Feeling better this morning, so after I found a new PSU for the CPC, I had a look at the original Spectrum + board to try and figure out what the video issue was. A bit of scoping and all the signals looked good at the LM1889, so I then wondered if there was a cold joint on one of the CHROMA pins where I'd removed the LM1889 and fitted the socket.
It turns out there wasn't, but I noticed pin 1 looked like the track over to C39 had a break. Sure enough, doing a buzz test and cross referencing to my spare board, and there was indeed a break, but very difficult to see. I can only assume I did it desoldering the LM, it's a super fine track, so it wouldn't take much.
Bodge wire time:
PXL_20231004_111418524.jpg
And it is working again. The + is now reassembled and running a soak test:
PXL_20231004_113503088.jpg
So far all good :) EDIT - 0ver 200 iterations later and all is well.
So it looks like this is the end of the Spectrum saga. I thought it would be just the one machine to begin with, but it ended up as 3. As a bonus, I now have a fully working spares board just in case, that turned up dead from eBay (I expected it to be dead) and I was able to revive.
It really has been a full Spectrum of emotions!!!
On a side note, doing all this stuff this morning has just made me feel ill again, so back to bed for me, at least I can write this and then go have a nap...