Cap reformation question

[Dwight K. Elvey]

Hi
 My experience with the variac is that it is too uncontrolled.
The idea of using a light is good but what wattage should
one use. Remember that the light has quite a bit of positive
temperature coefficient. When cold the resistance is a lot
less then when glowing.
 If it is a switcher, you shouldn't use a variac at all.
You should disconnect one lead of the capacitor and put
a series resistor to a power supply. I recommend using
a voltage on the supply of at least 50% of the rated voltage
and a resistor that if the cap is shorted will supply
a maximum of 10 ma to a low voltage cap that is about 1.5
inch diameter and about 3 inches long. Smaller caps should
have less current. Many computer capacitors may draw more that
10 ma as lower voltage, higher capacitance caps. If
after 10 or so hours at 10 ma, you don't see any change,
increase the current some. For high voltage capacitors,
you should limit to about 1 or 2 ma.
 Even on a linear supply, I recommend disconnecting the
capacitor and bringing it up with a resistor and supply.
Switchers can blow up with too low a voltage. Most newer
supplies will be OK but you won't get good forming of
the output capacitors since the supply will remain
off until the voltage is high enough to start the switcher.
This usually means full voltage to the output capacitors.
 For this and other reasons, I still recommend isolating
one of the leads of the cap and using the limited supply.

>From: "Robert Armstrong" <bob at jfcl.com>
>
>JC,
>
>  AFAIK the issue with reforming old electrolytics is that the caps will
>have very high leakage currents until they reform.  The goal is to limit the
>power dissipation of the cap to something low enough that it won't vent (or
>explode!) until it's recovered.
>
>  Lots of people use Variacs for this purpose and they'll certainly do the
>job, but they don't give you any way to measure the leakage current.  I
>prefer the simple technique of putting a light bulb in series with the AC
>side of the supply - when it's bright, the current is high and as it
>gradually dims and goes out, the current is low.  
>
>> maybe leave a junk card in for load?)
>
>  If it's a linear supply then it's probably happy without a load, and
>there's no sense in wasting even a junk board.
>
>  If it's a switching supply then it depends on the design, but most likely
>it will require some kind of load on at least some of the outputs to
>maintain regulation. 
>
>Bob
>
>> -----Original Message-----
>> From: cctalk-bounces at classiccmp.org 
>> [mailto:cctalk-bounces at classiccmp.org] On Behalf Of J.C. Wren
>> Sent: Tuesday, December 06, 2005 7:13 AM
>> To: cctalk at classiccmp.org
>> Subject: Cap reformation question
>> 
>> 
>>     I've got a system with a couple large electrolytics that 
>> hasn't been 
>> powered up in a number of years.  I have a 5A Variac.  What's the 
>> typical procedure for the reformation process?  Remove all cards (or 
>> maybe leave a junk card in for load?), start the Variac at 
>> 0V, increment 
>> by 10V every so often?  How long per step, and/or is a different step 
>> suggested?
>> 
>>     I've seen lots of people say "you'll need to reform the caps, 
>> first", but never a prodecure for it.
>> 
>>     --jc
>> 
>> 
>
>
>