ESR meter reading implications

[bigbeark]

In the past, when measuring ESR on certain caps on certain boards I have seen the ESR reading trickle down to zero.

Never thought much about it, but have been reading a bit on electronic theory lately (OK so it's Ohms Law). If there is no resistance, there is no current(?). So I'm thinking these zero readings indicate a problem. The capacitor is offering no resistance, so there is no flow.

I have noticed that when I recap a board, the new capacitors will NEVER show zero ESR, lowest maybe .02.

Also, in the past, I've been able to isolate bad caps with ESR readings higher than spec. I have observed that if the cap is 1500uf or more, that is it is not bulged or domed, the ESR reading is likely to be OK (or Zero).

It seems to me that the 1000uf 8mm dia caps cause a lot more problems than the bigger ones. Their ESR can measure OK, but when I take them off
the board and measure their capacitance, they are lower capacitance thn their label. Replacing these which are common around the PCI slots seems to restore many boards to life.

So is the zero ESR reading a red-flag or a red-herring? Is the cap in fact "shorted"?

I think there's more to using an ESR meter than just reading the numbers.
Anyone care to offer some insights on using the meter?

[Radio Fox]

Re: ESR meter reading implications

In the past, when measuring ESR on certain caps on certain boards I have seen the ESR reading trickle down to zero.

Never thought much about it, but have been reading a bit on electronic theory lately (OK so it's Ohms Law). If there is no resistance, there is no current(?). So I'm thinking these zero readings indicate a problem. The capacitor is offering no resistance, so there is no flow.

I have noticed that when I recap a board, the new capacitors will NEVER show zero ESR, lowest maybe .02.

Also, in the past, I've been able to isolate bad caps with ESR readings higher than spec. I have observed that if the cap is 1500uf or more, that is it is not bulged or domed, the ESR reading is likely to be OK (or Zero).

It seems to me that the 1000uf 8mm dia caps cause a lot more problems than the bigger ones. Their ESR can measure OK, but when I take them off
the board and measure their capacitance, they are lower capacitance thn their label. Replacing these which are common around the PCI slots seems to restore many boards to life.

So is the zero ESR reading a red-flag or a red-herring? Is the cap in fact "shorted"?

I think there's more to using an ESR meter than just reading the numbers.
Anyone care to offer some insights on using the meter?

First, I think you have got your Ohm's Law muddled up. I=V/R, so the lower the resistance, the higher the current, not lower.

Also from your post, you appear to be measuring some caps in circuit & others out of circuit. When you say that the reading trickles down to zero, is that only when measuring caps while still on the board? The only true ESR reading is when the cap is out of circuit. Have you ever measured zero on a capacitor out of circuit?

Capacitors can change their value over time, but remember that a lot of electrolytics have quite a large tolerance, so make sure that a cap than can appear to be "wrong vale" is not just on the limit of it's tolerance.

They can also go short circuit. This can be double checked using a regular resistance meter found on your multimeter.

[bigbeark]

Re: ESR meter reading implications

First, I think you have got your Ohm's Law muddled up. I=V/R, so the lower the resistance, the higher the current, not lower.

Yes but if R is zero does current still flow?

Also from your post, you appear to be measuring some caps in circuit & others out of circuit. When you say that the reading trickles down to zero, is that only when measuring caps while still on the board? The only true ESR reading is when the cap is out of circuit. Have you ever measured zero on a capacitor out of circuit?

The zero rating is observed in-circuit on the old caps. When I replace all the caps with new ones and measure the replacement caps in-circuit I never see zero. I think the zero is indicating a tired group of caps. True, when I measure those caps out-of-circuit, I don't get zero.

Capacitors can change their value over time, but remember that a lot of electrolytics have quite a large tolerance, so make sure that a cap than can appear to be "wrong vale" is not just on the limit of it's tolerance.

Yes, that's my point. When I remove the old caps from the circuit, they meter OK in lots of cases, but collectively, on-board, they don't work. When I replace the whole lot, the board comes alive. I'm saying that the ESR meter is useful for finding fried caps, but has trouble zeroing in on an entire circuit of bagged-out caps.

They can also go short circuit. This can be double checked using a regular resistance meter found on your multimeter.

I will try that! I've been a little leery of using DMM in circuit for fear I burn something out with excess voltage.

[Radio Fox]

Re: ESR meter reading implications

Yes but if R is zero does current still flow?

Yes it does...lots of it! In theory, when R=0, I = infinity! In practice it is limited by how much current the circuit can supply. However, this just confuses things when talking about ESR as it's an imaginary (equivalent) resistance, in series with the capacitor, not a real one.

The zero rating is observed in-circuit on the old caps. When I replace all the caps with new ones and measure the replacement caps in-circuit I never see zero. I think the zero is indicating a tired group of caps. True, when I measure those caps out-of-circuit, I don't get zero.

Hmmm...I can't explain that. I was going to say that perhaps 1 out of a bank of caps had gone short circuit, which would explain the zero reading while in-circuit, but if they measured OK when taken out, then none of them could be short. And if the replacement caps worked OK & didn't show zero Ohms while in-circuit, then there isn't a short on the board. Strange.

Yes, that's my point. When I remove the old caps from the circuit, they meter OK in lots of cases, but collectively, on-board, they don't work. When I replace the whole lot, the board comes alive. I'm saying that the ESR meter is useful for finding fried caps, but has trouble zeroing in on an entire circuit of bagged-out caps.

That's the problem when using an ESR meter on caps that are connected in parallel on a circuit board. For example you could have a bank of 4 caps. 3 of them could have really low ESR, but the 4th might have sky high ESR. Because they are connected in parallel, the 3 good ones will mask the bad one & unless the bad one is showing some physical signs, then there is no way to know which one it is without taking them off the board.

In my experience a cap that has changed it's value past it's tolerance usually shows a high ESR reading as well, although I don't know if that is always the case.

I will try that! I've been a little leery of using DMM in circuit for fear I burn something out with excess voltage.

I would remove any suspect caps first before testing them, although a DMM set to it's lowest Ohms range shouldn't do any harm to any other components on the circuit board (make sure the circuit board isn't powered of course, before doing any resistance measurements).

[tjohnson4]

Re: ESR meter reading implications

Actually, if R=0 then I is undefined. I wish I had a nickel for every time I've been corrected on that one. However, as R->0, I->infinity... if V is positive.

I usually prefer I*R=V. Then I don't get anyone telling me that I divided by zero.

[PCBONEZ]

Re: ESR meter reading implications

If R were ever actually =0 then V would also =0 because it would be the same point electrically.
If V and R both =0 then so does I.

How ya like them apples?
.