SV: NST resonance

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

Original poster: "Jan Ohlsson by way of Terry Fritz <twftesla-at-uswest-dot-net>" <jan.ohlsson-at-mbox319.swipnet.se>





> Original poster: "Malcolm Watts by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <m.j.watts-at-massey.ac.nz>
> 
Measuring transformers of 
> any kind with a scope and sig gen is an exercise fraught with 
> difficulties and traps and you have to be aware of these in order to 
> apply the correct techniques for the transformer you wish to measure. 
> The most insidious problem that will cause trouble is distributed 
> capacitances in the windings giving resonances and false readings, 
> particularly if the windings are unloaded. 

Yes, perhaps I am not measuring anything that has relevance. But it is not
easy to say whether the NST is loaded or not in a live Tesla coil
situation. If the spark gap is conducting there is definitively a load, but
what is the case the moments when it is not? Perhaps it could be argued
that these moments are so short that the resonances in the lower audio
spectrum that I was measuring have no time to build up.

On a more practical level I had great problems when starting my first coil
with the safety gap taking over the main gap regularly after a brief moment
every time I started the coil. The safety gap was set so that it never
fired before the main gap without the primary connected. I am an audio guy,
so it took me a while to conclude that the only reason why there could be a
vastly different voltage at different points of short and very heavy
conductors was that there were large amounts of RF of at least a couple of
hundred MHz present to cause standing waves on the short connections. 

I then built a simple RC filter with non-inductive resistors in series with
the primary and capacitors of  500 pF to earth across the NST. That should
take care of the RF, I thought. The safety gaps still took over the main
gaps. I did not understanding anything at first and was at the brink of
giving up the whole coil project after having fried one NST after beeing
stupid and having widened the safety gap. 

Then I tried the same filter across the NST without the primary connected.
The safety gaps still arced violently. It was then I realized that it was
probably two mechanisms at work at the same time , RF feeding back from the
primary and NST resonance with my filter caps. I tried series resistors
between the NST and the filter, and voila, the arcing disappeared. 

That was a bit puzzling, so small capacitors should not cause resonance,
according to what I had read  on this list and elsewhere. That is what
prompted me to try to measure the resonances.

Right now I am building a 12 dB/octave RF-filter consisting of two RC links
in series. If  there is RF of several hundred MHz, that would be a much
more efficient solution than just one filter. I donīt really know anything
about the frequency content of the RF for sure, it is just a speculation
based on my observations of the arcing problem before I had any filter at
all. I donīt have any RF instruments at the moment, just instruments for
measurements in the audio spectrum. 

My goal is to use a bank of  seven 8/45 NST:s, each with a series resistor
before beeing parallelled and connected to the filter, to get rid of the
NST/filter capacitor resonances.

Regards,
Jan