[TCML] Racing arcs and primary resonances

[Paul Nicholson <tcml88@xxxxxxxxxxx>]

Prompted by Brandon Hendershot's report of racing arcs, I set
up a model of the coil using tssp software in order to estimate
the amplitude and voltage gradients produced by HF transients.

A secondary voltage/current animation is here

 http://abelian.org/tssp/tmp/bh1.anim.110402.gif

Radial and axial voltage gradients along the secondary,

 http://abelian.org/tssp/tmp/bh1.grad.110402.gif

There's a small transient occurring during the first 3uS but
this doesn't seem to produce much voltage gradient, with the
axial gradient peaking at about 5km/cm.

I have to conclude that this type of transient (which is an
unavoidable by-product of non-uniform coupling to the secondary)
isn't strong enough to account for the racing arcs.

In fact, the only problem apparent from the model is the high
radial voltage gradient near the top of the secondary, caused
no doubt by the topload being too far above the coil.  This is
surely high enough to produce breakout from the top turns.

There's also a high gradient at the bottom of the coil due
to the primary proximity, but this is normal and is usually
a highly stressed part of the secondary.

Back to square one then with the racing arcs.

The report describes arcs moving all over the secondary, having
length about 1/3rd to 1/5th of the secondary length, occurring
more or less uniformly on the surface but with a preference
for the top 2/3rds.  This sounds very much like high frequency
activity in the secondary.  We might guess from the description
than the frequency is around 1Mhz or upwards.

The arcs occur when the variac is turned up to 140 or 145V
with the static gap beginning to fire at about 80V setting.

Since the secondary response doesn't seem to naturally develop
any problematic transient HF, I think attention has to be
turned towards the primary and spark gap as the source.

What can be going on with the arc and the primary to produce HF,
especially when 'overdriven'?

The primary will have a spectrum of resonances which will
receive energy from the bang - these are usually visible on
measured waveforms as a bit of 'scruff' lasting a short time
after the gap fires, and although brief they are sometimes
quite strong.  These are not modelled in the animations.
They occur in the range 1 to 10Mhz or thereabouts, depending
on the size of the primary, and could couple quite well to
the secondary.   They can be modelled by treating the primary
as a transmission line rather than just a coil.

For an example, see the measured primary current waveform in
Marco Denicolai's Thor System - 2nd waveform plot in

  http://abelian.org/tssp/md110701/

There is strong HF primary oscillation which decays by 5uS
and has components in the range 2Mhz to 7Mhz.

Presumably all primary and gap combinations will produce some
amount of HF, but only in a few cases is enough coupled to
the secondary to show up as racing arcs.  In the troublesome
cases, it may be that one component happens to coincide with
a secondary overtone, or there is excess HF production due to
some behaviour of the arc and primary transmission line.

More questions: Does anyone have anything like a spice model
of the non-linear gap behaviour?

When the variac is turned up beyond the level required for
normal firing, without opening out the spark gap, is this
likely to excite extra HF oscillation, perhaps through negative
resistance arc behaviour interacting with the spectrum of
primary resonances.

Are there any other coils out there that produce this type of
racing arc?

I think we need more measurements to know where to go with this.
Perhaps we can have a look at some primary current waveforms
and look for the HF components.

--
Paul Nicholson
--
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla