Currents, Meters, Circuit Breakers

Further to my ramblings about circuit breakers.

The question arose when 30A or so on the meter did not trip a 20A or
so Circuit breaker.  Now maybe the CB misperformed.  Or.....

Lots of things influence the readings on a meter.  Its easy to trust the
meter.  In a relative sense, (30a indicated is more than 20A indicated) thats
fine.  In an actual sense, welllllllllll.

There will be small scale variations, manufacturing tolerances in magnets &
coils  & scaling resistors.  Under 5%, with any luck.  There are also larger
scale effects, which can have significant effects.  Notably, each kind of
meter responds differently to the wave shape of the applied current.
And, driving a gap, the current is NOT sinusoidal.  For a simple fixed gap,
the curretn will be zero until the secondary makes sparkover voltage, then
go up FAST.  May saturate, depending on the transformer & cicuit details,
then go to zero FAST as the gap quenches.  Repeat for each half cycle.
Gets REAL complicated for a rotary gap.

Each type of AC meter will respond differently to the wave shape.  And each
Tesla-type circuit will have its own wave shape.  So its REAL hard to
(off site) generalize on how close 20A indicated is to 20A real.  There are
four principal classes of AC ammeter.  (5? lets see how well i do...8)>>)

	1) Rectifier Type: Stndard DC movement, with FW bridge.  Usable, but
	the bridge is vulnerable.  Not commercially available in the 20Aish

	2) Moving Iron: An iron slug moves in a magnetic field.  Rugged,
	common.  Non linear, & shows it in the dial markings.

	3) repulsion: Similar to moving iron.

	4) Hot wire: Not common.  Arguably best, as immune to wave shape
	variations.  Not rugged, generally.

	5) Thermocouple: NOT rugged, not common.  Also immune to wave shape

	6) Electordynamic: Not common. OK On wave shape, reasonably rugged.

And thats just the "needle type" meters....

DVM type instruments are likely to be suspect in the high RF fields.  Some are
effectively rectifier types.  Some use a wave-shape immune "true rms" element,
making them equivalnet to hot-wire or TC types.

When a workng electrical engineer (these days) has to work with complex
waveshapes, the weapon of choice is a digitizing scope, which can capture
current waveshape & will then (internally) compute & display the "real" (RMS)

I can list out if wanted, but any good EE text will talk of this in more or
less detail.  _i_ will go into more detail if interested, but, as above, its
not practical to say "meter foo will read mumble% high in circuit diddle".

Treu Story:
(I was there....)
Back in my student days, we had a motors lab, that ended up with us trying to
measure high currents (turned out to be an instructor error, details omitted
for brevity...).  We ended up with 3 meters, paralleled to try to measure the
currents (ca 70A?  its been a while...8)>>).  ANYWAY, the instructor could ALSO
read the current on the lab swichboard.  Team one got, say 65A.  The instructor
said, try again, i see 55.  Team one got 65A AGAIN.  Team two (Vic, others, me)
got 63A.  Instructor got 54A.  Instructor read OUR meters & agreed with our

He finally figgered out that we were pulling so much current we were either
overloading the metering transformers on his board & that distorted waveshape
was fiddling HIS meter, or we were modestly overloading the lab supply and BOTH
sets of meters were being fooled.

Moral:  Meters are no more goofproof than any other piece of a design.