With the memory repaired, the serial interface working correctly and most of the bugs out of the machine, it is time to load the Maindec programs, get them working and then load the CHESS program to give the machine a chance to beat me one more time.
I say most of the bugs out of the machine, because the remaining few proved to be some of the nastiest. The PDP-8 refused to load BIN programmes using the binary loader, because more than one of the Accumulator cards still had obscure bugs in them. Again David Gesswein was very helpful in providing RIM format Maindec diagnostics that I could load into the machine and continue the troubleshooting process.
Following a hard failure of the Maindec program I would check the listing, also kindly provided by David, to see what was happening. I might add that prior to running Maindec this the machine was successfully running all of the various ‘fat finger’ diagnostics that are available.
Being reasonably confident that it was an Accumulator fault, I would swap three of the twelve boards with the adjacent three boards and see if the fault moved. If it did, then I would then proceed to narrow the problem down to the failing board, if it didn’t I would repeat the process between the remaining six boards.
For the last few Accumulator faults the defective diodes all looked good using the DVM. I decided to bulk replace all of the approximately 200 diodes on the cards that Maindec identified because I knew if one was bad, it meant that eventually others in the batch would follow.
With Maindec running successfully, I moved on to load CHEKMO-II.BIN and as you can see from the video link below, the machine performed flawlessly.