Re: Critical rise time (RE: Terry's New Plane Wave Antenna)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Steve Conner <steve@xxxxxxxxxxxx>


> The time it takes for top terminal "energy" to reach a maximum has 
> been noticed with DRSSTCs.  Apparently, faster energy rise times 
> give better streamers.

The strange thing is that the top terminal energy is pretty 
insignificant compared to the total energy delivered to the load. The 
DRSSTC I've experimented with will deliver 7 joule bangs even though 
the terminal capacitance stores less than 1 joule, and I expect other 
DRSSTCs are the same.

So I'm not sure how the energy stored in the discharge terminal can 
be that important in the scheme of things, except in so far as it 
needs to be above a certain amount to keep the secondary Q reasonable.

Of course it is likely that streamers need high frequency (10s of 
MHz) pulses of RF to grow, and the top capacitance provides a low 
impedance source for that. IOW, it stores energy that is put in 
relatively slowly at the coil resonant frequency and gives it out as 
small fast pulses.

Although these days, I prefer to think of the energy as being stored 
in the whole volume of space around the coil rather than "on the 
toroid", and good coil performance being a result of maximising the 
stored energy density over the whole volume where you want to produce 
streamers.

The relationship between resonator size and spark length kind of 
falls out of that: small resonators produce small sparks because they 
can't "throw" the E-field so far. The air breaks down 
catastrophically near to the resonator before you can generate enough 
E-field for streamer growth in further away regions. You can overcome 
this by increasing burst length and repetition rate, but then your 
efficiency takes a hit. A sensible limit seems to be about 3x the 
smallest clearance dimension (for flashover) on your resonator. I 
have got to about 4.4x but had to use stupid amounts of power: over 
3kW into a 14" tall coil.

It also follows that those pulses of RF that the streamer appears to 
"draw" from the top capacitance are just a symptom of the stored 
energy relaxing as the air gives way. There is probably plenty more 
current flowing in the streamer channel than the conduction current 
that you would measure flowing from the breakout point to the end of 
the channel.

But then I don't see how that links in with load energy rise time. 
According to my view of things, all you need to do to create a given 
length of spark is to apply the required electric field strength over 
a big enough region of space, for long enough to let the spark grow 
to completion. If you don't have enough E-field, enough space, or 
enough time, you won't get the desired spark length.

The HF AC output of a Tesla coil must complicate things compared to a 
Marx discharge, due to the way it interacts with the space charge. We 
are fairly sure that the spark grows a little more around every peak 
of the output voltage. But I don't believe it makes any fundamental 
difference. It just means that you need to apply the voltage for 
longer than you would a DC voltage of the same amplitude, since the 
spark is not growing continuously.

Also, the repetitive nature of a usual TC discharge changes things 
since there can be ionisation left over from the previous bang. I 
recently saw a video of a power line flashover taken with a Daycor UV 
camera (at www.seeing-corona.com) After the flashover has happened 
and the power cuts out to clear the fault, you can see the arc 
channel "fall off" the insulator and float away, still emitting UV 
from residual ionisation. It seems to maintain more or less its 
original shape until it drifts out of the frame.

It would be very interesting to get one of these cameras pointed at a 
Tesla coil sometime.

Steve Conner
http://www.scopeboy.com/