I am amazed to find a board such as this. Such a wealth of knowledge!

I am trying to understand how to use the on-board clock generator of the 65C02. I think that all I need to do is connect a crystal from PHI1(OUT) to PHI2 with some load capacitors to ground from both pins. Using a specified 20pF load capacitance for the crystal gives me about 33pF for the load capacitors.

But the circuits I find on the web are much more complicated. Some circuits show a resistor (200K-2M) in parallel with the crystal and others show a resistor (1K-150K) in series with PHI2. Are these components necessary and if so why does everyone have a different idea of their value? I'm confused!

Different question: which is better, the WDC 65C02 or the CMD 65SC02?

If you just want to run at 1MHz, according to my books from different companies, you should be quite safe this way:

A. Connect the crystal from the phase 1 output to the phase 0 input. Sometimes phase 0 input is called phase 2 input, but it's not the same as the phase 2 output pin which goes to several other things on your board.

B. Add a 51pF capacitor (47 would be fine) from each side of the crystal to ground.

C. Put a 200K resistor (180K or 220K would be fine) across the crystal.

Deviating a little from these values shouldn't be any problem. This method of generating the clock is generally for systems where the speed is low and the accuracy is not terribly critical.

CMD v. WDC: WDC's 65c02 is available in faster speeds and has the extra bit-manipulation instructions. It also allows stopping the clock in either phase.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I saw the 51pF, 200K circuit in the CMD data sheet. I have done microprocessor designs before and have never used a parallel resistor in the crystal circuit. The load capacitors are generally set at twice the expected load of the crystal. Standard microprocessor crystals are speced for 20pF load so I would normally use 33pF assuming that there is some stray capacitance floating around to make up the difference. I don't see where the 51pF value is coming from. I will see if I can find someone at WDC to provide information and will post the response.

Thanks for the info on the WDC vs. CMD. I am redesigning a commercial application using legacy code and I need a part that I can count on to be available for many years. The previous design used the DIP part so CMD and WDC were interchangeable. Now I am going to PLCC and unfortunately CMD and WDC have a vastly different pinout so I have to choose one or the other.

I spoke with Bill Mensch at WDC and he said that they do not recommend using the on-board oscillator circuit because when they moved to the smaller die size, the gain of the circuit increased. In order for a parallel crystal oscillator circuit to work correctly the amplifier must have unity gain.

I then spoke with Fox Electronics (a major manufacturer of crystals) and they said that too much gain will overdrive the crystal and cause clipping and overtones.

So the answer seems to be: use an external oscillator.

All I've ever done is feed the Phase-0 (I guess that's now called Phase-2 IN) clock to the CPU and be done with it. The typical asymmetry of crystal oscillators screws up cycle timing so much that it's easier to take a higher frequency (and those crystals are smaller and cheaper) and divide it down to what you need. Generally that allows you to produce other clocks you need as well.

BTW, better is a difficult term to apply to unknown requirements.

Uli