Wow, it's been three years since I wrote up an MPS1 but done a fair few in that time.
One I have on the bench at present has been driving me demented. It came in as a non boot and there were a lot of crusty chips all around the battery area. To cut a long story short every chip 14, 16 and one 8 way plus the 4x 20way buffer chips around the CPU were all bad. Luckily the Async chips (TMS9902) were OK as was the CPU, PAL and RAM but the reel CPU was toast which was a bummer.
Anyway after changing all the chips out it was time to switch on. It came out of reset and ....nothing! Then followed about 3 days of trying just about everything. All the DATA and ADDRESS lines were checked for no shorts and that they all went to their correct places. The only clue was that the PAL chip was sending out invalid mem access?
Desperation was beginning to set in so I thought I'd ask AI and see what it would come up with. After explaining the issue, uploading diagrams and stuff I awaited an answer. Thinking it might just say '42' it eventually replied and said 'I'd jump in Thames if I was you'!
A few hours later and I was checking the DATA waveforms again and this time I had the scope set for different time period and I could see that the waveforms were slightly different either side of the buffer chip (74LS245). These buffer chips around the CPU are controlled by their enables on pins 1 + 19 and these in turn are driven from IC27 (7404) which had been swapped already as it was faulty and wasn't bring the enable lines down to 0v. Because the ADDRESS line chips were working I didn't look any further until I tested pin 19 on the DATA buffer chip and it was sitting at about 1v! Now pin 6 on IC27 was at 0v so there was a break in the track somewhere. Hurrah looks like I'm on to a winner. There was indeed a broken track which was rotten, you couldn't see it with the naked eye and not really under the microscope either. So in goes a patch wire and BINGO she boots!
So was it time to wrap it up....nope, it now looks like there's a switch input error on one of the PIO chips but at least I've got something to work with.
Watch this space.....