Re: Testing caps -> NST protection (fwd)

---------- Forwarded message ----------
Date: Thu, 30 Jul 1998 17:32:33 +1200
From: Malcolm Watts <MALCOLM-at-directorate.wnp.ac.nz>
To: Tesla List <tesla-at-pupman-dot-com>
Subject: Re: Testing caps -> NST protection (fwd)

Hi Terry, all,

> Date: Wed, 29 Jul 1998 17:40:54 -0600
> From: terryf-at-verinet-dot-com
> To: Tesla List <tesla-at-pupman-dot-com>
> Subject: Re: Testing caps -> NST protection (fwd)


> Hi Malcolm,
>         I suspect that there may be high voltage transients that are very
> high in frequency.  They may be so high in frequency that the gap cannot
> fire and block them.  A separate protection gap my offer more protection do
> to the fact that the gap may be cooler and smaller than the main gap and
> thus capable of higher speed firing.  What is really needed is a good filter
> to absolutely block the transients.  
>     This subject is confused somewhat because some neons survive the transients
> while others fail.  Thus, there are people that say theirs works fine with
> no filters.  I don't think running without filters is a good idea.  I think
> the people that don't have problems are just lucky.  Measurements and models
> show that neons need good filters if their specifications are to be even
> loosely adhered to.

Well, if filters are to be used my final preference for some time has 
been an RC combination. I am highly suspicious of chokes. The thought 
of wasting power in resistances is not appealing but neither is a 
cooked transformer.
     This whole business seems to be a two-edged sword. If one plants 
the transformer across the primary cap, the transformer sees some 
humungous ringing at Fr *but* at voltages limited to the initial cap 
voltage at most. On the other hand, placing the transformer across 
the gap forces it to see high voltage spikes whose amplitude may be 
considerably higher than the initial peak voltage on each zero 
current crossing in the gap. At least a filter can deal with 
situation two relatively easily and it does leave unwanted impedances 
(strays aside) from becoming part of the tank itself but filtering in 
either case leads to a supply loss. Perhaps going to DC charging with 
suitable snubbing is a better option. I think you may be right about
the oscillations preventing a gap arc as their frequency is becoming 
commensurate with ion transition times in the gap although I would 
still expect ions to be present during this time. I'm not too sure 
about "lack of current" argument as the equivalent series circuit 
still includes the primary cap, even if it does include the primary 
inductor which is not a stunningly high value. Still, the gap is 
dancing to the tune of these two components so I guess it washes.
     I still have some reservations about this being a complete 
explanation for the safety gap behaviours I have observed. I would 
expect a hot gap to conduct far more readily than a cold one. I would 
also expect whichever was set to the lower voltage to conduct first.
I suppose it becomes academic with proper filtering but it seems that
such a filter must take into account the inductance present in the 
transformer leads if it is to be sufficiently damped. Comments