Maker Pro
Maker Pro

Constant AC Current Regulator

I got a call from a company who was having problems getting an AC
current regulator working. They just needed some basic ideas. I often
get such calls and I try to help as long as it doesn't take up too
much of my time.

As near as I can tell, they are using the current to test and calibrate
current transformers. They needed to route 10 amps of pure 60Hz sine
wave AC current through a single heavy wire, as part of a calibration
circuit for one of their products. They said the total resistance of
the loop was about 0.025 ohms. Their first approach used a small
variac and a toroid transformer, to produce the needed current.
However, it was not well regulated. The variac approach worked fine,
as long as the line voltage remained at 120vac. But since they wanted
to keep the 10 amps to an accuracy of +-0.5%, any line voltage change
would certainly change the current. They considered attaching the
Variac to the output of a ferro-resonant constant 120vac supply or use
a motorized Variac but they really wanted something much more compact.
They were also concerned about distortion of the 60Hz waveform coming
from wall outlet, especially at the waveform peaks. I suggested that
perhaps a cleaner 60Hz signal could be synthetically generated and used
as a reference. The signal would be fed to an audio amp, which would
drive the winding of a low voltage transformer, perhaps 24vac. A
custom wound secondary of the transformer could then generate the
needed high current. By using another current transformer in line with
the test loop, in conjunction with an op amp circuit, the AC current in
the loop of wire could be kept constant. Can anyone think of another
way this could be done?

David A. Johnson, P.E. --- Consulting Engineer
http://www.djandassoc.com
http://www.imagineeringezine.com Home of the Imagineering on-line
magazine.
Also, http://www.discovercircuits.com A collection of over 11,000
schematics.
 
J

John Larkin

Jan 1, 1970
0
I got a call from a company who was having problems getting an AC
current regulator working. They just needed some basic ideas. I often
get such calls and I try to help as long as it doesn't take up too
much of my time.

As near as I can tell, they are using the current to test and calibrate
current transformers. They needed to route 10 amps of pure 60Hz sine
wave AC current through a single heavy wire, as part of a calibration
circuit for one of their products. They said the total resistance of
the loop was about 0.025 ohms. Their first approach used a small
variac and a toroid transformer, to produce the needed current.
However, it was not well regulated. The variac approach worked fine,
as long as the line voltage remained at 120vac. But since they wanted
to keep the 10 amps to an accuracy of +-0.5%, any line voltage change
would certainly change the current. They considered attaching the
Variac to the output of a ferro-resonant constant 120vac supply or use
a motorized Variac but they really wanted something much more compact.
They were also concerned about distortion of the 60Hz waveform coming
from wall outlet, especially at the waveform peaks. I suggested that
perhaps a cleaner 60Hz signal could be synthetically generated and used
as a reference. The signal would be fed to an audio amp, which would
drive the winding of a low voltage transformer, perhaps 24vac. A
custom wound secondary of the transformer could then generate the
needed high current. By using another current transformer in line with
the test loop, in conjunction with an op amp circuit, the AC current in
the loop of wire could be kept constant. Can anyone think of another
way this could be done?

David A. Johnson, P.E. --- Consulting Engineer
http://www.djandassoc.com
http://www.imagineeringezine.com Home of the Imagineering on-line
magazine.
Also, http://www.discovercircuits.com A collection of over 11,000
schematics.


We do this. We use a homemade sine generator (uP and a DAC) to
generate a stable 60 Hz sine, although there are lots of cheap
synthesized waveform generators around these days, so I wouldn't go
home-brew again. The homemade generator does generate voltage/current
waveforms with phase switchable in precise 45 degree steps, which is
handy for power testing.

The sig gen drives a 10-turn pot and a power amp, a Peavey 400
watt/channel 2-channel PA amp. We do two channels, one for current and
one for voltage, to calibrate electronic power meters. The current
channel has the power amp feeding an old radio-station filament
transformer rated 5 volts at 190 amps. The secondary loop includes a
current shunt to monitor the actual current. Active feedback would be
nice, as the current drifts as things warm up so you have to ride herd
on the pot. But you can hold 0.1% pretty easily.

We had to change some of the capacitors in the Peavey to reduce phase
shift at 50/60 Hz. They kindly sent us a schematic and pointed out
which caps caused the most rolloff.

DON'T use anything ferrous in the loop. We used a "brass" rod to slip
the CT over, but it was actually plated steel, and it caused massive
waveform distortion.

Working off the AC line is nasty, unless you do a bridge-type thing
that doesn't care much about the actual drive level. Even then,
harmonics and phase shifts will be a nuisance.

Pearson makes superb current sensors, CTs with integral burden
resistors, probably better than using a shunt. The Danfysik DCCT's are
even better. Expensive.

John
 
T

Tony Williams

Jan 1, 1970
0
....................................... I suggested that perhaps
a cleaner 60Hz signal could be synthetically generated and used
as a reference. The signal would be fed to an audio amp, which
would drive the winding of a low voltage transformer, perhaps
24vac. A custom wound secondary of the transformer could then
generate the needed high current. By using another current
transformer in line with the test loop, in conjunction with an op
amp circuit, the AC current in the loop of wire could be kept
constant. Can anyone think of another way this could be done?

Yes, that is the CT test setup we used to supply.

To reduce magnetic fields use a toroidal stimulus
transformer. Make sure that the AC amplifier you
choose can drive a transformer, or provide external
coupling capacitors.

We use two (carefully measured) reference CT's on
the loop... one measures the loop current, and the
other has it's secondary wired in parallel with the
unknown CT, but in antiphase. The resultant current
output is the difference between the reference and
the unknown.

The backing-off CT is a production CT, measured and
turns-adjusted to be in the dead centre of the spec.
We use traceable instruments for this and usually make
about 6-off, for future spares and for routine checking
of the working reference CT.
 
H

Harry Dellamano

Jan 1, 1970
0
Tony Williams said:
Yes, that is the CT test setup we used to supply.

To reduce magnetic fields use a toroidal stimulus
transformer. Make sure that the AC amplifier you
choose can drive a transformer, or provide external
coupling capacitors.

We use two (carefully measured) reference CT's on
the loop... one measures the loop current, and the
other has it's secondary wired in parallel with the
unknown CT, but in antiphase. The resultant current
output is the difference between the reference and
the unknown.

The backing-off CT is a production CT, measured and
turns-adjusted to be in the dead centre of the spec.
We use traceable instruments for this and usually make
about 6-off, for future spares and for routine checking
of the working reference CT.
Neat stuff Tony. Maybe a class D switcher amp would be a good choice here.
Great efficiency, power and size.
Regards,
Harry
 
T

Tony Williams

Jan 1, 1970
0
The backing-off CT is a production CT, measured and
turns-adjusted to be in the dead centre of the spec.
[/QUOTE]
Neat stuff Tony.

The trick is in that reference CT on the loop Harry,
used for backing-off. It means that the loop current
need not be so precise or stable and only one difference
measurement to be made. In fact the OP could even
consider using the idea on the mains energised setup.
Maybe a class D switcher amp would be a good choice here.
Great efficiency, power and size.

PWM'd audio amplifiers don't seem to be used much for
testing wound components. Possible concerns maybe about
hf ripple shock-exciting the inductance, and/or possible
trouble with hf fuzz affecting the accuracy of AC DVMs.
 
Top