PDP-8/L power supply prints and "reforming" capacitors

From: Tony Duell <ard_at_p850ug1.demon.co.uk>
Date: Sun Aug 17 10:42:00 2003

> Also, does anyone have experience in "reforming" catalytic capacitors for

electrolytic, surely?

> the PDP-8/L power supply. Is it necessary to do it ? What are the possible
> dammage if the "reforming" step is not executed.

OK, as you are probably aware, the dielectric in an electrolytic
capacitor is an oxide film on the postive plate. The electrolyte is the
negative 'plate'...

Sometimes the oxide redisolves into the electrolyte, whereupon the
capacitor becomes an effective short-circuit. Reforming is a process to
redeposite the oxide, and is perfomed by connecting the capacitor to a DC
power supply, with a voltage a bit higher than the working voltage of the
cap, with a resistor in series to limit the current. The current flow
should decrease fairly quickly -- if it's still drawing significant
current after a few hours, then I guess you have to replace the cap.

Of course if the capacitor is shorted, there can be 2 problems. One is
that a dead short on the output of the transformer/rectifier stage may
well damage those components, or at least blow fuses. Also, the capacitor
will pass enough current to get hot, and the electrolyte may even boil.
At which point the darn thing will explode. Most caps have safety vents
just in case this happens, but if not you don't want to be anywhere near
it! Even worse, the flying bits of capacitor, or the jet of electrolyte,
may damage other bits of the PDP8

However, my experience suggests that shorted electrolytics in computer
PSUs are _very_ rare. I've only ever seen one, that was in a Canon CX
laser printer. All my PDPs, etc, are running on the original caps still,
and I've not reformed them.

It's worth testing them by connecting them us as though you were going to
reform them (see above) and checking the steady current flow. If there is
such a current, you need to reform or replace them.

-tony
Received on Sun Aug 17 2003 - 10:42:00 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:35:47 BST