31 Aug
2014
31 Aug
'14
8:53 a.m.
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-bounces at classiccmp.org]
On Behalf Of Robert Jarratt
Sent: 29 August 2014 21:27
To: 'General Discussion: On-Topic and Off-Topic Posts'
Subject: RE: Mysterious rtVAX 1000 Failure
regulator, all
at zero.
I have compared a known good PSU with the bad one and I can see that all the outputs from
the PSU, including P OK and DC OK are shorted to Common, whereas on the good one they are
not. Any suggestions where to start looking in that case?
Regards
Rob
-----Original Message-----
From: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org]
On Behalf Of Robert Jarratt
Sent: 29 August 2014 12:04
To: 'General Discussion: On-Topic and Off-Topic Posts'
Subject: RE: Mysterious rtVAX 1000 Failure
board.
-----Original Message-----
From: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org]
On Behalf Of Robert Jarratt
Sent: 29 August 2014 08:40
To: 'General Discussion: On-Topic and Off-Topic Posts'; 'General
Discussion: On- Topic Posts Only'
Subject: RE: Mysterious rtVAX 1000 Failure
up.
>
>
> Yes, it is a BA23 and I do have the "bad" wire harness. In fact I
> think all my
> BA23 machines have the bad one, but I don't run them that often, and
> they are not particularly loaded, so hopefully that will be OK.
>
> The fans are powered from the half of the power supply that contains
> the input filter and the rectifier board. The bulk of the power
> supply is in the other half and does not appear to have any
> involvement with the fans. As the fans don't work, this would
> suggest that the fault is either on the input filter board or the rectifier -----Original Message-----
From: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org]
On Behalf Of Chuck McManis
Sent: 28 August 2014 23:09
To: General Discussion: On-Topic Posts Only
Subject: Re: Mysterious rtVAX 1000 Failure
If it actually made a 'bang' sound the most likely culprit is an
electrolytic capacitor. A visual inspection should find it,
although I've known them to fail pretty amazingly leaving the case
intact except for the circle on the top where the plates came
flying out and landed
somewhere remotely.
The system is waiting for P_OK (power OK) before it will fire up,
and the fans are on the non-mains side of that PSU as I recall. Is
it the
BA23 case? (most common case) there was an issue with the power
distribution harness on those where DEC used several wires in
parallel to carry the 5V current. Except they were not all the
same length and the
short one would some times vaporize.
> The 'fixed' harness you can see all the wires are the same length
> which makes it harder to tuck in behind the back plane but it
> doesn't blow
Is it safe to connect up the first half of the PSU to the mains,
switch it on and try to probe around to see if I can see where the fault lies?
Answering my own question, I decided it was safe to connect up the
filter/rectifier half of the PSU, and indeed it was. I have found the following:
The large capacitors, which I think are smoothing the output of the
rectifier, are charging up. The V+ and V- connector that goes to the
other half of the PSU has about 320V across it. However, there is
another connector that output +15V and -15V, these are both at zero.
The LM317T linear regulator appears to have no input and no output, it
could be that I am measuring it incorrectly though. I connected one
probe on my DMM to the chassis and then tried the three pins on the
So I am guessing the output of the rectifier is not reaching the
regulator, which I am guessing runs the fans. It would seem that the
rectifier itself, and the input filter are all fine.
I still can't see any failed components, but the next step is going to
be see if I can work out the schematic by examining the board, so that
I can narrow down where the failure might be.
Regards
Rob
I have done some reverse engineering on the board that houses the rectifier
and which has the outputs to the fans. I had been assuming that the fans were
powered by this board alone, possibly through that linear regulator, and not
through the SMPS, which led me to think that if the fans don't work it must be
this board rather than the SMPS (I am not an electronics person, so pardon me
if that was a silly thought).
However, I now see that it is impossible for this to be the case. The rectifier
portion of the board is isolated from the portion which powers the fans by an
optocoupler, so the power is not coming from the rectifier. In which case it
must be coming from the SMPS on the other half of the clam shell, through a
connector which I had assumed was an output from the rectifier board, but
appears to be an input.
As power is getting sent to the SMPS half of the clam shell from the rectifier,
that means the fault has to lie in the SMPS side of the clam shell. This fills me
with dread, as the board is much more complex on this side, I don't have
schematics, I am no PSU expert, and I still can't see any physical damage,
although there *must* be some.
Regards
Rob