CAP DESIGN (FWD)
* Original msg to: Kukkonen-at-snakemail.hut.fi
KU> I decided to mail you this message again, as I haven't heard of
KU> you for a while..
My mail was spotty for a week. There are still some bugs in this BBS.
KU> I just re-read your description of the rolled 20nF 12kVac cap
KU> a few times again.. I'd like to ask you about a few little things
KU> on the design:
KU> Does one not roll the poly-Al-poly-Al sheets on some tube?
KU> Why is the cap so large in outer diameter? Is this to enable
KU> efficient oil circulation?
I am not sure what you mean about the finished capacitor roll being
a "large diameter". The finished capacitor roll is less than 5 inches
in diameter (12.7 cm) and neatly fits inside of a cylindrical tank
6 inches in diameter (15.24 cm)
KU> What is the inner diameter of the rolled sheets?
About 2 inches or 5 cm.
KU> You do think that one cap like that will do for my purposes -
KU> using it with a pile of 10kV oil-burner x-formers..
Yes. This design is perfectly suited for 10 kV AC rms operation in
a spark excited Tesla tank circuit. The cap does require some
sitting or "rest" time before firing, and should be brought up to
full voltage slowly using an variable voltage autotransformer
(variac) on the primary windings of the step up transformer.
KU> Another text-file about making hv-caps states that oil will
KU> diffuse through the pvc - have you perceived this effect?
I build polyethylene caps, and there is no diffusion through this
plastic by the oil. PVC is slightly permeable.
KU> You mentioned that one should wash the ldpe-sheets with
KU> alcohol - I've found that pe does not endure methanol -
KU> is iso-propanol suitable?
Iso-propanol is fine.
KU> I really haven't found any reasonably priced paraffin oil
KU> to be used for filling the cap.. I guess that standard
KU> Shell Diala x-former oil will do.. (?)
KU> If I won't be able to find any ldpe with sufficient thickness -
KU> could I use hdpe-plates (1,5mm thick) and make a stacked cap
KU> of them - if (as:) so, would this assembly sound reasonable:
KU> I'd use two 6 meters long (16 cm wide) sheet of .1 mm copper-foil
KU> and a lot of 18*18cm hdpe-plates and use "continuos" plates aka
KU> I'd connect the foil from all four directions - with two of
KU> these connected together and the others together too.. Did you
KU> get the picture?
You can use two layers of thinner ldpe to construct a rolled cap.
Flat, stacked, plates of hdpe may be mounted sideways in some sort
of tank (glass fish tank or aquarium possibly?) and covered with
oil so that air can bubble out from between the plates, or; the
flat stack may be placed in a sealed tank (plastic food or storage
container with snap lid) and pumped down to a fairly hard vacuum
then backfilled with oil. Somehow, air must be removed or allowed
to escape from between the plate construction, either by natural
bubbling or forced pumpdown. Tiny trapped bubbles of air will
form hot-spots on the dielectric and will cause failure.
KU> (about the primary)
KU> What would be the relation between the "angle" of the cone
KU> and the size of the toroid? I might as well make a bigger
KU> toroid as that is not difficult..
This is one of those things learned from experience. A larger
toroid allows more energy to be processed by the secondary. If
you use a larger toroid, one of the ways to take advantage of the
potential gain is to snug up the coupling between the primary and
secondary. By lifting up the outside turns of the primary coil,
the coupling between the primary and secondary is increased. This
tighter coupling forces more energy into the secondary and gives
better power processing and greater resonate rise.
KU> What exactly have you used as a frame for your primaries?
KU> I've thought about fiber-glass tubes but they are not suit-
KU> able for a cone-shaped form as they can't be bend..
I cut some strips of plastic. I cut grooves into the plastic
strips where the primary conductor is layed, and held, into
place. These plastic strips are like combs, with spaced slots.
These plastic combs may be glued onto wedges of plastic giving
the correct angle. The plastic wedges with attached combs may
then be glued to a flat circular base.
KU> I used Teslaci.exe to approximate what a 10kV 80mA tran-
KU> sformer could run and the program stated that 18 nF would
KU> be maximum.. Does this sound reasonable?
KU> (I'd need 6 oil-burner transformers to use 20nF)
My tried and true calculator tells me that this power supply
will fully charge a .0212 microFarad capacitor at the rate of
120 times per second (if using 60 cycle line input). Suffice to
say that the 10kV 80mA power supply will work great with one
of the .02 uF rolled caps that were detailed in my CAP.TXT
file. It will also work fine with a .018 uF (microFarad) cap.
... If all else fails... Throw another megavolt across it!
___ Blue Wave/QWK v2.12