Re: More Coupling...

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

Original poster: "by way of Terry Fritz <twftesla-at-uswest-dot-net>" <paul-at-abelian.demon.co.uk>

Terry Fritz <twftesla-at-uswest-dot-net> wrote:

> The effects of the conical primaries, as opposed to the flat, are
> very interesting too.  We now try to avoid cone primaries since the
> higher coupling is believed to help create racing arcs. However,
> perhaps we are wrong and the people who have preferred cone
> primaries for that last 100 years knew what they were doing
> after all ;-))) 

If we adopt the hypothesis that a cone primary spreads the induced EMF
more evenly along the coil than say a flat primary does, then the
tentative evidence from acmi profiles would support this. This might
offer a way to achieve higher coupling before breakdown, or could help
cure a breakdown problem by redistributing induced EMF without loosing
coupling coefficient.

> If higher coupling can be achieved while avoiding racing arcs with
> cone primaries, the efficiency of a coil can be improved. 

Yes, and perhaps further improved by the higher primary self
inductance of the cone leading to a better primary gap efficiency.

However I should point out that the cone primary may not always be
appropriate. A low h/d secondary at high elevation may well exhibit
a peak voltage gradient point (of the free secondary resonance) at say
35% along the coil. The peak gradient point of induced EMF from a flat
primary would be comfortably below this. Switching to a cone primary
might cause a problem in this case by raising the point of primary
induction peak gradient. At the same time, the additional external
capacitance presented to the secondary by the primary cone would tend
to reduce the height of the gradient peak of its free resonance, and
thus the two stress points converge.

So it is not a case of one type of primary or another, but is a
question of choosing a primary shape which will tailor the induced
EMF profile to make the best use of the secondary.

> Although an order of magnitude beyond that, it would also be neat
> to have it do the full arc to ringdown transient responses of
> everything.  In otherworlds, the voltages/current from the moment
> the gaps fires to when it all rings down (maybe with a streamer
> load!?) as a function of time. Perhaps theoretically not to tough
> in light of all this new data but the computing power needed to do
> that would be "big" ;-)) 

Yes, monstrously big - I tried it once and my computer was on its chin
strap - and that takes some doing!! 

> Since it is Christmas I figure I can ask for anything ;-))) 

I'll write on tssp list with some suggestions as to how to make next
christmas a happy one!  Interesting times...

Regards All,
--
Paul Nicholson,
Manchester, UK.
--