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

Re: Residual Cap Voltage

Hi Steve,

At 12:31 PM 01/01/2000 -0700, you wrote: 
>
> Hi folks,
>  
> The following is a question for those who can base their answers on actual
> measurements (scope traces, etc.):



What!!  No wild, unsubstantiated, arm waving, armchair guesses!!! ;-)))


>
>  
> I am interested in knowing what the approximate typical residual voltage is
> immediately after a "bang" of the spark gap.  Assume the TC has operated for
> a half a minute or so to achieve typical spark gap conditions (temperature,
> ions, etc.).    


I have never noted much difference between a cold gap and a hot gap in relation
to such things...


>
>  
> I visualize if the cap charges up to V volts, then right after a bang (gap
> not conducting until the next bang) it will be less than a tenth V, since the
> gap is quite a low impedance while it is conducting.  



That appears to be quite true...

>
>  
> This information should be useful for determining the peak current one can
> expect in the cap, which could be useful in designing MMCs.  If the cap
> discharges to near zero volts, then the peak current will be about V/Z for
> each bang, where Z is the impedance seen by the cap during the start of each
> bang.  But if the residual cap voltage ends up at, say, half V, than the peak
> current could be up to 1.5V/Z on the next half-cycle of the AC charging. 



Remember that the gap and associated wiring have capacitance across them that
tend to make a very powerful multi GHz noise spikes when the gap first fires. 
These spikes are also seen (to a much lessor degree at every zero current
crossing during the firing waveform.  The actual peak is probably unknown since
it appears to occur at frequencies higher than our high voltage current
instruments can measure. The GHz+ guess is based on the way it saturates 1GHz
BW fiber optic LEDs used in such measurements and reading from 100MHz+ antenna
systems.  However, one could argue that the "power" in those spikes is far less
significant than the typical RMS current that flows in the tank at just the
fundamental frequency.

As usual, I'll refer you to my paper "Modeled and Actual Voltage and Current
Waveforms within a Tesla Coil" at :

http://www.peakpeak-dot-com/~terryf/tesla/experiments/experiments.html

This is the best paper I ever wrote in that it is a great reference for all
such questions.  Specifically, the following real scope traces...

These are the primary capacitor voltages and currents in 5mS/div and 25uS/div
time scales.  The top traces are voltage (20kV/div) and the bottom traces are
current (100A/div).

http://www.peakpeak-dot-com/~terryf/tesla/experiments/modact/Image170.gif

http://www.peakpeak-dot-com/~terryf/tesla/experiments/modact/Image172.gif

It is not unusual to see a few thousand volts getting left on the cap (about
10%) after a firing cycle (note the middle two cycles that seem to have ~1-2Kv
still on them after the firing burst and the offset in the second picture). 
Note that these are ADDING to the firing current as you suspect!  However, The
"residual voltage" is usually very close (hundreds of volts) to zero.  I would
think you can simply use the firing voltage for your purposes since that
voltage usually "jitters" a few thousand volts anyway.  The residual voltage
just adds a little more noise into the whole picture.  I can guarantee that you
will not see 40%, 60%, 80%, or any percentage like of the firing voltage
getting left on the primary cap unless you do something really bazaar in the
way of quenching and the quenching gets far "too" aggressive.  These waveforms
or very typical of all such waveforms I have "actually" seen.  There have NEVER
been any significant exceptions.

Hope this is what you were looking for.

Cheers,

        Terry

>
>  
> Thanks in advance for comments, and welcome to another great (and safe!) year
> of coiling!
>  
> --Steve Young
>