Re: Static Spark Gap

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

Original poster: "Stephen Conner by way of Terry Fritz <teslalist-at-qwest-dot-net>" <steve-at-scopeboy-dot-com>

At 11:27 25/04/03 -0600, you wrote:


>         I don't understand why your transformer got very hot unless you had
>your gap set much wider than that required to "just break down" with
>rated input voltage and no capacitor across it.

My gap was set in just the way you described, plus 0.025" for luck ;) I 
think it fired around 15-17kV. The transformer was a 10/25.

>What's happening to
>cause the low spark rate is that the voltage across the
>transformer/capacitor is increasing due to resonance and, since the Q is
>of the order of 15 or 20, it takes a number of cycles to build up to the
>gap firing voltage.

This is what I believe happened: I set the variac so the voltage in the 
steady-state was just less than the gap firing voltage. It took about 10-20 
cycles to ring up to 15kV peak or so, however the gap didn't fire 
immediately, instead it stayed at that voltage for a second or two (because 
power input equaled losses) until some random event (Cosmic ray etc) made 
the gap fire. So the cap spent most of its time with 15kV peak AC across 
it, hence (from conservation of energy) a correspondingly high current was 
in the NST secondary.

I calculated that the circulating power was about 400 VAr which is 4 times 
the rating of the transformer in neon service (by max power transfer 
theorem: rating=10*25/2=125VA)

>I wonder if the transformer failure was due to too
>wide a gap?  Was the failure a short in the secondary?

You guessed it :( I think it died due to my poor spark gap design with very 
uneven firing voltage, and the resonant tank cap that allowed stupid high 
voltages if the spark gap failed to fire. You can see that poor gap design 
at http://www.scopeboy-dot-com/tcspec.html for an example of how not to build a 
multi-section gap

Steve C.