Re: Optimum toroid size

["Jim Lux" <jimlux-at-jpl.nasa.gov> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

I would think that there is sort of a balancing act going on, as you've
addressed.

you want the topload overall diameter big enough to shape the field along
the secondary so it's even, preventing breakdown from the middle or top
windings. 

You want the topload overall diameter big enough to give enough topload C
to store enough energy to make the desired size spark. (this implies that
you have lots of primary C, otherwise the voltage will drop).

You want the voltage high enough on the topload so that the field is
appropriately nonuniform, so you get leader formation, as opposed to an
overall corona.

You want the ROC big enough so that there isn't a field concentration at
the edge of the topload all the way around.

You want a small ROC at just one point, to start that leader forming.

I suspect that the way to approach it is to design for a given topload size
and voltage that can develop the spark length you want.  Then, size the
primary capacitor so it holds that much energy (and then some, to account
for losses).

Tesla list wrote:
> 
> Original poster: "Lau, Gary" <Gary.Lau-at-compaq-dot-com>
> 
> The discussion so far has been how to maximize top-load voltage prior to
> breakout.  Increasing top-load size results in higher radius of curvature
> (ROC), which increases breakout voltage, while increasing top-load size also
> decreases maximum attainable secondary voltage by virtue of conservation of
> energy Esec (0.5 CpriVpri**2) <= Epri(0.5CsecVsec**2).
> 
> I have to wonder though whether maximizing pre-breakout top-load voltage is
> a necessary condition for optimum performance.  Many, if not most of us use
> a breakout point on our top-loads.  They serve to direct the streamers to
> where we want them, force breakout to occur, and also to concentrate all
> available energy into (hopefully) one long streamer.  But it would seem that
> the use of breakout points renders the ROC consideration a mute point.
> 
> If I have a toroid with a breakout point and streamers don't occur, then I
> conclude that the top-load capacitance is too large.  If I have a toroid
> with a breakout point and *multiple* streamers occur (I want only one), then
> I conclude the top-load capacitance is too small.  But I don't believe ROC
> of the top-load is a major consideration.  Is this thinking incorrect?
> 
> Gary Lau
> Waltham, MA USA