Re: Chokes vs Safety Gaps

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Malcolm Watts" <m.j.watts-at-massey.ac.nz> 

Hi Antonio,

On 12 Feb 2004, at 20:15, Tesla list wrote:

 > Original poster: "Antonio Carlos M. de Queiroz"
 > <acmq-at-compuland-dot-com.br>
 >
 > Tesla list wrote:
 >  >
 >  > Original poster: "Malcolm Watts" <m.j.watts-at-massey.ac.nz>
 > ,
 >  > RLC? That is exactly what I tried to do about 8 years ago. I
 >  > constructed a balanced-about-ground multipole filter with a
 >  > Butterworth characteristic based on the macro component values that
 >  > apparently counted and killed a NST in less than a minute (the only
 >  > one I ever lost incidentally).
 >
 > What did you consider as loads, or terminations, for the filter?
 > These filters require resistive terminations at both sides with
 > correct values for proper behavior. With low damping they will
 > resonate.

I considered a voltage source (approximating the gap) at the gap end
and the transformer itself at the transformer end. I fed it with a
range of frequencies from the gap end close to the coil frequency.
What I did not do (and should have done) was feed it with step
functions :(

 > As the objective is to protect the transformer from pulses
 > coming from the spark gap, it's probably possible to consider
 > the gap as a low impedance source, and the transformer as a
 > high-impedance load, and design a filter between these, that
 > obviously must include resistors, that make the role of the
 > terminations. Note also that the transient response of the
 > filter must be considered, since it will be operating against
 > pulses, and that the inductors must be constructed so they
 > can sustain the high voltages that will appear. Parasitic
 > capacitances may have to be considered in the design.

That was my mistake - I didn't shock-excite the filter at the gap end
so none of the very high frequencies which would have excited the
spurious components was present. The parasitic tuned circuits most
certainly do have to be considered. It is much the same situation as
treating a PC board trace as a piece of wire in an amplifier and then
wondering why the equipment oscillates in the MHz range.

Malcolm