Electrolytic reform school, how long a period of inactivity to be necessary?

[cdev]

How long does an electronics device need to be off to need "reforming" its capacitors before full power is applied?

I am thinking specifically about a few power supplies I want to use.

One of them is a power one MAP110-S148 (110VAC) multiple voltage supply and looking at it it uses ten nichion caps. It hasnt been turned on since around 2009 and I'd hate to lose it. I do have access to the caps from the bottom, but removing the largest group of six of them or lifting one leg is impractical. OTOH, the three biggest ones and the smallest could be removed.

The supply's full load output is 5.1 v at a bit more than 9 amps, 12.5v at 3 amps, and -13v at 1.5A

I have a small Variac which I hope to use to gradually bring the voltage up but I m wondering, do supplies of this kind behave when this is done. This would be under no load.   The Variac is very small, its only rated at 100 watts. Its a clasisc old school Variac.


The other PS is a pretty generic but nice quality 12v and 5 volt external supply that used to be used on an APS external hard drive.

 its much like the one that I had which blew up when it was plugged in. In ts previous life it was very reliable and always ran cool.

How long should the ramp up time be when doing this, in particular.

[floobydust]

If your power supply is not under full load (low ripple current) then I would use a low voltage soak and then a nominal voltage (idle) before loading it up.

I find electrolytic capacitors have a half-cell potential, you can measure some in your parts bins.
If the device has resistive loading (bleeder) that ensures the capacitors are completely discharged right to 0V, the oxide layer seems to degrade faster. Measure the DCV on the filter caps and if they are flat zero, then I would fully reform them.

"attention must be paid to the fact that after storage the leakage current may be up to 100 times higher than normal during the first few minutes following the application of power."

In VFD's where the capacitors will pop if ESR is high, they (ABB) have guidelines:
Under 1 year storage, no reforming necessary.
After 1-2 years storage, 1 hour idle at nominal bus voltage.
Beyond 2 years storage, reforming is necessary.

Variac with 25% for 30 mins, 50% for 30 mins, 75% for 30 mins, 100% for 30 mins to 1 hour per year of storage.

Cornell Dubilier reforms:
1. With full rated DCV but 100R (<100VDC) or 1k (>100V) series-resistor for 2-4 hrs.
2. Discharge with that resistor for 1hr.
3. With full rated DCV+extra and series-resistor for 5 minutes+5RC tau.
4. Check leakage current

Note electrolytic new capacitors are "formed" at up to 200% of their rated DCV.
"When the forming voltage VF is exceeded, the forming process restarts and large amounts of gas and heat are generated."


ref.
https://www.cde.com/resources/catalogs/AEappGUIDE.pdf
https://library.e.abb.com/public/9f1befcd62ce445ec1257466003c02e2/3BFE64050629.pdf

[cdev]

Thank you! This is exactly the info I was looking for!

This is so useful, This particular info should be a sticky post in the Beginners or Repair section (its probably a very common problem that destroys a lot of equipment before its time.)

Neat to see the internal structure of whats going on too.

This is an example of where I am really glad I still have my old Radio Shack (Well, Micronta) analog multimeter

Monitoring the voltage drop (proportional to current) as I apply the voltage I see this voltage drop (the current leakage level) going up fast and then settling down with each increase in the voltage, (but the curve isnt perfectly linear there is a tiny bit of noise in there.Popcorn noise's DC equivalent )and it is a very reassuring thing to see it going down.

But I want to make sure I am doing it right.

Thank you!

Dave, have you done a video on reforming capacitors? I bet you have, somewhere..  You should do one if you havnt already.

I have a large stash of top quality but older capacitors,  Properly de-rated (by a lot) after doing this, I can still use them.


Floobydust, There is one sentence in what you said - the one about the half-cell, that I want to ask you what you mean.

Ive noticed that even after being discharged with a resistor, immediately after removing the resistor, the capacitor starts behaving like a battery, as its not fully discharged yet, its still storing some of the previous energy.So I may have to do it several times.

I assume thats kind of an interaction between its charge holding ability and its internal resistance..

I have a DE-5000 which will happily measure all sorts of parameters that I still don't understand but of course unfortunately it cant do that under high voltages.. So even after shorting a capacditor (through a hefty resistor) I double check its all gone and often have to do it twice or even more.

This situation is a great demonstrator for my SO as to how useful it is to have all this "junk" as she sometimes calls it around.

If your power supply is not under full load (low ripple current) then I would use a low voltage soak and then a nominal voltage (idle) before loading it up.

I find electrolytic capacitors have a half-cell potential, you can measure some in your parts bins.
If the device has resistive loading (bleeder) that ensures the capacitors are completely discharged right to 0V, the oxide layer seems to degrade faster. Measure the DCV on the filter caps and if they are flat zero, then I would fully reform them.

"attention must be paid to the fact that after storage the leakage current may be up to 100 times higher than normal during the first few minutes following the application of power."

In VFD's where the capacitors will pop if ESR is high, they (ABB) have guidelines:
Under 1 year storage, no reforming necessary.
After 1-2 years storage, 1 hour idle at nominal bus voltage.
Beyond 2 years storage, reforming is necessary.

Variac with 25% for 30 mins, 50% for 30 mins, 75% for 30 mins, 100% for 30 mins to 1 hour per year of storage.

Cornell Dubilier reforms:
1. With full rated DCV but 100R (<100VDC) or 1k (>100V) series-resistor for 2-4 hrs.
2. Discharge with that resistor for 1hr.
3. With full rated DCV+extra and series-resistor for 5 minutes+5RC tau.
4. Check leakage current

Note electrolytic new capacitors are "formed" at up to 200% of their rated DCV.
"When the forming voltage VF is exceeded, the forming process restarts and large amounts of gas and heat are generated."


ref.
https://www.cde.com/resources/catalogs/AEappGUIDE.pdf
https://library.e.abb.com/public/9f1befcd62ce445ec1257466003c02e2/3BFE64050629.pdf

[CaptDon]

For capacitors of your vintage and ratings it is a 99.999%
chance you can plug-n-play. They are not under any real
critical stress. We have had HP voltage/current regulated
supplies surplused out of labs that sat in storage for 20 years
and they were fine!! Some of the Sorensons had caps that
were apparently not well sealed and dried out and became
basically a leaky open circuit. No amount of reforming would
help them anyway. The only capacitors in home consumer use
I ever did reforming with was old table radios from the 50's
and it was a total waste because they were so far out of spec
even after reforming they had to be replaced.