[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Tesla Coil Blunderbusses



Original poster: "R.E.Burnett by way of Terry Fritz <twftesla-at-uswest-dot-net>" <R.E.Burnett-at-newcastle.ac.uk>


Hi Malcolm, all,

I have been pretty busy recently with real work,  but I am following this
thread with great interest !

Over the last two years I have done quite a lot of simulation work on
AC resonant charging behaviour.  My observations have been that the
ballast inductor (or leakage inductance of NST) is primarily charged
during the first portion of the capacitor charging cycle, (when the 
tank capacitor is empty.)  Then during the later portion of the cycle,
the ballast inductor becomes a "source" and releases this energy to charge
the tank capacitor to a higher voltage.

(For those who find this explanation a little too concise,  there are some
graphics and explanations of this effect on my web site.)

I understand what you are saying Malcolm,  about the gap conduction
re-directing the energy from the supply into the ballast during that time.
In my opinion the amount of energy stored in the ballast as a result of
the gap being "closed" would be comparatively small compared to say a 5ms
charging time.  However if your static gap is firing at a high repetition
rate,  then I appreciate that the duration the gap is closed could become
significant.

My gut feeling is that the boost due to the gap conduction time won't make
more than a few percent difference to the peak voltage.  But, it is
definately this "inductive-kick" from the ballast which allows widely
spaced static gaps to fire smoothly.  The switch-on transient can be
enough to initially breakdown the gap,  and then the inductive kick from
the previous firing of the gap boosts the voltage to achieve the next gap
firing !

I have not had time to do so yet,  but I will run some Microsim 
simulations to check out this theory about the quench time.  The results
are of great interest to me,  as I had always considered the charging
system and the coupled resonant circuits as two seperate entities during
simulations so it would run quicker.  Now you're telling me they might
interact ! ;-))

Looking at the double-resonant behaviour,  I did find one interesting
fact when experimenting with tuning and coupling some time ago.  I found
that if the primary and secondary resonant circuits WERE NOT tuned
accurately to the same resonant frequencies,  then the energy transfer
from primary to secondary was incomplete.  (This was in the simulation
world,  without any accounting for corona loading,  etc.)

The result of this inaccurate tuning,  was that the primary current
notches were not as deep.  Specifically,  the envelope of the RF current
in the primary circuit went close to zero amplitude,  but did not quite
reach zero.  My conclusion at the time,  was that accurate tuning is 
important to get a deep primary notch,  and therefore give the spark gap
a decent chance of interrupting the current.

All of this was without considering the drop if Fsec due to corona
capacitance.  I also thought that good spark breakout loads the secondary
heavily.  As a result good spark breakout actually dissipates the
secondary energy quickly, and little is coupled back to the primary
circuit.  My conclusion here is that good spark breakout leads to an early
quench in the primary gap,  (not the other way around !)  If you suppress
breakout then the primary gap does not quench ???

Out of interest,  my CW inverter driven TC also displays intermittent
spark breakout when the driver is tuned too low compared to the natural
frequency of the resonator.  I have noticed that the Fres drops by 3 or 4%
during CW spark loading (8" spark.)  But if the driver is detuned further,
then sometimes the resonator would ring-up and pull itself into tune,
resulting in a spark.  Other times it would not ring-up so it doesn't pull
into tune,  and no spark would breakout.  This says to me that some
detuning occurs BEFORE the visible spark breaks out ???

I'm not sure if the SSTC behaviour is really the same behaviour you are
discussing,  but there may be some clues in there somewhere ;-)

A very interesting area.  I will let you know what the sims turn up.

						Cheers,

						-Richie Burnett,
						(Newcastle, UK)

> Original poster: "Malcolm Watts by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <m.j.watts-at-massey.ac.nz>
> 
> Hi Terry,
> 
> Pondering further I realize that for air streamers where energy 
> trades continue, tuning to the streamers would definitely improve 
> performance. For a single pri-sec trade, I can't see that it should 
> matter. Which still leaves the question of what I clearly observed 
> when tuning the primary to the LSB. I don't have the modelling tools 
> handy. Would you mind trying it in uSim and posting the results? Any 
> info that links pri-sec tuning to gap firing conundrums would also be 
> welcomed.
> 
> Regards,
> malcolm