Bad Meter or Bad cap?

[jimkarl]

Had a 4700uf 25V power filter cap with a ballooned head. Was sure it was bad. Took it out and did a cap measurement reading 5200ish, compared against 2 others in the circuit and had similar readings. Then learned a cap may measure good that way but really needed an ESR meter to tell for sure. So ordered a MESR-100 V2 off of EB*Y. While waiting for that to arrive, I decided to check what I saw with an ohmeter. The other 2 - assume good caps - read inf in one direction and about 40 ohms in the other. This one, the suspected bad one had a reading of near 0 and rising with time in one direction and 20 in the other. Given the differences in ohmage and the swollen top, I was sure this had to be bad.

My ESR meter came today and surprisingly didn't show any problem with the ESR and shows it has a low ESR. Checked against the other two witht he same result. I have no other "Bad" cap to test (had some weeks ago but threw them away DOAH!) but did use the meter on a variety of brand new caps and all seems fine and measurement show good caps with low ESR readings.

So why would the cap have a swollen top, weird (or at least different) ohmage readings, yet still have good esr? Is my brand new meter not functioning correctly, or has this cap failed in a way that doesn't affect ESR, or has it not actually failed at all?

[mariushm]

Re: Bad Meter or Bad cap?

You can't measure capacitors in circuit and expect to receive accurate results. Capacitors connected in parallel on the motherboard will behave like one single capacitor with capacitance equal to sum of the capacitances and the esr will also drop with the number of capacitors
You would have to desolder at least one leg to make sure the capacitor will be measured correctly.

The esr meter may say it can measure in circuit, but there's a distinction ... it CAN measure, doesn't mean it does it ACCURATELY. It more or less means the meter won't damage other components in the circuit, if you try to measure a particular capacitor (because the test voltage is very low).

You can't measure ESR with a multimeter on the resistance range (or an ohmmeter). I don't know what value the multimeter reports but it's useless to you.

The esr meter works at high frequencies when testing capacitors (up to 100kHz), a multimeter uses low frequencies (60-120 Hz, something like that).
The esr meter may not be able to measure capacitance going so high.

A bad capacitor can report a higher capacitance. 5200 uF is way too much for a 4700uF capacitor, that alone tells you it's bad. The swollen head is another reason to replace the capacitor without even bothering to check.

Test the mesr esr meter with known good capacitors .. get some capacitors from Digikey Mouser or Newark or whowever you like, check the datasheet and compare the results of your ESR meter with the column that says Impedance @100kHz. The ESR meter should report a very close value to that column.
Get a few capacitors, for example 100uF, 470uF,1000-1500uF, 3300uF.. this way you have the whole ranges checked.

[ramr]

Re: Bad Meter or Bad cap?

Puffed up caps are caused by poor electrolyte which deteriorate with age.

[mariushm]

Re: Bad Meter or Bad cap?

No shit Sherlock... thank you for posting such an elaborate and detailed and to the point answer, I'm sure it will help him a lot.

[stj]

Re: Bad Meter or Bad cap?

Lol

[jimkarl]

Re: Bad Meter or Bad cap?

You can't measure capacitors in circuit and expect to receive accurate results. Capacitors connected in parallel on the motherboard will behave like one single capacitor with capacitance equal to sum of the capacitances and the esr will also drop with the number of capacitors
You would have to desolder at least one leg to make sure the capacitor will be measured correctly.

The esr meter may say it can measure in circuit, but there's a distinction ... it CAN measure, doesn't mean it does it ACCURATELY. It more or less means the meter won't damage other components in the circuit, if you try to measure a particular capacitor (because the test voltage is very low).

You can't measure ESR with a multimeter on the resistance range (or an ohmmeter). I don't know what value the multimeter reports but it's useless to you.

The esr meter works at high frequencies when testing capacitors (up to 100kHz), a multimeter uses low frequencies (60-120 Hz, something like that).
The esr meter may not be able to measure capacitance going so high.

A bad capacitor can report a higher capacitance. 5200 uF is way too much for a 4700uF capacitor, that alone tells you it's bad. The swollen head is another reason to replace the capacitor without even bothering to check.

Test the mesr esr meter with known good capacitors .. get some capacitors from Digikey Mouser or Newark or whowever you like, check the datasheet and compare the results of your ESR meter with the column that says Impedance @100kHz. The ESR meter should report a very close value to that column.
Get a few capacitors, for example 100uF, 470uF,1000-1500uF, 3300uF.. this way you have the whole ranges checked.

"You can't measure capacitors in circuit and expect to receive accurate results." - I don't disagree - and I didn't. I said at the binging of the post "I took it out". All testing I did, with all 3 caps, was out of circuit.

"I don't know what value the multimeter reports but it's useless to you." - I tend to disagree with you here. I never said an ohmeter can tell me if I have a good cap or not. What I said was I did a comparitive test and 2 of the 3 caps, not swollen and assumably good, returned the same results and the swollen cap, expected to be bad, returned completely different results. It seems probable, with such different results on the third cap that it is clearly in some different state that the other two are. They all be "bad" and the third is just "more bad", but they are different. Or the other two may be good and only the swollen one is bad. The ohmmeter doesn't answer that conclusively, just that the third one is very different.

