I have been working on a System 80 MPU off and on for about a year. It had an IRQ stuck high problem that I couldn't seem to resolve. The 6502 as well as the three RIOT chips were replaced and the problem remained. Chip select circuit double/triple checked no problems found. All other logic chips were checked via a logic probe. All traces, new chip sockets checked double checked, triple checked for continuity and shorts.
I got out the Bugtrap a logic comparator and tested all the chips possible. The clock circuit was tested and no problems found. I was using a Quickscan tool for testing.
I used another test tool to test the RIOT chips multiple times and the chips always passed testing. The 5101 was also tested and it tested good. The only chip I didn't have full confidence in was the 6502. I replaced it with other in my parts supply with no change.
The Quickscan tool would always indicate a stuck IRQ, and of course when checked a logic probe it was stuck high every time.
Recently I loaded a set of game roms into the board and checked it out and found that it appears to be booting correctly with the IRQ signal looking good - oscillating not stuck. What?
I then suspected my Quickscan tool as being a liar. I had previous problems with the ribbon cable and TC1 connector due to wear and tear over time. I double checked continuity from the pinout on the connector to the bottom side of the Quickscan board at the solder points and found all good continuity with no shorts. It's a working ugly ribbon cable.
The Quickscan would reset and the LEDs would all go green with the IRQ going green then red. With a zero indicator in the digit display. It was consistent with previous boards that did have an IRQ problem.
A number of years ago I had a self induced problem causing the 27C512 Eprom to fail. I fixed this with a new programed chip and all was well until today. I pulled the Eprom from the Quickscan, hoping this was the problem and not the GAL chip as I didn't have a replacement for the GAL chip. I loaded the programmer buffer with the correct binary file for the Quickscan Eprom and the chip failed the verify. I burned a new 27C512 Eprom and that fixed the Quickscan as the System 80 MPU now passed all tests. I must have done something to cause the Eprom to fail, or my original burn was weak and it lost memory over time - maybe static damage.
I learned years ago that if you mis-install the Quickscan connector to the TC1 socket, such as pin1 going into pin 2, the Quickscan Eprom will no longer work - and the Quickscan will be functionally dead. So I am careful in the placement of the connector into the TC1 socket. It's easy on the bench, but the trouble is when the MPU board remains mounted in the machine. This was Not the problem as the Quickscan tester appeared to be working.
I no longer have any System 80 machines but had this nice clean non-working board I wanted to get working - it wasn't working and likely was fixed by a replacement RIOT but didn't know because of the Quickscan problem. Now I know for sure it is. Probably close to 30 hours of work on this board.
