Re: Certain s.s. phenomena; to J.F. et al

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

Original poster: "by way of Terry Fritz <twftesla-at-uswest-dot-net>" <FutureT-at-aol-dot-com>

In a message dated 3/13/01 2:38:41 PM Eastern Standard Time, tesla-at-pupman-dot-com 
writes:

> 
>  It may have been lost; I didn't see that comment.  You are saying, I take
>  it, that such a short coil, vs. the spark-length, would not tolerate an
>  attempt to establish a horizontal spark.  The spark would just jump the
>  coil to ground.

Ken, all,

It is true that the short coil may not tolerate a horizontal spark, 
but I was refering to the presence or absence of a breakout
point.  I meant that if no point was used, and a toroid was
used, the voltage would build up so high that the coil would 
break down or arc-over, probably.  
>  
>  > 
>  > In any case, it would appear that tube coils of the typical design,
>  > have to have a breakout point for longest spark output.  This seems
>  > to be because of the double-tuned design, the resultant frequency
>  > splitting, the changes in Q and bandwidth with breakout, and 
>  > probably related Z matching factors also.
>  
>  Well, perhaps...but a) spark-gap systems are also double-tuned, b) I do
>  not know what "frequency splitting" is, and 

frequency splitting is the split freq response that occurs in a double-tuned
over-coupled circuit.  I suspect that spark-gap systems have much
less of a tendency to build up corona on the secondary and primary,
despite the double-tuned design.  It may be because of the short
amount of time that the energy is present compared to a CW coil.
Also, tuning is not the problem in a spark gap TC, that it is in a
CW coil.  In the spark gap TC, you tune the coil, the energy 
transfers from primary to secondary, and all is well.  But in a tube
CW coil, with constant drive with changing impedances, and the
peculularities of the system, seems to demand a point on the
toroid for the longest spark:

If I tune my coil for breakout from a toroid, then I get short sparks.
If I tune for longest sparks from a point on top of the toroid, then the 
coil cannot break out without the point.  The required difference in
tuning is surprisingly large.  The coil has to be severely mis-tuned
for operation with no breakout point, so this reduces the spark.
Since the resonator is a tuned circuit, and the tank coil is a tuned
circuit, they both seem to fight for control of the oscillator to some
degree, and this makes the tuning situation complex.

>  c) the secondary's Q should
>  diminish essentially the same amount regardless of the manner in which
>  the spark is induced to initially break out.  I'd think that the
>  differing relative impedances during the build-up and sparking events
>  (Z-matching, as you say) would likely constitute a major factor in
>  determining the differences in performance of tube, s.s. and spark-gap
>  coils.

Are you saying that the impedance changes a larger degree in a tube 
coil when it breaks out compared to a SSTC?  I hadn't thought about
that.  I agree the Q's diminish the same amount in either type coil,
and this Q-drop is certain a big factor.

>  I raise the mains current merely by increasing the rep-rate--up to what
>  my current-limited power supplies are set for.  I've tweaked it up to
>  40/second now, producing a very satisfying buzz, but it's a lot of effort
>  to change back to the 1- or 2- turn configuration in order to make any
>  kind of a close comparison.  Have to unplug a whole lot of connectors,
>  removing the entire primary bundle, and substitute the other one.  Since
>  I don't >have< to, I just may not.

I don't blame you.  I guess varying the pulse rates adds 
another variable too.
>  

>  >I became intrigued & can now offer something
>  >quantitative for a change:
>  
>  Spark rep-rate & duty cycle, power input: the same.  
>  
>  Emitting radius:                   3"      1.5"      a sharp awl-point
>  E-field deflection on the scope:   8 cm    5 cm      4 cm
>  
>  Spark length, appearance:  Essentially unchanged: relatively fat,
>  branched.
>  
>  Now that's interesting!!  The sharp point cuts the peak voltage in half
>  (surely reducing stress on the coil) but no more than that!  I'd never
>  have believed it.  At least as observed visually, no sparks reach any
>  conductor to ground during this test; they just end in air.  

Duane Bylund did say that his SSTC also saw the voltage cut in
half with the point.  But the ball he used was only 1" dia, and his
sparks were just 7" long.  His sparks were also the same length
with or without the ball I seem to remember.  Your results are
interesting because they show that it is only a tube coil that
demands a breakout point for longest sparks, or a coil similar
to a tube coil in its characteristics.  Your results are also
interesting because they show that you don't need to build up
such high voltages which perhaps stress your MOSFETS to 
obtain a given spark length.  If you run with the breakout point,
the impedance won't shift that much, and will shift at a lower
power level.  This may permit better impedance matching, and
better efficiency maybe.   Another interesting test would be to
install a much smaller toroid, maybe just 1 foot in diameter or
so, instead of the 26" toroid, (sort of a corona ring), and see if
this has any effect on the spark length and fatness.  Maybe it
would make the spark output longer.

Regarding the branching, maybe a coil that is always in tune
produces such branched sparks?  My coil which sort of pulls
into tune at least to some degree can produce straight sword-like 
sparks.

>  But perhaps
>  it is the presence of nearby conductors, nevertheless, that accounts for
>  the only 2:1 difference in attained voltage between the toroid and a
>  sharp emitter. 

I seem to remember that Terry F also saw this same relationship
in a spark gap TC (2:1 voltage difference after breakout).  It seems
the sparks are not a terrificly heavy load.

>  Marco DeNicolai had a relevant posting, on 1/31, on the subject.  

I can't remember the posting.
>  
>  Very roughly, since I am unable precisely to measure my input power, I
>  measured around 1.5 mains KW going into 13/second sparks drawing power 6%
>  of the time.  Each spark, therefor, would contains 1.5/6% or 25 KW--over
>  about 5 ms duration. That's what gets me 2.5-3 ft. sparks.  I'd like to
>  know how to "massage" that set of criteria so as to push that spark out
>  further!

The only thing I can think of to improve things is to somehow improve
the impedance matching of the system.  If you're willing to make the
sparks appear pulsed visually, then you can reduce the pulse rate
even more, to reduce the input power.  The spark length will be 
reduced some, but the power will probably be reduced even more.
Once this happens, then you can put more peak power into the
coil, by using bigger MOSFETs if their available at a suitable cost.
Maybe using more primary turns, such as 5 or 8 will give better
results.  I realize the effort involved in building such an elaborate design.

It would be good if the impedance matching issue could be analyzed
by computer to see what the best setup is.

My 36" spark tube coil drew about 7kW at 60PPS, so at 13 PPS,
it would have drawn 1.5kW.  But the sparks would no longer have been
36" at that pulse rate, they would have dropped to maybe 28"
or so.   Do your sparks shorten if you reduce the pulse rate?

John Freau

>  
>  Ken Herrick 
>  
>  > 
>  > John Freau
>  > 
>  > >  
>  > >  Ken Herrick
>  >