Don't worry about power up. The engineers worried about that. They
didn't use 100,000 uF, don't use 100,000 uF.
As far as a tester based on a PC soundcard, I have never heard of one. I
will think about it. I used a scope for a long time with a 1Khz square
wave and it worked like the dickens.
Actually, I could probably draw the responses from good and bad caps of
a bunch of different values and explain what has to be looked for.
Aftert hat you can do a mental FFT and ficuge it out. I guess is 20Khz
bandwidth is not going to cut it, but I could be wrong. What's more
there could bea a workaround.
Interesting you mentioned Soundcard.
Use one channel OUT from the soundcard and two channels IN. The single
output channel, Vout, is monitored by channel 1 and channel 2 monitors the
other end of the resistor connected to the capacitor under test.
I just simulated the soundcard technique using LTspice and by simply using
some simple V1-V2 and V2 it is possible to use real() to get the esr and
imag()to get the capacitance value independent of the frequency [as long
as you use 2 pi f for the capacitor calculation]
I just started writing the C/C++ code [actually a modificaiton of
something I've already been doing] that can slightly self-calibrate and
then start running, as it runs the display screen will show the two
values, whether you attach a cap, or not.
Later I'll add a check for noise and/or ranges of signal levels to
estimate the accuracy, like in some ranges should be better than 0.1% for
the cap and 1% for the esr, but as the cap gets too small or too large,
those numbers change and it would be nice to have some level of confidence
show on the screen too, like...
C = 47.0 uF within 1%
Resr = 33.0 milliohm within 5%
something like that, I haven't calculated how accurate the numbers will
come out, I just guessed at the above accuracies.
The actual accuracy will be based upon how good a resistor that's used.
like a 0.01% 10ppm or such, or don't care and use some way to calibrate,
but that takes some very good loads.
With some more work, it would be easy to make this a true RLC meter, for
right now only need a cap/esr meter