I've got a big 560 uF, 400 V capacitor and I've tested it with a meter and the capacitance is within spec, same the ESR and it has a vloss of 0.8%. So all seems fine but then I came across this video geizer test'in. In his case there were obvious signs of capacitor damage so it was easy to see what was wrong even though the meter said it was fine.

He did the same and appeared to have a good capacitor capacitance and the ESR was good but when he connected up to a power supply as he cranked up to voltage the more it leaked.

So what tests should be performed?

And given that I do not have a power supply that will go up to 50 V how can I test the leakage at higher voltages or do some test in substitution?

  • 1
    \$\begingroup\$ Do you have some reason to suspect that the capacitor is bad? If not, and it reads OK on the meter it's likely fine. You could measure the leakage current at lower voltage to see if it's excessive, but testing a component thoroughly requires proper test equipment (made or bought) and procedures. \$\endgroup\$ – John D Apr 2 '19 at 15:46
  • \$\begingroup\$ You should check to make sure it actually makes a difference in your application if it's a low leakage and it's not consuming a lot of power then it probably doesn't matter. \$\endgroup\$ – Voltage Spike Apr 2 '19 at 15:49
  • 2
    \$\begingroup\$ Find a well-maintained Heathkit IT-11. \$\endgroup\$ – jonk Apr 2 '19 at 15:51
  • \$\begingroup\$ @John D there are three of them in all and I know one is bad because the current drawn on power up was enough to blow the connection wire. I could just change them all but then it would be nice to know which one is bad. \$\endgroup\$ – onepound Apr 2 '19 at 15:57
  • \$\begingroup\$ @jonk will it test other parameters than vloss, ESR and capacitance. I do have a good wayne kerr LCR in addition to a cheap chinese meter thats good for most things. \$\endgroup\$ – onepound Apr 2 '19 at 15:58

This circuit from Max Robinson shown below has worked well for me only instead of the transformer I used the primary of an old switch mode charger that provides a stable 340vdc and isolation but there is no reason the circuit below will not work just as well. Also I used my cheap dvm he suggests a parallel 10M but I am not sure why.

The leakage voltage is ILMeasured = V / R.

In my case the leakage was 30V after 10 minutes for a 560u capacitor with 220k series resistor. test bench

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.