Here are postings I made on this topic on this group on August 1, 2001
and April 22, 1999. I'm sure you can find a lot of other relevant
postings with a google groups search.

--------

You apply a variable DC voltage through a current limiting resistor.
Adjust the voltage so that the capacitor draws a few mA of current.
(What you want to avoid is excessive heating of the capacitor.) As the
dielectric layer thickens, the current will slowly drop. When it does,
increase the voltage again to get the current back up to a few mA.
Continue this until you've applied the full rated voltage and the
current at that voltage has dropped to a very low level. Or you can use
a fixed voltage source equal to the rated capacitor voltage, and vary
the resistance to maintain the desired current.

--------

One point that hasn't been mentioned here is that the voltage-handling
ability of electrolytic capacitors declines if they're stored for long
periods without use. So a capacitor might be bad after long storage,
but can be brought back to life.

The insulator is a coating which is plated onto one of the electrodes
by an electroplating process. This process naturally occurs due to the
applied voltage when the capacitor is in use. When stored for long
periods, the coating deteriorates and must be re-formed to restore the
capacitor to its original voltage rating. (An associated problem is
that when in use, the dielectric coating will, over a long period,
naturally degrade to match the applied voltage. So if changes are made
in the circuit which raise the capacitor voltage, say by replacing an
old rectifier tube, the capacitor can leak or explode due to the
suddenly increased leakage current.)

An electrolytic capacitor is re-formed by applying a voltage through a
current-limiting resistor. The voltage is slowly increased until the
leakage current reaches some specified value, then left at that level
until the leakage drops to a low value as the dielectric layer forms.
Then the voltage is increased until the leakage current again
increases to the limit and left at that value for forming, and so
forth, until the leakage is adequately low at the rated voltage.

There are various rules of thumb for how large a leakage current can
be tolerated during re-forming. One reference I have for modern
capacitors gives 5 mA for small PC-board-mount capacitors, and 10 mA
for large computer grade capacitors. It seems to me the important
thing is to limit the temperature rise, so the tolerable current would
really be a function of capacitor size and the voltage across it. A
quick look at the _ARRL Handbook_ index didn't reveal anything about
this topic, but I know it was in the Handbook many years ago. The
folks on the .boatanchors newsgroup should be able to provide a lot
more practical information on the topic for those who are interested.

Roy Lewallen, W7EL

Steve Nosko wrote:

"Per-Ake Andersson" wrote in message
news
I have a powersupply which have been stored for 15 years.
Now I want to use it. But I think the high volt(500 V) capacitors
need to be refreshed (right word?). (To slowly increase the voltage.)

Questions are for ex.: how fast, how long time at each level.
Cant find any info on the net or in the handbook.
Is there any info on the net ? Or here ?

/Per-Ake



Wish I could remember where, but there was something within the last 6 mo
where it talked about monitoring the CURRENT. I would think that HOT caps
are a BAD thing and should be avoided. My guess is that t15 years isn't
long, but better safe...