ESR for 1000v 10nF Poly Capacitor

[NZST205]

I am trying to troubleshoot a Sony SMPS and have come across a Poly Cap 1000v 103J with an east of >40. The chart I have doesn't show voltage this high. Th cap looks like it's been hot at some stage. Is this ESR too high ?

[dacman]

I think that it is unusual to test the ESR of a cap under 1 uF (or of a non-electrolytic).  What frequency are you testing it at?  At 0.01 uF, your ESR meter would need to separate Rs from Xc to be accurate.  Xc alone at 100 kHz is 159 Ohm.

Edited to be technically correct.

[NZST205]

Thanks dacman, using Peak Meter, but looks like ESR is irrelevant based on your response.

[wraper]

Measuring ESR is basically useless for non electrolytic capacitors. Usually ESR of electrolytic capacitors is specified at 100 kHz and many meters use this frequency or even lower or some alternative techniques. And most ESR meters actually don't measure ESR itself but just total impedance, assuming that most of this impedance consists of ESR. The issue is that at such frequency capacitive reactance and therefore total impedance of the 10nF capacitor is much higher than actual ESR. Therefore the figure what you see on the display can be order(s) of magnitude higher than actual ESR. For particularly 10nF (103) capacitance, capacitive reactance at 100kHz without ESR is 160 Ohms already. To measure actual ESR, you would need proper LCR meter, not ESR meter. For film and ceramic capacitors it is much more useful to check actual capacitance and if they are not leaky. For electrolytic capacitors usually there is no big need to check capacitance, checking ESR is enough.

[TimFox]

I think that it is unusual to test the ESR of a cap under 1 uF (or of a non-electrolytic).  What frequency are you testing it at?  If it were a perfect cap, the ESR would be 159 Ohm at 100 kHz.

159 ohms is the reactance, not the equivalent series resistance, for that case.

[dacman]

I think that it is unusual to test the ESR of a cap under 1 uF (or of a non-electrolytic).  What frequency are you testing it at?  If it were a perfect cap, the ESR would be 159 Ohm at 100 kHz.

159 ohms is the reactance, not the equivalent series resistance, for that case.

I admit that is badly worded and is technically incorrect, but an "ESR meter" that doesn't separate the vectors is in reality not returning ESR, but is instead returning Z (magnitude of the impedance). If an "ESR meter" has a floor of 1 uF and works at 100 kHz, that is a clue (in my opinion) that it is not a true ESR meter, and is most likely returning Z.

[Addicted2AnalogTek]

What about comparing it to a similar value new poly cap?  It's obviously still not going to give you the actual ESR, but it will give you the reading of a good cap to compare against.

[TimFox]

Typically, 10 nF capacitors are judged according to Q or D = 1/Q.  A good polypropylene capacitor should have Q > 1000 at 1 kHz.  I have found old paper capacitors that absorbed moisture and the Q fell to < 20, so I replaced them.  I have used old-fashioned impedance bridges as well as a DE5000 LCR meter.