Re: Homemade stacked plate cap. *Need* a vacuum pump?

From: 	Thomas McGahee[SMTP:tom_mcgahee-at-sigmais-dot-com]
Sent: 	Thursday, October 30, 1997 12:03 PM
To: 	Tesla List
Cc: 	absmith-at-tiac-dot-net
Subject: 	Re: Homemade stacked plate cap.  *Need* a vacuum pump?

> From: 	Adam[SMTP:absmith-at-tiac-dot-net]
> Sent: 	Wednesday, October 29, 1997 8:26 PM
> To: 	tesla list
> Subject: 	Homemade stacked plate cap.  *Need* a vacuum pump?
> After reviewing several capacitor options for my 8" coil setup, I have 
> found only one solution that I can realistically afford: A stacked plate 
> capacitor.  I know a maxwell cap would be nice, but I can't justify to 
> myself or to my family that it is fiscally responsible to pay $600-$800 
> for a 0.025uF 50kV capacitor.  I have hand build every other aspect of my 
> coil with the exception of the neon transformers ($30/each), so why not 
> this part. 
> I considered building the rolled LDPE and Flashing capacitor, but this 
> too seemed cost prohibitive, primarly with regards to time and money 
> spent finding the right materials.  So, I have settled on the following 
> flat plate cap, to fit an existing rubbermaid container that I have 
> already (15"W by 21"L by 5"H - $4.99):

I built one of my stacked caps in a 13" x 21" x 5" Rubbermaid container. They
are excellent, and very affordable. Go with it!

> Dielectric:    25 layers of LDPE sheet (14.75" by 19.2") Each layer is
>                90 mils thick (3 * 30mils)  Total= 5 30mil sheets -at- 
> $12/sheet
> Plates:        26 (12" by 20.2") pieces of Reynolds Heavy Duty Aluminum 
> foil,
>                Three inches will stick out each end of the capacitor for
>                connection to the buss wire.  $1.99 for one 50ft roll.

My recommendation here is to make the long dimension of the aluminum foil
17.2+5+5 inches. The 17.2" is the length that is inside the poly. The first 5"
is to allow for the foil coming UP the entire side of the cap. The last 5" is
to allow wrapping around the flat buss connector. I will explain in detail.
Yes, I know that that is "too much" foil for many of the layers, but believe
me, you will be GLAD you did this by the time you get done. Aluminum foil is
dirt cheap and readily available. 

After assembling the plates in the container, carefully bend all the foil ends
so that they come up the side neatly. I use aluminum flashing as my buss
connector. For 12" wide foil I would use a strip at least 1" wide and having a
length of 12+12+9 inches. 

For 1" wide strips I would make sure that I trimmed the aluminum foil tabs such
that when they are bent over the top of the cap they are 3*flashing width+1".
5" in this case.

After trimming the aluminum foil, mark the flashing at a point 12" from one
end. Lay the flashing down on the trimmed foil (at the very end) such that the
12" mark lines up with one edge. Now flop the flashing and the foil under it
over towards the edge about three times, wrapping the flashing inside the foil
set. Stop when there is about 1" of foil left (you will need a little "play").

Bend the 12" segment of flashing over tightly. You now have 12" of flashing
both inside and outside the foil package. I usually use a hole punch now to put
three holes through the flashing/foil combination. Then I use short bolts and
nuts and washers to secure the whole thing tightly at those three points (ends
and middle). The remaining 9" section of the flashing will be bent over and
punched so that it can receive a bolt of my choosing that will act as the final
outside world connection. I usually use at least a 1/4" diameter bolt with
really big flat washers to make good electrical connections and a nice
compression seal. You can have the bolt either on the side of the case or the
top. I usually do it at the top, but if you do it at the side, you can then
stack caps on top of one another.

I prefer aluminum flashing to wire, because it can handle the RF currents
better, and the flashing is very flexible. Modify the design as needed.

> Fill:          2 gallons of Mineral Oil. $???

Use THIN mineral oil. Transformer oil is the best. Some mineral oil is very
thick, and you want to avoid that, as it makes bubble removal harder.

