Re: Parallel and Series LCR Circuit Qs

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

----------
> From: Tesla list <tesla-at-pupman-dot-com>
> To: tesla-at-pupman-dot-com
> Subject: Re: Parallel and Series LCR Circuit Qs
> Date: Thursday, August 10, 2000 6:04 PM
> 
> Original poster: "Gomez ADDams" <gomez-at-netherworld-dot-com> 
> 
> on that fateful day 8/10/00 12:28 PM, Tesla list uttered:
> 
> > Original poster: "Jim Lux" <jimlux-at-jpl.nasa.gov>
> 
> > The propagation velocity of a spark is around 1E7 m/sec,
> 
> What?!?  10 million meters per second?  Where did that number come from?
> That's like 3% of the speed of light!  I was under the impression that
long
> sparks propagated rather slowly.  Certainly lightning does...
> 
> - Gomez

That's the speed (varies an order of magnitude either way, depending...)...
for lightning as well... Lightning is kind of strange though.. It goes in
jumps, roughly 100 meters at a crack. The jump moves fast, then it pauses,
then jumps again, so the overall rate is somewhat lower.  Even so, lighting
moves pretty fast. There are lot's o' theories here, as to why... perhaps
it is time for the charge to flow from the source into the leader that has
just been established in the jump and to raise the voltage at the end of
the leader above breakdown for air...   the overall field for lightning is
quite low (several kV, say 10-20 max, per meter, well below the 3000
kV/meter breakdown for air, in general).

Interestingly, nobody is quite sure why long sparks, lightning, etc. form
forked and erratic channels, unless it is just a random thing and one
direction is as good as the other at each step in the leader formation.  As
you can imagine, making measurements about what is going on is, to say the
least, challenging, at those speeds, esp if you don't want to perturb the
fields or phenomena.  

I have been theorizing about setting up a large array of e field probes
with very fast digitizers around the streamer formation area, digitizing
at, say 1 GSPS, then inverting the observed fields back to the leader shape
and size as it develops.  You'd correlate the E field data with optical
images recorded simultaneously from several directions. 1 Gigaframe/second
framing cameras aren't cheap either...
The instrumentation alone would be pretty pricey, not to mention the huge
computational load (which is actually pretty cheap to process, if you
aren't in a hurry).