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

Re: Neon filtering (long post warning)



Malcom:
	I have gone through the same simulation you seem to
have conducted, and reached somewhat similar conclusions.
First of all, it is certainly possible to damage the transforme
r with an L-C filter.  Some of the conditions I simulated
caused me not to hook up such a filter as I had intended.
I had wound a couple of 20 mH chokes on powdered iron core,
and had intended to use 500 uufd 15 kV filter capacitors across
the neon terminals to ground.  After running the simulation
I decided that the proper thing to do was to put 5 k resistors
in series with the chokes, as per your 1).  That seemed 
quite suitable to me, with no significant power loss if
the thing is operated with a reasonable primary capacitor
(in my case about 0.01 ufd) and the spark gap is kept at
reasonable length.  I'm using 12 kV, 60 ma transformers and]
at the moment have enough of a supply that I have not yet
installed the filter, waiting for a failure.  So far no failure
in several total hours of running.  The gap is set so it
just breaks down without the primary capacitor connected.  This
is somewhat conservative, b ut I am getting 24" + "sparks" 
from a little coil and not particularly well-designed coil,
so I'm satisfied until I get a bigger space to work in.
	I also looked at a couple of configurations which you
did not.  One involved paralleling the choke with enough 
capacitance to resonate at the operating frequency of the coil.
In my case it was round 30 uufd, as I recall, and I thought of
using sections of RG-8/U coax.  Other involved series resonating
the shunt capacitor, with resultant loss of attenuation at
higher frequencies.  Wish someone would try some of these
combinations and report on results.  I'm too lazy and also too
short on time......
Ed