[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [TCML] Spark gap Resistance
In a message dated 11/19/2007 10:14:29 P.M. US Eastern Standard Time,
bartb@xxxxxxxxxxxxxxxx writes:
I think Chris brings up a "very" valid point! The problem is, it is
experience that drives the idea of a higher surge impedance. There's a
reason it's termed "surge" impedance. There is a difficulty at quenching
after the first notch. Those who have done this have reported in the
past (from my memory) that they had better sparks lengths on 2nd or 3rd
notch quenching as compared to 1st notch quenching.
Those experiments which obtained 1st notch quenching did so by increasing
losses by using multiple gaps. It was a bad trade off. I think the reason
it's better to let a coil quench at 2nd or 3rd notch if it "wants" to, is
because
most of the spark growth occurs during the first transfer. So the extra
transfers
(during a single bang)
don't do much except waste some power and steal some cap charging time.
If one forces the 1st notch quenching by using lossy gaps, then even that
first
important transfer is weakened. This reduces the spark length. There is
evidence that large powerful coils more easily quench on the first notch.
For example Ed Wingate's magnifier quenches on the first notch at full power.
But at low power I think it quenches on the 3rd notch. I seem to remember
that the "effective" coupling for his magnifier is about k = 0.2 maybe less.
It's more
or less the same as a classic coil. The 12 point series rotary may be
helping
the quenching. It would be interesting to monitor the quenching using a more
typical 4 point series rotary. In any case the lower frequency of a larger
coil
makes the notches wider, so it's easier for the ions to de-ionize during the
notch
and let the gap quench.
Chris, several years ago, 1st notch quenching was the assumed ideal and
we tried to do that for all the reason's you stated. What was found is
that 1st notch quenching was not easy. Then, we found that when it
occurred, it wasn't "wonderful". How could that be? Well, losses of course.
Chris is trying to quench faster by using a fast rotary. That won't happen
because the quench is not controlled by the mechanical dwell time.
The spark at the gap will arc before the electrodes line up, and stretch
out if needed after the electrodes pull apart. To really stretch the spark
and force the quench would require a rotary gap construction well beyond
what anyone has tried. Generally rotaries are more for timing when the
spark occurs rather than for quenching purposes. Quenching is more
about draining the energy out of the system quickly. For the most part,
air streamers don't drain the energy that quickly, and that's the main
problem preventing a fast quench. In any case, as you said Bart, a fast
quench is not really very important. It's interesting to consider the
various DRSSTC designs and SISG coils which have lower switching
losses. They may be more efficient because of that, but the improvement
is not huge compared to a spark gap coil. This shows that the gap losses
(at least the part of the gap losses that matter) are not that large.
It's difficult to figure out and in my mind, it's still "not" figured
out. We do know that when the surge impedance is increased due to higher
inductance, we can get better spark output. But, there of course is a
limit. It is counter-intuitive to physics when all the pieces of the
puzzle are not accounted for. The only way "I" personally can explain it
is that the losses incurred during energy transfer in a single notch
arena are huge.
If the single notch quenching is "forced" by deliberately increasing the
gap losses, then yes the gap losses will be high. It's best not to do that.
It's best to accept a 2nd or 3rd notch quench, but keep the gap losses as
low as possible.
Forcing the quench in that way is really just wasting the energy in the
gap so there's mostly none left by the time the 1st notch arrives.
Now, are those losses in the gap?
If first notch quench is forced by using a lossy gap, then yes the losses
are in the gap. But in all coils the losses are shared between the gap,
the primary and the secondary, in some proportion.
Are they also shared
in the secondary or primary to a large degree? The question is "where
are the losses and what is their distribution" in this 1st notch quench
situation?
Again, if 1st notch quench is forced by making a lossy gap, then
the extra losses are in the gap, or mostly in the gap. If however the
quench could be forced
by an air blast or magnetic field or something similar, then there would
be a benefit because the caps would have more time to charge. But
the spark might not get longer, because energy is not being
transferred faster. We can transfer the energy faster by increasing the
coupling, but then the sparks won't be able to drain out the energy
fast enough. It will again become more difficult to quench on the 1st notch,
and an even stronger air blast or magnets, etc., will be needed.
The reason that tighter coupling increases spark lengths is because
the energy is transferred faster, before too much of it is wasted in gap
losses.
It's all a matter of getting as much energy to the secondary during the
first transfer as far as I can tell. This energy will produce the spark
length,
then if some energy reflects back to the primary, the spark length has
already
been produced.
Were talking about high energy pulse currents. If there is an escape
route, high energy pulse currents will find it.
It seems to me that what we are doing is increasing energy transfer time
to a degree in which the secondary and spark gap can "handle" the energy
as a combined system. I believe that when we attain first notch
quenching, we are simply releasing energy that is not being accounted
for. It's not getting to the sparks, so it's a loss somewhere else.
Yes, when the quench is forced by building a lossy gap, then the gap wastes
more energy. Since the energy is now used up more quickly, it lets the
gap quench more quickly. But if first notch quenching can be obtained
by using air blasts or something similar but without making the gap
lossy, (and without reducing the coupling) then a benefit may be seen.
Reducing the coupling has a similar effect on spark length, as using a more
lossy gap. Both waste energy before it can get to the secondary.
John
Take care,
Bart
FutureT@xxxxxxx wrote:
> In a message dated 11/19/2007 6:54:37 P.M. US Eastern Standard Time,
> list@xxxxxxxxxxxxxxxxxxxxxxxxx writes:
>
> Chris,
>
> If the current is less overall, then the gap losses are lower. Using a
high
> impedance
> primary results in less overall current and less overall losses. When
more
> inductance
> (more turns) are used in the primary, the inductance increases more than
the
> resistance increases, thus the primary losses are reduced. The Q is
higher.
> The result is that
> both the gap losses and the primary losses are reduced. Of course this
only
> works up to a point. At some point the secondary wire will be too
> thin and will show high losses.
>
> Generally low frequencies are believed to be more efficient in producing
> long sparks.
> Maybe something in the range of 30kHz to 150hHz. Also at higher
frequencies,
> it's harder to achieve a first notch quench. The sparks themselves may
grow
> better at low frequencies.
>
> Large coils are generally more efficient than small ones.
>
> Tank caps generally are able to provide their current fast enough for
> TC operations.
>
> Generally high breakrate coils need more input power to produce a given
> spark length. It's not known exactly what breakrate is best. It may vary
> somewhat among coils. Somewhere between 100bps to 200bps usually
> works well.
>
> John
>
>
>> Sorry for the amount of "ponders" in this mail. It is just my 2cents
>>
> worth
>
>> that a higher frequency with less primary turns and a faster RSG
would
>> overall reduce losses far more than anything else.
>>
>
> Chris
>
>
************************************** See what's new at http://www.aol.com
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla