Update and Pictures: It's been a hell of a day! HP 9845/2x

[Tony Duell]

> 
> Hello,
> 
> the PSU's of the 9845 are always a pain.
> 
> Esp. for the 9845s produced in Germany once, at first all capacitors 

Argh! I think my 9845 is a German one....

> have to been checked before applying power. Several of them are of an 
> epoxy based package and like to produce a failure of the PSU (I killed 

Which capacitors are these? Are you thinking of the snubber networks on 
the main chopper transformers -- the 2 large non-electrolytic caps on the 
mainboard?

> two PSUs up to now and I'm not daring to power on the third).
> 
> >>   I checked the HP 9845B and it's absolutely dead. The fan doesn't even
> >>run. However the fault should be easy enough to locate. It's odd though,
> >>    
> >>
> >
> >Be warned that thr 9845 PSU is _very_ complicated, even worse than the 
> >PDP11/44 PSU (!).

Perhaps I should clarify that although it's complicated, it's not too 
unfriendly to work on.

The little PCB at the front contains the 4 chopper transistors (2 
push-pull chopper circuits). These, of course, are on the mains side of 
the PSU, but the little pot-core transformers on this board isolate the 
base drive signals. Therefore the 4 chopper drive transistors (on one of 
the output PCBs) and the entire control circuit (on the complicated PCB 
at the back) are isolated from the mains.

Moreover, there's a nice, simple, linear startup supply for the control 
circuit. 

So checking waveforms and debugging the most complicated part of the 
supply shouldn't be too unpleasant.

> >
> As you seem to know the PSU quite well, do you have schematics?

Only reverse-engineered ones. THese will end up on a later version of the 
HPCC CD-ROM, after I've figured out the rest of the machine...

-tony