[volt-nuts] current-nut question .. total waste of ones time type question

Tue Apr 29 19:16:01 EDT 2014

Hi Pete,

Retired utility engineer here with lots of metering experience.

If you want the right answers, don't stray from revenue metering
equipment.  FERC requires that revenue metering be within 1% of actual.
 That means that even the tail of the error distribution curve has to be
better than 1%.  Most utilities I've worked with shoot for 0.25% and use
0.1% components.  CTs, PTs, revenue meters, revenue watts and VAR
transducers and so on.

The first thing you need to do is to set up your service entrance so
that you can experiment later.  That means being able to kill power to
your breaker panel.  What you want is the utility feed, a meter base, an
outdoor 200 amp service disconnect and have that feed your panel
indoors.  That way you can open the disconnect breaker and kill your
entire breaker panel and not have to try to work things hot.

The second thing I'd do is install a second meter base INSIDE between
your service disconnect and your panel.  That way you can install either
your own watt-hour meter or install jumper bars if you decide you don't
want your own meter.

You can get meter bases and meters for practically nothing here:

http://www.hialeahmeter.com/

That will be the most accurate power and energy measuring device you can
get your hands on for a reasonable amount of money.  Typically 0.1%.

Since you own the meter you can modify it.  You can install a
photo-interrupter to count disk rotations.  There is always a hole in
the disk called the anti-creep hole. (causes the wheel to stop under the
potential coil if there's no load but the wheel is still barely turning).

The Kh value printed on the meter face, normally 7.2 for 200 amp single
phase meters, is the number of WATT-hours (not KWH) that one turn of the
disk represents.  So you can total watt-hours by counting revolutions
and multiplying by Kh.  Take the rate (first derivative) and multiply by
7.2 to determine the current load in watts.

Every mechanical meter I've ever seen has 100 graduations around the
periphery of the wheel.  Some are printed and some are engraved.
There's a "zero flag" 5 divisions wide to mark the start of another
revolution.  If you rig up a retroflector sensor, you can count these
and have a 0.72 resolution.

If you want to measure voltage to revenue standards, use a voltage
transducer.  Typically the output is 0-1ma DC.  Many moons ago I settled
in on Ohio Semitronics as my supplier of transducers.

I've installed many energy audit systems in large factories where the
sum of several large load centers is compared to the input from the
utility.  The utility input is typically a pulse per X kWh.  In a large
plant such as a paper mill, the pulse would be per megawatt-hour.  The
individual measuring loops had to be quite accurate for their sum to
match the utility input to within less than a percent.  Ohio Semitronics
hardware never let me down.

The Chinese are making knockoffs that can be found on alibaba.com and
alimarketplace.com (or something like that).  I haven't used enough of
them to know their quality.

If you don't mind spending a little money, a revenue-grade 400:5 CT and
this instrument

http://www.ebay.com/itm/350423261101

Is what you need. It is a 3 phase meter but works just fine on single
phase.  I now work part-time designing induction heaters and this is my
lab standard for measuring heater power.  I matched my analog 0.1% lab
standard watt-meter exactly.  And it's true RMS.

Measuring the Edison three wire setup that we all have requires a little
setup.  You'll put the CT in your breaker panel with both hot leads
going through it.  One lead will be reversed from the other.  That way
120 volt to neutral loads are measured once while 240 volt loads are
measured twice. The meter's potential input is straight 240.

The key to making this work is to program your CT factor to HALF the
actual value.  So the real value is 400:5 or 80 but you program in 200:5
or 40.

It works like this.  A 240 volt load is measured twice so the signal is
240 X (2Xamps) * cos(theta) so the results would appear to be twice the
actual load.  But you've halved the CT factor so the indication is correct.

For a 120 volt load, the  signal 240 * amps * cos(theta).  That would
seem to be twice what the 120 volt load is pulling but again your CT
factor of half makes the reading correct.

For data processing, the meter has several digital output options.
Standard are RS-485 and RS-232.  Once a second it outputs a data
sentence similar to an NMEA signal in plain text that is easy to parse
and process further.  Optionally, the meter can have an ethernet
interface with a built-in web server.  I didn't find any value in that
so I didn't pay extra.

The revenue-grade CT and a revenue-grade meter is the only way that
you'll be able to make measurements accurate enough to compare to your
utility bill.

Especially stay away from the "solutions" that put a CT on each leg.
That creates the classical "small difference of two large values"
problem.  For example, if your heat pump is drawing 100 amps and a CF
lamp is drawing a quarter amp, the CF signal will be lost in the error
band.  Even a 0.1% 200 amp CT is +- 200 ma.  That's for each CT.  And
that's not counting the errors inherent in the circuitry that tries to
measure to that kind of resolution.

There are only two bad things that I can say about the above digital
meter.  First, the blue digits are very hard to read.  There is no
filter over the displays and the lit segments aren't very bright so it's
hard to tell which ones are lit if there is much room light.  I had to
make a hood out of black paper for mine.

The second thing is the manual.  The worst chinglish I've ever run into.
 I waded through it enough to get my meter set up for my needs but it
was a chore.

Finally stay FAR away from electronic revenue meters.  Especially the
Iton brand.  I've consulted to one utility regarding electronic metering
and know the details of a couple more.  All have been unmitigated
disasters.  Failure rates are very high and accuracy is all over the place.

My rural co-op learned from these debacles and did it right.  They use
something called the Turtle system.  It involves a photo-interrupter
board that is added to the mechanical meter to count wheel turns.
Somehow it phase modulates the data at a very low baud rate the 60 hz
signal in a manner that goes back through transformers and bypasses
power factor correcting capacitors.  The data rate is very slow, hence
the name "turtle".  There's a patent describing the system but it's
probably the most obtuse document I've ever tried to read.

This system works extremely well.  I've never had a dispute between my
bill and my own meter reading.  See, even the smart utilities are
staying with mechanical meters!

Hope this helps.

John

On 04/29/2014 01:25 PM, Pete Lancashire wrote:
> Next spring I'm going be rewiring my house. And for 'fun' I want to drop a
> couple current sensors on the input side of the main panel.
> 
> Being a beginner 'nut', I'm looking for more accuracy then needed.
> 
> Voltage, Frequency, waveform will be taken care of later, but accurate
> current has be a bit flustered.
> 
> I've been reading up on CT's Iron core and Ferrite, and on Rogowski coils.
> And the many new IC's on the market that take care of a lot of things.
> 
> Has someone done this before ? And if so any experiences to share ?
> 
> -pete
> _______________________________________________
> volt-nuts mailing list -- volt-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
> and follow the instructions there.
> 

-- 
John DeArmond
Tellico Plains, Occupied TN
http://www.fluxeon.com      <-- THE source for induction heaters
http://www.neon-john.com    <-- email from here
http://www.johndearmond.com <-- Best damned Blog on the net
PGP key: wwwkeys.pgp.net: BCB68D77