.

So: give it 5V and it will run at any speed you like.

TTL chips never give 5V or 0V outputs. CMOS will get close, if the output isn't loaded. But it doesn't matter: the inputs only care if the voltage is above or below some threshold. At 5V, the 65C02 considers anything below 1.5V to be a logic '0' and anything above 3.5V to be logic '1'.

Don't go by the 'typical' values on the datasheets. You want to listen to the guaranteed voltages. For the SN74LS00N, that's only 2.4V. Don't mix TTL and CMOS. Use the 74HC series instead.

> Does this mean that if I want to run 5V, I should be running 14Mhz, or 1.8V for 2Mhz???

The higher the voltage, the faster you can go; but at any given voltage, the clock frequency can be anything from 0 (yes, you can stop the clock) up to the maximum frequency for that voltage. (Actually WDC's are pretty conservatively rated.) The speed rating for a given voltage is generally for the top end of the temperature range too, since high temperatures reduce the maximum speed.


> If I can run 5V at 2Mhz, how does this affect the timings?

It just means you have really wide timing margins, making it easier to make parts of various speeds work together.


> I was looking at the 74LS30 and 74LS20 (8-1 NAND and 4-1 NAND)

There's really no reason to use 74LS anymore when the CMOS parts are faster, use a lot less power, and can drive the outputs near the supply rails.


> and the datasheets for both specify a typical high output of 3.4v - is
> this 'standard' for logic?? I thought the high state was supposed to be 5v.

As John said, CMOS can drive its output voltages much closer to the rails, and you will actually get practically 5V if the load is light enough (ie, it's only feeding CMOS inputs). 74LSxx can usually output high enough voltages to drive CMOS inputs to a valid logic high level, but it's not guaranteed so "use at your own risk" as they say. I personally have used LS to drive CMOS for personal projects only, and when a parts inventory crisis on a weekend night required it. I was lucky and did not experience any problems.


> At 5V, the 65C02 considers anything below 1.5V to be a logic '0'
> and anything above 3.5V to be logic '1'.

My older databooks (Synertek, Rockwell, and GTE, which later became California Micro Devices) all show .8V and 2.0V for input thresholds for their CMOS processors and peripherals, but my WDC datasheets are mixed. When I'm rich and famous (that's definitely not an ambition of mine) I'll have DSL so it'll be more practical to check things from huge datasheets online.

Thanks guys