65c22 handshakinng

[Dan Moos]

EDIT: I may have been misinterpreting which edge was witch on the scope. Basically, it looks like it works exactly as the timing diagram says it should.

So the timing diagram on the datasheet shows that, with read handshaking, the CA2 signal on the 65c22 is active low (like CA1).

It seems to be outputting active HIGH signals. Is there a flag to control this? It doesn't matter necessarily, but I don't like un-explained behavior.

Also, the CA2 signals "data taken" basically simultaneously when the MCU signals "data ready" to the 65c22. Like, ~ 3 ns after. The timing diagram says it doesn't signal "data taken" until the input register is actually read, which at least on the diagram is a full clock cycle. I'm running at 1 MHz, so that would imply at least a full us before CA2 responds.

My setup "mostly" works, but is reliably unstable under certain circumstances. These two behaviors of the VIA seem like a clue. Thoughts?

Side note: "Reliably Unstable" would be a great name for an all female rock band. I jest of course...

[GARTHWILSON]

That's one of the few features of the '22 that I have not used; so I cannot speak from experience on it.  As far as a flag to control it goes, I would just say make sure bit 3 of the PCR is a 1, then choose an option in bits 2 and 1.  You'll probably want the 00 there.  I imagine you'll have a forehead-slapper enlightenment, followed by a happy report that you got it going.