[Prev][Next][Index][Thread]
Re: More Tuning/Debugging
Richard Quick wrote to Julian Green:
> Whoa. If you put two 11KV windings in series for 22KV out to
> the coil you are talking about some serious power. Not the
> kind of setup you want to be guestimating breakdown voltages
> for, your system had best be designed for 22KVAC inputs, and
> worked up to the peak voltage in stages over time.
Quoting Julian Green <julian-at-kbss.bt.co.uk>:
> Richard, have I over done the voltage? Should I scrap my plans
> or should I proceed with extreem caution. I was going to use a
> variac to limit the output voltage with resistive balast to
> limit the current.
In general when you move up over 20KVAC inputs into the tank
circuit you are looking at sophisticated gap systems, commercial
pulse capacitance, very heavy RF grounds, and proven control and
line filter circuits. You can squeek over a spot or two, but you
can't cheat all down the line. One reason is gap quenching and
control (dwell times and break rates).
If you are over 20KV you will require a minimum gap distance (or
opening) in order to secure a decent break. This goes whether or
not you are using static gaps in series with a rotary. In order
for that minimum gap setting to fire, your input voltage will
have to be pretty high. Quenching the gap once it fires at these
voltages in Tesla systems is a lot of challenge. I have turned
even quality vacuum gap electrodes into copper slag in seconds,
even when they were in series with a rotary. The rotary even
shows signs of arc strikes to the wheel rim, and even tank
circuit arc strikes to the ground shield next to the motor when
the tank circuit did not wish to quench.
Need I mention what a strain this puts on the capacitor di-
electric? Only the best homemade caps, placed in series, need
even think of applying here. I personally would not put my
homemade units on line at these voltages, even though I may have
a few that would stand up for awhile when placed in series.
Freshly made or weak homemade units will fail quickly.
Count on flashovers and safety gap events until you get things
settled in. Loud reports that sound like high powered big game
hunting rifles accompanied by brilliant white-hot flashes are the
norm when an arc jumps a primary turn, strikes a ground rail, or
the safety gap fires at 20KV+. The RF ground system must be heavy
enough to route this energy away from the mains wiring and the
operator. I have seen _neon_ core failures at 12KV (~3KVA) during
flashover events that resulted in HV entering the power cabinet.
An arc blew through the insulation in the 240 volt wiring just
before the line filter mounted between the variac and step-up.
The arc scored the polyurethane coated base board of the cabinet
in a pattern that was the size and shape of a man's hand, where
the "fingertips" terminated along a 1/4" ground strap stapled in
a parallel path to the 240 volt wiring.
Sure you can tone the energy down by loading resistance in series
with the step-up xfmr primary, and refraining from opening the
variacs up... But what is the point here? The best way to keep
the input voltage down is to use a reasonably rated xfrmr to
start with.
> What is the break down voltage of LDPE per mil when used for TC
> work? The best I have so far is 1KV per mil, is this correct?
The rating depends on the film thickness, cleanliness, and
quality of construction. If you are using 60 mil LDPE under oil
and everything is spotless you can get a reliable cap that will
hold up with 9-10KVAC inputs into a spark excited tank circuit.
Thinner films will hold up to proportionally higher voltages
(I know this sounds backwards, but it is quite true), but a tiny
flaw, speck, or air bubble will cause failure, and thinner
dielectrics are more difficult to work with. Capacitor types,
construction, and characteristics are constantly discussed here:
keep you ears open.
> Your reaction to my plans worries me, it makes me think that I
> have overlooked something. Can you provide some advice.
Yeah, start off with a single 11KV transformer. There is plenty
of voltage here to get you off and running. The equipment specs
are easier to meet; meaning everything will be cheaper, easier
and faster to build, and it will hold up longer. Once you are up
and running you can upgrade and test the system a section at time
before you consider a higher voltage. I would go to 14,400 before
I would leap to 22KV.
I have a 22,890 volt two bushing pig. This power supply is used
exclusively on my larger coils with extremely rugged commercial
pulse capacitors and advanced gap systems. Operation of these
setups require that they be either outdoors or in a commercial
type structures.
In my lab I use 14,400 volts ballasted and variac controlled as
you indicated, and this is plenty of voltage. 11 KVAC is very
workable, and is probably where you should plan on starting.
Richard Quick
... If all else fails... Throw another megavolt across it!
___ Blue Wave/QWK v2.12