Here's a crop from an image attached to post #9.
Worth experimenting with the routing of the PSU and monitor interconnect cables I would have thought.
Alan
I know from personal experience that the Western Design Centre W65C21 and the Motorola MC6821 are suitable replacements for 6520 PET PIAs although I'm sure there are other equivalents.
Alan
I've just read the first page of this thread again. A couple of observations:
1. The PET didn't display an image at all using the original character ROM.
2. Things seemed to go seriously awry when the 2114s were first used in the video RAM positions.
Alan
My guess is that you have an intermittent connection between the transformer and the VDU. The voltage rise will reflect a change from load to no load due to the intermittency. Clean the connection to the VDU board and see what happens. The random character screen is a different issue but at...
That's an acceptable Hi signal as far as I'm concerned and it doesn't look like there's a problem with the reset circuitry.
Overall I'm a bit confused and wouldn't be testing the RAM chips with the Tynemouth board in test mode. Also I think it's worth sticking to using just one CPU.
Alan
No, the characters' positions will not be identical to the example image and they will change on restart. If you're sure you've identified all the faulty 6550s and that both CPU's are ok try removing the PIAs if you've not already done so.
Alan
I assume this means you've replaced the character ROM and that you still get the garbled screen when using the Tynemouth board's ROM & RAM. Presumably this is the early static RAM logic board with socketed RAM chips. If so it looks like there's a problem with at least one of the video RAM chips...
Bitfixer's ROMulator includes a NOP generator and Dave's PETTester V4.
You may struggle to find a schematic specific to the 8032089 PCB. However you should find that the 8032090 schematic is close enough. Any differences are likely to be very minor.
Alan
Before going any further please try abrading a few rubber keypads with plain printer paper before testing with the keyboard assembled. This is a tried and tested method of reviving keyboards of this type. I think you'll be surprised at the results.
Alan