Optoisolated HV Monitor (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 parts to begin breadboarding Opto-isolated HV
Monitor (OHVM) are in hand.  Using a NEC PS2501-4
as used in Woodward's Power Meter (Electronic Design
10/14/1994), with middle two channels of optocouplers
leads removed (broke off flush with case).  I did this
to increase creepage distance and indirectly HV
rating of device.  Mulichannel device is being used to
help insure that Current Transfer Ratio (CTR) is more
closely matched for two active opto channels versus
two randomly procured single channel optocouplers.

Circuit Vin >> Vo transfer function is linear from ~750V
to FSD of circuit using Multisim DC sweep, and frequency
response corner frequency is ~10-15Khz, with useable
response to >100K; error in the useable range once
calibrated for 0 and FSD is <<2%.

Component cost to date is ~$50 as predicted.  Making
two PCB's using "Iron and Peel";

(1)  A HV current limiter board consisting of 2 banks
        of 30 200K 2W carbon film power resistors (600V
        continuous rated, 1.1kV pulse peak each).  LV/gnd
        end is bidirectionally clamped with (2) 1N4728
        3.3V zeners in series to HVCT/equipment gnd for
        each row of current limiting resistors.  Also placing
        a NE2 neon lamp in parallel with zeners to provide
        an additional fail safe clamp for optocoupler. Max.
        dissapation at full power for each resistor is 0.5W.
        Maximum safe design input voltage of this array is
         26.8kV pk (19 kVrms) with 2X thermal safety
         factor.  Maximum service input Vin will be
        24kVpk (16.8kVrms).  Approximate size of
        PCB: 7.5" x 10.5".
(2)  An aluminum 'L' bracket will be mounted above
        PCB (1) to act as a Faraday shield and ground
        plane for the optocoupler V sensor PCB. Opto,
        LF412 FET op-amp, and comparator will be
        mounted on board.  +/- 15V, 5V,COM will run
        this board, PCB-BNC will allow 50 ohm line
        direct to o'scope to reduce likelihood of EMI/RFI.
        Approximate size: 4" x 7.5".

Design issues thus far.
- Input current FSD should be <2mA into optocoupler
   (i.e. size currrent limiting resistor at 1000 ohms / applied
   rms volt.  Circuit has two non-linear regions, one when
   opto is turning on (0-750V) and upper region where
   opto transitions from a linear to logarithmic current and
   op-amp saturation approaching the + rail (12.5-13.5V).
- 3 resistors may have to be scaled to allow adequate
   dynamic headroom for transients or operating a Variac's
   in a Vboost (overvoltage) condition.
- Operation of 0 cross detect comparator may be also
    triggered by SG operation. that would actually be good,
    no need for RF detector (as radio).  SG operation may
    effect operation of peak detector.  Worse case may
    require integration to help circuit "ignore" transient
    overshoots.

But thus far nothing that can't be optimized or adapted.
Will keep list informed on progress...

Regards
Dave Sharpe, TCBOR
Chesterfield, VA. USA