"5200 uF is way too much for a 4700uF capacitor, that alone tells you it's bad." I don't know what the tolerance is on these caps but if it were a 20% tolerance, it would be well within the limits.

"Test the mesr esr meter with known good capacitors" - As I said, I had also done that. With several brand new caps at different values (all except the 4700 because my replacement is still on order). For all the caps I checked, they were all well with the recommend esr range. The recommended esr range for this 25V 4700uf cap is .12 at the high end. These caps measure .019, 029, and .032 with the latter being the swollen one - but all a order of magnitude under the recommended range, and they manufacturer and various sites say that a bad esr with be several times larger than that high recommended range.

Please don't get me wrong - my gut has been telling from the beginning that this cap is bad. It looks bad visually, it reacts RADICALLY different on a ohmeter then the other two (although you state that is a worthless test, respectfully I disagree), but the cincher to me would have been getting a bad ESR measurement - and I'm not and cannot explain it. This brand new ESR meter could be bad, but then why does it report good on some many other know good caps? And if it can't detect what would seem to be an obvious bad cap it's certainly going back.

[jimkarl]

Re: Bad Meter or Bad cap?

Followup to this. I did a number of ohmeter related tests I found on the internet - One is using the resistence meter to charge the cap. It works on this one and the other two as well. Smaller value caps see that charge much faster than these larger caps, but there is a charge building up in them/
Did a 9 volt battery charge test - hooked it to the battery for a few seconds and then measured the voltage across the caps after disconnect - had the full battery value charge held.

Then the kicker - My replace 4700 caps came today. Same brand but slightly higher voltage rating (35 vs 25). On a cap meter they also measure 5200uf, confirming my belief the capacitance value is well within spec. They also measure extremely similar on my new ESR meter, with the slight difference easily explainable by the slight higher voltage rating.


Summary - the cap may well be bad, but if it is, beyond the visual it has very odd characteristics for a bad cap, because in most every respect I've tested, it acts identical to it's old siblings and it's new cousins. It does not measure bad capacitance, it does not measure bad esr, and holds a charge although it would be difficult for me to say if it is holding as long as or as well as the others, capable of reading it's full volt potential, or charging as fast as it should. I would expect that for some aspect of those problems and Bad esr would be obvious - but here it is not. Based on the visual alone, and the fact I now have a replacement for it anyway, I will replace it, but this was a bit of a mystery with not a clear cut common test results that would normally back up a "bad looking" cap.

[SteveNielsen]

Re: Bad Meter or Bad cap?

I've had bulged caps that tested ok too. Of course I replaced them anyway, but it is curious and worth noting anyway I think.

[mariushm]

Re: Bad Meter or Bad cap?

Then the kicker - My replace 4700 caps came today. Same brand but slightly higher voltage rating (35 vs 25). On a cap meter they also measure 5200uf, confirming my belief the capacitance value is well within spec. They also measure extremely similar on my new ESR meter, with the slight difference easily explainable by the slight higher voltage rating.

Or your multimeter is faulty or not capable of measuring capacitance correctly above a particular value.
Check your multimeter manual, some multimeters can only measure as much as their counts (ex 4000 count -> 3999uF maximum, anything above is not guaranteed)

5200 is way too far from the 4700uF - a good capacitor will be +/- 100uF or something like that. I have tens of 3300uF or 2200uF here for example, they're all within 5% of the actual value.

[SteveNielsen]

Re: Bad Meter or Bad cap?

Standard tolerance for caps is 20%. 5200uF is within that tolerance for a 4700uF cap.

[momaka]

Re: Bad Meter or Bad cap?

Did a 9 volt battery charge test - hooked it to the battery for a few seconds and then measured the voltage across the caps after disconnect - had the full battery value charge held.

That test is not very conclusive. I've had caps with bad ESR and good capacity hold a charge okay as well.
However, the above test can be made fairly-useful. After charging the cap, take a piece of wire or metal (preferably something you don't care much about) and slowly and carefully short the terminals of the charged cap. You should either hear a pop or see a spark (or both). May have to try this a few times, depending on how you short the leads. Of course, this is done with the capacitor removed from the board.
You have a pretty high capacity cap on your hands, so you should see some good sparks. I can hear pops / see sparks even on caps as small as 470 uF.

If shorting the cap leads is silent and no sparks, the ESR is likely high.
And yes, most electrolytic caps have 20% tolerance, so the high capacity you got is reasonable.

I've had bulged caps that tested ok too. Of course I replaced them anyway, but it is curious and worth noting anyway I think.

Same here. Had a pair of JEE 200V input caps from a PSU. One is truly bulged (not just its plastic cap). Yet both read good ESR and capacitance. Passed the spark test too. I even tried charging it up to 20V and short its leads - now that was loud!

I've also had bulged caps from overheated Xbox 360's. After pushing their tops in, I decided to re-use them for fun in some non-critical junk computer hardware. All are still working and none of them have bulged again.