Re: HV Measurement II

["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>



Tesla list wrote:

 > Original poster: "Jim Lux by way of Terry Fritz <teslalist-at-qwest-dot-net>" 
<jimlux-at-earthlink-dot-net>
 >
 >  > If you put your rectifier bridge at the low end of your divider string,
 >  > remember that diodes are nonlinear devices.  They will not conduct at 
all,
 >  > and your meter will show a zero deflection, until the voltage across the
 >  > bridge is about 1.2 volts.  This means that the supply will be delivering
 >  > 15 kV * (1.2V / 100V) or 180 volts before the needle even begins to move.

I'm in process of rebuilding my HV monitor on my new controller.  It was 
originally
built on a perfboard about 10 years ago.  Jim L. comments are right on, the 
HV +
a multi-megaohm resistor becomes a current source of approximately 0-1mA for
span of HV.  Effects of rectifier at high or low level becomes about 1% 
error term
which is less then +/- 5% for a moving coil analog meter.

In my original design, I had basically a NST on steroids (12kV -at- 300-400mA).
To monitor this, I used (2) ~10 megaohm seried resistors (epoxy encapsulated),
which drove a pair of 25V bidirectional TVSS, which were tied to HVCT/Equip
ground.  LV rectifier, filter, and meter scaling was on downstream side of 
TVSS.
But, then issue is if the series string opens or if TVSS should fail, you 
could end
up with at least 1/2 HV to gnd at meter, with attendent safety concerns.

My new design uses an isolated voltage sensor posted in IEEE PELS newsletter
Aug. 2002.  Basically a dual channel optocoupler with a low offset high speed
FET opamp on transistor side rebuilds FW rectified signal.  A NEC PS2501-2
allows 2.5kV isolation, which at least is better then '0' with original 
circuit, and
worse case error appears to be in the 3% range.  Meter side circuitry is hard
grounded to safety ground so metering circuit should never exceed a safe
potential, allowing monitoring of HV with better instrumentation safely 
(like an
oscilloscope).  Output from signal conditioner drives a precision peak detector
to drive meter, meter displays output voltage as peak AC.  If one assumes
SG firing window is from 45-135 degrees, RMS value is ~.90 not .707 for full
half cycle, and this can be monitored with an oscilloscope.  Circuit at 
least on
surface appears much safer, more robust, and scaleable to _very_  high 
voltages.
Modelled on MultisimV2000 and EWB, bandpass appears passable up to
nearly 50kHz, which is plenty good for me.

I know, I'll hear "but it's not RC compensated".  YEP, I agree, but <$50 for
something I can watch the power output safely with a scope versus >$1000
I'll follow any day of the week...  Besides, I have a 20kV PT to test and
calibrate it  (and I will... by running at 30kV peak for 1 hour before risking
instrumentation or me on it, nominal input will not exceed 24kV pk).  In this
area designing to the "envelope" is begging for something to blowup, or risking
one's life or equipment; I'm designing a generous safety factor into circuit..
Will post design info within the week...

Regards
Dave Sharpe, TCBOR
Chesterfield, VA. USA