reforming PS caps w/primary-side components

From: Tony Duell <ard_at_p850ug1.demon.co.uk>
Date: Tue Jun 3 14:02:00 2003

> Anybody want to offer advice on reforming filter
> caps? I'm planning to power up some stuff that
> hasn't seen voltage in a long time. My plan is
> to put several low-wattage light-bulbs in series
> with the thing, to limit the amount of current it
> will get, and hopefully allow those old caps
> to reform instead of, say, exploding in my face.


Light bulbs will limit the maximum mains current in the event of a
catestrophic failure on the mains side of the PSU (this is particularly
valuable when repairing SMPSUs), but I doubt they'll have much effect
here. You really need to remoce the capacitors and connect them in series
with a fairly high value resistor across an adjustable DC power supply.
And then increase the voltage while keeping the current fairly low until
the capacitor will stand the rated working voltage.

>
> It would be better to put the light bulbs (or
> whatever kind of resistor) in series with the
> caps themselves on the secondary side of the

Yes


> Also, what kinds of things could be damaged by
> getting less voltage than they were designed for?

Switch-mode PSUs for one. These things appear as a constant _power_ load,
and draw more current as the input voltage decreases. If you're unlucky,
this will kill the chopper transistor, etc.

> I could imagine hard disks spinning too little to
> lift the heads from the surface, for instance.
> (This is moot, because there are no hard disks
> in any of the equipment I'm looking at.) How
> about CRT's? Could too little deflection make
> the beam hit something it shouldn't? Anything

CRts are an interesting case. A low supply voltage will almost certainly
result in a low EHT, which will mean a less stiff electron beam. So the
deflection angle might actually increase!. However, I think the main risk
would be damage to the cathode by trying to draw too high a beam current
from it while it's too cold (the heater voltage would be low too).

What are the devices in question? My experience suggests that most stuff
from the late 1970s and 1980s is fine -- the capacitors are still good
(OK, they'll be the odd failure, but..). This particularly applies to
computers and test gear which used good quality components in the first
place. Older, valved, stuff, needs to have the capacitors removed,
checked and re-formed. More modern stuff, particularly consumer-grade
stuff, may well have capacitor problems too.

Incidentally, all my HP desktop calculators, my Tekky 555 'scope, and
most of my minicomputers still all the original capacitors!

-tony
Received on Tue Jun 03 2003 - 14:02:00 BST