When is an electrolytic capacitor reformed?

[stevehertz]

The past few days at intervals I've been playing at reforming old electrolytic caps. I've had a stash of old ones recovered from scrap kit for years, but having recently bought a load of new ones (again) I got to thinking about how expensive new ones are, and the overall cost when you're doing a lot of restorations. Hence my interest in reforming old ones. Using an HP 0 - 50v power supply I tried a few different caps. The usual procedure, monitor voltage and current, gradually increasing the former while keeping the current low. Some caps reformed to the point of the current being, or near enough, zero. Others don't seem to want the current to reduce below about 10 uA. Is that 'ok', is that cap now ready for use? I suppose it matters what role or function the cap is fulfilling. As a smoother I guess it would be ok, but in a DC block role, perhaps not. Any views or comments on reforming electrolytic caps? Thanks.

[joebog1]

I think you have it in one. If its a DC blocking role. I re assign it to the bin. That simple.
If its a filter or "smoother " as you suggest, expect a little more than whats nice as far as hum and ripple goes. If you're doing the rebuilds commercially, CHANGE em!!!.

Otherwise make sure your insurance covers failed transformers, expensive unobtainium, rectifier valves and worse case, a house burning down.

Joe

[stevehertz]

Seems wise words to me Joe, thanks. Oh I don't have time for commercial 'radio' repairs or restorations but I do guitar repairs, mods and setups so I suppose these could be considered 'electrical' as more often than not they have pickup systems, but not a lot of use for my huge, khaki coloured, 1945, 8uf 350v electrolytic, or others in that 5 litre paint tub.

[GrimJosef]

Electrolytic capacitors are one of very few passive components that are specified with a life expectancy. Depending on how hard they're worked they will eventually fail. They generally have a specified leakage current too.

So if they're used in a circuit which, if they fail, could destroy other much more expensive components then you'd expect the designer to have accepted that when he made his component choices. If the cap failure could cause a house to burn down then that sounds like the circuit doesn't include adequate protection. Because sooner or later electrolytics will fail.

As far as I'm concerned reforming electrolytics isn't about making an inherently imperfect component perfect. It's about extending the life of a historic component which has been subjected to a procedure (a long period of disuse) which is known to degrade it. I don't think I've ever had to reform a cap which has been in regular use (say a few times a year for HT reservoirs and smoothers) .

Cheers,

GJ

[joebog1]

GJ, some equipment HAS adequate protection, but there are also experts that will up a fuse rating to 3 amps from 1 amp. It just takes a bit more time for the transformer to catch fire.

AS Steve has acknowledged, IF its commercial equipment, it shoud be "perfectly correct ". At least as it was designed. If its a radio or TV set for a friend, they are more likely to ring you and tell you its failed again. So you fix it. I would think most "friends or family " jobs will fall under the latter category. Its when the show must go on, you tend to run into problems.


Joe

[Craig Sawyers]

There is a somewhat famous capacitor reformer circuit here https://worldradiohistory.com/UK/Rad...RC-1969-05.pdf

pages 664-667

Needs a 250V @ 35mA secondary mains transformer, some WW resistors and a 2p12w rotary switch. It neatly uses a voltage doubler to generate up to 600V, but can handle down to less than 50V too. Uses a neon relaxation oscillator, which stops flashing when the capacitor is nominally reformed. The 12th position of the switch discharges the capacitor into a suitable resistor.

There all sorts of ways this simple unit can be adapted, with a meter to measure charging voltage and leakage current. And of course not use selenium rectifiers!

Craig

[Diabolical Artificer]

Whilst looking for electrolytics for a repair the other day I found some 470u 10v caps,they tested ok for ESR but read high capacitance - 556u. I tried reforming them and though the leakage current dropped to an acceptable level they still read high C, so it's hi ho, hi ho, into the bin you go.

Andy.

[Maarten]

They could be perfectly fine given the capacitance is within +20%. Of course it's always a judgement call with old electrolytics.

[Bob_Moss]

I've built the capacitor reformer from the 'projects' section at the front of this forum. It works very well and would recommend it. It is surprising how well some capacitors that have been in storage for decades can be brought back to spec.

[Valvepower]

Hello,

With regards leakage current in electrolytic capacitors, I make reference to the Molloy Book Radio and Television Engineers’ Reference book. I’ve attached a couple of scans of the relevant pages.

I replace or reform depending on whether it’s a general restoration for the collection or something where peak performance with reliability is required; however, I’ve had to persevere with some multi section capacitors (3 or more per can) as replacing these can be a pain in the proverbial. If these fail, they get re-stuffed or replaced with a dual can, and what can’t be stuffed or isn’t available in the can are fitted on a tag strip on the underside of the chassis.

As already been said a commercial job like guitar amplifier for professional use if there is any doubt at all replace!

I use my Farnell 350V supply for high voltage and Farnell 2x 30V for low voltage with a series resistor to limit current. I leave them to ‘soak’ even after they have reformed.

Regards
Terry.

[stevehertz]

Thanks gentleman for some great input, appreciated. I may even make a reformer, either the one Craig links to, or the 'projects' one here.

Were/are commercial ones available?

[GrimJosef]

I'm not sure about commercial units, but I guess the military 'Reforming Unit, Electrolytic Capacitor, No 1' would have been mass-produced.

Cheers,

GJ

[Uncle Bulgaria]

I have made the Everyday Practical Electronics reformer (thread here) and the documentation includes a table for different capacitor types and values, and the expected leakage when 'reformed' for each type.

Attached is the redrawn table I made to double as the front plate of my reformer, using the UNO typefaces as uploaded here by Chris55000 from Chris Flippard.

[Craig Sawyers]

I have a Lafayette TE-46. This measures capacitance, resistance, turns ratio etc using a magic eye, but also has a very useful reformer function.

Manual is here https://bama.edebris.com/manuals/lafayett/te46/

Also reproducible, but would need to be adapted to not use a magic eye, and use a sensible sensitivity meter for leakage current - the one that is in there is a weird 60uA movement.

Circuit adaptation is shown here from someone who had an open circuit meter. http://air-ministry.uk/TE46.htm

Craig

[Herald1360]

To get back to the original question.....

When it behaves as it should

[trobbins]

I would comment that testing for leakage current is the same as reforming for any e-cap that has not been so tested for longer than a year or two, and that should be done at the rated voltage of the e-cap.

Even when testing e-caps that have been in service, they won't have sat at their rated voltage, and reforming is needed to confirm they still have a suitably low leakage current at rated voltage.

There have been some reports of statistically significant batch testing of old stock parts to determine how much they spread in parameters like leakage current, and the spread to be anticipated is significant (as was ESR).

The oldest reference to a commercial part's leakage current spec I have come across is from Aug 1960 for UCC, which was shown in an advert as (CV/10 + 100)uA (with the expectation that it is about half of that), and gave the recommended re-aging process as 90% of rated voltage through series 10k.

The three measured parameters of capacitance, ESR, and leakage current are what I use to assess whether a vintage cap could be kept in use.

[Craig Sawyers]

I've had capacitors that have supposedly reformed, but when measured have little evidence of there being a capacitor at all.

So I agree - reforming, ESR and capacitance value are necessary to confirm that a part is suitable for use/reuse.

Craig