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

Re: Input Current Measurement, effects of waveform distortion



---------- Forwarded message ----------
Date: Thu, 16 Oct 1997 22:04:46 -0400 (EDT)
From: FutureT-at-aol-dot-com
To: tesla-at-pupman-dot-com
Subject: Re: Input Current Measurement, effects of waveform distortion   

Fr. Tom, Jim, All,

Thanks for the thermocouple and shunt info.,etc.  I've lashed up the
thermocouple unit and found that I needed about a 0.4 ohm shunt, 
this is to give about 70uA output with 3A input.  A giant
1 ohm tappable resistor works well here.

I installed the unit onto the TC, ran it and observed the reading, and
the reading of an ordinary ammeter which is connected in series,
then replaced the TC with a pure resistive load that gives the same
reading, then read the ordinary ammeter.  This avoided the need for
precise calibration of the set-up, yet resulted in an accurate 
comparison.

Bottom line is, there was only a 1% higher reading on the regular
ammeter when running the coil compared to operation using a
non-distorting resistive load. In other words, the regular ammeter 
reads 1% too high in this TC due
to input ac (wallplug) current waveform distortion.  This is of course
using the neon sign tranny and sync-gap at ~ 720 watts as 
measured on a wattmeter.  Since the effects of the distortion are
only 1%, I can therefore trust the reading of the wattmeter.

As another test, I compared the wattmeter readings, both with the 
TC running, and with the TC replaced by a resistive load,
and kept the ammeter reading the same in both cases.
The variation in wattmeter readings in the two cases suggests a
power factor of 0.88 which agrees closely with other previous power 
factor measurements.  The wattmeter reading was 720 watts with
the TC running.  The TC is powered by a 360 watt rated neon tranny
(12kV, 30ma).  Losses in the shunts are estimated at about 5 watts.

These tests show that distortion of the input current waveform in a
neon tranny (or potential tranny, as we shall see) sync-gap TC do
not significantly affect the accuracy of ordinary wattmeter and
ammeter movements, since the distortion is minimal (or in the case
of the potential tranny...of an acceptable type)*.

Next, a potential tranny was installed in place of the neon tranny,
and the tests re-run.  Again, the readings for an ordinary ammeter
and the thermocouple ammeter agreed within 1%, regardless of
the setting of the ballast choke.  Only choke settings that gave
smooth efficient operation were used.  The potential tranny, due
to its robust design, gives stronger output sparks for the same
power input level than the neon tranny.  The input current waveform
was observed using a scope and this showed that there was more
ac input current waveform distortion using the potential tranny, than
the neon tranny.  Lesser amounts of inductive ballast resulted in a
greater degree of waveform distortion.  Yet, this type of distortion
did not appreciably affect the accuracy of a typical ammeter.  To
visuallize the distortion that appeared using lesser amounts of 
ballast, imagine a mountain range with three peaks...the highest
peak being in the middle, and 2 lesser peaks on either side...this 
is how the positive ac current half cycle appeared.  The negative 
half cycle looked the same, but upside-down. 

TCs using higher break-rates may perhaps display much greater 
input ac current waveform distortions, which may perhaps radically
undermine a typical ammeter's accuracy.  *It is possible that 
distortions would be greater at higher power levels also since this
work was limited to below 800 watts.  Or there may be other 
distortion producing factors which may come to the forefront in 
certain other TC systems.

John Freau