Re: Inductive Kick Effects

At 09:35 PM 6/20/99 +0200, you wrote:
>Hi Terry, John, Malcolm, Richie, all,
>And now....., I DONīT ;o)) get it anymore. How can the energy stored
>in the xfomer be of help? Let me draw two schematics (use courier to
>1.Charging mode:
>X                   P
>X                   P
>X                   P
>X                   P
>X                   P
>X                   P
>Current flows from the xformer (X) through the cap, primary
>and then back to the xformer. I ommited the gap here,
>because it is of no importance, as it isnīt firing, yet. So, we
>now have a lot of energy stored in the cap, some energy
>in the secondary of the xformer and very little (low ĩH)
>energy stored in the primary.

In my case...
Neon Secondary		23.75 J
Primary Cap		5.97 J
Primary Inductor	0.0J

>2. Discharge:
>X--->    >------||-->----P
>X   |    |               P
>X   |    |               P
>X   o    o               P
>X   o    o               P
>X   |    |               P
>X   |    |               P
>X-<--    <------------<--P
>You will note that I have "installed" two gaps in the above
>schematic and disconnected the xformer from the tank
>circuit. As soon as the gap fires, (hopefully) all the energy
>is fed into the priamary as our switch (= the gap) is now
>closed. At the same time, the "second gap" (which is in
>reality the same gap) also shorts the xformer. Any energy,
>that was stored in the xformerīs secondary is now drained.

Stop there!!  If it were a cap, shorting it out would instantly draw a
giant current and drain it.  However, the inductor will tend to keep
supplying the same current it did before!  In order to instantly remove the
energy from the inductor, you would have to OPEN the circuit (causing a
giant voltage which may damage the transformer*?)!  By shorting the
inductor, you are maintaining the circuit and thus maintaining the energy
stored in the inductor (as current).

* I ran a model of what would happen if one were to suddenly open the
output of the neon at a current peak.  The voltage across the transformer's
output shot up to 272kV!!!  Although that should never happen and the
protection filter would stop it, if it did... the neon would die!  With the
protection filter, it only shoots up to 88kV :-)  However, the voltage rise
takes a millisecond to reach that level which is more than enough time for
the safety gap and MOVs to stop it.
	I modeled every fault condition I could think of for my new LTR coil but
this is one situation that I did not consider.  However, looks like it is
not a problem.  My coil should still be bullet proof (famous last words :-))

>However, how can this energy enter the tank circuit? I think
>(point out the error to me, please), the only thing this extra
>shot of energy can/will do, is increase the time in which the
>gap conducts. 

The filter resistors we use in the RC protection filters, reduce the
current so the gap can easily stop conducting.  At that point the current
flowing in the transformer secondary is switched right back into charging
the primary cap again.  The advantage is that this current is larger than
one would normally see (in a purely resonant system) and it is capable of
charging a larger cap.

>The inductor (xformerīs secondary) will try to
>resist the cut-off. I.e: it feeds itīs energy into the still conducting
>spark gap. This will increase quench time, which in turn will
>not allow for a first notch quench (of which we do not know
>if it is really necessary for nice long sparks). 

Don't forget the filter resistors, they effectively limit the current
feeding the gap but still easily allow enough current to flow to keep the
stored energy in the transformer.  The value of these resistors may be more
important than we think due to this.  They are now limiting the current
keeping the gap from over conducting but still have to maintain the current
from the transformer.

>The voltage at
>which the gap fires and accounts for the number of Joules
>being "injected" into the primary is (mainly) a question of gap
>spacing. I.e.: if the gap is set to fire at 15kV, the equation
>0.5*V^2*C will give you the Joules. I can see no way in which
>the voltage (and the Joules) will rise above this level. 

It won't.  The key is that once the gap stops conducting, the transformer
is able to supply larger currents to the primary cap allowing for faster
recharging.  In the case of LTR systems, that extra current is used to
charge larger primary caps and thus get more joules stored in the larger
cap even though the firing voltage still stays the same.  As John found, If
your caps are too small, the extra current will cause over charging.  If
the caps are REALLY small, this stored energy in the secondary coil will
charge a small cap to 100kV+ volts.  Although there are losses that prevent
the voltages from going ultra high, consider if the neon were able to dump
all of it's 24 joules of stored energy into a 10nF cap.  That works out to

>course, I am ignoring the fact that a static gap will not always
>fire at voltage "x", (due to preionization, etc) so there WILL
>be variations, but not due to the energy stored in the inductor.

For this effect to really work, the gap needs to fire late.  I think that
means you need a sync rotary gap like John and I use.  However, there may
be ways to use a static gaps too.  My gap fires about 30 degrees late while
John's fires 50 degrees late.

>If the voltage "kick" that the inductor (xformer) produces, after
>the gap opens, charges the capacitor up to firing voltage "x",
>the bps rate (in a static gap) will increase. This might account
>for the fact (i.e: the real mechanism behind it?) that one
>actually can get a BPS that is many times Fmains in a coil
>using a static gap. The statements above are made for a
>coil using a STATIC gap.

I am sure my system fires at the 120BPS rate and no faster.  I can't
imagine a 28nF cap hooked to a 60mA neon would charge faster than 120BPS
(although I used to not thinks it could charge anything higher than 10.6nF
:-)).  I think John has tried higher rates?

>If we now use a RSG, instead of a static gap, the picture changes
>somewhat. The RSG can NOT "fire at will", so there indeed may
>be a considerable voltage rise during the non conduction time. Up
>to a certain point, this would mean the lower the breakrate, the
>higher the voltage can rise, simply because there is more time in
>between the breaks. If I am thinking straight, the voltage rise is
>dependant on three things:
>1.) The inductance of the xformer:

I think that is about 3700H per leg!!

>a.) The more inductance, the higher the possible voltage.
>b.) The higher the inductance, the more time is needed.
>2.) The quiescent time:
>a.) The longer the time (up to a limit given by 1), the higher
>the voltage can rise.

The gap actually has to fire significantly after the voltage peak.

>3.) The setting of the safety gap.

The safety gap should not fire and thus not be a factor?

>However, all this would also suggest to me, that each coil setup
>would only be "happy" with "itīs" specific break rate, because
>the xformer inductance (and primary cap) is different in every
>setup. I remember John tried "ultra low" bps rates (60 bps)
>and the results werenīt very encouraging. John, did you (or
>could you) retry those experiments using once a xformer
>with a low secondary inductance and once using one with a
>high inductance. The setup with the high inductance *might*
>benefit from the low bps.

All high voltage transformers have a giant secondary inductance...  I think
one would have to rectify the AC for lower than 120PBS operation since the
transformer voltage would go through 360 degrees.  The net charging current
would be zero over that cycle...

>Comments welcome.
>Coiler greets from germany,

Hope I have shed some light here...  This effect may open up a number of
new opportunities besides just the LTR coils...