Dubilier Yellow Electrolytics from NVCF B&B

[Pamphonica]

I bought 10 of the yellow Dubilier 16/32uF 350V wire-ended caps from the B&B at NVCF yesterday. NOS but unknown condition.
I put the first one on to re-form overnight and by this morning leakage on both sections was down to 250uA at 400V. So they look very good indeed for use up to their working voltage of 350V. Leakage was circa 5mA before reforming. I'll probably re-form all the rest now.
They were a bargain - but definitely DO need re-forming before use.
I hope everyone else has similarly good results.

[Pamphonica]

mods - title mis-spelling, sorry.

[Duke_Nukem]

I bought ten too; I figured if half of them didn't work then it was still a bargain! Gut feeling is 250uA sounds a tad high, but then it was slightly over-voltage. Be interested what results you get from a few more of them.

TTFN,
Jon

[Pamphonica]

At the working voltage (350V) the leakage is now 100uA - fully acceptable I reckon.
I just did a quick ESR check on the rest. One is definitely duff - very high leakage. But all the rest show roughly 40uf/0.4R and 20uF/0.6R with no reforming yet.
I'll put the next one on the reforming rack!

[Chris55000]

Hi!

As a rough guide to what was considered normal for that type of component in those days, C.A. Quarrington in Vol 3 of Radio & Television, 1963 Edition, stated in Chapter 5, "Testing Components" :-

"An electrolytic of the type used in receiver power-packs has a typical leakage of 0.07mA per uF when new, a figure that may be exceeded by about three times before serious thought need be given about replacement,", so by Mr. Quarrington's reckoning, a new 32uF electrolytic should reform to about 220-250uA when operated at rated working-voltage.

I think most data-sheets give today's leakage as roughly:-

Leakage in uA:- 0.07 CV approx. when new but they don't specify an end-of-life figure!

Chris Williams

[threeseven]

Jeremy used his Jedi mind powers on me too, so I bought 10 also, but I haven't messed with them yet.
What procedure did you use Jeremy?

[Diabolical Artificer]

"What procedure did you use Jeremy?" I used to think reforming capacitors was some mystical art, but all you do is bung some current limited voltage up it's pipe while monitoring current (Crudely put I know ). Or to be more technically correct you connect said capacitor to a PSU or some description, whilst limiting the current. You start at a low voltage, 50v say for a 350v rated capacitor and work you way up in 50 - 100v steps, a tad past it's wkg V. There are a few threads on the forum on this subject.

It might be an idea to put the cap in a box or similar in case it blows. Keep an eye on it for an hour or so at first and when upping the voltage.

Andy.

[Pamphonica]

Some while ago I got fed up lashing up cap reforming circuits and boxed up my own version. (Details here). The response was not very positive. Mostly saying I had over-complicated it.

But it allows me to set a maximum voltage and a maximum current and let it do its stuff unattended.

For these caps I set 350V initially and 5mA max. I left it on overnight on just one section and the current fell from 5mA to 250uA and the volts raised to the maximum, 350V. A bit later that leakage was down to 100uA. I then did the other section and that also dropped to 100uA. I then parallelled the sections and upped the voltage to 400V where the leakage was 500uA total, 250uA per section.

Emboldened by this success I quickly checked all the other caps on an ESR meter and got good results on all but one, which I chucked.

I have now done two more, but this time I have paralleled the sections for speed. I am getting a consistent 100uA per section at 350V and 250uA per section at 400V after 8 hours or so of reforming.

I have a theory that a limited maximum current (2-5mA) is better for the cap and a lot safer - much less heating at <1W max.

Also I set the max voltage and current on my unit, clip on the leads and get on with other things.

Each to his own...

I hope others get good results from these caps. They all went, at NVCF, so there are probably 200+ of them out there!

Jeremy

[Pamphonica]

All caps have now been treated.
1 very high leakage (useless) so scrapped
1 high leakage (nearly 1mA/section) so scrapped
8 now giving 100uA - 200uA per section at 350V. Average about 125uA/section. More than adequate.
I hope all the others come up this well
Thanks to Jeffrey for digging these out of the Dulwich museum's overflowing stores to make a nice fund-raiser for the museum.
Jeremy

[David G4EBT]

Are those the ones that look like the Dubilier one below?

If so, in my experience they're excellent, but the ones that I bought had tags - they weren't wire-ended, so might be a different animal.

I bought some from John Birkett's at Lincoln some years ago and am down to my last one. I built the Cap reformer that Stef Niewiadomski designed which featured in RB and later in The Bulletin some years ago, and they've all reformed very well. I missed those on the B&B at bargain basement prices in my brief sorties to the B& B from my own little table or I'd have snaffled some in very short order!

[Pamphonica]

These are quite small, with wire ends. about 1" by 2". Some evidence of damp storage. Although cans are bright, the paper label under the plastic is crinkled.

[David G4EBT]

Bargain!