Re: Planning a magnifier

["Antonio Carlos M. de Queiroz" <acmq-at-compuland-dot-com.br> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

John Freau wrote:

>...
> There may be a way to get a little more performance from a
> magnifier which is described at Antonio's webpage, but this has
> never been demonstrated in a working coil.
>...

Actualy, it was. Not in a coil intended to produce sparks, but in a
high-power pulse generator. See the references in my site. The patent
can be obtained from:
http://ep.espacenet-dot-com/
Patent US4833421
The papers by F. M. Bieniosek are available from someone's site
(Terry?).

The ideas are, in sequence:

1) The objective is to obtain fast energy transfer, reducing the losses 
   due to current flow in the primary gap during many cycles.
2) This requires tight coupling in a conventional Tesla coil: k=0.6.
3) Excessive proximity between primary and secondary results in
sparking.
4) Split the secondary in two, moving away the third coil with its
   high-voltage terminal. This ends the design of a conventional
   magnifier. The relation L1*C1=(L2+L3)*C3 holds, approximately.

But you can do better:

5) The tightly coupled (k>0.6) primary-secondary transformer adds 
   significant capacitance to the secondary windings. The energy
   stored there is not transferred to the output, causes stresses
   in the system after the main discharge, imprecision in the tuning,
   and inefficiency.
6) Add -more- capacitance to the secondary coil, and design the system
   so -all- the energy is in the third coil capacitance and terminal
   when the output voltage is maximum. The required design equations
   are described in my site:
   http://www.coe.ufrj.br/~acmq/tesla/magnifier.html

And I could add:

7) The energy in the self-capacitance of the third coil is not available
   at the terminal when the output voltage is maximum. 
8) Split the third coil in two (!), add more capacitance at the 
   connection, and end with a system that squeezes a bit more of 
   energy into the terminal. Actually, the proccess can be repeated 
   any number of times. The only fundamental limitation is that the 
   last coil has always some self capacitance, and the energy stored 
   on it cannot be made immediately available. The theory behind this is
   described in a paper that I wrote recently, and that is available
   in the papers section of my web site ("Synthesis of Multiple
   Resonance Networks").

Antonio Carlos M. de Queiroz