permissible leakage current values for capacitors?

[maxhifi]

I am going to build a test jig for measuring capacitor leakage current. Basically just a variable power supply which supplies up to 600V. It will supply voltage to the capacitor under test, and the capacitor will be in series with the 50uA scale of my Simpson 260. The 260 will be bypassed by an spst switch, which will be opened once the capacitor is charged then allowing the leakage current value to be read.

What I can't find are pass/fail values of leakage current for coupling and filter capacitors. I want to be able to test old bathtub and paper in oil capacitors, and also to be able to test NOS paper in oil Russian capacitors before use.

Something like, "if leakage is less than 10uA at 600V, the capacitor is good" is what I am looking for. Higher current scales will be used for electrolytic capacitors.

[Wildon]

Suggest you add a series resistor in inline with your 50 u.a. Meter for meter protection.(not to exceed full scale current)of meter movement. I built such a circuit to check capacitor leakage,to use with my digital capacitance meter .

[maxhifi]

Originally Posted by Wildon

Suggest you add a series resistor in inline with your 50 u.a. Meter for meter protection.

Excellent idea. I will look at the manual and see what the maximum allowable meter current is, would be a shame to ruin the simpson because of a shorted cap.

[old_coot88]

I would start the test with the 260 at a higher ma setting, then back it down in steps to the 50uA while the test is underway.

[maxhifi]

Originally Posted by old_coot88

I would start the test with the 260 at a higher ma setting, then back it down in steps to the 50uA while the test is underway.

That's the plan!

The power supply will just be a line voltage operated voltage quadrupler, I can plug it into my isolation transformer and plug that into my variac. Then use the variac to adjust voltage. This will only cost me a few 10uf capacitors and some 1N4007s.

[pearsonk]

I suggest you look at: www.electrojumble.org/reforming.htm
I have used a variation of the circuit found at: http://www.vmarsmanuals.co.uk/newsle...s/big_bang.pdf, or search "capacitor reforming-or how to avoid the big bang". I incorporated a 0-1ma meter, activated through a momentary spdt switch when neon lamp goes out ----shows leakage less than 1ma. Combined, these sites should give you some good info.

[Gregb]

Hey Max:
You want to just borrow my Sprague Tel-ohmike? It works great and I rarely ever use it as all old caps are bad. This machine will show you that my last statement is true.

Gregb

[Electronic M]

Originally Posted by Gregb

This machine will show you that my last statement is true.
Gregb

That statement only goes for pre-1960 equipment.

[Gregb]

I will clarify, all old paperwax capacitors.

Gregb

[maxhifi]

Originally Posted by Gregb

Hey Max:
You want to just borrow my Sprague Tel-ohmike? It works great and I rarely ever use it as all old caps are bad. This machine will show you that my last statement is true.

Gregb

That would be awesome - what I am doing is recapping my big RCA Theatre amplifiers, and I want to see if the original paper in oil filter caps are still good (they do work well and don't get warm), I also want to test a bunch of old stock USSR made paper in oil caps I just ordered on eBay before I use them. They're 25 years old, not 65, hermetically sealed, and they look fantastic, like old spragues or pyramid oil caps, but I just don't trust them without some pre-use testing.

[Electronic M]

If those oil capacitors are in places and of values typical of paper coupling/bypass caps, then their leakage specs probably should be comparable to a new paper cap, and if they are being used like lytics they should be as good or better than good lytics.

[maxhifi]

Originally Posted by Electronic M

If those oil capacitors are in places and of values typical of paper coupling/bypass caps, then their leakage specs probably should be comparable to a new paper cap, and if they are being used like lytics they should be as good or better than good lytics.

That's my thinking too. With the Russian ones I am getting ten, so I can also compare how they test to each other as well as to an arbitrary standard. The big old filters I am not that worried about.

[Tubejunke]

OR in opposition to a Sprague Tel Ohmike in which the better models (TO-5 and TO-6) are getting pretty pricey for vintage test equipment; you may opt for a Heathkit C3. To a large degree these old units have been all but forgotten about, but are essential for anyone doing what most of us here do in relation to testing capacitors. A Heathkit C3 can be obtained for less than $50. Of course it's the luck of the draw to obtain one that doesn't need a recap its self.

The great thing about the Heathkits is that they are super simple and you can replace all of the caps for beans and have a great piece of test equipment that they just don't make anymore. Most all of the old units no matter the manufacturer utilize a "magic eye" tube as the primary indicator of both value and leakage. The Spragues' use the "magic eye" along with panel meters which are a big plus.

Heathkit evolved to the final model IT-28 (made well into the 70s) which offers test voltages as low as 3 VDC. The older units only went to (I think) 25 VDC. Some of the more dated Heathkit units used a neon lamp (C-1, C-2) and the "magic eye. I like those the best or finally old Solar units can be obtained and have a really neat vintage look with exposed tubes and work well, but are way more complicated to restore.

Just ideas I think well worth looking into. At least you wont have to design and build your own.