Continuous, very small reverse voltage on electrolytic capacitor

[ccurtis]

I designed and built a circuit that works as desired, but I am concerned about a several nanovolt reverse voltage that is continuously developed across one electrolytic capacitor. The minus side of the cap is tied to the base of a NPN BJT. There is a 10K resistor also connected from the base of the BJT to ground. There is an unavoidable, finite leakage current from the collector of the BJT through to the base that develops the small reverse voltage at the minus side of the cap. The plus side of the cap is normally at zero (and I do mean zero) volts, except when the circuit is very occasionally (a dozen times a day) triggered with a positive voltage level.

Is this tiny reverse voltage any problem at all in terms of long term reliability of the cap?

 

[Tony Stewart]

If it was -5% of Vmax, I might be concerned.

 

[Nigel Goodwin]

You might try posting the actual circuit, as a description is a bit vague.

But a small reverse voltage is no issue anyway, and I wouldn't e en be measuring down in the nanovolt range - in fact, this is probably the first time I've ever seen 'nanovolt' used .

 

[danadak]

https://www.cde.com/resources/technical-papers/AEappGuide.pdf

Seems like your tiny Vreverse not issue, unless over extended time causing
oxide to build up, raining ESR, lowering C.....


Regards, Dana.

 

[ZipZapOuch]

The negative voltage on a cap has to be enough potential to cause electrochemically induced changes to the aluminum/aluminum oxide layers of the cap. A nano-volt or microVolt or even a millivolt is not enough.

 

[danadak]

 

[ccurtis]

Thanks, folks. Good stuff. I opted to lift the voltage on the + (input) side of the cap a few millivolts to counteract the reverse voltage, just to be sure.

 

[Tony Stewart]

You can test any caps with 10% Vreverse with a current limiting resistor and limit the leakage current to 100 uW in the cap and verify the data provided by danadak . e.g. -10V for a 100V cap with 1.1Meg to -10V measured with a 10Meg DMM. for voltage of 1 Meg R measured in mA scale. This cannot create any significant temperature rise to cause thermal runaway. The reverse voltage will have and exponential avalanche effect but not explode, but you can learn how degradation occurs by accelerating it with heat and aging effects for future as I have done for solid tantalums. Not all caps will behave the same.