RE: Spark gap voltage transients

To: tesla-at-pupman-dot-com
Subject: RE: Spark gap voltage transients
From: "Lau, Gary" <Gary.Lau-at-compaq-dot-com> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)
Date: Fri, 23 Jul 1999 10:20:05 -0600
Approved: twftesla-at-uswest-dot-net
Delivered-To: fixup-tesla-at-pupman-dot-com-at-fixme
Original Poster: "Robert Alwyn Jones" <alwynj48-at-tesco-dot-net> 


This may be old news to some of you expert coilers. Here it is anyway. 

I believe the Tesla secondary behaves as an open circuit transmission line
due
to it distributed structure. In which case presumable the primary will
behave
similarly except that it s a much shorter transmission line and both ends
will
appear to be shorted to the transients.  So that when the spark gap fires it
excites the transmission line characteristics of the secondary and a damped
osculation will be produced superimposed on what I will call the lumped
parameter response of the L and C combination. OK so what. Well I had
assumed
that what looks like a damped oscillation in the voltage waveform was due to
the stray capacitance and inductance of the interconnections and the C.
Therefore this could be reduced by better connections and a C with a minimum
of
self-inductance. Yes part of the transient my be due to that effect but the
dominate one will be due to the transmission line characteristics of the
primary. I don t think this has any major practical significance to the
design
of a system, although it may be possible to minimise the power loss due this
transient by selecting a particular primary configuration such as spiral or
helical. 

In the above I have ignored the power filtering. Particularly in the
configuration were the spark gap appears directly cross the power input, any
filtering circuits will be exited when the gap fires and the resultant
transient may dominate the above effect. Presumable the configuration with
the
C directly across the input is the preferred configuration for this reason.
I
have always put a bit of series R in the power supplies lines to reduce the
ringing of the power filter. A power supply filter trick is to put the R in
parallel with the L to eliminate its DC effect and reduce its power rating.
The
value being calculated such that the resonance is critically damped. It has
the
disadvantage that the filtering performance is reduced. A lossy core
material
has a similar effect but difficult to design.

Constructive comments anyone.

Regards Alwyn

I have also been thinking about primary circuits from a transmission line
perspective.  No one has, to my knowledge, attempted to match circuit
impedances between circuit elements and interconnects.  As such, there will
be signal reflections at impedance discontinuities.  While I doubt there are
significant performance issues due to this (I could be wrong), it may well
explain the EMI effects that appear to be due to primary circuit features.
Consider a primary coil, tapped 15 feet from the open (unterminated) end.
Imagine the reflections going back and forth there?  Unfortunately my
knowledge falls short of being able to model and simulate such a distributed
parameter system of lossy transmission lines, but that would make for some
interesting simulations.  Does anyone want to chop their primary at the tap
point and see if EMI is reduced?

Regards, Gary Lau
Waltham, MA USA