> The plates will have a 1.375" border on each side, and 2" on the end that 
> does not connect to the terminal.  Thus, there is always at least 2" 
> distance from aluminum to aluminum without going through the PE sheet.  
> Total overlap area of the plates will be 12" by 15.2".  I have choosen to 
> use 90mil of PE, as I will be running at 12kV or 15kV input, and I only 
> want to build this thing once.  
> My calculations tell me this cap will measure between 0.022 and 0.025 uF, 
> depending on whether I choose 2.0 or 2.2 for the dielectric constant of 
> PE. Total cost should be under $100.
> OK, here's the dilemma: I have no vacuum pump, and I don't intend to buy 
> one.  Is this going to be a big problem? Is there a way do get the air 
> bubbles out of the cap without one?  According to my Electromagnetics 
> Textbook, it would take 90,000V peak to punch through 90mil of 
> polyethylene.  Shouldn't I have enough of a saftey margin here that a 
> little air bubble isn't going to break the thing? Also, wouldn't a vacuum 
> pump just crush my rubbermaid box down and break the seal? This thing is 
> going to have a good 2" of airspace above it when I'm done (I could not 
> find a box shorter than 5").

Pumping the air out of such a container would cause the container to collapse.
On my flat plate designs I use one of two methods. Method one is to use
oil-soaked kraft paper at each poly surface, and that includes between the
three layers of poly. (kraft paper will be on BOTH sides of foil). Note that I
said OIL-SOAKED kraft paper. I either soak or paint the paper with transformer
oil just prior to assembly. This ensures that oil gets where I want it. The
size of the kraft paper should be at least as large as the poly sheets. Do NOT
try to use dry kraft paper. SOAK it liberally in oil FIRST!!! Yes, it is
slightly messy, but well worth the trouble!

Method two is to forego the kraft paper and use a paint brush to liberally coat
both sides of each poly sheet as it is built up. No need to coat the foil as
the adjacent coated poly sheets will supply the needed oil.

Methods one and two both ensure that you start out with more oil than air
between sheets. Once the cap is built and the busses attached, then you can
slowly pour in the oil. Cover by about one inch. With time this will subside,
and you will have to add more oil as air is eventually expelled and oil takes
its place.

Pressing down on the top of the cap can go a long way towards getting initial
bubbles out through the sides of the poly. My own personal favorite method is
to put a board on top of the top plate and then apply my body weight on and off
many times by pushing hard against the board and then releasing the pressure
and doing that over and over again. This will expell little air bubbles. Than I
jiggle the container rapidly to break the bubbles loose so they rise to the
top. After about ten minutes of this activity the cap is ready to use! No
vacuum pump required!!

> Right now, I plan to just put as much weight as I can on the cap (to help 
> squeeze air out), fill the box enough to cover the cap with 1" of oil, 
> release the pressure and wait a few weeks for the oil penetrate as much 
> as possible.  If anybody else has a better idea, please let me know!  

No need to wait a few weeks with the above methods. I *DO* however recommend
that you run them at reduced power levels for a few minutes before applying
full power. After the break-in runs, apply pressure and jiggle as before. The
applied AC will cause the plates to vibrate slightly, and this aids in bubble

Check oil level daily for at least the first month, and keep it filled 1" above
the top poly plate.

My Rubbermaid container had walls that were slightly sloped, so I made
plexiglass wedges to help keep the thing centered. Wood is also OK, as the oil
will soak in and make them excellent insulators. If using wood, be aware that
as time goes on the oil will be absorbed more and more by the wood, so you will
have to add oil occassionally. I use plexiglass scrap that I buy by the pound.
Wood will initially want to float in the oil, but plexiglass does not have that
problem. Four wedges are generally adequate, and they do not have to be very
large to do the job.

> -Adam
>  absmith-at-tiac-dot-net
> --------------------------------------------------------------------------
> Adam Smith
> absmith-at-tiac-dot-net
> Epoch, Inc. Digital Music Project

Adam, I hope the above stuff helps you as you build your cap.

There is one additional suggestion that I would make. Instead of using three 30
mil poly sheets between plates, assemble THREE caps using single 30 mil sheets,
and then connect all three in series within the same container. There are a few
additional gotchas that come into play with this method. Obviously you have to
interconnect the three caps. That is easily done by just wrapping the aluminum
foil end tabs together at the proper place. Let's call the caps A B and C, with
A on the bottom and C on the top. Assume AL means Left tab of A, and AR means
Right tab of A.

To connect the three caps in series, use AL as a main electrode. Connect AR and
BR together. Connect BL and CL together. CR is now the other main electrode.
You must ensure that no spark can occur between connected sections. In
addition, the AL electrode needs to be totally insulated from BL/CL section.
This requires some forethought and usually some Poly sheeting to accomplish.

Such a capacitor is more robust, actually has a higher voltage rating, but does
require a bit more attention to detail and insulating.

Hope this helps.
Fr. Tom McGahee