Re: Suggestion on Power Supply?

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

Original poster: "Terry Fritz" <teslalist-at-qwest-dot-net>

Hi Jeremy,

At 11:57 PM 5/2/2003 -0700, you wrote:

>I've got a secondary coil that's 25.5" long by 6.625"
>in diameter, 22 AWG.
>
>I've got ducting for toroids 6, 7 and 8 inches cross
>section, and about 10 Maxwell .03uF 35KV pulse caps
>that I can arrange in any order for just about any
>capacitance or voltage. The primary is now going to
>be 2" brass strips totalling .045" wide, approximately
>15 turns max.
>
>I've ran all the simulations electronically in
>microsim, my Tesla perl module, as well as JAVATC,
>and any number of possibilities exist for topload,
>power and tank capacitance. JAVATC says I can
>expect sparks up to 9 feet in length, which would be
>just dandy. (does this program consider resistance
>from the spark gap?)

The spark length is pretty much determined By John Freau's formula:

L = 1.7 x SQRT(power)

So for 9 foot arcs:

9 x 12 = 1.7 x SQRT(P)  P = 4040 watts input power.


>Of course, in the real world, inputs aren't as easily
>chosen as they are for simulations.
>
>The two major remaining variables in my design are
>selection of the power supply and spark gap.
>I initially wanted two 15Kv/60mA NST's in parallel,
>but decided that was not going to be enough current
>to charge a .03uF tank capacitor. From my
>calculations, I figured it would take at least
>15Kv/270mA to charge the capacitor in 8.3 ms.
>(AC half cycle), which I think can only be handled
>by a pole pig. (that's 4KVA) A PT just won't do
>because it'd always be run over it's ratings.

If you want 9 foot arcs you need 4040 watts.  4040 / 15,000 = 270mA as you 
say.  That would be five 15/60 NSTs but you might as well go for a 5KVA 
Residential Distribution Transformer (Pole Pig :o))

Watts = Joules / second

Joules = 1/2 C x V^2

Assuming a 15kV transformer, you will get 21000 volts peak.  Thus a 30nF 
cap will have...

1/2 x 30e-9 x 21000^2 = 6.615 Joules for each firing.  Just divide to find 
the number of firing per second needed to get 4040 watts

4040 / 6.615 = 610 BPS  However.........  This would be an asynchronous 
gap.  It will fire over the average of the voltage not just at the 
peaks.  Therefor you need to fire at twice that BPS to get the power you 
need or 1220 BPS


>Now, the biggest circuit in my house is 240VAC wired
>to a 30Amp double breaker, which provides 7.2KVA to
>a sauna heater rated for 6KVA. (Hey, this could be
>a ballast too!, now where'd my ammeter go....)

I think just a 500 foot roll of 12 gauge wire will work too???  Might be 
easier and cheaper.  You may be able to put in a 50 amp breaker too, but be 
carful not to violate you panels ratings or anything.  But your sauna 
circuit sounds fine.


>I guess I've been a little hesitant to make the
>move from 120 to 240...but I think my setup warrants
>it whether i like it or not :) It may actually be
>safer because running wires for 240VAC -at- 30Amps
>is a lot less riskier than running 120VAC -at- 60Amps.
>(I am afraid of wires burning in the walls...)

Your 12 gauge wall wiring will not carry 60 amps for long ;-))  You need 
like 6 gauge.


>So, should I try and get 240VAC 5KVA pole pig or
>reduce my capacitance to something smaller? (Which in
>turn would affect the frequency, which would affect
>the topload, amperage->spark gap etc...)

You will need the pig (I mean Residential Distribution Transformer, pigs 
are what bacon is made out of :o))  You need about 4000+ watts if you want 
9 foot arcs.  That is a given...  A 5KVA PDT can actually do like 10KVA for 
a short time so it will be all the power transformer you ever need.

>
>Also, I would like to run the coil with one of my
>15KV/60 NST's -- or even smaller, as a method for
>low power tuning and adjusting. Then, switch a few
>cables around and run it at whatever it wants :)
>(270mA...) Does anyone do this, surely you don't
>go full blast from the start...

Hooking up a 15/60 NST with a cap is trivial compared to everything 
else.  But a variac may do it all for you anyway.  I hope you have a variac??


>Next, I've got a bunch of 1/8" thoriated tungsten rods
>left over from my first rotary spark gap. Are these
>too small for the power levels involved now? I'm not
>worried about them melting or anything like that, it's
>just that 1/8" seems awful puny to carry 500 Amps.
>I've seen SRSG's made with them before, but I would
>think their "on" resistance is high...why have a large
>capacitor when it's energy is going to be choked by a
>high resistance gap?
>
>Would 1/4" tungsten be better?

1/4 inch would be better.  The 1/8th really is small and will get pretty 
hot.  Many gaps have big brass mounts and such to protect against the 
heat.  You could use the 1/8th, but it sounds like you are going for a 
pretty nice coil with plenty to spare, so no reason to use 1/8th inch.


>I have an idea for rotary gap that involves this:
>four 5/16" diameter tungsten balls pounded into a
>1/4" thick sheet of G10. They're not flush, 1/32"
>sticks out from the disc on both sides.

I think John mentioned that they will get too hot and carbon arc so near 
the surface.  You really need the electrodes to stick way out in the wind 
to cool them.

>This rotates
>between four stationary electrodes, two of which are
>connected to each other. For the stationary electrodes
>I would like to use the 5/16" or 1/4" TIG welding
>electrodes that screw in to torches. (They have small
>.025 diameter holes in them), but I am unsure of what
>they are made of...it looks like copper, but I don't
>think it is...does anyone know? I used them in a lower
>power coil once before, and they appeared to work, but
>I never ran it for very long or at high power.
>
>The other option is a sucker gap...but that works best
>with resonant sized capacitors...which brings us back
>to the capacitor/power question...

At your power level, the rotary gap is well tried and trusted.

PDT (pig) systems really don't have a "resonant" cap value.  Since you can 
control the current limiting, you can drive a very wide range of caps.  You 
could probably consider like 150nF! if you wanted.  At 120 BPS sync your 
cap could be:

  P / BPS = 1/2 x C x Vfire^2       4040 / 120 = 1/2 x C x 21000^2   C = 153nF

Cheers,

         Terry



>Hope that's not too much! Have a good night everyone!
>



Cheers,

         Terry