[Prev][Next][Index][Thread]

Re: Ritchie Burnett's website

Hi Nick,

At 12:11 PM 12/31/1999 +0000, you wrote:
>
snip....
>
>1    Ritchie strongly argues for an early quench which probably demands a
>rotary spark gap.  However, the real arc to ground occurs after only 2.5 rf
>cycles, well before the first quench.  Practically, we're all after the
>longest, brightest arcs to ground so do we need rotaries and do we need to
>worry about time to quench?

Quenching does not seem to matter in some systems as much as others.  In my
experience, static gaps really benefit from early quenching where rotaries
do not.  I have used a wonderfully quenching 16 gap rotary (still on loan
to me :-)) that can quench anywhere in the power cycle it wants (even at
current peaks!).  However, the increased losses hurt overall performance.
If you have a static gap, go with the multi gaps such as the famous Richard
Quick design.  If you have a rotary, I would go for as few gaps as possible
(two) and get the gaps to come as close ass possible to give the lowest
loss possible.  "I" do not consider quenching as a factor in rotary gaps.

Actually, an arc to ground does not hit immediately.  The streamer grows
upon the ionized path set up by the PRECEDING streamers.  Thus, getting
maximum energy into the arc is still very important.  When the arc to
ground does occur, it will be before much quenching can have an effect on
THAT cycle.  However, the PRECEDING cycles still may benefit from a good
quench in order to allow the streamer to grow to the longest distance
possible before the final ground strike cycle occurs.

Ideally, we want great, perfect, zero-resistance, and first notch
quenching.  However, different gaps will begin to get too lossy and you are
better off not worrying about it.  The higher loss is still acceptable in
static gaps but rotaries seem to get too lossy if you try for much
quenching.  Of course, one gap may perform better than another and this is
still sort of a "seat of the pants" area of TC design...  I find rotaries,
especially sync-rotaries, are far superior to static gaps.

>
>2    Before the gap quenches, there will be two output frequencies and
>nothing at the Fn, the natural frequency of both the secondary and primary
>circuits when uncoupled.

At any given instant of time, All the energy is at one frequency (if you
get technical it is actually a rate of change instead of a "point"
frequency (like a location vs speed thing...).  The two frequency spectrum
is due to the fact that the spectrum analyzer displays the frequency over a
time span much lower than the frequency being displayed.  Thus, you see a
spectrum of frequencies since the actual frequency wobbles around or is
slightly Frequency Modulated (FM).  That is totally an artifact of the way
the analyzer displays the data.  At a give time it is really just in one
place but the analyzer "smears" that data over time (there is a knob to
select the time span if I remember right).  The same smearing is seen if
you look at the output of an FM radio station on an analyzer.  The Corums
my disagree with me on this, but "I" think you can forget about the dual
frequency energy storage thing...  It does not exist other than in the
spectrum analyzer's design tying to display events that are far to brief
for us to see directly...  Storage scopes (or simulations) show the real
thing which looks pretty much like a sine wave (and is for practical purposes).

>
>3    I'm interested in the frequency plot of the operating coil.  Was this a
>real measurement or a simulation?  It shows that the output power at Fn is
>fully12dB higher than that at Fu and Fl, ie, most of the power is in the
>long secondary voltage decay after the gap has quenched.  I guess this means
>most of the power is dissipated in losses rather than an exciting arc or
>corona.

About 1/2 of a well designed TC's energy is burned up in the gap and other
losses.  I have been able to get near 70% with LTR coils.  What we NEED is
a perfect gap.  I recently looked into some gap designs with new high
voltage IGBT's and such but the losses would be far too high.  There are
some high voltage/current triacs that are optically fired that may be
getting closer but their cost is too high to "play" with...

I did not quite see what you meant here but I can assure you that the
simulations (if done right) will give a very accurate representation of a
Tesla coil and all the fancy analysis tools in those simulations work very
well and are just as accurate.  You may want to see my paper on such things at:

http://www.peakpeak-dot-com/~terryf/tesla/experiments/modact/modact.html

Most of these things can now me measured directly too.  The "true dynamics"
of what arc impedances are is only now beginning to really be measured...
The 220K + 5pF estimation or arc impedance is very good in general, but is
rather simplistic when we get down to trying to match a Tesla coil's output
impedance to the arc impedance.  Such work takes all the computers, probes,
test equipment, and "tricks" we have, but only limit we have is time to do
the work...  A number of people now have fiber optic probes and plane wave
antennas so I no longer have much of an equipment advantage over them.  So
I hope I am not going to be the one playing this game ;-))

>
>Ritchie, thanks for your work - is it research for a PhD?  I wished I'd got
>that topic for mine!

Richie's wonderful site is GREAT!!!

Cheers,

	Terry

>
>[BTW, don't bother going to my website for Tesla stuff - it's all my
>professional marketing stuff at present.]
>
>Nick
>
>-------------------------------------------
>Dr Nick Slaymaker     Tel   +44 1462    457    582
>Hi-Tech Marketing     Fax   +44 1462    457    582
>5 Standhill Close, Hitchin, Herts, SG4 9BW, UK
>Email             nick.slaymaker-at-lineone-dot-net
>Home page    http://website.lineone-dot-net/~nick.slaymaker
>-------------------------------------------
>