Mac Classic II - Only Works On Isolation Transformer

[TheMightyMadman]

Hi all,

I have a bit of a weird issue with a Macintosh Classic II that I'm trying to repair.
At first, the analogue board was seemingly completely dead, and the digital board was giving a "stripy" screen.
The digital board now seems to be working OK, following a recap and a CPU replacement (some of the pins fell off due to electrolyte corrosion).
The analogue board still has problems - the system seems to work OK, but only when powered from my isolation transformer.
When powered from direct mains, the system tries to boot - however, when cold the system constantly resets and the high voltage cuts in and out; when warm, it seems to usually boot, but the display is very wobbly and the hard drive doesn't spin up immediately, so the system doesn't boot off it.

The analogue board suffered major electrolyte leakage and subsequent damage.
I've replaced all of the electrolytic capacitors on it (except the line filter cap), and I've replaced the two mains filter capacitors.
I've removed and cleaned most of the larger components including the transformers and heatsinks, and the PCB underneath them all.
I've replaced several of the actives and passives, particularly those in the low-voltage filtering area which was most affected by electrolyte leakage.

Does anyone have any ideas as to why this might be happening?
Is it the current limiting of the isolation transformer, or the decoupling from mains earth?
Could there be a problem on the mains input side of the analogue board?

Many thanks,

Adam

 

[ymk]

What about powered from mains with ground lifted?

 

[stepleton]

I think there is one salient difference between running off of an isolation transformer and running off of the mains, which is that the earth pin will be connected to nothing on the isolation transformer, while on the mains it will be connected to neutral (usually) at a point near your breaker box.

I think it will be useful to work from a theory that the connection between earth and neutral leads to the problem you're seeing here. A problem with the mains input side seems likely to me, but I'm not expert enough about power supplies to suggest what it could be.

Anything that has strange behaviour relating to earthing makes me nervous. I would be extra careful around this analogue board.

 

[TheMightyMadman]

I think there is one salient difference between running off of an isolation transformer and running off of the mains, which is that the earth pin will be connected to nothing on the isolation transformer, while on the mains it will be connected to neutral (usually) at a point near your breaker box.

I think it will be useful to work from a theory that the connection between earth and neutral leads to the problem you're seeing here. A problem with the mains input side seems likely to me, but I'm not expert enough about power supplies to suggest what it could be.

Anything that has strange behaviour relating to earthing makes me nervous. I would be extra careful around this analogue board.

Thank you I agree, I'll definitely take extra care when working on this one.

Mains earth *seems* to only be used on the power input section (see schematic snippet attached).
CP19 and CP38 (XY class) have already been replaced.

I'll go ahead and pay close attention to the other components in this part of the schematic.
I also didn't pull the components in this area to check and clean underneath them, so there may be some electrolyte underneath causing problems.

Mains earth is connected through some kind of choke? Is this just the ferrite bead?
CP30 is between live and earth.
CP29 is between neutral and earth.
Components between live and neutral are RP46, LP3, CP16, LP2, CP28, and the bridge rectifier.
CP19 (replaced) couples the board ground to earth via CP38 (replaced).
CP22 couples the DC output of the bridge rectifier to earth via CP38 (replaced).
RP18 is across the DC output and common of the bridge rectifier.

I'll keep you all updated.