Re: Parallel and Series LCR Circuit Qs

["Malcolm Watts" <M.J.Watts-at-massey.ac.nz> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

Hi Richard,

On 20 Aug 00, at 10:33, Tesla list wrote:

> Original poster: "Richard Kircher" <richard.kircher-at-worldnet.att-dot-net> 
> 
> Hi All,
> 	It sure sounds like a lot of things are important in TC design.  Maybe the
> prime question is "what's most important".  I suggest the key to that
> answer is $$$'s (component cost) and your source of energy.  Certainly,
> without a high Cp/Cs ratio there won't be much voltage or possibly a
> decline in voltage as an extreme.  Voltage is squared in the Energy
> equation while capacitance is not.  I did not mean to say that a high Cp/Cs
> ratio should be obtained by making Cs small.  I agree with all that say Cs
> should be large, but that only means that Cp also has to be large in
> proportion to maintain the high Cp/Cs ratio.
> 
> 	I really don't know where the compromise should be made when Cp size is
> limited by $$$'s for the primary capacitor and power source (AC Transformer
> or DC Supply).  I would take my hat off to anyone that has that answer
> nailed down taking all variables and non-linearities into account.  Of
> course, we all learn quickly that it is not just dollars and Cp that limits
> a TC's performance, there are other factors like Lp (primary coil) physical
> limitations, proportion, silver plated copper, efficiency, voltage
> break-down ratings, cost, cost, cost, etc. too.
> 
> 	Efficiency and Impedance matching are also intriguing and important
> subjects.  I still think efficiency, (aka Q) is of prime importance FOR THE
> OVERALL COIL.  The secondary dissipative losses are not that important
> because the reactive impedance levels are high by comparison, so who care
> about 100 ohms.  On the primary side, losses in the coil, spark gap, and
> capacitor are very important.  The 10,000 amp primary current also flows
> thru the capacitor and makes me a little concerned about the MMC made from
> the interconnection of a lot of "thin" (compared to copper tubing) wire
> leads.  MMC's are very popular these days so the thin wire lead worry and
> internal construction may not be valid - another "don't know" for me.

You are absolutely right to be concerned about this. There is 
no getting around it: #of paralleled MMC strings must be 
proportional to primary current and because there is a 
substantial time difference between the rate of heating vs the 
rate of heat loss and bearing in mind that power loss is 
proportional to I^2 I'd rate mine on (Ipk X BPS)^2 for a 
disruptive coil. Anyone else have a view on the validity of 
that formula? Forced air cooling might alter this somewhat. 

> 	Most confusing to me is impedance matching to the load/streamer.  Some
> folks think that the most power transfer occurs when the source impedance
> is matched to the load impedance.  Don't try that idea with your car
> battery.  The most power transfer occurs when the source impedance is ZERO
> ohms and this applies to all lumped component circuits.  Of course,
> absolute zero is not achievable especially when the Cp/Cs ratio is high in
> the world of realizable coil parameters.  Given that the streamer impedance
> can't be changed or matched, it would seem the only thing left to do is to
> make the coil output resistance as low as possible (high efficiency) and
> match the reactive component.  

To what exactly? Suppose that the secondary contains all 
energy when the output discharge is initiated and further 
suppose that the gap has been quenched. At resonance, there is 
no nett reactive component in the secondary is there? Wouldn't 
the streamer choose to do what the secondary allows it to? 
 
> 	Terry's work indicates that the streamer impedance is capacitive (1pF per
> foot)  Therefore it sounds like the coil should be inductive to tune it
> out.  Alternately, it would seem that the streamer de-tunes the coil and
> lowers the secondary resonate frequency.  So, I guess the primary should be
> tuned lower in frequency than the secondary to start with (?).  I wonder
> how much lower?  Maybe the answer is to just tune the coil by trial and
> error until the maximum streamer length is reached and forget about the
> grid dip meter.  Any ideas about this (or any ramblings) would be welcome.

The streamer capacitance does pull the secondary lower. Tuning 
the primary to a lower frequency does help but might have more 
to do with altering the apparent coupling between pri and sec  
and it also modifies the sideband peaks, tending to favour the 
lower of the two at the expense of the upper.

Regards,
malcolm