[Prev][Next][Index][Thread]
HV oil probe details
From: terryf-at-verinet-dot-com[SMTP:terryf-at-verinet-dot-com]
Sent: Tuesday, September 23, 1997 11:03 PM
To: tesla-at-pupman-dot-com
Subject: Re: HV oil probe details
Hi Tom,
First of all, let me make it clear that I have not actually built
and tested this type of probe. It has to be categorized as theoretical and
experimental. I don't assume any responsibility as to this design working,
being safe, etc. Obviously, this should only be attempted by experience
high voltage experts who can handle the known and unknown risks and problems
that may be encountered.
The probe would consist of a 4 inch diameter PVC pipe that would be
placed up the center of the Tesla coil (probably a six or eight inch
diameter coil). The top end of the pipe would be in the center of the
toroid assembly where it would be surrounded by the electrostatic fields of
the toroid. The bottom of the PVC pipe would extend down past the bottom of
the coil and extend a few inches lower to safely clear the primary coil.
Inside the pipe is a 1/4 copper tube that would travel through the
center to a point level with the top turn of the secondary coil. It would
have the needed supports. The bottom of this copper tube would exit the PVC
pipe at the bottom and would travel a safe distance (maybe 10 feet) from the
Tesla coil. This end would be connected to a very good ground independent
of the Tesla coil's ground.
This 1/4 inch tube contains a thin coax cable. The top of this
cable goes up the 1/4 inch tube and has its bare center conductor protruding
about 3 inches at the top. This would act much like an antenna that would
in effect be capacitively coupled to the intense fields at the top terminal.
The other end of the coax would extend a distance to where the test
equipment would be located (maybe 40 feet). This coax would be grounded to
the test equipment grounding at this end. This ground would be separated
from the Tesla coil's ground and the 1/4 inch tube's grounds.
The PVC pipe would be filled with oil to insulate the grounded
conductor's inside it from the high voltages. This layer would be about
1.75 inches minimum at any point in the PCV (except the bottom). The top
antenna would be covered with about three more inches of oil to prevent
arcing to it. The PVC tube is capped at both ends and the exiting copper
tube and coax is sealed to prevent leaks.
So we now have an oil insulated antenna at the top of the coil that
is connected through a double shielded path back to the instruments. The
1/4 inch copper tube absorbs the vast majority of the "noise" from the coil
and carries the capacitively coupled charges on it to ground. The coax
cable's shield can protect the signal from the relatively low-level noise
remaining.
At the instrument end of the coax are two 0.01uF poly caps (the
common 50 volt types) connected to ground. The two poly caps are in
parallel (fail safe) along with a 10M ohm resistor to bleed off any DC
voltages (a high value inductor may be a better choice?). The antenna end
inside the coil will act as an ~0.5 pF capacitor to the high voltage
terminal. The two 0.1uF caps will act as the lower end of the capacitive
voltage divider at a value of 0.005uF. This gives about a 10000:1 AC
attenuation ratio. This divided signal is sent though a 1k ohm resistor and
clamped by transorbs (just in case) to the equipment.
There are two known dangers that would need to be addressed:
1. There is a possibility that the oil insulation layer could
be punctured by too high of voltage. This would send heavy currents and
voltages to the test equipment (if the copper tube and coax shield didn't
take the hit). The capacitors would have to short to ground and the
transorbes would need to be able to clamp this event. This would be tested.
2. A voltage puncture could break the PVC and possibly ignite
the oil causing a fire. None flammable oil would prevent fire but there
could be a mess. The oil would be tested to insure it's insulation
properties and determine a safe operating voltage. (400kV??)
3. The losses in the probe may cause heating of the probe.
This would have to be monitored. The probe needs a tiny vent hole at the top.
There are four known errors that would need to be addressed:
1. The actual attenuation ratio would need to be measured. The
primary voltage could be used to charge top conductor to a know and
measurable voltage for calibration (with a typical high voltage divider).
You would need detune the coil to prevent any resonate amplification. This
needs more thought. I have equipment that can do this so I didn't give it much.
