[Prev][Next][Index][Thread]

Re: Dead Capacitor

Subject:       Re: Dead Capacitor
       Date:   Sun, 11 May 1997 14:16:04 -0700
       From:   Bert Hickman <bert.hickman-at-aquila-dot-com>
Organization:  Stoneridge Engineering
         To:   Tesla List <tesla-at-pupman-dot-com>
 References: 
            1


Tesla List wrote:
> 
> Subject:  Dead Capacitor
>   Date:   Sun, 11 May 97 04:22:31 UT
>   From:  "Larry Robertson" <LWRobertson-at-msn-dot-com>
>     To:  "Tesla Builders" <tesla-at-pupman-dot-com>
> 
> Hi all ...
> 
>         I decanted the dead capacitor today - doesn't mineral oil with
> vaporized
> plastic smell great - and found brown stains on the plastic all along
> the
> edges of the aluminum plates. Some was possibly carbonized oil as it
> wiped off
> easily, but some was definatly damage to the plastic. There was one big
> hole
> about 3/8 " in diameter surrounded by carbonized plastic. There were no
> signs
> of trauma elsewhere.
> 
>         I used two sheets of .030 HD Poly per side and rolled it up a la
> RQ's
> instructions. I rather think this is not enough dielectric, as I am
> using two
> 15 KV neons. Richard suggests as much.
>         In reading spec sheets from plastic manufacturers, and from
> Sargeant and
> Dollinger's book it appears decreasing the dielectric thickness by 4
> doubles
> the dielectric strength per mil. I shall try using 3 layers of .020 HDPE
> ,
> however I'm beginning to wonder why I don't go to the local Home Depot
> and buy
> vast quantities of .010 LDPE for 1/4 the price.
> 
>         Also as I bought 48" by 96" sheets, cut somewhat sloppily into
> 22 inch
> pieces, I now have 6 pieces 4 inches wide left over. I shall try adding
> 4 of
> them over the 4 edges of the aluminum, although this may not enhance the
> rolling characteristics.
> 
>         The smell from the oil I got on my hands does not come off
> easily. The wife
> won't stand nearby and the dog refuses to talk.
> 
>         I suspect maybe rubber gloves could help.
> 
> LR

Larry,

Sounds similar to the line about loving the smell of Napalm in the
morning... I prefer the smell of fresh coffee and ozone myself. :^)

You're seeing "partial breakdown" damage to the dielectric fluid and the
LDPE. Even if you've sanded the aluminum plate edges to round them off,
60 mils of LDPE will not relaibly withstand sustained 15KV RMS Tesla
Coil operation - the E-fields at the plate edges are simply greater than
the dielectric strength of these materials. Throw in the fact that
disruptive discharging typically results in >85% voltage reversals, and
you're seeing effective peak-peak RF voltage swings approaching 40 KV.
Although LDPE is a very good dielectric material, 1/16" of it is just
not up to the task of withstanding this kind of stress. You need to go
to at least 90 mils of LDPE to reliably handle this abuse. 

Also, fabricating a number of equal-valued capacitor modules in series
is a MUCH better approach than having all 90 mils in a a single
dielectric layer between plates. The E-fields at the plate edges are
substantially reduced, as is the incidence of partial-breakdown damage
at these edges - this is one of the many tricks used in the design of
commercial high voltage pulse caps, and is one you can easily adopt in
your design.  

The only real "downside" of using thinner material from Home Depot is
that it generally comes in a maximum thickness of 6-7 mils, and, since
it is often made from recycled LDPE, it may not be as consistent in
quality as virgin plastic coming directly from a plastics vendor. Also,
there's a fair amount of labor associated with cutting and using
multiple thicknesses of this stuff. 

However, if you do choose to build caps out of this material, 90 (total)
mils of thinner material should be virtually bulletproof at 15 KV _once
you've removed all the entrapped air_. You may want to reference some of
the earlier posts from Fr. McGahee, Gary Weaver, or David Flinn on
building rolled or flat plate caps out of this material, particularly
using kraft paper to aid in wicking. 

Furthermore, if you fabricate these as a group of 3 or 4 caps tied in
series (easily done for plate caps), you should end up with a tank cap
that'll take any abuse that you can dish out! For example, instead of
making a single cap with 90 mil multi-layer dielectric, make several
identical caps using 30 mils each, with each smaller cap having 2 or
more layers of thinner dielectric, and connect them in series using
heavy, low inductance connections. While the overall material usage is
roughly the same, the reliability will be much greater. This approach is
very easy to do with flat plate caps, and the identical capacitor
modules can all co-reside in the same container. For wound caps, this
may mean using multiple containers. 

Safe cappin' to you!

-- Bert --