[Prev][Next][Index][Thread]
Grounding Rebuilt Neons
[The following text is in the "ISO-8859-1" character set]
[Your display is set for the "US-ASCII" character set]
[Some characters may be displayed incorrectly]
----------
> From: Mike Hammer <mhammer-at-misslink-dot-net>
> To: Thomas McGahee <tom_mcgahee-at-sigmais-dot-com>
> Subject: I got a question
> Date: Friday, February 07, 1997 8:38 PM
>
> Fr. Tom,
>
> I read your excellent article about neons on Chip's
> list a while back. I learned a couple things and there
> were a couple points I disagreed about but it was informative.
>
Mike, Fr. Tom here...
Yes, there will always be differences of opinion. I try to give equal time
to all theories that are at least probable. I believe that there are
actually *several* things that are going on with neon transformer
destruction. I need a bit more data from destroyed neons before I would say
that definitely this or that is *the main* destruction mechanism. The thing
to do until then is to let people know all the scenarios. Then if they want
to disregard what has been said, at least they can't complain that no one
ever told them!
> Since you seem to know a great deal about neons I thought I would
> ask you a question and get your opinion.
>
If I know much of anything at all about neon transformers, it is because:
a) I have used the surly beasts myself.
b) I have destroyed my fair share of the expensive things over the years.
c) Others have done the same and shared their experiences.
d) I have tried to look carefully at what others have said about neons and
then integrate that with my own knowledge of electronics and my own
experience.
e) I have tried to keep my mind open and not champion this or that theory
because it was my own pet theory.
f) I have tried to gather useful information from others who have rebuilt
neons. This is one of my main sources of my information on breakdown
mechanisms.
g) When It comes to theories, I try to subject them to scrutiny and
experimentation.
h) I share the theories and my own views on them with others and I get
feedback which I try to evaluate and pass on.
i) I try to distinguish between known fact and conjecture.
> I run almost exclusively rebuilt neons. I have a good pile
> of good neons but I put my wear and tear on dead ones I have
> revived.
>
In certain ways a repaired neon is actually better than a new one. Let me
explain. The dumb tar creates more problems than it solves. It may be OK
for the original intended use of neons in causing a pretty, colorful glow
from excited gas molecules, but in an RF environment the tar just messes
things up royal. The tar is supposed to insulate and draw heat away. What
it most often does for us is melt under RF influence (mostly at the first
and second layer of the secondary coil) and then promote secondary
breakdown due to a combination of inductive *and* electrostatic effects.
Tar appears to be quite lossy at RF levels, and may be one of the reasons
why some people have noticed that as the frequency of the Tesla tank
circuit increases, so does the RF current that wants to flow back into the
transformer secondary. This appears due to capacitive effects more than
inductive effects (one would expect inductive effects to increase across
the secondary windings as the frequency in the Tesla circuit increased...
but inductance alone would cause *less* current, and we see *more*)
If you repair a neon by removing all the stupid tar and then plop the thing
in an oil bath and get the air bubbles out, you have a *great* HV
transformer! If you try to replace the tar with another form of solid
insulation, it may be better than tar, but it will never equal what you get
with oil. Oil is self-healing! Tesla himself repeatedly stated that in RF
work oil will always be better than even the best solid insulation. That
still holds true today. Do you notice how pole pigs take a licking and keep
on ticking? I attribite that to the fact that they are *oil filled* more
than to the fact that they are not center tapped on the HV side.
> My question concerns the use of a mid point ground.
> What would be the pros and cons on grounding the neon core?
>
> Once transformers are unpotted that midpoint connection to the core
> can be removed. You could connect the midpoints of the 2 secondaries
> together but not to the core. So you could have 3 configurations.
>
> 1. If you ground the core and the secondary midpoints are tied to the
core
> that makes the core 0 volts. This is the normal configuration.
