I'd get a can a freeze spray for electronics before shotgunning it. If it really is thermal (i.e. stops working when warm), you should be able to spot it fairly quickly. You can also use a piece of paper, etc. to help shield individual components when you spray. Helps to narrow down the specific part.
pasted_image (resized).png

Get some freeze spray, really!

It could also be a bad solder joint. After it is acting up, give the board a little bit of a press in several areas. If the sound comes back, then start looking for a cracked solder joint, broken trace, etc. Might be hard to spot, to the point of being easier to just touch up a bunch of solder connections.

Nah, you didn't say/do anything wrong. I'm just an old fart that has used freeze spray sporadically over the years. And chasing down a thermal problem is a little harder with air than freeze spray. Freeze spray REALLY gets cold instantly. Like frosty cold! Good deal w/the cardboard tube for isolation.

Just make sure you didn't cool any possible culprit right next to the audio chip. Try cooling down parts next to the audio IC, excluding the audio chip to be sure.

Sounds like you are on the right track tho!

Also, does the audio chip have a heatsink with thermal grease? It the thermal path used to cool the IC is so-so (loose heatsink, dried out thermal grease), that can cause the problem too. The IC might just be going into thermal shutdown.

I just looked up the LM3900..that's just a quad op-amp. I was thinking it was the output audio amplifier. So obviously my heatsink question is a 'whoops, nevermind'. LOL.

What pin is this in?

Sorry, I should have asked the last time...Is there a Ux by the chip? Like U1 or something similar? I'll look up the schematic and take a peek at the circuit.

Yeah - the heatsunk transistor is probably the actual output device that drives the speakers (haven't looked at the schematic yet, was just looking it up). I was thinking the LM3900 was the output amplifier, until I looked it up and saw it was just a lower level audio amp - my mistake.

OK, looked up the circuit.

Not much going on there. I doubt R5 is bad. R4 - the same. Not impossible, but unlikely. C3, the 68pF cap - I'd blame that before the resistors - but I'd blame U8 first.

The schematic shows a couple of voltage levels you can check when it acts up.
pasted_image (resized).png

Link: https://www.ipdb.org/search.pl?any=space+invaders&search=Search+Database&searchtype=quick

Scroll down a ways to the Space Invaders by Bally section, the look a little bit lower to Documentation. Full schematics there.

U8 is up next then. I should have the resistors if you want to replace just for the heck of it... Just PM me your address.

NP, it's just the cost of a stamp...