Capacitor has different readings if stored upside down

[ImpulsiveJames]

I'm working on an amplifier and have removed all of the large 100v 4700uf caps for testing outside of the board. I tested all of them last week and they were all within 20%....

The super odd thing is.... This week, while they've been sitting upside down (legs up on my desk), they all measure 1pf or 2pf. I have a brand new one, so I know the DE-5000 is correct and on the right settings.

Has anyone seen anything like this before?

[Kim Christensen]

Those capacitors would have gone tits up whether they'd spent the night with their legs up or not. 

In seriousness though, it's odd that they tested OK the week before and now are completely open. I guess they've got to fail sometime.

[CaptDon]

Sounds more like the test leads went open circuit. Or the capacitors have a D.C. bias fooling the meter.

[Psi]

- Confirm you are making good contact to the cap terminals, oxide coatings can be thicker than you think.

- Check the meter wires don't have an intermittent break in them somewhere

[Harry_22]

I have a similar situation. When carrying out measurements on the board, the capacitance tends to zero. When you unsolder it everything is back to normal. Good capacitor on the board shows 9uF out of 10.
I take measurements by R&S LCR meter.
Apparently my capacitors recover after warming up.

Try warming up yours and measure again.

[helius]

If a capacitor "recovers after warming up", it suffers from high ESR and is in fact bad.

[factory]

And what type & brand of capacitors are these??? Might be useful to know for other people working on similar equipment.

The other comment, you need to wait for the parts to cool down after desoldering, for meaningful measurements, as warming them up will make the ESR go down and make you think they are good parts, when they are not.

David

[Doctorandus_P]

I once took out some big bulging electrolytics out of an audio amplifier. I think they were 22000uF 80V and they tested OK, which I found strange. Then I took out a knife an cut the top of. It turned out the top of the capacitors was perfectly flat, but there was an insulating plastic disc on the top of the capacitor (kept in place by the crimping foil) and that plastic disc distorted over time.

[amyk]

Could the stress of being upside-down caused the terminal connections to fail? Perhaps the inner roll was heavy enough to put a large enough tension on the thin ribbons connecting it to the terminals to break them?

[m k]

I once took out some big bulging electrolytics out of an audio amplifier. I think they were 22000uF 80V and they tested OK, which I found strange. Then I took out a knife an cut the top of. It turned out the top of the capacitors was perfectly flat, but there was an insulating plastic disc on the top of the capacitor (kept in place by the crimping foil) and that plastic disc distorted over time.

I'm quite sure some models have bulky plastic top from the beginning.
Can't say how obvious or not that is.

[TERRA Operative]

Turn them back over so they are sitting upright for a while and test again to see what they say.

[iMo]

An electrolytical capacitor is a jar with electrolyte inside, the electrolyte is a liquid. Between the electrodes there is a material soaked with that liquid. The concentration of the electrolyte may not be uniform between the electrodes (the highest on the "bottom" because of gravity). So putting the jar upside down may cause the electrolyte to slowly move in the direction of gravity and thus changing the capacitance..

[Psi]

Also, in one direction gravity is trying to squash the plates closer together, but 90deg from that it is not.

However the affect of this I would expected to be very very small. Probably quite measurable but not with a typical accuracy DMM.