[Prev][Index][Thread]

# Measuring Secondary Current

• To: tesla-at-objinc-dot-com
• Subject: Measuring Secondary Current
• From: wesb-at-VNET.IBM.COM
• Date: Fri, 31 Mar 95 13:51:07 EST
• >Received: from vnet.ibm-dot-com by ns-1.csn-dot-net with SMTP id AA21138 (5.65c/IDA-1.4.4 for <tesla-at-objinc-dot-com>); Fri, 31 Mar 1995 14:45:38 -0700

```Ed;
I'm sorry your work on doing the measurement isn't going as well as you
might like it to be. I can make several suggestions and point out some
consequences of measuring the current in this way, and you can decide where
to go from there.

First of all, the method works by measuring the magnetic flux that surrounds
the wire as the current flows through it. The magnitude of the flux goes up
in proportion to the current. The voltage induced in your toroidal winding
will be proportional with the derivative of the flux with respect to time.
That is, the faster the flux (and hence the current) changes, the greater
the voltage will be induced. This will make it interesting to calibrate
such a setup. The fact that a derivative is involved means that a
high frequency, low current event will produce a signal of the same
magnitude as a low frequency high current event, and a simple AC voltmeter
won't be able to tell the difference. Still, assuming that the frequency
remains the same, the signal will increase as the current increases, so
you should be able to get some indication as to when the current is
increasing or decreasing. To get a decent indication of the actual current,
the waveform from the sense coil will have to be integrated over time.
Decent integrator circuits have been around since the days of analog
computers, but the setup is somewhat more complex than just hooking up an
AC voltmeter.

As far as your seeing no signal goes, the signal from a mere 70 turns may
be rather small. I'm digging deep into old memories, but I seem to remember
seeing coils with far more turns inside commercial clip-on-ammeter probes.
Since the output is a voltage to be measured by a high impedance amplifier,
the wire needn't be so beefy as #26. The coil seemed to be a little wad
of the hair-fine stuff found in some relays. The coil was small, but the
fineness of the wire made for many, many, many turns.

This brings up an interesting point. I suggested wrapping the wire around a
plastic toroid because it seems rather rude putting anything that's
ferromagnetic near the secondary. However, if you want to attatch your
sensor some distance from the secondary, you could replace the plastic toroid
with an iron loop of some sort. This would concentrate most all of the flux
within the iron, giving a stronger signal.

Commercial clip-on-ammeters have a ferrite loop with all of the wire wound
around one end of the ferrite loop. This allows the loop to be opened and
clipped onto a wire conveniently. You might be able to do something
similar with some bits of soft iron and the coil taken from a relay. This
would give you lots of turns, which would induce a stronger signal.

On top of all this, I did say in my original post that you may still need an
amplifier. That is, the signal may not be anywhere near what a 5VAC voltmeter
can detect. Then again, your individual meter may not be able to measure
a signal at that frequency and at that very low duty cycle. Without being
there with an oscilliscope to make some measurements, I can only guess.
Still, the procedure does work well enough for many vendors to produce
ammeters to measure current like you are trying to do. If you don't have
an oscilliscope, but can manage to borrow one, a quick measurement of the
peak-to-peak voltage of the waveform across the sense coil would tell us
most of what we need to know to get a decent signal. I hope this helps.

Wes B.

```