--- In 7x12minilathe@..., "Ian Foster" <fosterscons@...>
wrote:
John was concerned about the effect of the waveform
on meter accuracy. I don't think it is an issue because the
industrial drives with which I am familiar (up to 1500HP) have
used
moving coil meters connected across shunts. I think the armature
inductance smooths the current enough for the meter.
Hi Ian,
I think I'm being misquoted or at least misunderstood. Let's try to
baffle with detail again. Check me Uncle Rabid!
My concern over waveform affecting meter accuracy were only in
regard to electronic meters - like the ubiquitous digital meters.
This is especially so at low speeds measuring PWM with FET
controllers. The peak value is much bigger than the average and with
the meter on what appears a suitable range (even auto-ranged) the
peak signal gets clipped internally in the meter electronics prior
to digitising. If you're using such a meter over-ride any auto-
ranging. On full load increase the meter range until increasing it
further doesn't change the measurement result (just reduces the
available number of digits of resolution). Then stay on that scale.
Putting crest factor issues to one side, there are 2 main classes of
non-electronic meters in everyday use - moving coil and moving iron.
Both will accurately measure waveforms involved in FET and SCR
controllers as long as 1) the currents flow through their magnetic
windings (not shunted at HF by stray capacitance), 2) waveforms
aren't so spikey that the magnetic path saturates on peaks (similar
to the electronic meter clipping except not as abrupt) and 3) the
frequency components are within the performance capabilities of the
magnetic path materials. BTW, it's not motor inductance that does
the smoothing but the ballistics of the meter movement.
The big difference you need to keep in mind is that moving coil
meters are AVERAGE responding. Moving iron meters are true RMS
responding, on their supplied non-linear scale.
Actually, moving coil meters respond to the product of a coil
current and a constant permanent magnet field. Once averaged by the
meter ballistics you get a deflection proportional to the true
average. OTOH, moving iron meters respond to the product of a coil
current and a magnetic field generated by that same current through
a second series coil. Yes, I do know that's a gross simplification
of the moving iron meter principle! That means they respond to the
square of the current through them. Once averaged by the ballistics
they deflect proportional to the mean square and the scale is
conveniently graduated in a non-linear fashion to give the square
root of that deflection. Yes, it's truly Root Mean Square.
Now, which meter do you actually need? That depends on what you want
to understand. If you want energy transfer into the motor then it's
further complicated by the motor voltage being rectified (but
unsmoothed) AC. Best use a real power meter that puts the voltage
and current on 2 windings, somewhat akin to the moving iron
principle, giving a true average power. Or log both voltage and
current waveforms to a PC and crunch the numbers - that's what I'd
do in my lab. If you're concerned about copper loss in the motor,
that's just I2R and a true RMS current that hasn't got clipping
issues will serve you well.
By now I've no doubt confused you. Leaving aside the math behind
this, I think for your purposes any non-electronic meter will serve
as a relative indicator. An electronic one requires more care. An
RMS indicating one will relate particularly well to fuse opening. I
suspect anything more refined is more suited to the lab. In short,
I'd have used a moving iron meter - but I wasn't gonna tell you that
after you bought the MC and I'm sure it'll serve fine as a relative
indicator. And all you really wanted was a simple recommendation! <G>
John
