It's actually not exactly 2/3 but 1 - 1/e where e = 2.7182818284..., so 1 - 1/e = 0.632120558...
Nigel proposes to charge instead of discharge the capacitors via a resistor, but charging to 1 - 1/e times the voltage or discharging to 1/e times the voltage is completely equivalent , except for the effect of leakage currents. Those would lead to a somewhat too high value with Nigel's charging test and a somewhat too low value with my discharging test.
Anyway, all the values you measured in post #33 are a bit higher than the capacitor's nominal value. That could be due to dielectric absorption, see Dielectric absorption - Wikipedia , or maybe the capacitors have a skewed tolerance interval, like -10 %/+50 %. In any case, it shows that they are not dried up.
Nigel proposes to charge instead of discharge the capacitors via a resistor, but charging to 1 - 1/e times the voltage or discharging to 1/e times the voltage is completely equivalent , except for the effect of leakage currents. Those would lead to a somewhat too high value with Nigel's charging test and a somewhat too low value with my discharging test.
Anyway, all the values you measured in post #33 are a bit higher than the capacitor's nominal value. That could be due to dielectric absorption, see Dielectric absorption - Wikipedia , or maybe the capacitors have a skewed tolerance interval, like -10 %/+50 %. In any case, it shows that they are not dried up.
! minute is still near the time constants involved and does not tell us much about *health*; if anything about *capacitance*, nice to know but not the main issue today.
Yes.
Ok, that tells me thay can be safely used at 28V rails, where LM3886 is very comfortable.
Of course, at +/- 22V even more 🙂
I saw elsewhere in this forum that we need to keep the voltage 20% under the rated voltage. Right on at 28v. 🙂
I need to take note of all those formulas. 🙂
''skewed'' may be in the way of fabricate say 12,000uf capacitors and print 10,000uf on them to be certain people don't complain they have less. I think Roederstein was a reputable compagnie at the time and they want to keep it that way.
Anyway I can go to work with them... or play , it's a hobby that I like. I find it relaxing to make something with my hands, not just the electronic but working on a case too and trying to understand what's going on in all those little composantes is interesting too.
Have a good night/day