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

Re: Best cap size for a sync gap



Original poster: Terry Fritz <teslalist-at-twfpowerelectronics-dot-com>

Hi Gary,

At 08:52 PM 4/28/2004, you wrote:
>I finally "got around to it" to an experiment that's been nagging me for a 
>very long time.  I wanted to determine the optimum cap size for a coil, in 
>terms of maximizing the power pulled from the NST.  Tank frequency issues 
>come later.

A defining experiment indeed!!!!


>With a newly constructed sync RSG and a dummy load consisting of three 
>500W halogen lamps in series in place of the primary inductor, I scoped 
>the gap voltage.  As is typical with sync gaps, I adjusted the phase such 
>that the bang occurs somewhat after the peak charging voltage.  The later 
>I phased it, the brighter the lamps and the higher the bang voltage. If 
>the timing is too late, the gap stops firing altogether, so I brought it 
>just to the brink of this point.
>
>To measure the voltage, I use Terry's fiber optic probe, which sadly, I 
>have yet to accurately calibrate.  But for the purposes of finding the 
>best cap size, even a qualitative measurement is adequate.

No problem in this case at all.


>I use an unmodified 15/60 NST, cranked up to 144VAC.

Note that the tranny may act "odd" at 20% over voltage input.  Might be 
best to run at the name plate voltage.

>I had available two .02uF caps, and one .01uF cap.  With these I measured 
>the bang voltages using .02, .03, .04, and .05uF.  The peak-to-peak bang 
>voltages measured were 688, 618, 552, and 482 mV respectively, as directly 
>indicated on the scope.  FWIW, if I scope just the unloaded NST secondary 
>-at-120VAC input, I get 598mV p-p.

It would be nice if you could have a bit more variability for the 
capacitance,  10nF is a big jump....  There may have been peaks that such 
large steps missed...


>If I calculate the relative bang size with a simple scale-less formula of 
>C*V*V (mV*uF*uF), I get
>.02   9,467
>.03  11,478
>.04  12,188
>.05  11,616
>
> >From this I conclude that using a .04uF cap with my 15/60 NST will 
> result in the highest power throughput.

Facilitating!!!


>Has anyone else performed such an experiment?

Nope!! ;-))

>Just trying to understand why my result is so at odds with the widely 
>suggested value of .028uF for the same power supply.  Hmmm, wonder what 
>I'd have gotten if I had tested at 120VAC?

The question I would have is did the tranny "hum" loudly indicating shunt 
saturation???  If an NST's shunt go into saturation, all bets are off.  In 
your tests you make have gone into that black area where we know 
zilch...  But a Saturating NST (SLTR) system should be able to charge 
pretty big caps as your test suggest!!!!

We know that if the shunts go into saturation, very odd and wonderful 
things can go on.  But we know no details....  Maybe a cheap kill-a-watt 
meter on the input would provide more crucial details for those of us that 
try to get the computer models to match reality ;-)))

I think your shunts are saturating and thus blowing the normal "old" models 
out of the water.  But things seem stable an "happy" in your experiments 
which suggests that such regions are reasonable to work at!!!  I am worried 
that such regions my be very finicky depending on the brand and 
characteristics of an exact type/brand of NST....

Cheers,

         Terry



>Regards, Gary Lau
>MA, USA