Re: On sparks

["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 3/30/01 3:00:59 PM Eastern Standard Time, tesla-at-pupman-dot-com 
writes:

>  >No, I'm still skeptical of that charge in the top-load.  Upon overnight
>  >reflection, I'm beginning to conclude (and this point has no doubt been
>  >made by others, before) that the big difference between s.s. and
>  >spark-gap systems--exhibited in the difference between the natures of the
>  >sparks obtained--lies in the power-vs.-time relationship of the primary
>  >excitation.

Ken, all,

I agree the power vs time relationship is the key.  

>  >
>  >In my s.s. system, I establish a spark duration of 7 ms but I only
>  >presently generate the primary's magnetic flux from about 36 A x 3 turns
>  >= 108 ampere-turns.  That situation produces relatively short sparks
>  >(although they're 3' or so: not too bad, I will say, from the 160 V power
>  >source) but relatively fat, bushy and branched ones because there's time
>  >for those qualities to develop.

Yes, systems which supply a low power over a long time period, seem
to give time for branches and fatness of the spark to appear.  I have
seen this in my own work.  Very long durations give very fat branched
sparks.

>  >
>  >In a spark-gap system on the other hand, I conjecture that the flux is a
>  >whole lot higher--many more amperes, maybe orders of magnitude, through
>  >more-or-less the same quantity of turns--and the duration of that flux,
>  >per spark, is a whole lot less.  Thus, the spark "gets out there in a
>  >hurry"; it's longer but it doesn't have time in which to develop as much
>  >thickness or as many branches.  But the power taken from the mains is
>  >much the same in the two cases; it's just that that power is applied to
>  >the primary in shorter, more-intense bursts in the spark-gap system.
>  >
>  >I think that the relatively small effect of the size of the top load is
>  >clear from the fact that I can readily generate enough voltage to break
>  >down the smooth 6" x 24" Landergren toroid but I still can push the spark
>  >length only to 3' or so, even though I keep pushing for all of 7 ms.  It
>  >has some effect, certainly since--all else remaining the same--I find
>  >that my 7 ms sparks as emitted from the toroid are at least half-again as
>  >long as those emitted from a surface having half the radius. 

That is very interesting, so the ititial voltage is having a great effect.

>  > But the big
>  >factor that gets the spark 'way out there is that initial ZOT!! of
>  >primary flux.

In light of the fact that I now realize that your spark was 50% longer
when emitting from the big toroid, yes that makes sense.  I don't know
why I had the idea previously that your spark was the same length
whether from large toroid, small ball, or point.  

In a related issue, I attribute the benefits of a low break-rate to a
similar cause.  For a given power input, a low break-rate demands
a larger "bang size" (a bigger dump of energy from the cap for
each gap firing of a spark gap TC).  It would seem that using a 
high break-rate to produce long sparks is similar to using a longer
duration application of lower-power bursts like a CW coil.  In constrast, 
a low break-rate spark is formed from a powerful, fast burst of energy,
repeated at a less frequent interval.   There seems to be no
substitute for fast powerful bursts for obtaining longest sparks
for a given input power.   I would go as far as to 
say that high breakrate coils are tending towards CW coils, and that
is why a high break-rate tends to be less effective at producing long
sparks for a given input power.  High break-rate sparks tend to be
fatter, but shorter than low break-rate sparks for a given input
power to the system.  I should say here that I define a low break
rate as something around 120 bps.  I define a high break-rate as
something around 240 bps and higher.  The effectiveness of the
input power to produce long sparks seems to decline as the
break-rate is increased beyond 120 bps or so, at least in the
work I've done.  Of course if the break-rate is too slow, then
the benefits of the growth of sparks due to partial ionization of
the air, is lost, or severely diminished, and the spark length
is reduced.

Again let me say that my break-rate comparisons assume
a constant input power.  Certainly if the break rate is raised,
but the input power is permitted to rise, then the sparks will
grow longer.

I agree that the MOSFET's characteristics are not hurting the
coil's performance.  It just a matter of bang size per unit time.

John Freau

>  >
>  >I'm sure this subject needs more clarification & perhaps correction.
>  >
>  >Ken Herrick
>