OHVM Status

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "David Sharpe by way of Terry Fritz <teslalist-at-qwest-dot-net>" <sccr4us-at-erols-dot-com>

All

Ran into a minor setback to circuit, but have a
functional solution.  As alluded to earlier in
thread on HV voltmeters, any kind of conventional
rectifier, even with super schottky's (or any tech
you chose) will result in non-linearities and
offsets that will hose the low level linearities
of the circuit.  As designed on my board, with
1N4148 diodes, best "linear" (joke) operation
was a range from 1-4 volts output; ~5-20kV
peak as scaled, with 4% linearity end to end.
For the effort (and money) expended
this was a joke...  :^C

I'm now using a high speed precision full wave
rectifier (absolute value circuit) published in
EDN 3.14.96 "Two Op-Amps make a fast
full wave rectifier".  Successful operation to
~0.02 V or less has been observed, and output
is highly linear using LF412 op-amps. Some
slight offsets are seen in waveform, I'm using 5%
parts at this point, in final circuit 1% parts plus
a Symmetry adjust pot will be added to one
op-amp to compensate for slight circuit and DC
offsets.  This should allow circuit operation down
to approximately 10mV (HV detection to ~50V,
you are rapidly approaching circuit noise floor
at these levels...   :^)  )

These measurements suggests linear operation to
<=100V HV input to 50kV; arguably 3 decades
of range, which  is quite impressive for a circuit
that could be built at home using leaded
"Radio Shack/Digikey" level components.

So bad news is I'll probably have to rebuild signal
conditioning PCB scaling circuit to add revised
iso-power supply and new high speed full wave
rectifier, good news is it will be working
end-to-end, and predicted accuracy of better
the +/- 0.5% appears achieveable.  Board
costs to this point is ~$150, but once design
is finalized, I believe a 0 - +/-50kV system
could be built for ~$100.

Regards
Dave Sharpe, TCBOR
Chesterfield, VA. USA