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

Re: NSTs and Cap charging tests (fwd)





---------- Forwarded message ----------
Date: Fri, 16 Apr 1999 09:42:12 EDT
From: FutureT-at-aol-dot-com
To: tesla-at-pupman-dot-com
Subject: Re: NSTs and Cap charging tests (fwd)

In a message dated 99-04-16 05:01:51 EDT, you write:

<< John, Reinhard, Tony, Terry, Malcom, and All.
 
> John, your practical results seem very close to my simulations,
> I really wish I had a chance to do more practical tests,  as I think
> there is much to be learned in optimising the charging system for AC.

Richie, all,

Yes, and I just posted some more results using pfc vs. no pfc which
also support your simulations.
 
> John,  Are you using a 120 BPS rotary for your tests ?
> If so,  

Yes, 120 BPS.

>I have not done many simulations with 100BPS (or 120 BPS for
> 60Hz.)  I concentrated mainly on 200BPS for 50Hz,  because I found
> that the voltages are about 30% less,  and I can get similar power
> throughput with smaller capacitors due to the higher break rate.
> (Will this give similar performance in practice ?)

That's the big question.  Recent work I've done suggests that 120bps
outperforms higher break rates.  However I used non-sync rotaries for
the higher break-rate work.  I've always thought that that at higher
break-rates, there would be little difference in sync vs. non-sync
operation, efficiency-wise.  I could be wrong.  I have some
evidence that the relative underformance at higher break-rates is caused
by charging inefficiencies (at least in my coil) rather than by spark
streamer dynamics.
 
> Can you get a measure of the firing angle of the rotary by any means ?
> I usually record the time from the zero crossing of the supply voltage
> to the actual gap firing.  Maybe the orientation of the motor or
> electrodes indicates the firing angle.

I could do this with a scope, and I seem to remember that i checked
it once.  I'll have to check my old notes.
 
> The posting from Tony Greer is very interesting because this displays
> something which I have seen in my simulations but have not heard of in
> practice:-
 
> Tony Greer wrote:
> > ..... But when 96 volts was reached, input current
> > SUDDENLY droped to about 16 Amps, and the output current droped from 200
> > mA to 100mA, and became very steady....
> > ..... but most noticeable was the different sound of the gap and
> > streamers. At between 96 and 98 Volts, the noise became a steady buzz,
> > like a single note. Probably firing in sync with the line, because
> > raising the Voltage past 98 Volts would again cause the current to
> > double and the noise would regain it's raspy sound.....
 
> In one of my previous posts I commented on a few "special" capacitor
> values
> which would cause the static gap to fire smoothely as Tony described,
> at all other values the firing was fairly chaotic.  I never tried
> simulations to find the effect of different supply voltages though.

Yes, that is interesting, I haven't seen a voltage range effect like that
in my coils.  It could be the cap size, combined with the certain
voltage as you suggest, but I don't really know.

> I also found that it is easier to get a high power factor by using
> a correctly set synchronous gap,  than by using a static gap.  The
> power factor for the static gap only seemed to approach that of the
> rotary gap,  when the gap is set to fire at a VERY high voltage.

It makes sense, considering what we now know about the charging.
I wonder if some reports of quench failure at high powers using static
gaps was really the result of an abysmal power factor instead?
(it's time for the wattmeter!)
 
> Like John, I found that it is possible to find certain firing angles
> which draw alot of current but produce little spark.  I found these at
> timings near to the zero crossings of the HV voltage,  where the
> rotary begins to miss fire,  and the safety gap kicks in !  In these
> cases the simulation shows that the voltage builds due to resonant rise
> and the voltage is near to zero when electrode presentations occur,
> therefore gap firings take little energy out of the system.  Current
> is mearly sloshing in and out of the HV capacitor.  Under these conditions
> I noticed the voltage get very high between gap firings,  (Another
> reason for changing to a 200BPS rotary.)

I agree about the bad sync phase settings.  I've done some work with
higher sync rotary break-rates in the past; 240, 360, 480, 720 BPS,
but this work was not carefully done.  Also, I used an unusual bunched
electrode method, that concentrated the power into the peaks of the
input sine wave.  As expected, this method drew a lot of power, but
I didn't see any advantage efficiency-wise.  My general concern with
using 240 or higher BPS for a sync gap is the shorting of the power
supply during gap firing (due to firing angle), which i think would
cause lower efficiency of charging.  
 
> For my new TC I have chosen a 200 BPS (50Hz) synchronous rotary, and
> have already operated it with a 6kV 125mA neon transformer.
> The firing angle was adjusted to obtain two equal energy bangs per
> half-cycle.  The capacitor voltage was checked with a scope also:-

Again, my big concern with 200 BPS is, will it give the same spark
production efficiency as 100 BPS.  
 
> Tank capacitor voltage waveform over one complete supply cycle:-
              __
             /  \                    Both bangs are equal amplitude
      /!    /    !                   but there is always a voltage
     / !   /     !                   overshoot before the second firing
    /  !  /      !                   of the gap.  This is where the cap
   /   ! /       ! __                is charging on the downside of the
  /    !/        !/  \               AC supply voltage.
                      \   !\       !
                       \  ! \      ! Since the rotation of the electodes
                        \ !  \     ! is synchronised to the supply freq.
                         \!   \    ! the charging waveform is always the
                               \__/  same for every mains cycle.
 
> Several parameters such as current draw, peak cap voltage, and firing
> angle were all found to be within 10% of the simulated values.  It's a
> pity that PSpice cannot estimated spark lengths !  This system charged
> a 44nF cap to 11400 Volts,  200 times per second,  and produced 2 ft
> streamers into the air.  However I did not have time to try adjusting
> the primary tuning. :-(
 
> Does this sound reasonable in terms of performance ?  Supply VA was
> measured at 668VA.

It's certainly reasonable but I am getting 42" using about the same VA,
at 120BPS, (but I did adjust my primary tuning).  Have you simulated
the power factor issues for 200 vs. 100 BPS synchronous?  In any case
I think it's important to try different break-rates to compare. 
 
> I have got bigger sparks with the same transformer using static gaps,
> but the firing was erratic and the supply VA was more than double !
> With the rotary, the TC is running so much smoother,  and the sound is
> totally different..  It produces a buzzy wine rather than white noise.

It would be interesting to compare the efficiency of 200 BPS sync
operation with non-sync operation at a similar break rate of 180,
or 220 BPS, etc.  It would be interesting if the poor efficiency I'm 
seeing at higher break-rates was due to the non-sync rotary, not
to the high break-rate!

John Freau

 
> I plan to run the system from a 10kv / 224mA Radar power transformer,
> and have detailed information if anyone would be kind enough (and bold
> enough) to predict performance.  I would also appreciate having someone
> more experienced check it over,  before I toast something expensive
> due to a silly error.  Any offers ?  Please contact off list.
 
 
 >			Richie,
  >>