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

Re: Calibrating E-field probes by simulation?

Original poster: "Malcolm Watts" <m.j.watts@xxxxxxxxxxxx>

Hi Steve,
           Needless to say, measurement of TC output voltage has
occupied my mind for many years. I do have a suggestion:

On 4 Apr 2005, at 8:48, Tesla list wrote:

> Original poster: "Steve Conner" <steve.conner@xxxxxxxxxxx>
> Hi all,
> I was just thinking about ways of measuring the output voltage of a
> coil. E-field probes like Terry's planar antenna are a nice solution
> but the problem is how to calibrate them. It struck me that if you
> were to use a probe with an easily modelled geometry (a small toroid
> or sphere?) then the calibration could be done using a finite element
> modelling program like FANTC.
> All you would need to do is have the program calculate the capacitance
> between your coil's topload (plus some fraction of the secondary self-
> capacitance I suppose?) and the probe. The division ratio of the probe
> is approximately (capacitance between probe and coil)/(capacitance
> between probe and ground including cable and scope input cap)
> This may only have to be done one time to generate a calibration
> table. For instance:
> To get a 100000:1 ratio, place the probe
> 6 feet from a 12" toroid
> 9 feet from a 24" toroid
> 11 feet from a 36" toroid
> etc.
> Can anyone comment on the feasibility of this, and likely sources of
> error? Streamer capacitance is an obvious one but I assume you could
> use a breakout point to divert streamers as far from the probe as
> possible.
> An interesting way of calibrating E-field probes by experiment might
> be to drive the base of your secondary from a Marx generator. You
> could assume the topload will "resonant charge" to twice the Marx
> output voltage.
> Another possibility for voltage measurement might be to build a
> capacitive divider using a long string of small capacitors down the
> inside of the secondary itself, rather like the load capacitor Terry
> built for his magnifier. Maybe 100x 100pF 10kV ceramics in series,
> with a 0.1uF to ground, to give a 100k:1 ratio. The divider would be
> shielded against stray capacitance errors this way since the voltage
> gradient down the middle of the secondary should be almost the same as
> the gradient down the divider (both in space and in time). IOW, you're
> using the secondary as a guard.
> But I don't know what you would ground the bottom of the divider to.
> High dv/dt on the toroid would induce high currents through the
> capacitor string and maybe spike the ground to dangerous levels.
> Anyone have any thoughts on this?

Why not simply hook a known voltage source (e.g. a stiff 50Hz HV
transformer) to the sphere on top of a resonator in situ? Without
resonant effects to upset the output voltage calibration it should be
easy to physically measure both directly and by capacitive coupling.
In fact I see no reason why 5kHz or even 50kHz couldn't used as long
as the frequency is well below the self-resonant frequency of the
coil. I do like the idea of modelling but think that it _must_ be
verified by measurement no matter how solid the reasoning. I have
reason to suspect that unverified modelling is behind problems we
have found with hypotheses promoted by other researchers ;)


> Steve Conner