I have a filter cap here for a rectified 18V rail, it 's 6,800uF /25V, and at @100kHz it reads 0.03-0.04 ESR and ~16uF, at DC it measures 6,190uF

So that's probably nearly bad ? It would be ok for now, but I have others here anyways.


What about 47u/25V, some charts say 2ohm is the limit, others say 3.5 ? I have a few well over, so I should ditch any near 2 I guess, there's some that read 1.8, but they're all the same brand in the same audio sys. so I guess might as well replace as many as I can now anyways.

 

That 6800uF cap is close to its tolerance, yes. If it looks good and isn't bothering anything, I may use it if it's already installed and just keep an eye on it, otherwise I would look for another one and relegate that one to general experimentation.

 

If you have a digital multimeter you can do a charge holding test that will help understand the performance a bit more. Just connect the volt meter across the capacitor and charge it to your anticipated working voltage, and then disconnect the supply and observe the voltage decay. That will tell you about the leakage current versus capacitance relationship. If the voltage drops rapidly then the capacitance is lower or the leakage is higher and the device is less useful.

But another consideration is the application.The capacitor provides the power during the part of the input voltage cycle where the diodes are not conducting. So if your voltage to the load is not showing any mains frequency ripple at full load then the capacitance is enough. for that application.