After laying out and etching two boards and populating them with parts I now remember a problem I had a year ago trying to use two PIC16F84A's together on the same board. I finally got so frustrated with the business that I switched to another device with more pinouts (a 16F877, I think.)

Last night I finally connected a secondary board with a PIC16F628A to the ground and 5VDC buses of a primary with another 16F628A. The primary has an LM7805 on it to source both PIC's. When tested apart, both boards run and test out perfectly.

When I connect the secondary to the power buses of the primary, the PIC on the secondary board fails to operate. I have checked the Vss and Vdd pins on the secondary and I get 5VDC on Both!

Here are the design facts:
- Both 16F628A's are set to run on their respective internal clocks at 4MHz
- Both 16F628A's have !MCLR enabled (pulled up with a 4700ohm external resistor) to be used as a manual pushbutton reset.
- Comparitors ar disabled on both PIC's
- I always solder a filter cap right under the socket across Vss and Vdd, in this case a mono. ceramic 1uF, 50v and did so on both PIC's.
- Finally, I did not make the error of putting 5VDC on both pins of the power connector, and there are no circuit paths on the secondary board which would cause that. I am certain that all ground traces/pins end up at the 7805 center pin, and there are no shorts. The filter caps are the only bridges.

This is really bewildering. I can't believe an industrial application of plc's wouldn't have a situation where multiple devices would be used together in the same sub-system so device isolation must be an art.

Does anybody know why my PIC's don't ever play well together? Is there some harmonics problem between the two clocks that the filter caps. don't overcome? I've often wondered about this. I don't have the years of experience to recognize things like that (the art.)