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

Re: Skin Effect & Primary Current?

Hi Richard,

> How did you capture the damped wave?  You should have a good calibrated 
> Rogowski coil and mesure the exact peak current for a true idea of the 
> spark gaps actual resistance which in most tela circuits amounts to as 
> much as ten ohms!  The more normal value is 1-4 ohms.  Arcs of the low 
> value you measured are usually not normally encountered until the current 
> is over ten thousand amps.  The decrement is, indeed, used to calculate a 
> number of loss factors in an LC circuit, but without a very accruate 
> value of peak current it would be tough to use. The ultimate would be to 
> get a current transformer reading as well as a second channel voltage 
> reading across the gap and integrate the two to yeild the actual power 
> loss and resisatance of the gap over time.

You're right. I didn't include enough data to make sense of this. Cp
was 0.1uF, Lp about 12uH. The waveform was captured on a digital 
storage scope with an aerial hung a few feet away from the primary.
I repeated this with a range of L/C's. I was stunned to find that in
all cases, the decrementing waveform gave a Q of about 10. The range
of capacitances tested was 25 to 100nF and L's ranged from the one 
given (the resident primary for the coil I mentioned elsewhere) to
a coil wound from high quality 80 strand Litz measuring 330uH. Not
only did all give that Q, but it was peripherally verified by the 
number of oscillations recorded until the gap extinguished (about 20 
- 22). I was able to changed the scope timebase and superimpose the 
captured waveform on top of a previously captured waveform. The match 
was amazing. In all cases, the gap was set for an identical firing 
voltage. In fact I was using the voltage decrement to do the 
calculation.
    Looking at Q = 1/R x SQRT(L/C) showed that for the higher L/C
ratios the gap resistance was much higher than that earlier 
combination. It was clear then that gap resistance was inversely
proportional to current. For 330uH and 25nF then, this Q gives a
gap+circuit resistance of 11.5 Ohms which fits well with your finding 
I think.
     The most dramatic realization for me was that changing L/C ratios
in the primary had no noticeable effect on dynamic Q. The other one 
was seeing just how much power is being lost in the gap. Looking at 
these results from several angles suggested that the gap had all the 
characteristics of a pair of back-to-back zeners with a voltage 
drop of about 60 which seemed a reasonable conclusion for a gas 
discharge.

Further thoughts on this welcome.
Malcolm