>From: Richard Quick <richard.quick-at-slug-dot-org>
>Subject: Re: Capacitors
>Pardon me for jumping in. I have never seen corona under oil.
>Oil is like the _ultimate_ corona supressor.
>> I'm asking because I've blow my 4 home rolled caps. Each was
>> evacuated to <10^-2 torr, filled with oil and then pumped down
>> again. I am guessing that I am just over voltageing my caps but
>> will watch for corona if it is observable.
>I doubt you will see much under oil. The caps likely failed
>because they were overvoltaged, or they were not fully broken in.
>Even pumped caps will survive higher voltages longer after the
>oil has a chance to permeate the entire capacitor. Even a
>pumpdown and an entire month under oil does not mean the
>capacitor has reached it peak breakdown voltage rating.
>Realistically it takes a pumpdown, 30 or more minutes of
>intermittent operation, and about 6 months under oil for the
>capacitor to reach maximum breakdown voltage. Until these
>qualifications have been met, I run homemade caps in series.
2 weeks ago in response to a thread on paper-oil caps I
started to test some butcher paper (no kraft paper at home and EIS
wants $100 min order -- I don't want THAT much paper around;)
Using wrinkled but smoothed out Al foil, I found that the paper
punctured through at the ridges on the Al foil. No surprise here given
the voltage gradient a point will exhibit. But I lead me to thinking:
"I wonder if my poly caps are blowing because of small 'points' in
their surface? The recommended method of running the caps for weeks to
break them in, which I did not do because I had vacuum outgassed them,
was to remove the air bubbles. What if it's effect really was really
to EDM (electrical discharge machining) off the sharp points that
sanding missed or created?"
After an hour, I could not see any difference in the surface texture
under 50 power magnification.
Has anyone examined or proposed this hypothesis?