Checking ESR in large electrolythic capacitors.

[Skogsgurra]

I am testing a simple method to check the ESR of capacitors in VFDs. I will be using a low voltage high current sine generator (a few hundred millivolts and 20 A) which can be set at any frequency from around 25 Hz to 10 kHz.

My thinking is that, by testing at low frequency, I can find the capacitance and at a considerably higher frequency, the ESR. The data (voltage, current, frequency) will be entered into a simple Excel sheet and there, C and ESR will be calculated.

The measurement will be carried out at zero voltage across the capacitor (only a few hundred millivolts measurement voltage will be present). I have not been able to find at what voltage the Al2O3 starts degrading and I certainly hope that it is above 500 mV or thereabouts, preferably higher. At lest if the measurement only takes a few seconds.

I hace tested the technique on smaller drives, up to 4 kW and it seems to work just fine. It responds to artificial ESR added to the circuit and it also seems to produce reliable uF numbers. At least, they correspond to the data that I get from a discharge measurement.

My question is: Can one safely run 200 - 300 mV AC voltage across an electrolythic capacitor? Any experience? Any references?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

I found a reference:

http://www.rubycon.co.jp/en/products/alumi/faq.html

It mentions that a reverse DC voltage more than 1,0 or 1,5 V may reduce the oxide layer over time.

So, I think that it should be OK to apply a few hundred millivolts for five or ten seconds. I have tested prolonged exposure to both Mallory 480 uF and RIFA 22 000 uF and there's no difference in DC leakage before and after 500 mV AC exposure for a couple of hours. I will definitely test this IRL.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

Any thoughts on this? I get plausible data on C and ESR. It seems to easy to be true - there must be a catch somewhere?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[itsmoked]

I think you're right-on with it all. The ~1V seems correct to me too. EVerything I've ever seen said 1V or less to 'not cause damage'. I've also seen use 4V thru 100k ohms measured both ways thru a cap to find the positive terminal. The lowest current is the 'correct polarity'.

My favorite cap handbook is: [URL unfurl="true"]http://www.cde.com/catalogs/AEappGUIDE.pdf[/url]
but it doesn't talk about abusing the polarity.

I suspect AC is less damaging than DC in regards to oxide damage, remembering a recent discussion we had on electrolysis.

Neat idea. You should make a little box that does that. Add a "former" feature too.

Keith Cress
kcress -

http://www.flaminsystems.com