Original poster: "Barton B. Anderson" <bartb@xxxxxxxxxxxxxxxx>
Hi Gerry, Steve,
The key, I believe, is "the stress on the coil".
Those words, in my opinion, is the racing arc situation in a
nutshell. The relief of course is to find what causes stress. What
type of stress? Well, a voltage stress in a part of the coil that is
high enough to breakout. In the case of breaking out on the side of
the coil, it finds it's path right down the surface of the coil (and
sometimes may breaks out even after the surface arc has started and
strikes the primary or strike ring = I've seen this). Most of the
time, it's simply a path down to the bottom winding where RF ground
impedance is low. If the stress is at the top (usually when the
toroid is positioned too high above the top of the secondary), a
racing arc can occur there, but most of the time, breakout from the
top winding. Although there are frequency nodes that can aid or
oppose the situation, there is 2 common situations.
1) Tuning: Near resonance, but not perfect for the power. Power is
involved here. A poorly tuned high power coil can produce racing
sparks very easy. So can low power coils, but not as easily.
2) Coupling: Plays right into tuning. Increase your coupling and
you'll find out just how well your coil is tuned. It's important in
my book to tune the coil and increase k to until you can no longer
physically tune either any better - if you want to get max spark
length. Tesla coil couplings can be somewhat high, but much depends
it's tuning. I do something a little different than most. I like to
set k relatively high and tune in the coupling vs. using a low
coupling and tuning in the coil. When the latter is done first, it's
hard to tell if k could have been higher. This is simply a view of
my own to process as much secondary power as possible.
Tuning and coupling are one and the same. Both should be "played
with" for maximum performance. I realize there are multiple
frequency nodes that can spell trouble (higher voltage stress in an
unwanted part of the coil), but any coil can be tamed, and that is
the key. If we shorten our coils, we decrease the creepage distance
(not a good thing). If well make the coil too long, too much R that
sacrifices power. We find ourselves in that 5:1 h/d ratio (or near)
because of the voltage stress limitations we encounter. Geometry is important.
Take care,
Bart