Help with Custom Cap thoughts??
From: Barton B. Anderson [SMTP:mopar-at-mn.uswest-dot-net]
Sent: Friday, April 03, 1998 9:00 PM
To: Tesla List
Subject: Re: Help with Custom Cap thoughts??
Tesla List wrote:
> From: djQuecke [SMTP:djQuecke-at-worldnet.att-dot-net]
> Sent: Wednesday, April 01, 1998 4:53 AM
> To: Tesla List
> Subject: Help with Custom Cap thoughts??
> Hello Coilers,
> I wrote to the list several months ago about a material I want to try for
> building custom capacitors. As usual, I have taken what seems to be a
> fairly simple task and turned it over in my mind for so long that I'm now
> thoroughly confused about the basics.
> If you would bear with me I'll try to explain myself and perhaps get a
> couple of rather simple questions answered.
> The material is called ALPOLIC. It is maybe a 3/16" thick piece of
> polyethylene sheet with perhaps a 1/16" thick sheet of aluminum bonded to
> each side. It comes in 4' x 8' sheets and you can use wood working
> equipment to cut it, etc.
> Pluses: zero air gap between poly and AL. Poly is thick enough to withstand
> high voltage. Works fairly easily.
> Minuses: The AL sheets have a special paint bound to their surface which
> doesn't remove easily. Poly perhaps too thick. Electrical connection
> between plates has me confused.
> The following ASCII drawing (sorry) shows 3 plates connected together. The
> easiest manner to connect them would be with UMW rods (E1 and E2) and use AL
> bushings between plates for the necessary electrical connection.
> Assume the conductive AL plates C2 and C3 as well as C4 and C5 are connected
> by AL bushing placed in between them. The UMW rods slide thru the centers.
> Each plate will have two holes 3/8" diameter for the rods to bind them.
> I'll remove a 5/8" ring of AL on center with each 3/8" hole to avoid arc
> thru holes. Between twinned plates I will place a 1/2" ID x 5/8" OD x 1/4"
> thick UMW bushing over the 3/8" UMW rod and then place a 5/8" ID x 3/4" by
> 1/4" OD AL bushing over that.
> I'll remove the paint only in a 1" circle center on the holes between
> twinned plates.
> Each plate is about 5" square. I'll remove a 1/4" strip of AL along all
> edges on each side of plate to avoid arcover around edges.
> I'll remove paint around holes/rods on end plates and connect leads to each
> end. The UMW rod ends will be threaded and I'll use nylon nuts to hold it
> all together.
> In this case then the conductive plates C1, C4 and C5 are positive plates
> and plates C2, C3 and C6 will be negative (don't forget these "conductive"
> plates are actually bound to the dielectric plates and it can be confusing
> to refer to them separately).
> This requires that an odd number of plates be connected together so you a
> pos and neg end plate.
> I figure 25 plates per cap. Should be about 12-13" long, 5" square around.
> I'm hoping for .0025uF per cap and a voltage rating of at least 50kV and
> perhaps actually more like 100kV???
More like 0.0043uF at 19KV. It's wise to design for 2.5 times the max voltage
you will be running. You could run 7kv ok, but I wouldn't push it beyond this.
> I've got to get a micrometer and find out how thick this stuff actually is.
> Q1: Does this have chance of working? <g>
> Q2: In this scenario do I get to count the surface area of both conductive
> plates C2 and C3 or just one because they are in reality joined?
Just one. It's the "area" over the dielectric, not the thickness that counts.
> Q3: Do plates C2/C3 have to be physically connected to the end plate C6 and
> does C4/C5 need connected to C1.... or does the current pass thru C1 then
> D1, C2 & C3, D2, C4 & C5, D3 and finally C6??
C2/C3 are physically connected and must be physically connected to C6. Same with
> Q4: Doesn't my design actually just connect 3 capacitors (plates) in
> parallel? (I hope it is really that simple and I've just made another
> mountain out of a molehill, my specialty.)
No. Your design makes one capacitor (as it should). For a parallel config, C1-C3
should be connected and C2 separated from C3. C2-C4 should be connected and C4
separated from C5, etc...
> Q5: Assuming it works at all, is there a chance they will work without an
> oil bath. I just don't see the necessity for oil since AL is bonded to poly
> (no air gaps there) and corona would have to jump either 11/16" inch of air
> around edge to avoid 3/16" poly or jump 7/16" thru holes to avoid the 3/16"
> poly. Or is that what can happen by "surface effect" and the oil somehow
> alleviate that??
They would probably work without an oil bath, but I wouldn't recommend it. I
think corona would still dominate causing potential flashovers. Air can be
ionized easily. Oil removes the air and dissipates heat. Two important quality's
for a long lasting cap.
> Q6: I got this material though scraps at a construction site. I need to
> check pricing but wouldn't it be really easy to just buy a 4' x 8' sheet and
> rout off a 1/2" strip of the AL on both sides, place one electrode on each
> side at opposite ends, get about .0085uF and call it finished?? A tad
> unwieldy but hey, I like the labor level......Of course you pole pig types
> will need to stack a half dozen of these....
Yes, it would probably work. I can't imagine the size not being a real bother
though.Good luck with this design if you pursue it. Keep us posted how it works