Re: TESLA COIL REVISED

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

Original poster: Paul Nicholson <paul-at-abelian.demon.co.uk> 

Hi Terry, All,

Terry wrote:
 > [7Mhz PA] the designers may not have had the slightest idea what
 > they were doing....

Could well be true, since it's my own homebrew.

 > As the frequency goes up, the losses go way up since currents in
 > surrounding junk goes way up.

Very true. All you can do is make the surrounding junk conduct as
well as possible.  At least as the frequency goes up, the shielding
of the H field improves and the overall containment becomes more
effective.

 > [silver plating] They tend to turn black in these days of sulfur
 > dioxide pollution :-p

Not my field, but I believe the conductivity of the tarnished
surface is almost as high as the original. Others will correct this,
but I think that's the major benefit of silver plating.

 > So one may perhaps conclude that streamer length per given input
 > power falls dramatically at higher frequencies.

Yes, John's formula does very well doesn't it - over a wide range
of power levels.   It must start to error at higher frequencies
because, presumably, for a fixed discharge capacitance the
heating effect will be proportional to Fres and the relatively
cool plasma of the trad TC discharge gradually turns into the hotter
plasma of the HF discharge.   This was mentioned recently in
connection with magnifiers with the suggestion that the mode 3
content was contributing to the plasma heating.   I suggested that
someone might want to fire up a cap discharge TC to streamer
breakout at the same time as injecting CW into the base at one of
the higher modes.  One could then document the character of the
discharge as a function of the relative proportions of LF and HF
content of a measured discharge current.  Wonder who's got the
technology :)   Ideally we might expect to be able to vary the
brightness of the streamers to some extent independently of the
streamer length.

As regards coil losses, shielding makes a big difference as the
frequency goes up.  Whether we like it or not the coil will couple
to its surroundings - the H field stirring eddy currents and the
E field stressing dielectrics.   These get rapidly worse with
increasing frequency and the only cure is good shielding.  You get
large eddy currents in the shield, but that's ok because the loss
resistance of the shield is low.

Terry wrote:
 > http://www.mksinst-dot-com/cgi-bin/product.exe?pid=MWH-100

The commercial chaps must have got this stuff down to a fine
art.  Dare I ask what the DC bias output is for?  They're driving
into quite low impedance loads which makes the job a lot easier.

 > http://advanced-energy-dot-com/Upload/wp_16_tuner_topics2.pdf

The L network in here is matching down from 50 ohms to a lower
impedance, which makes sense with their plasma arc chamber.
Although, getting the gap to fire in the first place could be
quite exciting.

 > http://advanced-energy-dot-com/Upload/wp_18_impedance_match3.pdf

A nice summary of impedance matching issues.

The coiler faces a tougher challenge than these commercial
plasma applications because the discharge terminal is usually quite
isolated from the other terminal and a very asymmetric discharge
results - discharging 'into air'.  This must surely be less
efficient, overall, as a way of generating volumes of plasma.

At least there are no striking difficulties with this mode :)

I suppose that as the frequency is raised (for a given terminal
and fixed load power) the discharge becomes shorter and hotter.
But does its impedance go up or down?

Most RF engineers can find a tale to tell involving the interesting
consequences of an accidentally unloaded resonator or an open
ended waveguide or coax.   My own favourite is my neighbour reporting
blue flames from my 145Mhz antenna - was running the 'full legal
limit' at the time :)    Goes to show that HF plasma is all too
easy to generate and suggests that poor results from attempts by
coilers to produce HF HV discharges are due to resonators too lossy
to achieve the required impedance transformation at reasonable
efficiency.   Most coilers prefer the long low frequency streamers.

Perhaps we need a competition to stimulate HF efforts - for eg a score
based on product of frequency times load power, or something?
--
Paul Nicholson
--