Re: High-Bandwidth Primary Circuit Behavior (fwd)

---------- Forwarded message ----------
Date: Wed, 29 Apr 1998 00:57:56 -0600
From: terryf-at-verinet-dot-com
To: Tesla List <tesla-at-pupman-dot-com>
Subject: Re: High-Bandwidth Primary Circuit Behavior (fwd)

Hi Gary,
        Thanks for you comments!  I never expected the spark gap to be as
complex and interesting as all this.  There are still many details to be
worked out.

>From: Gary Lau  28-Apr-1998 1428 <lau-at-hdecad.ENET.dec-dot-com>
>To: tesla-at-pupman-dot-com
>Subject: Re: High-Bandwidth Primary Circuit Behavior
>>From: terryf-at-verinet-dot-com
>>Hi All,
>>        I have been spending most of my time looking at primary circuits with
>>my probe system.  As a result, I have written another paper that tells about
>>what I have seen so far.
>>The HTML and the original Word97 documents are at:
>>  http://www.peakpeak-dot-com/~terryf/tesla/experiments/experiments.html
>Fascinating stuff!  It would seem that the simple models we've assumed
>for a gap's behavior, being one continuous conduction for the duration of
>the primary ringdown, is far more complex, opening up with each
>zero gap current crossing and ringing each time.
>For the experiment involving "good" and "bad" tank wiring, did you scope
>the ringdown envelope for each?  If the "bad" envelope is much shorter,
>that would quantify and add even more weight to the obvious need for
>"good" tank wiring.

There are two pictures of the "good" and "bad" wires.  There are four scope
pictures before those.  There first two show the "good" wiring ring down and
the next two show the "bad" wiring ring down.  It would appear from those
pictures that the bad wiring ring down is about 25us shorter.  The bad
waveform was rather jumpy so it was hard to tell.  One thing I need to look
at is the harmonic content of both waveforms.  I suspect the noisy bad
wiring case may also have less energy in the fundamental waveform.  This
means that the secondary would not be able to receive as much energy.  When
the paper was converted to an HTML document it lost a little clarity as to
what each picture represents.  Sorry about that.

>During the gap zero current crossings and resulting HF ringing, I can't
>tell what the gap voltage is doing.  Does this have implications towards
>how one ought to protect their NST's?

The gap voltage stays very low and "tame" after the inital spike.  I used a
cap across the neon and 10K resistors that should protect it well.  I
haven't actually checked this, however.  

One interesting note:  I tried a simple multi gap (5 gaps) spark gap and the
initial spike went away!  Unfortunately, the losses went sky high.  Mutigaps
seem to have huge losses.  Still more to look at there.  I am surprised so
many people have such good luck with them.  Perhaps there are unknown
properties with those gaps that are benificial dispite their poor looking
performance.  Ie. the harmoics may be closer to the fundamental.

>Now, if only someone knew how to create a Spice model for this mysterious

I have been trying for the last four days.  I can use a bunch of switches
and get pretty close but spice was never made for this.  The high current,
high speed, very non linear stuff pushes the program to the limits of
stability.  It would help if I wasn't using the free demo version :-))
However the effects can be repeated on the computer.  The model is far from
elegant, however.

>Gary Lau
>Waltham, MA USA