2. The grounded center conductors and the oil insulated column
would require retuning the coil substantially. It would form about a 15pF
capacitor to the secondary coil. The 1/4 inch copper tube would act as a
shorted winding. This would mean that the coils true best output would not
be measured due to these losses.
3. The high voltage would cause corona between the secondary
coil and the 4 inch PCV. This may cause noise to be picked up by the
antenna. The top portion could be covered by metal connected to the top
terminal to prevent this and increase coupling.
4. The frequency response of the probe would need to be tested
to insure it is reasonable. I never got to the point of carefully
calculating this out. It would work at the 200KHz range.
This probe would eliminate the capacitive coupling of the signal
from areas other than the actual terminal area (this may be a problem with
your resistive probe? Can the resistors pick up stray E-fields? They
should be shielded but those shields attract arcs and filter high
frequencies out.) It would also load the coil reactively so that there
would not be much resistive loss. If it worked as planned the measurements
would be defendable as to their accuracy and others could verify the
results. It is based on sound theory and proven principles. I assume
nobody else has a 2 megavolt X-ray tube and the associated calibrated
standards laying around. I don't know what kind of losses the X-ray system
may have had. However, the reported measurements are fascinating and the
method is cleaver!
I was planning on making a special coil to accommodate this
assembly. It would have low turn-to-turn voltages to prevent arcing at the
coil or probe. The initial testing would be done using single pulses
without test equipment connected just to see if everything was ok. I would
start out at very low voltages and work my way up. My cheap scope ($50 easy
to fix Heathkit from the 1950's) would then be connected. The Tesla coil
could then progress to continuous output and I could move up the much better
$600 scope. After a reasonable amount of confidence is obtained, storage
scopes etc. could be used. I would go ahead and deliberately arc the coax
to destructively test the voltage clamp circuits to both the secondary and
primary circuits before I hooked up anything expensive. There is always
still some risk!
You can see why I stopped this project to pursue a fiber-optic
approach!
It is 95% simpler and safer.
It has very minimal effects on the coil.
It should work on any coil design.
It can very accurately measure the currents from an arc to an
object. ie. find the real resistance vs time of an arc!!
Zero risk to test equipment.
Measure voltage, current, and phase angle vs time. Ie. Spice models
that work!
No oil (I hate oil!)
It may also handle the 1 million volt coil that DR.RESONANCE told of. The
fiber probes should have very good frequency response They can be duplicated
and verified easily, cheaply and safely. There were very few responses to
the original post about this and the fiber-optic probes (I was glad to hear
that the fiber should not fluoresce! Thanks Sulaiman :-)) Maybe this is at
the "bleeding edge" of Tesla coil research! :-))
Terry (waiting for mail-order parts to get started!)
>Hi Terry
>
>I would be interested in knowing the details of your of your other
>probe design.Please include things like connectioning to test equip.
>
>You mentioned using oil, is this to make a couple of capacitors in a
>divider configuration? If so, have you considered its response over
>a large frequency range.
>
>All I have at present is a high value resistive divider that is a little
>less than linear ;)
>
>Any info would be appreciated
>
>Tom Weagant
>
>
><From: terryf-at-verinet-dot-com[SMTP:terryf-at-verinet-dot-com]
><Sent: Friday, September 19, 1997 2:46 PM
><Subject: Secondary voltage measurement.
>
><I was working on a more conventional voltage probe but it would add about
><15pf to the secondarys internal capacitance. This would require retuning
><and would materially change the coil's true output. It feel it would work
><with these limitations, however and probably give useful information. For
><the time being I have stopped work on this probe. It also requred about 1.6
><gal of oil for insulation (I hate oil!). I hope the fiber-optic probe will
><work. It would have many many advantages.
>
><If these do end up working I will post the plans and results. It all costs
><about $75 (not including the test equipment it plugs into). If anyone is
><intersted in the old voltage probe idea, I can also provide details on it.
><Someone may want to play with this. If the fiber-optic probes fry, it will
><be me :-)
>.................