# Re: Rotary Sparkgap

• To: tesla-at-grendel.objinc-dot-com
• Subject: Re: Rotary Sparkgap
• From: jim.fosse-at-bdt-dot-com (jim fosse)
• Date: Wed, 28 Feb 96 23:27 PST
• >Received: from bdt.bdt-dot-com (root-at-bdt-dot-com [140.174.173.10]) by uucp-1.csn-dot-net (8.6.12/8.6.12) with SMTP id AAA29740 for <tesla-at-grendel.objinc-dot-com>; Thu, 29 Feb 1996 00:27:44 -0700

```Malcolm,
You are CORRECT! We can argue about the magnitude of your
correctness later:)

I was sitting here composing my reply in my head: We are worlds apart both
physically and metaphorically because of unspoken assumptions on each of our
parts.... I then composed: Do this thought experiment: measure the output
voltage of your neon, connect up your cap, measure the output voltage. Did
it change?

Then I thought, don't tell him to do a thought experiment, go measure the
bugger yourself! So I did.

I connected my 20Kv voltage divider to 1 side of my 15KV 60ma. neon and case
ground. The divider is 220M ohms to 33K ohms. my Simpson 260 is 5K/volt, 2.5
volt scale. therefor 220M || 9K = 24.3:1 ratio. Connected to 1/2 the
primary, I measured .38V => 9.2KV. I connected up my 13.7nF cap across both
output terminals (I ran some numbers from your last post and they indicated
that my cap would resonate with my neon). I measured .57V => 13.8KV. Now
lets multiply my measured numbers by 2. Unloaded: 18.4Kv, Cap semiresonant:
27.6KV. Not bad for a 15KV neon. Q=27.6/18.4 = 1.5 but I'm probably not
really in resonance!

>Continuing on this from Jim Fosse....
>
>> But it most certainly does resonate
>> >at low-level, and at least three people here have encountered diffi-
>> >culties associated with resonance's under normal running conditions.
>> >I think the 360mA IS invalid. BUT, any resonance implies a higher
>> >current than the limited value into a resistive load. I have opened
>> >up a spark gap to well over the rated peak voltage and I know Ed
>> >Philips has too.
>>

OOPS! I saw the safety gap firing, not your main gap!

>> I'm sure I have also!
>> But is this due to the TC primary L and C or to the neon L and C?
>> I can't answer this right now, I tore everything apart tonight to measure my
>> neons.

My thought here was that the voltage rise was due to the LC of the TC
primary, not the neon's secondary inductance and the C.

>
>It has to be the transformer inductance and capacitor IMHO for
>two reasons : (1) The spark gap has to initially fire which it
>can't do if the transformer won't deliver sufficient energy to
>charge the cap up to that point (note that Cp and Lp of the coil
>is a mile off resonance at 50/60Hz), and (2), if the capacitor
>voltage at ANY stage does exceed that of the transformer, where
>does the EXTRA energy come from - bear in mind Vcp = SQRT(2Ecp).
>
>> >> If I remember correctly, the leakage inductance of a transformer is
>> >> measured by shorting the secondary and then measuring the
>> >> inductance of the primary.
>> >
>> >Shorting the primary should achieve the same end as shorting the
>> >secondary IMHO. The idea of using a low-Z generator was to achieve
>> >exactly that.
>

My thought was: when modeling the neon, a leakage/limiting inductance in
series with primary would more accurately simulate the operation of the
system than a limiting inductance in series with the secondary would.

>> NO, real world here! the secondary is so much BIGGER and future away from
>> the core than the primary (of the neon) that it will have a much larger
>> leakage inductance than the primary will (both cases with the shunts removed!
>> ).
>
>I must have misinterpreted. I was measuring the TOTAL secondary
>inductance, not just leakage. Sorry bout that.
>
I was looking to separate the "pure" inductance of a winding and the leakage
inductance due to the shunt for modeling/understanding purposes.

>Thanks for the tables.
no problem.

>Malcolm
>
jim

```