[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Racing Spark Prediction

Original poster: Paul Nicholson <paul@xxxxxxxxxxxxxxxxxxx>

Dimitry wrote:

> folded resonator as one of the possible cures from racing
> sparks

Since your earlier mentions of folded resonator, I've been
trying to find out what one is.   There are some references
in the archives from about 1998 (before my time on the list)
but I haven't had time to read them yet.

Taking a guess here... folding will probably just shift
around (raise) the existing overtones of the resonator, and
not actually remove any of them.

D.C. wrote:
> These reflected waves are always much more pronounced in coil
> systems that are coupled beyond the critical coupling point.
> They diminish substantially when you properly adjust coupling
> to slightly undercouple the coil.

This issue of 'critical coupling' needs to be examined.  There
is always an additional resonant mode appearing when the primary
is coupled, but when the coupling is below about 1/Q where Q is
the overall Q factor of the coupled resonators, the two
frequencies are no longer distinguishable as separate peaks
during a sweep.   What is happening is that the dual resonance
decays before it can complete a beat cycle.   What D.C. is
calling 'critical coupling' is properly referred to as the
'transitional coupling'.  A k below than the transitional
coupling shows up as a single peak in a sweep of the resonator.
Above the transitional coupling, a dip appears in the centre
of the response and as coupling is increased further, the
two resulting peaks move further apart.  All systems that I've
ever looked at operate above the transitional coupling.

The question is, does coupling relate directly to racing arcs
in some way, or is the elimination of racing arcs just a kind of
coincidence here?  TC systems that I've examined the waveforms
for all seem to be 'overcoupled', eg k between 0.1 to 0.2 and
unloaded Q factors of 40 or more, thus k > 1/Q and so gives a
distinct double humped response to both sweeps and pings.

> The energy will start to divide itself into the two
> off-resonant frequencies and the standing waves begin.

The division of bang energy into the two modes always occurs,
above or below transitional coupling, and there are always
standing waves so long as the overall Q is greater than about 6,
so I would disagree with that part of the explanation.

> Large toroids also reduce the tendency to operate in multiple
> harmonic modes, are at least surpress them to a very low level
> that does not create breakout along the sec coil.

Agreed.  The two main modes are shifted down in frequency by the
applied topload by a greater factor than any of the overtones, so
less of the bang energy ends up in the overtones.

Also: consider transients sent into the top of the coil by
step changes in the top voltage occuring when the breakout
makes sudden step advances.   A large topload will, for a given
size of 'sudden stored charge reduction', produce a smaller
voltage step than a smaller topload trying to support the same
breakout step.

Also: large topload gives a more linear voltage rise along the coil,
so that the highest volts/turn is reduced towards the ideal average
volts/turn given by topvolts/turns.   Any HF ripples travelling the
coil do so on top of the pedestal of the main 1/4 wave voltage
gradient, so the more linear that is, the more headroom the coil
has to cope with transients.

My own feeling is that HF signals are involved in some way or other,
because that would account for the various descriptions of racing
arcs, and also for the (sometimes reported) sensitivity of the
phenomena to small changes in tuning and coupling.  I see a number
of possibilities:-

a) Bang energy going directly into secondary overtones due to
   concentrated (localised, not necessarily high) coupling.
b) Primary overtones coupling strongly to secondary overtones.
c) Energy scattered into overtones due to non-linearity
   of the load (ie transients generated by breakout activity).

Depending on which of these is happening, racings arcs could appear
here or there, or everywhere on the coil.

tssp software is limited to looking at case (a) and when studying
this, I haven't found any particularly sharp sensitivity of overtone
content to variation of k or tuning.

If only we could capture some waveforms from a racing arc coil, we
would have something to go on - we would be able to identify any
significant HF components and determine where they originate.

Paul Nicholson