Capacitor leakage

[Boom]

What is reasonable when measuring leakage on (wax) capacitors? I don't realistically expect insulation to be up to modern capacitor standards.

Some that I have tested give a leakage of 20Meg Ohms on a digital meter. Is this servicable? What would they have measured when new?

Thanks

[russell_w_b]

Originally Posted by David Tilley

What is reasonable when measuring leakage on (wax) capacitors? I don't realistically expect insulation to be up to modern capacitor standards.

Some that I have tested give a leakage of 20Meg Ohms on a digital meter. Is this servicable? What would they have measured when new?

Thanks

I would be inclined to pass DC through them at - or near - their rated voltage with an AVO in seies to measure the current. Use a series lamp limiter in case anything catastrophic happens.

Or use a 'Megger'. If they leak in the slightest, they are not functioning correctly and I would chuck them, as the leakage will only get worse over time.

[paulsherwin]

I agree with Russell. You can't tell much about leaky wax caps by measuring their DC resistance with a DMM, as this often varies with voltage and temperature. Even a very small leakage will cause problems in some circuit positions, and as Russell says, it will only get worse.

They would have had very low leakage when new.

Rip 'em all out I say

Paul

[russell_w_b]

Originally Posted by David Tilley

What is reasonable when measuring leakage on (wax) capacitors? I don't realistically expect insulation to be up to modern capacitor standards.

Some that I have tested give a leakage of 20Meg Ohms on a digital meter. Is this servicable? What would they have measured when new?

Thanks

Just a small point, and do please say if I'm teaching my granny to suck eggs, but you're not gripping the exposed metal part of the test-leads and capacitor wires between your fingers when measuring, are you?

[Boom]

lol - No Russell. My fingers only measure 100k. Thanks for comments

[johnn]

This is a good thread!
When I started restoring sets, I simply measured the resistance of capacitors after changing the critical ones first. Then I realised this may not be correct.
I simply rejected all those which did not show open circuit. What current should I go for if I apply say 200v? Is it necessary to apply a voltage any higher than would be present in the radio concerned?
John.

[Station X]

Test waxies if you must, but in my opinion they should all be replaced. If they aren't faulty now they will be later. You soon realise that when you've removed the chassis from the cabinet for the fifth or sixth time.

Any capacitor which fails a low voltage ohmmeter test is absolutely useless. I generally test at the working voltage. Most will be either so good or more likely so bad that you'll soon see the difference.

Many capacitors can be tested in circuit. A faulty screen grid decoupler will pull down the screen voltage and a faulty coupling capactor will put a positive voltage on the grid of the next stage.

[Kat Manton]

Hi,

I've just had to deal with a couple of leaky capacitors. After removing them, just out of curiosity, I checked them on the 20M range of my DMM.

They measured as having no leakage whatsoever.

I think that counts as a practical demonstration of how utterly useless low-voltage DMM checks for capacitor leakage are

Cheers, Kat

[Kat Manton]

Hi,

I've just remembered another one. In this case a resistor was burning up when the set was powered. Resistance checks around the ciruitry concerned didn't show any possible cause, though.

What I did to find that fault was to temporarily substitute a higher value resistor and take voltage readings with the set powered. This showed that there was almost a dead short somewhere. It turned out that a small tubular ceramic capacitor hidden inside a screening can was failing, but only when there was a high voltage across it.

I'm with Graham on this one, it's usually possible to test capacitors in-circuit under operating conditions. Removing the odd valve or temporarily inserting resistors can help track down faulty capacitors while preventing damage elsewhere.

Cheers, Kat

[Sideband]

Originally Posted by Station X

Test waxies if you must, but in my opinion they should all be replaced. If they aren't faulty now they will be later. You soon realise that when you've removed the chassis from the cabinet for the fifth or sixth time.

Likewise I agree with this statement. There are some people who prefer to replace only those that are obviously faulty and then fault-find the others when they fail. OK if you like keep pulling sets to bits and risking further damage when doing so and OK it may help build up your fault-finding skills. However the risk to other parts (output transformer on DAC90A for example) is quite considerable and it's not just a question of popping down to the local components store and getting a replacement over the counter. Best bet is to ditch the waxies (or moldseals for that matter), use good quality polyester or similar and fix the set once-and-for-all barring the odd valve failure. The set will be the more reliable for it.


Rich.

[Tim]

I use a battery powered Megger tester for checking for leaky capacitors. It tests them at 250 volts DC.
Then of course there is the old trick of disconnecting the "earthy" end of the capacitor under suspicion and measuring the voltage between the now free end and chassis. This should of course be zero.

[Brased]

The PO Tester SA9083 used to be available very cheaply and perhaps still is. This tests insulation at about 140V. I used one of these before getting my Megger which will test at various voltages up to 1000V. Although the SA9083's scale is somewhat cramped it is far more effective than a normal multimeter ohms range.

Edward

[kalee20]

I agree with Richard Newman and Station X. Whether waxies show leakage or not, it's best to replace them - for all the reasons given.

Having said that, you may get away with not replacing wax-covered mica capacitors, especially if they are weird custom-valued types in RF or oscillator circuits.

[Sideband]

Yes going back to my post, I should have said all the waxy paper capacitors. Most micas' seem very reliable and rarely need attention.


Rich.

[johnn]

Many thanks, appreciate the discussion. I shall go back to changing all wax types, but will do the voltage test on those removed from a friends Bush AC41.
John.