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

Re: Radical Flat Cap design



Tesla List wrote:
> 
> >From DavidF4797-at-aol-dot-comSun Oct 20 21:40:33 1996
> Date: Sun, 20 Oct 1996 13:13:51 -0400
> From: DavidF4797-at-aol-dot-com
> To: tesla-at-poodle.pupman-dot-com
> Subject: Radical Flat Cap design
> 
> It has been recommended that "multiple" layers of dielectric be used in order
> to minimize the chance of  breakdown due to flaws in the dielectric.  While
> philosophically correct, this practice usually causes problem when
> implemented.  Most home brew cap builders build with what is available.  What
> is typically available at most home supply and hardware stores is roll
> polyethylene with a maximum thickness of 6 mills.  In order to build a robust
> cap with a dielectric thickness of 90 to 120 mills, 15 to 20 layers of
> dielectric material between plates is required.  This implementation is
> seriously flawed!
> 
> First, polyethylene sheets trap air bubbles between them, the more layers,
> the more air is trapped.  A 120 mil dielectric of twenty layers will trap ten
> times the air of a 120 mil dielectric composed of two layers.  There is
> therefore a trade off between reducing the likelihood of failure due to flaws
> by using multiple layers -and- using the minimum number of layers to reduce
> the total quantity of trapped air pockets (which can also cause failure due
> to heating effects).  I propose that two or three layers of thicker
> dielectric -two 60 mill sheets, for example- is sufficient to reduce the
> potential flaw line-up problem significantly while at the same time *greatly*
> reducing the amount of trapped air.
> 
> Second, in standard flat cap design, it is suggested that the cap be left
> idle for at least a week after assembly to permit "off gassing" to occur.  I
> would suggest that in a traditional multi layered dielectric design, one week
> is not nearly enough.  Low density polyethylene has a soft waxy consistency
> which tends to "weld" it together with other layers, particularly if the
> plates a "clamped" together with some force as they should be in any good
> flat cap design.  This "welding" of the polyethylene creates air pockets that
> the wicking action of the oil is not sufficiently strong to overcome.
> The first build of my cap which used fewer thicker plates (as per the first
> suggestion above) had this problem.  I constructed the case out of 1/2 inch
> clear acrylic and I observed that even one week after assembly, after all
> off-gassing had ceased, a 3 or 4 inch diameter air pocket in the center of
> each plate remained.  The size of the air pocket stopped diminishing after
> the fourth day, leading me to believe that the remaining trapped air at the
> center of each layer of dielectric would only gradually be eliminated after
> months, perhaps, if ever.

Some thoughts: 

1. Would the use of a partial vacuum help remove the trapped air. It
sounds like the housing you used could support drawing a substantial
vacuum.

2. Could the layers have been initially clamped too tightly?? I seem to
remember a post some time back where a failure in a CP capacitor was
traced to oil not being able to penetrate inside the capacitor roll
because the roll was too tight, and the dry area then failed...

> 
> I propose that this problem may be solved by interleaving thin absorbent
> paper sheets between each layer of dielectric to improve the wicking action
> of the oil and more completely and quickly displace the air pockets while at
> the same time having a negligible impact on the dielectric strength and
> dielectric constant of the assembly.  I implemented this approach on the
> first -and only- rebuild of the cap.  The result was a spectacular success.
>  *All* off-gassing stopped within 12 hours and *all* air pockets were gone by
> 24 hours.  The day after assembly, I cranked the cap up to half power -which
> is supplied by a 15Kv  -at-60ma neon xformer- for a minute or two and observed
> only one thin stream of about 10 bubbles.  After this stream of bubbles had
> stopped no more were apparent.  I slowly cranked the neon up to full power
> and ran it for several minutes with no heating and *no* further out-gassing.

Were these power-ups under heavy spark discharge conditions or just the
60 Hz HV? 

>  I suggest that two to three thickness of dielectric (at most) with paper
> wicks between each layer is a highly superior design approach.

With caps in Tesla service, only time will tell...! :^) 

One of the major advantages of using thinner film is the dramatic
increase in breakdown voltage (volts/mil) for the dielectric. Individual
sheets of the store-bought 4 and 6 mil material is "slipperier" than the
0.062" sheets. This should reduce the tendency to stick together.
Assuming you don't clamp too tightly, air should be able to work its way
out fairly quickly in Tesla service. I'd suspect, after outgassing, that
the resulting breakdown voltage for equal total thicknesses of LDPE will
be significantly greater with mutiple/thinner film sheets. Practically
speaking, at 120 mils and 15 KV, either approach will perform
adequately. At 20 KV, I'd opt for the latter construction... 

Any possibility of posting some pictures of your cap?
 
> 
> Sticks and stones welcome......
> 
> I apologize for the poor quality of my last post on foil-tab arc over.
>  Composing at two AM is apparently not workable for me.  I even forgot my
> name!  Although in that case it may have been fortunate ;)
> 
> Gas-less flat cap building to you.
> 
>  - David F -

And safe cappin' to you, David!

-- Bert --