Testing a capacitor with an ohmmeter

[mort]

I've read that after an initial swing, good capacitors will read in the Mohms, and that bad ones will read open or low resistance. I've tried it on a few new capacitors and get mixed results. Some of them read open after the swing. Does it matter what value the caps are rated as to whether this works or not? Does it also work on non-polarized caps?

The ones I've tried all read within their tolerance for uF, btw.

Thanks for any light that can be shed on this topic.

[Tom Phillips]

I've read that after an initial swing, good capacitors will read in the Mohms, and that bad ones will read open or low resistance...

Realize first that this is just a crude test. It does apply to all caps. However, a more correct way of thinking about it is after the cap finishes charging up at the very low voltage applied by the Ohm meter a capacitor that reads in the Mohms (or "open" by your meter) MAY be good. If it reads low resistance then it IS bad. When your meter reads "Open" it most likely means that the reading is beyond the capability of your meter. However, if there is no initial charge phase (The "swing" you mentioned) and the cap immediately reads open then the cap MAY be bad. The smaller the capacitance the quicker will be the initial swing. Below ~ 0.047uF you may not notice the initial swing at all.
Hope that makes sense.
Tom

[mort]

YEp, that makes sense. Is it also normal that once you've charge up a cap, let's say a 100uF 100v cap, that you can take the meter off and put it right back on and read OL instantly? Sounds like that could be normal. Just trying to get comfortable with various observations.

Is a VTVM the best way to check capacitors?

[Tom Phillips]

...Is it also normal that once you've charge up a cap, let's say a 100uF 100v cap, that you can take the meter off and put it right back on and read OL instantly?

Yep.

...Is a VTVM the best way to check capacitors?

Well... It's just one way to make one type of confidence test. There are many more levels of test such as a capacitor tester that can measure Equivalent Series Resistance (ESR) and Dissipation Factor (DF). Testers that subject the cap to its full working voltage and measure the leakage current under that condition are also available. There are more tests that can be done too such as capacitance, EST & DF at different frequencies, temperatures etc.

[Aloha_Mark]

Nowadays you can buy a cap tester for around $16 from China using Ebay. Not that I recommend this, but that testing for the value is the minimum you should do to check out if you need to replace the cap. Brand name multifunction testers cost over $100. The very good ones cost a lot more, but these can measure inductance too. Just because something is advertised as a cap tester does not mean it can figure out the value of all caps. Bumble bees, black beauties, Hytron and other old caps may give spurious readings on the cheap testers.

If it is an electrolytic, then you need a tester that can measure leakage at the rated voltage. In the good old days before everyone discovered the Sprague Tel-Ohmike, you could get these vintage and highly capable testers for under $30. Now the prices have skyrocketed like pre-CBS Fender Strats.

[mort]

This is the meter I use most of the time.

Fieldpiece SC53 Dual-Display Mini Clamp Meter with Temperature - at the Test Equipment Depot


It says it will read up to 40M and up to 400uF. I've been able to read down in the nF on the low end.

[R.G.]

Let me just speculate a bit.

A capacitor can be considered to be a perfect/ideal capacitor with some external "dirtying" bits attached inside where you can't get at them.

The most important of the internal sullying bits are the leakage resistance and the Equivalent Series Resistance (ESR). The leakage resistance can be conceived of as a resistor in parallel across the whole thing. The ESR is a smallish resistance in series with just the capacitance.

You can measure the leakage resistance and internal capacitance several ways. I would start like this:
(1) Calibrate your meter input resistance by putting your meter in series with a 1M resistor across the working voltage you're going to test the cap at. The voltage your meter shows after the 1M is the voltage BEFORE the 1M times Rmeter/(Rmeter+1M). Doing a little algebra tells you what Rmeter is. Note that your meter is probably 10M input resistance; some of them are only 1M.

Next, hook up the 1M resistor in series with the capacitor and connect the 1M+capacitor across the desired working voltage; monitor the voltage across the cap until it doesn't change significantly any more. When the voltage quits changing, the internal capacitance is all charged up, and the DC conditions are set by the external 1M resistor and the internal leakage.
You can then calculate the internal leakage as being set by the voltage divider of the 1M, and the meter input resistance and the cap leakage in parallel. The DC voltage will be
V = (Rmeter||Rleakage)/(1M + (Rmeter||Rleakage))
Since you know the value of the 1M and the meter, you can calculate the leakage of the cap.

(2) You can then measure the DC value of the capacitance. You hook your meter across the cap and connect a resistor in series with the cap. You let the cap run down to 0V, then apply voltage to it and note the voltage as it rises. This sets up the cap to be fed by a resistor network, and caps fed by resistors have a very deterministic voltage. A capacitor C fed by a voltage V through a resistor R will rise to 63.2% of the total voltage it will eventually rise to at a time equal to T = 1/RC. So if you know the R (from the leakage, 1M, and meter resistances) and you measure T to get to 63.2% of the total voltage, you can calculate C = 1/RT .

The V is the settling voltage of the 1M, cap, and voltmeter across it, and the R is the 1M, meter resistance, and cap leakage all in parallel.

[drewl]

Sounds like your using an old analog meter, which do tend to test caps better than a digital voltmeter.
Those old Simpson meters put out a little bit more voltage.
As mentioned the only way to test for sure besides in circuit is with a good ESR meter that can put out up to 5-6 hundred volts.

[Mick Bailey]

Capacitors will very often check out ok for value on the bench, even if the ESR is way too high, or they're bulging and oozing. I disregard value unless I'm trying to identify an unknown component. Similarly, low voltage checks don't always show up a cap that fails in-circuit with a working voltage applied. All that you can determine with a meter is that the cap measures OK at the test voltage applied to the meter probes. This is too unreliable for amp work as it can throw you off the trail - you check a cap on the bench and say to yourself "that checks out fine" and move on - but the cap could actually be bad.

It would make a nice project to build a tester with 200v/400v/600v switchable ranges and an inexpensive analogue meter.

[Enzo]

As Tom said, a hand meter is a VERY crude test of a cap. ABout all it tells you is if the cap is shorted.



A cap can be thought of as a crappy little rechargable battery. Your ohm meter puts out maybe a volt or so. COnnect it to your cap and watch the cap charge up - the reading goes up and up until it stops somewhere. Now flip the meter over to DC volts and see if the cap has stored any of the voltage from the meter. This is also VERY crude, but might give some insight into the cap.


I usually test caps by their function in the circuit.

[Steve Conner]

When using an old analog meter on the ohms range to test capacitors, bear in mind that the polarity of the voltage coming out may well be opposite to how the terminals are labelled. This has obvious implications for testing electrolytics. To find out the polarity of your meter, measure the voltage coming out of it with another meter, or test a diode on the ohms range.

Digital meters are the other way round. The polarity of the test voltage is the same as the polarity marked on the terminals.

[gbono]

When using an old analog meter on the ohms range to test capacitors, bear in mind that the polarity of the voltage coming out may well be opposite to how the terminals are labelled. This has obvious implications for testing electrolytics. To find out the polarity of your meter, measure the voltage coming out of it with another meter, or test a diode on the ohms range.

Digital meters are the other way round. The polarity of the test voltage is the same as the polarity marked on the terminals.

Mmmmm solid tantalum capacitors do not tolerate reverse bias very well. I would use a DMM when testing solid Ta capacitors.