I recently bought one of these ESR meters, a 'MESR-100':


ESR Meter Image

It claims to be able to test capacitor ESR with the caps left in-circuit. It appears to work well from my tests so far. However, how much can I trust the in-circuit readings? Are they going to be less accurate? Are there certain circuit configurations where it wouldn't be possible to measure ESR in-circuit accurately?

[EDIT: Added the model as requested] :

  • \$\begingroup\$ To the downvoter: I'd honestly appreciate some feedback as to why this is a bad question! \$\endgroup\$ – UpTheCreek Mar 16 '14 at 17:24
  • 3
    \$\begingroup\$ I think this is a decent question. It might help if you included the actual model and a link to the datasheet/manual of the particular ESR meter you're using (rather than a picture). Also, three close votes for the question being about the use of electronics? For questions about test equipment that would only be used in an electronics design context, I think the close votes are uncalled for. \$\endgroup\$ – JYelton Mar 16 '14 at 17:55
  • \$\begingroup\$ Here pa4tim.nl/?p=3775 is a nice article about ESR. There is a nice cirquit as well. In any case do not confuse real ESR meter with a square wave voltage drop detector...this is caused by impedance. ESR is quite complex issue. \$\endgroup\$ – GR Tech Mar 17 '14 at 7:01
  • \$\begingroup\$ @GRTech - Thanks, I'll read that. The meter I'm using claims to use a sine wave (rather than a square wave), however upon analysis it's more like a rounded triangle-wave. \$\endgroup\$ – UpTheCreek Mar 17 '14 at 9:34
  • \$\begingroup\$ Great video review of the unit over here youtu.be/qDABYKoVO4Q \$\endgroup\$ – jxramos Apr 6 '18 at 23:54

It's quite useful for repair purposes. The voltages are low enough that semiconductors don't come into play. When there are parallel e-caps, you should be comparing a total ESR of the parallel combination, but you're really looking for gross differences not even the last 2:1. The typical situation is an older piece of electronics where the electrolyte dries out, causing the capacitance to change a bit, but mostly the ESR to increase, so you get mains ripple or SMPS ripple, which eventually causes the thing to stop working adequately. E-caps do eventually wear out, and replacing them can often restore an older item. Especially the caps in the power supply- because they usually have less margin and because they tend to be in a hotter area and have more self-heating (due to ripple current and ESR-- so the higher ESR gets, the more self-heating occurs).

The other factor affecting usability is that a lot of recent electronics uses low-Z caps, so what's excellent for a standard e-cap is a bulging mess on a low-Z cap. You may have to look up the part number or guess from the schematic or circuit configuration (something like 1000uF/6.3V on a motherboard is just about guaranteed to be low-Z).

I wish I had one of those, there's a few times it would come in handy, but I don't do enough repair work to justify it. Usually I just tack a known-good cap across the suspect one and see if it makes a difference!

  • \$\begingroup\$ I like that; drop a parallel cap in to bolster the suspected defective cap. Nice experiments. What value do you choose to go with, 10% of the rated value or some other rule of thumb? \$\endgroup\$ – jxramos Apr 6 '18 at 23:47
  • 1
    \$\begingroup\$ The same value, many caps are rated at -20%/+80% and usually circuits are tolerant of much higher values, of course you might discover one that isn’t some day, but that’s what I do. \$\endgroup\$ – Spehro Pefhany Apr 7 '18 at 0:34

You can't rely on readings when the capacitor-under-test is in circuit because other components (such as other similarly sized capacitors) may be in parallel. I'm talking about power rail circuits because they are a common usage for large electrolytics where you might be particularly relying on ESR being low.

In places other than power rails, electrolytics are of course used but, is the ESR that important in these places? I'm thinking time delay circuits like 555s.

  • 1
    \$\begingroup\$ Thanks! Right, the parallel issue is one thing I was thinking about. Regarding your second point about whether ESR is important in other areas - I agree, probably not so much, but the main reason I bought the meter is I'm often dealing with very old audio equipment, where electrolytics are quite often faulty. From what I've read, ESR is a good indicator of ageing capacitor health - and the in-circuit ability of this test would be very convenient, if it works properly. \$\endgroup\$ – UpTheCreek Mar 16 '14 at 18:15
  • \$\begingroup\$ @UpTheCreek If it helps test old capacitors then good luck - my Mackie speaker setup (not that old) has just developed a fault that I tracked down to a faulty electrolytic on one speaker but it had gone leaky! \$\endgroup\$ – Andy aka Mar 16 '14 at 18:18
  • \$\begingroup\$ @UpTheCreek Usually the "in-circuit" ESR testers they have a just a DC isolation at the input. To my experience doing in-circuit ESR measurements you can not find a good capacitor. Only bad capacitor. Even worse you can not find bad capacitor if it is on a bank with good capacitors. Take special care if you are going to do this measurments in areas woth high voltage! \$\endgroup\$ – GR Tech Mar 17 '14 at 7:20
  • \$\begingroup\$ @GRTech - Interesting, but that's not my experience so far - I've yet to find a bad one with the in-circuit test. Is there a specific meter that you found gave false-negative results when using in-circuit? \$\endgroup\$ – UpTheCreek Mar 17 '14 at 9:32

I have used a relatively cheap Atlas ESR meter to test capacitors in a faulty Jamo E7 Subwoofer.

The meter accurately identified the faulty 10uf capacitor in circuit.

Measurement generated significantly higher uf readings as well as leakage.

I am happy to report that the sub is still working, 6 months after the cap replacement. (I was expecting the repair to be short lived.)


It depends on the design (excitation voltage) and the dut surrounding circuitry. If the excitation voltage is low, it may not turn on a transistor or diode, sand there isn't other resistors or inductors around, it can be fairly accurate.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.