Re: New type of home made cap almost 4nf -at-at least 5kv

["Metlicka Marc" <mystuffs-at-orwell-dot-net> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

hi tom, all
i have had some experience with flat plate cap construction in the past
and in looking for dielectric materials, i thought that PET
(polyethylene terpthalate, or close enough spelling) was a miracle
answer in my quest. the dielectric constant was good and the pierce
voltage rating was outstanding, so i ordered some from mc master carr.
after building a cap from it and soaking for a few days in oil, the cap
failed almost immediately. i took it apart and my findings are that
mylar (pet) will not tolerate ANY corona, the way pet is formed or it's
general property's are such that when it burns, it pulls back on itself,
so one little burnt pinhole instantly becomes a large, perfectly round
hole? in that i have found no way to completely eliminate corona in a
hawg capacitor, i abandoned pet in short order. i think in a commercial
cap that uses metalized mylar film, there is very little edge exposed
for corona to form, but i can't say for sure? all i know is it didn't
work for me. in my polypropylene sheet protector caps, the self healing
property (another thing i don't understand?) works very well, i pulled a
cell apart that i could see was carbonized, but would fire again after
letting set for a few minutes, the propylene seems to get "stringy"
around the hole and i'm guessing these strings must "melt" into a seal
over the hole?
a while back i had asked about using glycerin for a dielectric filling
material instead of oil, my thinking at the time was that this fill
might stop any corona from forming at the edges of the plates, because
the constant is 60 or more i think, thus stopping any burning of the
dielectric. since then, through the power of this list, i realize that
mylar is not a good rf material so i have gave up on it as a tank
dielectric.
hope this clears up a few points for you?
marc