RE: I'm wondering about this gap and tuning issues. Am I finally *beginning* to understand?

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

Original poster: "Lau, Gary by way of Terry Fritz <twftesla-at-uswest-dot-net>" <Gary.Lau-at-compaq-dot-com>

Hi Garry:

Comments interspersed...

>Original poster: "Garry Freemyer by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <garry-at-ndfc-dot-com>
>
>I've been reading here quite a bit, and a lot of what is written here
is WAY
>over my head. The hardest part I think is getting the terms straight.
>Usually, by the time my mind finally coughs up the definition of a term
I am
>reading, I've forgotten what was said about it. Then by the time I am
done
>reading it again, I've forgotten the definition and by the time I
recall the
>definition, I've forgotten what was said again. ;-)
>
>So, I try to visualize it in my head and I was thinking about NST's. I
seem
>to remember that (Reason I say this is because I don't trust my memory
any
>farther than I can throw a fully loaded and operating washing machine)
that
>when you have alternating current that the voltage is like a sign wave,
>going positive, then neutral at the point of reversal, and then
negative.
>There are peaks and valleys in the sine wave. I will rever to them both
as
>peaks.

I'll call them positive and negative peaks.

>So, I was thinking about how could a cap in a tesla circuit be charged
when
>the current is going negative and positive. I asked myself why it
wasn't
>canceling it all out and here is what I came up with. Someone tell me
if I
>am close?

If you were averaging things over several 60Hz cycles, you could think
of  things canceling out.  But the stuff we're interested in occurs in
the interval
between spark gap firings, often called "bangs".  The charge on the cap
just prior to a bang is either positive or negative.

>Are these statments true or close?
>
>The primary and the gap spacing in a tuned tesla coil is such that the
gap
>fires as near the PEAK voltage (Or is this current?) so as to discharge
the
>cap. The close to the peak the gap fires the better. When the charge is
>reversed, the gap fires on the reverse peak. If the gap fires early,
output
>is reduced because the charge is drained before it can maximize, and if
it
>fires too late, output is reduced because some of the charge has been
>canceled out by the opposite charge that is now flowing into the
capacitor.
>True?

The primary coil doesn't affect when or how often the gap fires.  The
things that affect it are the size of the NST, the size of the cap, and
the gap width.

It is largely true that "the closer to peak voltage the gap fires, the
better". But there are some important and non-intuitive things that are
widely misunderstood.  The following assumes the use of static gaps.
This may get long so please bear with me...

Let's say you have an NST that has a peak voltage of 10KV.  Ideally you
would choose a cap such that at the end of each half-cycle, the cap
would have a charge of 10kV on it, and you would choose a gap spacing
such that it would fire just at or under 10kV.  So at the end of every
60Hz  half-cycle, the gap would fire and it would fire 120 times per
second.  This would be nice if it were that simple, but it's not likely
to happen in the real world.

What's not obvious is that if there is a charge on the cap and the gap
doesn't fire at the end of that half-cycle, that energy will not be
lost, but will be carried over into the next opposite-polarity
half-cycle.  Let's say that you turned on your NST halfway through a
60Hz half-cycle, so there wouldn't be enough energy in the first
half-cycle to charge the cap all the way to +10kV.  Instead it charges
up to say 5kV, then starts to decline without the gap firing.  The
energy in the half-charged cap won't be wasted, but will be carried into
the next negative half-cycle, and roughly halfway through the next
negative half-cycle, the cap will achieve a charge of -10kV, and now the
gap will fire.  But the negative half-cycle isn't over yet.  After the
relatively short primary ringdown, the cap will again start to charge
from zero in the negative direction.  But since there's only about half
of a half-cycle left, it only makes it to -5kV, then starts going back
towards zero as the positive half-cycle begins.  And so on.

The other thing that's important to understand is the concept of
resonant rise as it relates to the NST/cap.  Caps are generally sized
such that they are _roughly_  resonant with the NST secondary at 60Hz.
It needn't be exact.  Let's say that again your NST peak voltage is
10kV, but let's say that you accidentally set your gap to twice the
right width and that it now fires at 20kV.  One might think that it
simply won't ever reach 20kV and won't fire, but it will!  Remember that
energy is not lost at the end of a half-cycle if the gap doesn't fire,
but is instead carried into the next, opposite polarity half cycle.  So
if you turned on your NST at the beginning of a half-cycle, at the end
it will peak at 10kV.  Then it will head back towards zero, and shoot
past -10kV, possibly reaching -20kV.  If it doesn't reach -20kV, it will
swing even higher on the next positive half-cycle, and so on, until the
gap firing voltage is reached (OR UNTIL YOUR CAP OR NST IS OVERVOLTED
AND SHORTS OUT).

An interesting thing to do is to note that if you turn your variac down
to 50% or 30%, your gap will still probably be firing, though at a much
lower BPS rate.  This is because while the open-circuit NST voltage is a
fraction of its normal peak, the voltage on the cap builds over
subsequent cycles until the gap breakdown voltage is reached.

The simulated waveforms on my web site may be easier to understand than
all of these words, see:
http://people.ne.mediaone-dot-net/lau/tesla/gapsim.htm

>The primary, acts as a kind of resistance to emf flow, halting it at
the
>right time (Quenching) and delaying the discharge to the right moment.
True?

The primary doesn't affect how fast the cap charges or when the gap
fires.  Quenching refers to when the gap *stops* conducting, following
the gap's firing, and is affected by the gap geometry and airflow,
pri/sec coupling, whether streamers have formed from the top load, and
probably other things as well. 

>So, that's where we get the 120 breaks per minute because the gap fires
on
>the peaks that are double the alternating cycle of 60 cycles per
second.
>True?

In a synchronous rotary gap one can achieve exactly 120 breaks per
second, but with a static gap, the gap firing interval will in practice
be chaotic.

>Than a BPS of 60 would mean the gap discharges on the peaks of only the
>upper or lower half of the sine wave or every other peak?

Exactly.  In fact I was once able to operate my static gap at 60BPS.  I
had to set my variac to a lower, very critical setting and the effect
was not at all stable, wanting to slip back into chaotic mode, but it
was surprisingly easy to make happen.

>So, rotary gaps have the advantage of better control. AC would do best
with
>a sync gap and DC would work with either sync or assync?

Probably a fair assessment.

>I recall folks speaking about voltage ringup, would that be kind of
like
>where the relationship between the frequency and the inductance in the
NST
>is such that the current gives a shove to the next cycle kind of like a
>person rocking back and forth on a swingset to get more swing
(Amplitude).
>Wow, this last one sounds a little far fetched to me. Take this one
with
>grain of salt if desired.

That's a great analogy for what I was trying to explain earlier.  Energy
on one side of the middle (zero) of the swing is carried into the motion
on the other side.

>So, after reading all your posts out there, am I finally getting some
idea
>of what ye all are saying or am I still in clueless canyon?

>PS: Warning! If you reply in really technical terms. I probably won't
>understand you. ;-)

Regards, Gary Lau
Waltham, MA USA