http://octopus.freeyellow.com/esr.html
For R2 I use a 1k, 1W resistor. R1 is 2 ohms. I have them mounted on a terminal block so that I can screw the capacitor leads in parallel with the 2 ohm resistor. Cut an old BNC-to-BNC cable in half, and use one end for the Source input end (ie Function Generator from the drawing), and the other for the analyzer measurement of the divided voltage (ie o'scope input from drawing).
You can calculate ESR at any frequency up to 140kHz. Indeed, I guess you could sweep the ESR using the software. Just thought of that! Anyhow, I'll set the source to 100kHz, 16V output. Measure the drop across the 2 ohm resistor. Then put your cap in series. Measure. Then, instead of using a table, or solving it out on paper, I just use this calculator here:
http://hyperphysics.phy-astr.gsu.edu/hb ... divac.html
It does the job of paralleling my load with R2 and re-calculating the divided voltage altogether. I just adjust my load impedance 'til it matches my measured voltage. This isn't difficult, just start at 1 ohm and just in 200 milliohm steps in the right direction and get as accurate as you want. Yes, the contacts of your jig will have some extra resistance, but it seems to be less than 100 milliohms in reality.
I wouldn't use this for in-circuit testing, but it is handy looking at old can caps, or polyprops, or modern electros. While low ESR caps are extremely important for switching power supplies, the whole "bypass your electrolytic with a polyprop" is a pretty common meme in the tube community. I think it is worthwhile to see if a bypass or power supply electrolytic is poor enough to need this extra bypass. General rule: the bigger size the modern electrolytic, the lower the ESR. So a puny electrolytic might benefit greatly with a polyprop bypass. On the other hand, the power supply caps in the diytube ST35, for example, are only ~300 milliohm up to 1MHz. I doubt polyprops would add a significant value and would just complicate the circuit.
Shannon
