I never tried it, but here's what i would do:
Put your cap in series with a resistor of a known value, say 100 Ohms. Now connect this to your Signal Generator using a rather high frequency of 100 kHz or so.
Now if you measure the Vpp coming out of the generator and the Vpp over the cap you should be able to calculate the ESR by using simple voltagedivider math.
Problably not very accurate but schould be enough to tell whether the cap is still ok or not.
Thanks, that seems to work. I get 4.8 Ohms for an old 160V 33uF cap, which is about right according to ESR tables.
ESR for a cap is most time given for Just one frequency. Most times somewhere between 100Hz and 1KHz. Many ESR meters meaure at 100 or even worse 200KHz.
For 33 uF at 100KHz the reactance is 0.048 ohm. If the ESR is 4.8 ohm a 100KHz that cap is so bad you can not even imagine. That is a D of 99.5 or a loss angle of 89 % ( 0% is an ideal cap, 90% an ideal resitor)
But if you measure at 1 KHz, reactance is 4,8 Ohm ( coincidence or are you just reading the meaurements the wrong way ?) a cap that meaures 4.8 Ohm is a D of 1, also not very good so to say ( 45 degrees loss angle) but if they state ESR for 100Hz Xc is 48.22 Ohm and then a ESR of 4.8 Ohm is a D of 0.099 ( or loss angle of 5.6 degrees) and that is not a bad cap.
So you see, three different frequencys, same ESR totaly different result. Now what was your frequency and what frequency in the datasheet of the cap ? ( or did you use those out of the blue tabels that are printed on ESR meters ?)
However the way with the squarewave works, i tested it too, but you must just measure the voltdrop over the ESR alone and that is often so small with regard to the measurement signal that it is hard to do. You realy must zoom in. But it gives an indication if you compare good and bad caps and use the right frequency it can be a quick go no-go test.
A cap has a selfresonant frequency, ESR is minimum there, capacitance and ESL are zeto Ohm. Below that ESR increasing upto almost infinity at 0 Hz and it incresase again after the SRF because the increasing skineffect. Besides that, your meauring an inductor above SRF.
100KHz or 200KHz is only usable if the cap is made for high switching speeds.
ESR is no magic, it is just plain network analysis and you need to know what you measure, to bad many people use those ( much to expensive) ESR meters without knowing what they measure. And everybodys happy because a real bad cap shows itself ( one you would have seen in a second using your scope to meaure ripple) but a lot of caps that are not bad get replaced too, because at 100KHz almost all caps measure bad, so the result:
This was a good deal, the monitor works again because I found a lot of caps with high ESR.
The truth, you replaced a lot of caos, most however probably had good ESR, but lost capacitance or were leaking DC. But most important your monitor works again
About in circuit meaurement i will not even start ( I make enough enemie allready this way probably
)
I measure ( and examine caps) ESR as a hobby, not while repairing, i designed an ESR meter, build several others, used vna's, bridges ect. Build capacitance meters upto 1 fF. Caps are very interesting.
I find bad caps most times by measuring with a scope for ripple ect. But after that I keep the bad ones and examine them for fun. From the many, many caps I measured ( the right way) only a very few ( like 5 out of 100) really had bad ESR. In smps most caps just lost capacitance, second place is DC leakage, third place ESR.
If a cap gets to hot ( bad design pcb or to hot housing) the water from the electrolyte vaporises and the electrolyte loses its capability to restore the oxide layer. Dielectrium dries out, capacitance decreases and it starts leaking DC, gets even hotter and pops open.
If there is dirt in the cap, or coroded electrodes ESR increases. It is important an ESR meter is not capable to measure Resistance direct ( like only a resistor), unless it measures phase angle because otherway you do not know what is R and what is jX. It is just measuring |Z|. And that is why they measure at high frequency, because most of them only measure |Z| and jX is small at those frequency.
A shorted cap will show a low ESR in that case. A good meter measures ESR upto pF ' s because the principle is the same. But the signal amplitude becomes high, |Z| is high. I need around 10Vtt to measure 10 fF capacitance at 1KHz with my homebuild fF meter. 1 pF is 160MOhm, an extreme good ESR for 1pF at 1KHz is 1600 Ohm. but at 100KHz Xc is just 1.6M so ESR for a real good capacitor is around 15 Ohm.
Now more realistic, most ESR meters stop at 1 uF. Xc is around 1.6 Ohm at 100 KHz, if D is good, say 1% loss angle, D is 0.017 so ESR is just 27 mOhm. But 1nF is a jX of 1600 Ohm and the same D ( not unusual for 1nF) is 27 Ohm. This is within the range of most meters, so ask your self why they do not measure ESR of a 1nF ? Because they just measure Z, Rest my case
Damm, when do I learn not to react any more on ESR topics.