Problems With DRY Caps (fwd)

---------- Forwarded message ----------
Date: Sun, 26 Apr 1998 12:49:40 -0400
From: Thomas McGahee <tom_mcgahee-at-sigmais-dot-com>
To: Tesla List <tesla-at-pupman-dot-com>
Cc: acmq-at-compuland-dot-com.br
Subject: Problems With DRY Caps

I changed the title to reflect the message contents.

> ---------- Forwarded message ----------
> Date: Sat, 25 Apr 1998 12:04:24 -0700
> From: "Antonio C. M. de Queiroz" <acmq-at-compuland-dot-com.br>
> To: Tesla List <tesla-at-pupman-dot-com>
> Subject: Re: Creeping Sparks and Homemade Caps... (fwd)
> Thomas McGahee wrote:
> > A few of my observations:
> >...
> Excellent post. Thanks.
> I have been thinking and experimenting with a way to make dry capacitors.
> What I can comment is:
> The tabs coming out of the plates for external connections are the worst
> problem. The simple presence of the thin plate material coming out of the
> assembly produces corona in all directions. It is difficult to charge
> a capacitor built in this way to more than a few 10 kV due to corona losses
> at these tabs. And the corona to air easily finds a way to the other side
> of the capacitor, producing that nasty surface tracking sparks.
> Maybe some form of enclosing of shielding of the terminal areas can help.
> For example by covering the terminal areas of a flat capacitor with metal
> foil, or metal tubes with rounded ends, without any sharp edge of corner.
> Departing from the flat capacitor form, a safe structure is a battery of
> Leyden jars, made with -solid- connections. It uses more space but is almost
> immune to failure.
> Antonio Carlos M. de Queiroz
> http://www.coe.ufrj.br/~acmq

A "dry" capacitor is always going to have a lot of corona leakage, especially
when the air is moist. That is one reason why immersion in oil is so useful.

If you make "dry" capacitors, be aware of the fact that corona discharge is
going to literally eat up your poly around the edges. My post-mortems on dry
capacitors show that the poly is riddled with tiny finger-like 
depressions in the poly where the corona has eaten away the poly. With
time these can get quite deep, and then they allow even more air between
the plates and promote even MORE corona. Eventually a surface spark crawls
along one of these channels and initiates a power arc. Then your beautiful
capacitor she is defunct. Kaput. Dead. Even with extremely large borders
around the edges the corona eventually does its stuff and destroys the poly.

Even with compression it is next to impossible to keep the corona down. Look
at your capacitor while running and you will see a bluish light outlining
the plates. That is the corona at work. And believe me, it IS working!!!
It is working to destroy the poly.

I have done a few experiments trying to find something that I could use to
allow me to make long-lasting "dry" capacitors, but to date I have not
been too successful. The idea would be to first build the capacitor 
under liquid (oil, varnish, something amenable to TC conditions...),remove 
air, compress, and then somehow get the edges sealed with something that 
will harden and seal the whole thing tightly. I have attempted this using
varnish, but air bubble removal is no where near 100%.

Perhaps others who have built "dry" capacitors and used them for a reasonable
time would be willing to chime in here and tell us what the end results

Hope this helps,
Fr. Tom Mcgahee