That Marx-like primary idea

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "K. C. Herrick by way of Terry Fritz <twftesla-at-qwest-dot-net>" <kchdlh-at-juno-dot-com>

On 9/20/02 I posted http://hot-streamer-dot-com/temp/marx-pri.gif showing my
notion of a Marx-like primary configuration.  Since then I've been
playing around with Newbury Technology's SiMetrix simulation program &
have come up with http://hot-streamer-dot-com/temp/marxpri1.gif,
http://hot-streamer-dot-com/temp/marxpri2.gif and
http://hot-streamer-dot-com/temp/marxpri3.gif.

...pri1.gif is the schematic of a set of 4 capacitors that is resonated
at 60 Hz, in parallel, with a series inductor.  On every 3rd 60-Hz
half-cycle, the capacitors are discharged in series via spark-gaps. 
...pri2.gif shows the waveform across the capacitors and also the
waveform across the simulated primary inductance.  ...pri3.gif is an
expanded portion of those waveforms, showing the primary-inductance
waveform in detail.

In the schematic, V1 is the mains voltage which I set at 150 V peak.  C1
thru C4 are put in parallel, for 60 Hz via L3 thru L8, and
series-resonated at 60 Hz with L1 (with R1 in the circuit).  The
series-discharge circuit consists of C1 thru C4, spark-gaps S1 thru S4,
simulated primary inductance L10 and circuit-resistance R2.  S5 and S6
are hooked up to simulate a back-to-back zener diode across L10--merely
to clip any turn-off spike greater than 2 KV so as to keep the graphing
manageable.

V2 pulses the gaps on every 25 ms for 200 us.  My complete notion
incorporates gap-triggering using a trigger-coil but I haven't gotten to
simulating that yet.

My complete notion also incorporates 24 discrete capacitors (sets of
capacitors), and 23 gaps, rather than the 4 that I show here.  Those
capacitors, in parallel, give ~31 uF of capacitance to resonate with ~227
mH of inductance at 60 Hz; and in series, ~1.3 uF to resonate with the
primary coil.  In the circuit here, I take liberties with the values &
the configuration just to show how it is supposed to work.

You will note, in ...pri2.gif, the 60 Hz resonant-rise up to ~600 V at
each 3rd half-cycle.  The voltage then very rapidly declines to near-0
during the time of resonance with L10.  In ...pri3.gif, I show the first
such occasion expanded.  Notice that the 1st half-cycle peak of C1's
voltage is ~300 V while that of L10 is ~1200 V (4 x 300).  Just what one
would expect!  In the real thing, that 300 V will get multiplied by 24 to
yield 7200 across the primary coil.  Not taking into account, of course,
gap- and circuit-losses which here are simulated with only 5 m-ohms in
each "gap" and 50 m in addition.

Now if I could just simulate the sparks-in-air themselves, I could do
without the bother of building the thing.  As to building, I have it "all
worked out", CAD-designed, on paper & just have to motivate myself enough
to get started.  I decided not to switch capacitors in & out for tuning
but instead to utilize a flexible copper-ribbon turn in the primary
inductor which I can expand or contract, by means of a lead-screw, to
change the tuning.

Comments appreciated...

Ken Herrick