measuring capacitor using Time Constant?

[microfarad_the_2nd]

I remember Time Constant theory from my Electronics courses 30 years ago.

I have been reading badcaps for a few days now.

I'm not yet understanding much about ESR yet, but I do understand capacitance.

This might be a really dumb question, but assuming a person had virtually NO money to spend on any new test equipment, for out of circuit capacitor testing, would a series resistor and a stopwatch (and good quality DMM) not be enough to make a "rough" measurement of a capacitor? I mean, using the 5 Time Constant principle?

If the resistor is carefully chosen, and if it's value is measured using a high quality DMM, the RC circuit could be designed to be quite accurate, would it not? example, a relatively long RC circuit Time Constant could be made such that 5 TC's are about 2 or 3 minutes.


, uh, I guess I just realized, this would only be good for DC current measurement and testing the cap using just DC charging current. Does testing capacitors in our computer power supplies require testing at high (relatively) frequencies? I notice many posts here at badcaps talk about 100khz testing.

? sorry if the questions are on the dumb side, just looking for a little guidance. I like finding cheap/free solutions to things....

[Kiriakos GR]

Re: measuring capacitor using Time Constant?

I am aware about this rough testing method but it is used for high capacity caps above the 20MF were most digital meters does not reach.

For small and regular caps this way of testing is not good enough in our times.
There is no cheap/free solutions when you need accuracy and productivity.

Few months back I watched an Youtube video, some one with an analog multimeter was trying
to show that you can charge and discharge one capacitor at the Ohms range and see the needle moving when discharging.
Yes one carefully chosen resistor it can slow down things, but this is for people who does not use a clock on their wrist.

[tom66]

Re: measuring capacitor using Time Constant?

You'd actually be best using a single time constant, or 63.2%.

5RC is nearly 99.3% charged, so very hard to tell apart from 6RC etc.

But another good way is using a 555 timer. An astable oscillator will vary frequency with capacitance in a 1/x relationship. I know of some older multimeters using this circuit, before they moved to charge-integrate-discharge schemes under control from a processor or logic.

[eccerr0r]

Re: measuring capacitor using Time Constant?

Most cap meters for caps > 1 microfarad use RC timeconstant to calculate capacitance. But it's not necessary to charge it completely, just need accurate time/voltage measurement.

You can still build an oscillator with a 1millifarad capacitor but you'd be using fairly low value resistors to get it to oscillate at a reasonable frequency, and if it's low enough, the transistors driving them will become an issue, possibly frying them...

[tom66]

Re: measuring capacitor using Time Constant?

That is why many older, cheap multimeters only had a capacitor range up to about 20~40uF.

[eccerr0r]

Re: measuring capacitor using Time Constant?

Unfortunately the two multimeters I have, have this exact issue... Apparently they do make oscillators but with large resistances and thus make measuring large capacitors take excruciatingly long (plus cap leakage current becomes a factor.) One of them goes to 100uF but it takes forever to get a reading.

My Sencore LC meter however appears to use RC timeconstant with a more appropriate resistance to combat leakage and does not completely charge the capacitor, making readings much faster. But it still "caps" out at 0.2F unfortunately...

(The one thing that I do wonder is that the Sencore can measure leakage at different voltages and able to discharge them fairly quickly as well somehow. This means I could charge one of those input caps to 200V... that's a lot of charge on that capacitor...)

[Kiriakos GR]

Re: measuring capacitor using Time Constant?

(The one thing that I do wonder is that the Sencore can measure leakage at different voltages and able to discharge them fairly quickly as well somehow. This means I could charge one of those input caps to 200V... that's a lot of charge on that capacitor...)

This is why God created the modern LCR meters..
No more .. feed them volts to see if they are blowing up.

[eccerr0r]

Re: measuring capacitor using Time Constant?

yeah the sencore cannot measure ESR/ESL... but sometimes you have to charge it to the correct operating voltage to see what the real leakage is.

Also it was designed to reform capacitors - though it's not a problem with most of today's electrolytics, older electrolytics needed to be reformed if they sit uncharged for too long.