Re: On sparks

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

Original poster: "Metlicka Marc by way of Terry Fritz <twftesla-at-uswest-dot-net>" <mystuffs-at-orwell-dot-net>

jim, all

Tesla list wrote:
> 
> Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-uswest-dot-net>"
<jimlux-at-earthlink-dot-net>
> 
>
> 
> The "optimum" long spark waveform might be a big spike at the beginning (to
> cause breakdown), and then substantial low impedance current to flow along
> the leader.  The end of the leader is pretty "pointy" so it probably doesn't
> take much voltage to create a big enough field there to keep the breakdown
> going, especially compared to the "top load" which has a big ROC, and needs
> a fairly big voltage to even start to breakdown.

isn't this what happens anyway? as the energy builds on the topload it
eventually reaches the breakout force, this puts almost all stored
energy into the leader in a burst. once formed it would be the following
current pulses from the secondary that maintains the channel and i would
think these will be a much smoother zero to peak consecutive pulse? a
too large roc topload seems to me just has much more area to charge with
a given amount of energy thus breakout force is never reached until more
energy can be stored, higher voltages or currents. wouldn't even the
biggest topload breakout given enough power input?
or maybe i'm just saying the same thing in a different way?


> 
> An interesting thought here.. If the topload breaks down too early, not
> enough charge will be stored to support the leader growth, so you kind of
> want a big smooth topload so that it can accumulate enough charge to support
> a big long spark.

is this because the smooth surface allows more charge to build due to no
bumps to centralize the force?
very interesting stuff.
marc m.

  However, given that you can transfer charge into the
> topload on every cycle of the RF, with the right circuit, a breakout point
> (to force earlier breakdown), combined with the right tuned circuits, might
> work even better.
> 
> The whole thing reminds me of a "peaking gap", as used in impulse testing.