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

Re: Frequency Split



tesla-at-grendel.objinc-dot-com On Mon, 18 Mar 1996 20:01:11 +0700, you
wrote:

>>From MALCOLM-at-directorate.wnp.ac.nz Mon Mar 18 14:12 MST 1996

>In reply to some queries....
>
[snip]

>>     1.  Why are you calling these K values sweet?
>>         Because they couple in an integer number of 
>>         cycles?
>
>The maximum amplitude of the two frequencies coincide in the secondary 
>when these values of k are used. The tighter the coupling, the fewer
>the half cycles of oscillation needed to achieve this. This means
>fewer pri gap fires with consequently fewer gap losses. For example,
>with k=0.6, just 2 gap conductions are required for this condition to 
>be met. This is typically what they use in SF6 surrounded 2-coil
>accelerator systems.
> 
>>     2.  Once I've picked a coupling constant, do I use it
>>         for setting my spark gap commutation time?
>
>Yes. The optimum gap conduction time (read switch closed time) is
>set by this. If the gap opens too soon, the maxima won't get to
>coincide and too late, and the energy starts oscillating back into 
>the primary circuit. At the time the maxima coincide in the secondary
>circuit, primary energy is ideally nil (zero Vcp and zero Ip). You
>can see when the maxima occurs by single-shot scoping the secondary 
>with no discharge rod. The relation is:
So,
	When I get a small signal k of 0.276  n=4.
>	  0.28   |   4   |  1.33
	I'm seeing say, 10 cycles. That means that I wasting power in
my spark gap in the 6 cycles that it's still conducting past the 4
that are needed. Aside: with the gap set at 0.030" I get an Req of 22
ohms based on the d=ln((V0/Vn)^1/n). I'm also seeing only 3 firings of
my gap per 60hz half cycle: at  watts=.5 * c * v^2 * n * f ,  Even
with 15Kv for V, I'm only using 460 watts of the neon's 900 watts. I
need to double my spark gap's firings rate and quench it sooner.

	jim