Testing Capacitors For Leakage

You can build a simple leak checker for caps.
You don't have to buy an expensive cap machine.

You need:

1) black box
2) wire
3) 2 banana jacks
4) momentary switch
5) 9 volt battery
6) batter clip
7) digital voltmeter
8) 2 alligator clips

here is the diagram:



here are some pics of the ten minute box i threw together:





Directions:

Wire the circuit
Short the alligators
Measure the voltage (you are reading battery volts)
Hook up a cap to test
Press the button to charge the cap
Release the button, let the voltage stabilize, and read the hopefully low voltage, in this case, I had 0.140 volts dc.
Now divide the low number by the batt volts to get leakage.
Don't forget to multiply by 100 to put it in percent form.

So, in my test-

0.140 volts divided by 9.45 volts (new 9 volt batt reads high)
so 0.14/9.45 = about 1.5 percent leakage for a lytic.
Try it on a spraque orange drop and you should see zero leakage.
Anything over 1 percent....is bad if a coupling cap.

Good to get a DIY project started and finished in the same day
.onward...

I've always wanted something like this. Cool!

I used to do it like this:



Any drawback?

Then I would just compare my old lytics to new ones that are not leaky for sure. Did you know that many oil caps gets pretty leaky over time? Depending on the construction, I mean, the cheap oils, not the bigger ones in metal cases... I´ve bought a bunch to use in tube circuits as couplig caps to grids (here´s where you want very low leakage because grid impedance is so high), and it was money badly spent. All were pretty leaky.

Thanks CJ :thumb:

Does this work for HIGH VOLTAGE Caps too?
... like 250v +

Is the 9v enough to charge a higher voltage cap?

awesome! must build one.
thanks CJ!! :thumb:
-grant

For the record, your test rig is not measuring leakage as a percentage of the applied voltage but the leakage current working into the input impedance of whatever volt meter you happen to be using. If you look up the input impedance of your voltmeter you can compute the leakage current using ohms law,

Leakage is typically specified relative to applied voltage and amount of capacitance. A useful tester would need to be able to apply representative working voltages. 9V may be useful for simple line level blocking, but would be inadequate to test phantom mic pre blocking caps that are 50v or higher.

Perhaps of interest, I have encountered leakage issues in mic pre front ends where the leakage current made audible noise. So not only is the amount of leakage an issue but the spectral content for low noise applications (note: I expect this is not a common problem).

Finally, be aware that some older caps sitting unused may need to be formed in and will express higher leakage until the dielectric re-forms properly.

JR

not meant to be a HP lab instrument, just something you can throw together from your junk drawer.

yes, checking leakage on a 500 volt cap with a 9 volt batt, well...

so for a better test on high voltage stuff, you can use tricks like leaving the 0.01 cap on the git amp, disconnect the low side, connect a meter from the cap to the chassis, and fire her up.

But this checker is cool for an eductaion tool, if somebody does not have a Sencore sitting on there card table.


You can try tants, atoms, orange drops, oil, paper and wax, to see how they compare.

Be careful on reverse polaritizing a 5 volt lytic, don't know if a 9 volt will pop it, but i would not try to find out.

[quote author="CJ"]... but i would not try to find out.[/quote]
naww... I know you'd never melt a cap. Or destroy or take apart anything...

Is this CJ I'm talking to?

I'm confused?

Help, where's my hacksaw.

the saw is over here.
i need it for tonite

31....no, make that 32...transformers will suffer unspeakable crimes of the hideous kind.

say i'm a punk or i'll cut ya...
.
.
what?

[quote author="rafafredd"]
Any drawback?
[/quote]

Yes, if you have a shorted capacitor, you may blow the fuse in the multimeter. Better add a 1K resistor in series...

Jakob E.

Yeah, Jakob. Thanks for that one. I´ll do it just like that for precaution... But generally, when I test for leakage, I´ve tested for capacitance already on the DMM, so I know it´s not shorted.

I'm fortunate enough to have a very well equiped bench and although I do have a Sencore LC53, I wind up using a neat little box made by Creative Electronics. It's an ESR meter and uses a 100 KHz signal to measure the resistance of a cap. It also does a good job at showing ones that are leaky, and at that point I can switch to the Sencore and investigate further. A very handy little device that can't have more than 10 bucks worth of parts in it. I've seen other ESR meter schematics on the internet, in case someone wants to built their own.
I generally only use the Sencore for matching caps and winding coils.