So the current state of play is that IC42 (which was a VC2069) has been replaced with a VC2023 borrowed from a model B. They’re pin compatible (and I believe functionally compatible).

I now have:

  • pin 8 – 16Mhz in
  • pin 7 – 8MHz out
  • pin 6 – 4MHz out
  • pin 5 – 2MHz out
  • pin 4 – 1MHz out

BUT, the 2MHz out, while it makes it to IC34 pin 2, doesn’t make it to IC28 pin 3, which results in no 1MHZE signal being generated – and that’s used by the VIA’s, they keyboard, and in various other places. So a bodge-wire corrected that, and now I get the 1MHZE signal in the relevant places.

With that in mind, I buzzed out the other pins from the video ULA (not exhaustively – just to another IC). This confirmed that at least all the signals are making it from the legs of the video ULA to at least one other part of the board.

I think the next step will be to drop the logic analyser on and see what the CPU is doing on reset – see what values it’s getting back when it tries to read the reset vector from the top of ROM, etc, etc.