Re: Improved Model for a Primary Charging CKT

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Antonio Carlos M. de Queiroz" <acmq-at-compuland-dot-com.br> 

Tesla list wrote:

 > Original poster: "Robert Jones" <alwynj48-at-earthlink-dot-net>

 > I believe Antonio comments are correct in the general case.
 >
 > However just inspection of  the equations provides considerable insight in
 > to the apparently chaotic behavior.
 > i.e. none integer ratios of the LC to supply frequency.  Where as simple
 > integer ratios much like the magic k values may result in simple time
 > traces.

This only causes the circuit to be nonresonant. The steady state
waveform is a sinusoid at the supply frequency. The transients
associated with the gap firing, these have another frequency,
function of the RLC elements.

 > For particular cases the problem may not be so intractable.
 > For example if you assume the sg fires at peak voltage, zero supply voltage
 > and zero cap current this may effectively make it a half cycle solution.
 > Which is applicable if deQ diodes are used and or a sync rotary gap/
 > triggered gap.

Yes, there are several simple cases, and this is probably the
simplest. Sync gaps simplify the analysis substantially.

 > In any case values like peak cap voltages and rms charging currents should
 > be derivable.

With a sync gap, yes. Everything is derivable without great trouble.

 > I have the fundamental equations in s and to as a mathcad file if any one
 > would like them.
 > They can be read with a free web plug in from mathsoft.
 >
 > Writing this it has just occurred to me that one approach to the general
 > problem may be to use power spectrums. Although it will not give you any
 > specific time traces it will produce rms values and even peak values with
 > probabilities in some cases.

Yes. With a sync gap all the waveforms are ideally periodical.

 > Yes circuit simulation is very usfull but you can spend a lot of time
 > exploring the parameter space. Maths can tell you were its best to look.

I agree. But simulators are good to verify calculations.

Antonio Carlos M. de Queiroz