>
Correct. Ground it and each side will equally get hit with HV inductive
spikes that will excite the outer secondary windings. But you can't stop
that anyhow. Grounding is the sanest approach, because it keeps the maximum
voltages in the transformer lower than any other method.
> 2. If you left the core ungrounded but the secondary midpoints still
> attached to it would the core remain at 0 volts? Or would it float to
> a voltage? And what would that voltage be? This is the way I normally
> run my neons.
>
Don't ground it and the core is going to float to 1/2 the total secondary
voltage. If you are saying that you don't connect the case to the core,
then there are two scenarios. If the case is grounded and the core isn't,
then that will force the core and case to have a voltage between them of on
average 1/2 HV. I say on average, because the core will actually have a
voltage across it that will vary wildly. It will couple capacitively to the
case, and due to the phase differences that will exist, you can actually
get a case to core differential of more than 12KV. That sounds like fun.
If the case is not grounded and the core is separate and not grounded
either, then the capacitive coupling will cause the case-to-primary voltage
to swing all over the place. Like about 12KV max. More fun.
But, also, my dear friend, are you grounding one of the sides of the
transformer? Then the *other* one of the HV terminals will tend to have the
FULL HV across it even though the insulator was only designed for 1/2 of
the total HV.
OK, so you say "I'm not gonna ground either side of my neon." Don't make no
difference. The nature of the RF Bang Bang Disruptive Discharge circuit
under those conditions will just alternately whack first one side and then
the other into wild inductive swings. Don't ground anything, and watch all
hell break loose as one side and then the other madly try to reference
themselves to *something*. And that *something* in this case is going to be
your RF ground. Which is most likely somehow connected to the AC mains
ground. Which you can translate as the primary of your neon. (You know,
those two wires that enter through your *ungrounded* case through the dinky
little porcelain insulators that were not quite designed for 12KV and
more). Hmmm, looks like a possible source of problems to *me*.
I'm not saying the above circuits don't *work*. I *am* saying they are an
accident waiting to happen. The fact that some people can run their neons
connected like this without instantaneous damage is a known fact. But when
they build their next one using a neon that to all outward appearances is
*exactly* the same and it decides to call it quits early, then the meaning
of statistical mortality begins to strike home. Not *all* persons who play
Russian Roulette die on the first try. But if you play the game long
enough, you eventually lose. So why play that game at all?
> 3. If you separate the secondary midpoints from the core and connect them
> together you would essentially create a center tapped secondary.
> Now the core is totally out of the circuit. How would things react if
you
> grounded this center tap or not?
>
Pardon me, but what on earth would you gain by such an arrangement? Now you
have the voltage between the bottoms of the two secondary coils residing at
6KV minimum with regards to the core. Unless this sucker is in an oil bath,
*I* for one would not want to be in the same room with it when you plug it
in.
Connecting the center tap to the core and then to the case was the
designer's way of cutting the high voltage tensions within the case in
half. Grounding just the center tap makes no sense at all. You have just
thrown away your core and case protection and gotten nothing back in
return. Don't do it!! It won't kill your neon or anything, but you will now
have these interesting out of phase voltages floating around the core and
case, just waiting to do some mischief. You don't need the grief.
Now if you decide to connect the secondaries together and then NOT attach
them to the core/case, and if the core/case is grounded, then you have an
arrangement something like what you would find in the marvelous pole pig.
Well, *almost*. You see, the voltage between the center tap and the
core/case is now 1/2 HV. (in your case, 6KV). Lets assume you're somewhat
sane at the moment and decide to ground one side of the neon to *some* kind
of a ground. Will the grounded core/case throw an arc to the center tap? It
will sure as heck do its best to try. Insulating the wires will be somewhat
futile, but will postpone the inevitable to some *future* date not yet
specified. Now, if you put this baby in an oil bath it will probably
succeed with this arrangement.
