My scope is not real fast one, So any tiny 'glitchwaves' might not be seen.

I agree with the grounding question as well, I did ground the scope probe pigtail to the CPU ground pin with no change. I did change the opto drive level with no visible change to the symptom. The way I see if there is a problem is in the PCA isr I check the level on the pin and if it's wrong I flip a port pin for scope visibility. (and increment a byte in ram for my serial datastream)

The opto is a PS2501, slow and cheezy. I am trying to capture frequency from 0(really) to 1khz using the PCA (extended in software to 32 bits)The opto is/was for noise immunity. I changed the ground path from the opto to the CPU (1.5 inches), added decoupling, added series resistance, changed the pullup, all didn't help. A .001uf at the cpu fixes it, but if the cap is at the opto it does not.... Seems (ground)loopy. Maybe if I turn the TV off...... ;)

I will try to borrow a fancy scope and see if I can find a glitch. I was just curious if there was some 'sensitivity' in these parts, I usually just use a P87C51FB...
Movin up to flash to stay 'modern'.

Thanks!!!!

Bob