Since I name this "AnyCPU5V40DIP tester", I thought I'll post it here. The tester is built on a prototype board for RC2014. The EPM7192S CPLD is bigger than needed but I do have plenty of it on hand.

The idea is connect every pin of DIP40 to CPLD's programmable I/O. Power is special case because CPLD output is unlikely to drive 5V to desired level and ground needs to be solid, so there are two jumpers, black for ground and brown for 5V, that can be plugged into the corresponding power/ground pins on DIP40. The CPLD provides clock, reset, 128 bytes of ROM and a serial port. It can exercise the CPU sufficiently to show it is working.

I'll start with 6502 since that's what I'm working on right now.
Bill

Nice idea! Should certainly help when you have some CPUs of known type but in an unknown state.

(Could it be set up to safely distinguish say a 6502 from a 65C02?)

Dare I ask, though, what kinds of things might happen if the device under test isn't as-marked? I'm wondering, for example, if it might be handy to put polyfuses on those power jumpers? Could you program the CPLD somehow to tickle each of the pins to try to sense which are outputs, to avoid contention?

Had I thought about it a bit longer, I probably would've added a 100 ohm series resistor for every pin to limit the current. Some LEDs are helpful as well, which I did add after the photos was taken. The failure mechanism for contention is mostly overheating so power supply current limiting and turning on only briefly are fairly effective in protecting the tester. I guess I'll find out soon enough--I have plenty of non-working 6502 to be tested. 6502 has so many variants that were covered up with re-labeled parts that figuring out which variant for the device-under-test should be an interesting problem.
Bill

Certainly will be interesting - do please keep us updated!