Pre-show testing found a couple of issues. Firstly, the game periodically froze or crashed. Secondly, the screen would flash red. I had a note in the binder from last years show that the power connector was bad and pressing on it did cause crashes and other effects that eventually resulted in the game not booting at all.
The plan was to replace only the +5V power and ground pins on the power connector whilst leaving the low power pins & housing intact.
The pins were removed using a thin wire to depress the catch on the pin to allow them to pull out. The old pins were then cut off and wires stripped back to solder on the new pin before sliding them back into the housing.
Testing with the replaced power pins appeared to have fixed the intermittent effects of pressing on the power connector however the game PCB was still not booting and removed for repair (in this photo you can see the score display from the Pong console aimed 2636 graphics controller. Arcade games typically only used the sprite portion of this IC).
On the bench the game PCB wasn't booting, no difference from in cabinet. Setting
up the Arduino ICT and a "ROM Check All" flagged "E: 8H 0001 01 1F". Going through
all the ROMs individually found:
There was still an intermittent red flash over the whole screen. Checking IC 13M (LS00) detected a corresponding low on pin 8. The low wasn't fully low, suggesting a low on another open collector output feeding the red output drive. Checking IC 13O (LS06) pin 2 found a full low output as a result of an intermittent high on input pin 1. Moving upstream found the flashing on IC 11O (LS02) pin 3 (C3/1). The flashing seemed to be present on IC 8F (2636) pin 33 (C3) output. Even with the ICT idle after a "RAM Write All Lo" the flashing occurred, thus not as a result of a bus access problem. Surprisingly replacing IC 8F (2636) did not fix the flashing so the new theory was that there was coupling through IC 11O. Indeed, I saw a noticeable level change on input pin 2 when pin 3 flashed the signal. Using freezer spray on IC 11O caused the flashing to stop and replacing IC 11O (LS02) fixed the flashing.
Though the game appeared to be running OK the Arduino ICT still flagged "E:2C 1803 0b 03" on the "RAM Check All" test. "RAM Write-Read" returned "OK: 01 02 04 00" confirming bit 0x08 was stuck low. A "RAM Write All Hi" did not show an observable low on IC 2C (2114) pin 11 (B3). A corresponding "RAM Read All" also did not show a low on IC 4E (LS244) input pin 8 but did show a low on output pin 12. I suspected IC 4E was bad and replacing it fixed the RAM test. With the last of the issues fixed the PCB ran OK on the bench for several hours with no further issues.
Re-testing the game PCB back in the cabinet confirmed everything was now working reliably except that the ship missile fire sound was missing. Suspecting an issue with the audio board it was removed for further testing on the bench.
On the bench with a pulse injector providing a substitute for the noise signal from the game PCB, IC 2 (LS04) output pin 2 looked OK as did the noise signal on IC 2 pin 10. IC 3 (555) pin 5 had a signal with a DC level that moved with the sound trigger. IC 3 output pin 3 was always high (no signal). Moving the pulse injector to IC 3 pin 3 didn't cause audio. Pulse injection on IC 4 (3900) pin 6 did cause audio. I suspected the 3900 was working but the 555 was bad. Replacing IC 3 (555) fixed the ship fire sound and a full size signal was observed on IC 3 pin 3. Back in the cabinet all the sounds appeared to be working properly.