New user - fixing an Atmos

[virtvic]

I first fitted RAM that I had from stock, then bought some new and tried that.

I have also continuity tested all the ram to each other and they are fine?

I removed the 6522 tonight and fitted another 6522, but it could be a dodgy one. I’ll find a known working one tomorrow and try that. The original 6522 had a bit of a hot spot on it when left for a few minutes.

[virtvic]

Nope. Just swapped in a known working 6522 (I think - from a vic20 that has a faulty keyboard?)

I’ll order a 74LS04 and also a 6502 and try those next.

This is just shot gunning now. I have a logic probe. How would I check the 6502 and 6522? Get the pinout chart and check HI / LO / pulsing pins?

[Chema]

For what I've seen from other people fixing computers, I'd say that until you fix the voltage, you won't be able to go further. I've seen people checking voltage at different points starting from the regulator and moving further to check if they can find where it drops.

But most of the time, they end up removing the main chips one by one. I would not plug in replacements. Just check if the correct voltage reaches them, because else they won't work at all.

In fact I think the 6502 could work with 4.75v, not sure the ULA, so if you have a probe you can start checking the clock signals, from the 12MHz oscillator, then the ULA (in pin 17), which generates the 1MHz clock to feed in the CPU (CPU pin 37) and the CPU phase 2 (CPU pin 39). As you have an image, I guess the oscillator and the ULA clocks are correct, checking the CPU could help.

Then, yes, you could check activity in the address bus or RAS/CAS memory signals.

[kenneth]

The voltage of 4.75v is critical for logic circuits and it is not normal. You must try to connect a big resistor at the output of the regulator (two 3.3ohm 5w resistors in serial mode) and disconnect everything else. We would have proof that the power supply is correct.