Dying VaxStation 4000 VLC - help?

From: Mark Tapley <mtapley_at_swri.edu>
Date: Sun Nov 17 00:13:01 2002

>From: ard_at_p850ug1.demon.co.uk (Tony Duell)

>I rarely bet except on certainties :-)

Yeah, me too. As you probably noted...

>>> Now *I'll* bet - Tony wants to know what the comparator *inputs* are
>>> connected to.
...
>>
>>I don;t think it's for overvoltage protection. I think this is where the
>>PowerOK signal comes from. And it's triggering at the wrong time. So
>>either the comparator is defective or there's a problem with the passive
>>components areund it, or there's a PSU fault.
>>
>>How hard is it to reverse-engineer the schematic of the components around
>>this comparator? Can't be _that_ complicated...

Yup. a sure bet. :-)

Complicated, probably not. Hard, yes. They are all surface-mount and near
microscopic. For which reason, all numbers and letters quoted below may be
wrong. Here's what I think I found:

Comparator INV Input B and Input B connect to ground (zero Ohms to chassis).

Comparator INV Input A connects to the middle pin (which points left)
        of a 3-pin device marked "88W". I couldn't find that the upper right
        pin connected to anything. The lower right pin connected to ground.
        That input stayed very steady at about 3.4 V.

Comparator Input A connects to an orange device marked "6192" and a black
device
                marked "1152". Both are 2-terminal surface-mount rectangles.
        The other end of the 6192 connects to ground.
        The other end of the 1152 connects to +5 (Red wire).
        That input *varies substantially and irregularly*
        It's at 4.2V and reasonably steady when the computer is running.
        It's at 2.9V, or wavering in between when the computer is hung.

There may be *lots* of connections I missed.

"Wavering" is what I see my needle do - for all I know, the voltage may be
popping rapidly between the two values, with varying duty cycle.

Here's what I think may be happening: 6192 and 1152 are voltage-dividing
the +5V supply down to some intermediate value for the comparator. But
either the resistors are flaky (do resistors *do* that?) or the comparator
is intermittently drawing current, and throwing off the voltage-divider.

Good so far? Any way to test those theories with my analog VOM? Should I
order a new LM193 comparator chip instead of a pile of capacitors for my
first soldering attempt on this machine?

If I *do* need to pull the comparator, I'll need to take the mainboard out
of the chassis. That'll mean removing 4 extremely nasty plastic rivet
things. How do those suckers come out in such a way as to leave them
intact? I wrestled with one for a while and got nowhere.

>It can't do any harm provided the new capacitors are good, and you get
>them the right way round (they are, of course, polarised.)

I didn't make it to the electronics shop today - Kids' soccer game and
wife's Newtonsday (sic?) shopping interfered. She promised to go for me
Monday if I wanted. I explained to her what a "capacitor" is so that she
could breeze in, look knowledgeable, and say "I'd like replacements for all
the electrolytic capacitors on this board." She practiced it three or four
times so she wouldn't stumble over "electrolytic capacitors". I also
pointed at them, just for good measure.

All that notwithstanding, the capacitors may get due process after all.

                                                        - Mark
Received on Sun Nov 17 2002 - 00:13:01 GMT