Hi Quench
So that was fun...lol
I got the scope out and first took the working board from a Pacman and put the 8 Ball ROM's on and jumpered. Tried to take measurements at those different points you mention and it was really hard as the signals where basically what looked like total noise except for pin 4 Op Amp output that was a great large square wave when pressing the switch. So that I could work with.
On Test point 7 when pressing a working switch you could see a very small change in signal to a little more defined type square wave but almost impossible to get any sort of proper measurement. But at least I knew what to look for from a working board...
So then stuck in the faulty board. Long story short I basically got the same signals at the different locations so moved on a little in the circuit to C36 2uF. Working board back in and measured and the signal was there but smaller from the resistors. Faulty board in and measured and the signal was there but not as defined. Almost just noise. So I figured **** it, it's 5p and needs replacing anyway - replaced C36 with a new 2.2uf 63v (closest I can find these days) and it now works! Both missing sounds are back.
I measured C36 in circuit and got 2.2uf (within 10%) but when I measured it out of circuit it was struggling to get to 1uf. So in circuit the reading may be getting affected by the other two closely situated caps. Threw me off totally as I had checked it early on...
What I don't understand is that with C36 being buggered - how the other sounds were working!!?? Does my head in Perhaps something to do with the different frequencies - who knows. At least fixed though...