OT: Smart meters...not very!

Oh they do it right allright. They overcharge without you being able to argue :-)

The Natural Philosopher presented the following explanation :

There is no valid evidence of this - ignoring that wild article mentioned.

They sometimes (don't know how often, I've only seen one example) have a shading coil which provides an additional magnetic field that's out of phase with the voltage coil. I believe this is designed to just overcome frictional losses so that very small loads are measured. But sometimes (again, don't know how often), this makes the disk creep. So some disks have a hole or slot drilled into them at a certain position, and they will creep round to that point at zero load.

I've only seen one example though, and that was a very old 240v single phase meter back in the 1970s (the meter would have been much older than that). So I could be talking about a very rare behaviour on a particular type of electromechanical meter.

The permanent magnet provides a braking force that's proprotional to the speed of the disk. As you say, it's stops the disk spinning like an induction motor and reaching 3600 rpm or whatever.

But it will have virually no effect at low speeds, because the eddy currents that it induces get lower as the disk spins slower. So I don't think the permanent magnets stop disk creep (if it's ever actually a problem in the first place, which I'm starting to doubt now).

Perhaps because very few people put a second meter upstream of the electricity company's one, and therefore there's no way for most people to know.

Perhaps this inaccurate measurement is not as common as the article states, but how would you know? I certainly wouldn't.

That's a good point - if there was a shading coil, and it was mis calibrated, I guess that would do it.

Its in the Register as qwell. So not as wild as you may think

Caecilius explained :

A known fixed electrical load on the meter would be a good indication, within a few percent.

I know my boilers maximum load and its gas demand is logged, so a means to check the gas meter is within reasonable bounds of accuracy.

The Natural Philosopher explained :

I think most people would notice the difference between paying say £40 per month for their Kw needs and £240 per month. As said above, it is not difficult to carry out a basic check of accuracy.

Except the type of load they claim was inaccurately measured is nothing like a steady resistive load ...

My memories of working with these smart meter chips are somewhat hazy, but as I recall, the problems are twofold, firstly the range of currents over which the current sensor is required to work and secondly the sampling rate/accuracy of the A/D converter in the chip. Not easy problems to solve. We had a lot of chip problems.

Capitol pretended :

I can imagine the output will not be very linear and will need some trimming in the firmware.

ISTR hearing about it when I was in the electricity board and it happened sometimes if the meter was not installed level. Never actually saw it.

I have a fish tank aerator works on the principle.

It would not be self starting but would run on once it was going. I have a fish tank aerator works on the principle. Drives a twin cylinder pump

Read meter. Turn (3Kw) immersion heater on for 20 minutes. Read meter again.

(Everything else to be turned off.)

Do it again using whatever else smaller load you have available over a longer period if necessary.