Mac IIcx booting intermittently, PRAM problem?

[cesare]

I've got a IIcx which has a booting problem - I can't work out a pattern, but when it fails to boot, the PSU gets going from the button on the machine or the keyboard button, but there is no boot chime.

The fix is to remove the battery, wait 30 mins or so, replace the battery, and then it fairly consistently boots. It might then work for a few days, or maybe an hour later it will fail to boot. Once it fails, it never starts working again without replacing the battery (i'm fairly sure of this).

I've checked the power supply rails, and they look great. I have recapped the PSU, and it seems to be strong and sensible, so i think I can rule that out.

The motherboard has been recapped, and I initially thought it was a bit grim, but i'd got away with it, since the capacitors by the battery had clearly started to leak onto the board, but it was mainly grim looking solder rather than anything more nasty. I did however find when the machine was first running that the clock was slow, and traced this to a corroded trace, so there is one bodge wire to replace this.

So, any thoughts on what to try next? I'm assuming that the RTC also holds the PRAM data? If so, and i've had one damaged trace here, is it possible that there is another issue which is allowing the PRAM to get corrupted, leading to a boot failure? Speculation on my part, but is there, for example, a write disable line which is missing, and so random code is managing to corrupt the PRAM? Some more info on where the PRAM is, and how it is managed would be handy. I can dig around with an oscilloscope once I have a plan of attack.

Thanks,

Cesare

 

[Addicted]

Are you certain the battery isn't nearly dead? Removing it and replacing it might allow a really weak battery to build up enough residual charge while it is removed to get the OS running again, but it would then be so very borderline that very shortly, it would no longer be able to repeat the process. Mouser Electronics and Digi-Key both still stock the type of battery used for the Clock/PRAM.

 

[just.in.time]

Agreed, sounds very similar to how my 6100 (which requires a PRAM battery to boot normally) behaves right as the battery is at the end of life.

 

[Arbee]

Low enough battery voltage will corrupt PRAM and the RTC. I misread what machine this was originally, the IIcx has the discrete PRAM/clock.

Datasheet for that chip is at https://www.brutaldeluxe.fr/documentation/cortland/v6/3430042B Real-Time Clock ERS 198411.pdf

 

[cesare]

The battery is new and is not the problem. In fact, my strategy for getting it to boot again seems to have stopped working, and I think was a total red herring. I've done some analysis on the bus and things don't look good at startup. I'm not totally sure what things should look like (i've little experience with 68k busses) and as I don't have a working board to compare with.

Anyhow, the current plan is to work through the first few bus cycles and see what is what. I've got 16 digital channels on my scope so I should be able to at least trace the main bus signals (NUCHIP seems to expose lots of interesting stuff). It'll end up being a dodgy trace or chip causing a chip select to be wrong, or something like that, but without knowing what things are supposed to look like makes it somewhat harder, but that's the fun right?