RE: THOR Bang energy vs. streamer length measured

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

Original poster: "Jim Lux" <jimlux-at-earthlink-dot-net> 

At 07:14 AM 7/28/2004 -0600, you wrote:
>Original poster: "Denicolai, Marco" <Marco.Denicolai-at-tellabs-dot-com>
>Hi Malcolm,
>
> > That is particularly interesting because the result shows a
> > deviation from a simple power dependence. Now the dependency
> > becomes voltage and/or charge availability,......
> > perhaps..... because we don't really know what part the
> > oscillations might play in spark initiation, propagation with
> > repetition and continuance. i.e. the oscillations occurring
> > before the secondary is drained by the output discharge are
> > continuously altering the instantaneous charge availability
> > at the terminal.
> >
> > Malcolm
> > <snip>
>
>Well, actually we do know something. From previous literature and from
>what I saw from probing the grounded stick current.
>
>1. The filaments start at the positive semiwave time, as the inception
>voltage for positive polarity is always lower than for negative one.
>
>2. The filaments elongate also during the following negative (and
>positive) semiwaves.
>
>3. The portion of the semiwave pumping charge into the gap grows from
>semiwave to semiwave. If we talk abot degrees of a sinewave (from 0 to
>180) regardless of its polarity, the first semiwave conducts, say, from
>80 to 100 degrees. The second from 70 to 110 degrees, the third from 60
>to 120 and so on. This reflects the gap (channel) charge accumulation
>and the lowering of its corona inception voltage. At the end, the whole
>semiwave will generate current i.e. charge conduction in the channel.
>
>The above are raw results that I still need to investigate (and
>document) thoroughly. That's what my PhD could possibly be about :)


This is quite fascinating.  When you're measuring the current, is it at the 
base of the leader?  Here's a sort of conceptual model.  the spark channel 
is an ever growing transmission line (that grows in jumps).  The charge has 
to propagate down the currently existing channel (keeping it hot in the 
process) to the tip, where the field steadily increases until breakdown 
occurs.  Then, a short new segment is added, with charge flowing into it 
(mostly from the end of the channel).  At some point, though, the voltage 
at the base starts to decrease, and charge will eventually flow out of the 
channel and back to the base electrode.

As you say, positive polarities have lower breakdown voltage than negative, 
so the time from zero crossing until new jump occurs should be lower for 
positive than negative, BUT, there's less charge in the channel when 
breakdown occurs, so the length of jump might be less.

A HV transformer that can be driven by a high power RF source might be 
quite useful to investigate this, because you could change the RF frequency 
(ideally on a cycle by cycle basis, as with an arbitrary waveform 
generator), which would change the relative timing of the channel 
conduction, leader jump, etc. processes.

One really needs a very high speed camera synchronized to the RF, but, 
perhaps suitable high speed current measurements and some carefully chosen 
efield measurement points might help.