[Prev][Next][Index][Thread]
New cap info
To All:
Warning: Huge Post!
I just blew up my cap - which had been working just fine for about 8 hours of
cumulative TC use by switching to a rotary gap with my neon. I had heard
that a rotary gap with a neon put a heavy load on the neon and would likely
burn it out, but since I had repotted my neon in oil I figured that it could
handle it and it did. It was the cap that couldn't :(
I then spent 5 or 6 hours (!!) with one of the capacitor engineers (Beau
Meskan) from a company here in Chicago called Plastic Caps which specializes
in High Voltage discharge caps for military and industrial use. He was very
helpful, took me on a full tour of the plant and was willing to share
everything he knew about HV cap design and construction. He was also willing
to put my cap (when I finish rebuilding it) in one of their vacume tanks and
"pump it down" over night along with one of their regular runs. They are
also willing to sell their HV AC caps to tesla coilers at cost! After the
great and informative experience I had with them I don't hesitate to
recommend them to anyone looking for a commercial cap for their TC.
(shameless plug) Their phone number is (773) 489-2233.
Plastic Cap has been experimenting with various materials and designs for the
last forty years and has settled on some general rules of thumb which apply
to high frequency AC caps these are as follows:
1) Polyethylene (LDPE) film is no longer used because it is no longer
available in "capacitor grade" film as of the 1970's.
2) The 1200v/mil breakdown voltage that is reported in the literature is
*only* for "capacitor grade" polypropelene or polyehtelene film. And since
capacitor grade polyethylene grade film is no longer made, the 1200v/mil
figure is no longer correct. The polyproplelne (PP) and polyehtelelne (PE)
materials that are currently available in sheet form from a plastics
distributor have breakdown voltages only in the 450 to 700v/mil range because
of impurities. The breakdown voltage for the PE sheeting that one can buy at
the hardware store ("uyility" plastic) has an even lower breakdown voltage of
about 250v/mil because of even greater levels of impurities.
3) One of the reason that the industry switched from PE to PP was that the
breakdown voltage variation in PP was much less than for PE. For example,
with 4 x 8 foot sheets of commercial grade plastic available from a plastics
distributor, the breakdown voltage of PE (same manufacturer) is roughly
450v/mil on the low end to about 700v/mil on the high end which is a 250v
variation The he breakdown voltage variation, however, for PP is roughly
550v/mil to 600v/mil or only a 50v variation. In other words, the PP is much
more consistent in its breakdown voltage characteristics that PE.
Also, the melting point of PP is much higher than for PE so that if "hot
spots" do occur in the cap, PP is more resistant to melt down than PE.
4) When building capacitors in mineral oil (without a vacume pump to remove
the air bubbles) the plates should be no more than 3 inches long in any
direction (for flat cap designs) or more than 3 inches wide for rolled caps.
This is because dielectric more than three inches wide will end up with dry
spots which will heat and eventually cause cap failure. This is because the
capillary action of mineral oil cannot easily penetrate more than 2 inches in
any direction between plastic sheets. (I noticed this effect in my flat cap
the first build (which has 8.5x11in plates)).
5) If you are determined to use a plate/dielectric size wider that 3 in, you
must then use paper to wick the oil farther between the layers (I conclusion
I came to myself when building mine). This works well up to a 6 inch plate.
Alternatively, If you use use *silicone* oil instead of mineral oil, you
may then dispense with the paper. This will also work well to about a 6 in.
plate.
6) If you use silicone oil (very expensive -at- $50.00/gallon) and paper, you
may then be able to get away with a plate size as large as 8in if you allow
sufficient time for the wicking action of the paper to completely remove the
dry spots.
7) While using paper for wicking has the advantage of elimination dry spots
and therefore eliminates melt through at those spots, it also, unfortunately,
has the disadvantage of raising the dissipation factor (heating effect at
high frequencies) of the total plastic paper sandwich. The bottom line though
is that continuous more gentle heating over the entire surface of the
dielectric (which is what happens when one uses paper) is preferable to the
intense "hot spots" that occur in dry pockets.
8) The best paper to use for wicking is Kraft paper which is available for
capacitor use in various widths. It is called "capacitor grade" Kraft paper
and is *very* thin. Plastic Cap uses 3 tenths of a mil thick Kraft paper
which they feel has the best trade off between wicking action and the
somewhat increased dissipation factor that results when it is paired with PP
or PE. If you can't get "capacitor grade" Kraft paper, use the *thinist*
possible regular paper you can find.
9) The Kraft paper should be interposed between each layer of PP or PE but if
this is impractical then at the very least it should be used between the
*shiny* side of the plate (aluminum foil or flashing) and the next layer of
PP or PE. The shiny side of anything, apparently, impedes capillary action
while "dull" surfaces (slightly rough) encourage it.
10) The cap, whether rolled or flat, should be built in several sections
connected in *series*. This lowers the voltage that occurs across each
section. The voltage at which corona begins to become a problem in cap
construction is approximately 1200v. This means that each "section" should
not be forced to see more that 1200v. A 15kv power supply, for instance,
would require at least 12 or 13 sections (smaller caps) wired in series.
This combination series/parallel design makes calculating the number of
plates for proper capacitance a bit harder but is well worth it since corona
(and the heat and loss produced thereby) will be eliminated resulting in much
longer life and better performance for the cap.
12) The aluminum foil available in the supermarket has many impurities such
as sodium (the worst) that causes "hot spots" on the aluminum. Once again
"capacitor grade" aluminum is the best but in lieu of that, thicker
dielectric and the use of polypro Propelene (which is more temperature stable
than polypro Ehtelyene) is acceptable.
13) When using non "capacitor grade" materials expect heeting of the
dielectric (PE/PP), plate (aluminum) and wick (Kraft paper). Leave plenty of
room for expansion of the oil in such a system or explosion is likely.
Whew!!! that was a long one.
Hope it is usefull
-DavidF-