Re: Vulcan IIgs Power Supply Repair

[Michael J. Mahon <mjmahon@aol.com>]

Polymorph <mike.a.stephens@gmail.com> wrote:
> On 8/07/2012 7:13 AM, Michael J. Mahon wrote:
>> Polymorph <mike.a.stephens@gmail.com> wrote:
>>> On 7/07/2012 2:21 PM, Michael Black wrote:
>>>> These filter capacitors operate at a much higher frequency, which means
>>>> if they go bad, they will do a much worse job than if 60Hz was
>>>> involved.  The higher frequency ripple that they have to deal with every
>>>> day may also help to make them go bad.
>>>> 
>>>> Those filter capacitors after the transformer and diodes are doing much
>>>> harder work than the high voltage capacitors on the AC side.
>>> 
>>> Thanks Michael. I guess I'll continue and try replacing all of the
>>> electrolytics after the transformer. I should note though, that none of
>>> the capacitors (even the ones I replaced) show any signs of leaking or
>>> bulging - but I'm guessing they're still the most likely cause of the
>>> failure. The caps after the transformer are all low voltage (I think) and
>>> easy to come by for a few dollars.
>>> 
>>> Thanks for your advice. :-)
>>> 
>>> I'll get the parts tomorrow and post back my results.
>> 
>> I'd check all four outputs for shorts.
>> 
>> The repeated shutdown/restart cycle can result from too much load as well
>> as too little.
>> 
>> And the line side capacitors you replaced do every bit as much work as the
>> secondary filter capacitors. They have to rapidly charge to the mains peak
>> voltage, then sustain the supply until the next half cycle.
>> 
>> No supply ever cycled because of an open filter capacitor on an output, but
>> it certainly would if an output were shorted--possibly by a filter
>> capacitor, but given the history of your repair, I'd look for a short.
>> 
>> -michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon
>> 
> 
> hmmm.... there could well be a short somewhere, as I've just replaced
> nearly every electrolytic capacitor in the power supply with no change in
> behaviour as yet. There are only 3 I didn't replace, as I couldn't find
> axial capacitors at my local electronics store (Jaycar). The axial
> capacitors are required because there is no room for a standard radial
> capacitor (they fit underneath the mains power socket coming into the supply).
> 
> It would seem this one's going to require more attention.  :-(
> 
> What is a good method for testing for shorts? Simply going from component
> to component and testing across the terminals with a multi-meter? My
> multimeter has the continuity test with the beep.
> 
> If I can't figure this out 'shortly' (pardon the pun), I might have to
> consider alternatives, like replacing the innards with one of those pico
> power supplies with the external transformer. I'd prefer to keep it
> original, but my hardware skills are pretty limited.
> 
> I'm still at a bit of a loss to understand how it could have got a short
> when I simply replaced the fan? (no soldering required, as it just plugs
> onto the circuit board). Was it just a case of bad timing for a component
> to burn out or what? Could it be that the fan stopping was a side effect
> of something giving out on the circuit board?

Think something mechanical--a component lead pushed against another or
against the case, a bit of conductive debris dropped into the supply, etc. 

Since the supply is cycling, you could connect a voltmeter to each output
in turn and briefly turn it on. You should see a pulsing voltage on any
unshorted output, but very little on the shorted output. 

Although a short is most likely, it may be an open as well--a tiny break in
a board trace or a broken component lead. 

I'm assuming that the supply was working fine except for the fan when you
replaced the fan. Is this correct? The fan runs on +12v, so that output may
be shorted. No disk drive will work without +12v. Were they working prior
to the fan replacement?

It's always wise to assume that when you do some work and a fault develops,
the fault was caused by something that happened during the work. It's not
*always* true, but almost always. ;-)

-michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon