Quoting dewzenol-at-inlink-dot-com as "DWZ"
Quoting Robert Michaels as "RM"

 RM> I could be off-base by a couple of spark-gap lengths, but
 RM> there are several things in your description which jump across
 RM> to me: 
DWZ> I've been running a very small tesla coil off of a 6kv, 1.2ma supply.
DWZ> I want to replace the supply with a 9 or 15kv neon xfmr (30ma)

 RM> Althought any number of Tesla Coils have been built by
 RM> experimenters using neon sign transformers - they are =not= a
 RM> good choice for this.  Many reasons.

But there are many, many more good reasons why beginning coilers start
out using them. It is true neons are not a good choice for this, but 
they are frequently pressed into service as power supplies for Tesla 
tank circuits. They can be very cost effective if the coiler is re-
building dead or weak cores that are scrounged for free. Properly 
grounded; and using bypass capacitance, ferrite toriod RF chokes,
and safety spark gaps on the HV side of the xfrmr; they work reasonably
well up to 1.5 kVA or so in Tesla work. I have extensive experience
running paralleled neon power supplies up to nearly 4 kVA.

1) Neons are readily available. And as I mentioned, failed or failing
cores can frequently be obtained for free, and with elbow grease they
can be unpotted and restored/rebuilt. 

2) They will hold up well on smaller coils provided they are properly
grounded, choked, bypassed, and equipped with a safety gap.

3) Power levels can be increased by wiring neons with matching voltage
into parallel banks. Pairs of matched xfmrs can be wired with the 
primaries in series with 240 volts across the two series primaries. 
This allows large power supplies drawing heavy current loads to be 
supplied with better rated wiring.

4) Power factor correction can be placed across the primary of a neon,
or a bank of neons, to reduce the current draw, and to increase
DWZ> I do realize that the neon xfmrs are center ground, and have 
DWZ> wired them as such... Am I missing something obvious???

 RM> Yike!  You're siphoning off all your rf energy (or most of
 RM> it). There is nothing in the =primary= circuit of a Tesla Coil
 RM> which requires grounding.

Not exactly true: the center post of the safety gap should always
be grounded to the dedicated RF ground. So should the centers of
the HV bypassing caps used across the HV bushings of neon power 
supplies. The transformer core should always be grounded to the 
dedicated RF ground, even if the HV windings are not center tapped;
the only exception being some potential xfmrs that are completely
potted in insulating compound.

 RM> The center tap of a neon transformer is grounded to permit the
 RM> suppression of interference when they are used in sign service
 RM> - get it???

No, I certainly don't "get it", and I thought I knew what I was
talking about.

 RM> The interference is caused by unwanted =rf energy= incidental
 RM> to the operation of a neon tube.  

What? How? Please explain, and reference if possible. Does anybody
else have references or data to support this statement?

 RM> In a Tesla Coil we =want= rf energy.  Lots and lots of it.  
 RM> Lots of good, wonderful, free-flowing, warm and juicy rf 
 RM> energy. We do not want to suppress any part of it. None. Nix. 
 RM> Nil. Nyent. Nada. Okay?

Not exactly: we want serious RF supression on the supply side of
the tank circuit to protect the step-up xfmr, especially if the 
power supply is a neon sign transformer. Also, we want RF supression 
on all of the supply and control circuits.
DWZ> I purchased a couple of G.E. caps  which say: DC Filter Service Only 

 RM> The capacitors of choice for Tesla Coils should be mica
 RM> transmitting capacitors, ceramic, or glass.  Note that the
 RM> voltage, capacitance, =and rf current handling= capabilities
 RM> must be appropriate.

Dead wrong. The capacitors of choice are plastic film dielectric
covered in mineral or silicone oil. Custom made commercial, or
homemade; the dielectrics of choice are polyethylene, polypropylene, 
or polystyrene. Glass is one of the WORST, and mica is not a whole
lot better. Plastic film caps are much higher Q, with much lower
RF dissapation factors. You have to really work hard to get a
plastic cap to warm up at all.

 RM> may get sufficient results to satisfy.  Else - go buy some
 RM> glass plates, aluminum foil and shellac at the hardware store
 RM> and stop messing with store-bought capacitors.  What was good
 RM> enough for Dr. Tesla should be good enough for you.
 RM> Robert Michaels - Detroit, U.S.A.

My god man! Tesla built the Colorado Springs machine in an era of
"stone knives and bear skins" (to quote Spock).

Be assured that if Tesla was alive today, he would be telling us
to: use the heaviest RF grounding available where required on our
systems, use plastic film capacitors covered in oil, build her-
metically sealed plastic coil forms with the wire never entering 
into the interior of the form, and use the largest toriod discharger
and primary inductance we can reasonably achieve.

Richard Quick

... If all else fails... Throw another megavolt across it!
___ Blue Wave/QWK v2.12