Oh, yeah, did I mention that the *other* side of the neon now has the full
12KV from its HV terminal to case, even though the terminal is not rated
for that amount of voltage? Again, if you have immersed the whole thing in
oil, moved the wires for maximum separation from the core, and grounded the
core/case, then the floating center tap will probably work just fine. But
you have double the tension on that one terminal, and that isn't really
necessary. No gain, all Pain.
Circuit #1 has the least problems associated with it.
> Do these questions make sense? I have thought about this for a while and
can't
> see any valid reason to ground the core. Except for safety in the event
of a
> RF flashover in the neon secondary. I think I may be missing something
here.
>
The questions are *good* questions. The following points may be helpful to
remember:
1) The stupid tar is more an enemy than a friend. Carbon tracking and
losses due to dielectric heating can be eliminated almost entirely by
removal of all tar and immersion in oil. If you consider oil too messy,
then removal of all tar and replacement with multiple applications of
varnish or other HV insulating medium with *low* dielectric loss will be
better than nothing. We need to keep air and moisture out. If you can apply
a layer at a time and remove air bubbles you may get decent results.
Remember also that the tar performed a heat removal function and provided
some insulation from the outer part of the secondaries and the case. The
outer insulation function might be able to be done with a good insulation
such as that used in capacitors. If the transformer gets too hot you might
have to use a muffin fan to flow air over the windings/core. Remember that
to gauge how hot the transformer gets you have to run it under rated load
for a reasonable period of time, and then turn it off and actually touch
the CORE, not the case. If you have removed the tar, then the case will
always be deceptively cooloer than the core. Don't be deceived! Check the
*core* temperature, not the case temperature.
2) Reduce voltage tensions to as low a level as possible. Center-tapping
and case grounding this center tap is an excellent way to do this. Floating
the center tap may *work*, but it really does cause much greater electrical
tensions to exist, and these can cause failure. Use the best low dielectric
loss insulation you can get between all parts that have HV, such as between
the secondaries and the case.
3) The damage that would be caused by inductive effects at the neon
secondary can be greatly alleviated through the proper use of L/C RF
filters. The coils connect from the transformer terminal(s) to the the
Tesla tank circuit. By the way, a coil without any capacitance to ground is
not really going to provide any useful filtering, because the inductance of
the neon will be greater. The best method is to use two small value
capacitors each connected between one of the HV terminals and the center
tapped ground. I put my safety grounds in parallel with these capacitors.
It appears that these protection caps work best if their dielectric is
somewhat lossy. I have not verified this independently, but it makes sense
because such a lossy capacitor will be able to divert more RF current while
remaining fairly aloof from the 60Hz currents.
4) Use safety gaps directly across the secondary terminals *to case
ground*. This will protect each terminal independently. We are talking
about neons here. Adjust them initially too close and then open them up a
little each time until you get minimal false firing. Use of balls rather
than points will cut down on corona losses and false firings.
5) Wire the Tesla coil main Spark Gap in parallel with the neon
transformer. Do *not* connect the capacitor directly across the
transformer, as this sets up an oscillatory circuit whose phase voltages
can cause breakdown, and whose phase currents can cause dielectric heating
of the insulation between windings of the neon secondary. This can cause
hot spots to develop and grow until there is destruction of the insulation
between wires or layers. Again, there are those who persist in connecting
the main capacitor in parallel with the transformer. Yes, the circuit does
*work* that way, but the chance of having transformer breakdown is
*greatly* increased when using this arrangement.
>
> No hurry needed on an answer to this. Do it when you have the time.
>
> TTYL!
>
>
>
>
>
> Mike Hammer
> mhammer-at-misslink-dot-net
Mike, I hope that the above discussion helps you in some way. I think some
of the new members on the Tesla list might find this discussion useful, so
I am also posting this to the list. Some may want to clip it and keep it
for future reference. This material will also be covered in detail in the
Guide article that I am writing on Transformers.
Fr. Tom McGahee