50 Hz watt-hour meter monitoring 60 Hz

[ike5547]

Customer provided a made-in-China small electronic watt-hour meter rated for 50 cycles only (though not expressed in those exact words -- just says 50 Hz and nothing else).

I'm concerned that we'll have misreadings when monitoring 60 Hz but am unable to explain how, why or in what way. I don't have the technical background.

This thing is pretty compact and pretty cheap so I don't think it can adapt to multiple input/output frequencies. Chinese-English documentation provides no info -- except 50 Hz.

Can you engineering guys enlighten me so I don't end up having to install (probably/possibly) the wrong equipment for this application?

 

[GoldDigger]

Customer provided a made-in-China small electronic watt-hour meter rated for 50 cycles only (though not expressed in those exact words -- just says 50 Hz and nothing else).

I'm concerned that we'll have misreadings when monitoring 60 Hz but am unable to explain how, why or in what way. I don't have the technical background.

This thing is pretty compact and pretty cheap so I don't think it can adapt to multiple input/output frequencies. Chinese-English documentation provides no info -- except 50 Hz.

Can you engineering guys enlighten me so I don't end up having to install (probably/possibly) the wrong equipment for this application?

If this were an analog watt-hour meter I would be very confident that it did not matter at all. A 60Hz meter design might not work as well at 50 Hz if iron core saturation were a problem (really value engineered design).
I would also not be too concerned about a meter that integrated the product of instantanteous voltage and current samples digitally, except for the possibility that the meter is using the power line frequency as a time reference. If it does that, the results would be unpredictable without knowing more about the internal details.

 

[gar]

161216-2024 EST

No way for us to know.

But you can determine by experiment.

We have a local energy show every year sponsored by our mayor. This closes our main street for one whole evening. If you want to exhibit there are a number of conditions that you must agree to. One is that you will only use recycled paper, but a nonexistent condition is competence.

One year there was one exhibitor that was showing that a CFL was using about the same amount of power as was a standard incandescent with the same light output rating. Totally wrong, but his meter showed this and was marked in watts as a wattmeter. This exhibitor was a PE and he advertized that. But he was a civil engineer that did business with the city and clearly did not know much about electrical engineering.

Whatever this meter was it was probably just an ammeter calibrated in watts at 120 V with a resistive load.

A true wattage measurement will not care what the frequency is. However, there are restrictions to the statement.

Instantaneous power is simply instantaneous voltage times instantaneous current. Anything beyond this and you have to define average power and over what period the averaging is performed. Energy use is always instantaneous power integrated (summed) over some time period.

.

 

[ike5547]

161216-2024 EST Instantaneous power is simply instantaneous voltage times instantaneous current. Anything beyond this and you have to define average power and over what period the averaging is performed. Energy use is always instantaneous power integrated (summed) over some time period. .

So could I assume that if the period is 1 second the sampling rate is probably going to get us close enough? I would assume if it is a 1 second period there's no problem, but if it is some fraction of a second calibrated to a specific frequency (like 50 Hz) there could be a misreading -- but possibly negligible. I could be way off track here. My knowledge of electrical theory is trade level.

 

[ike5547]

Basically I'm trying to break it down to my level of comprehension.

 

[GoldDigger]

So could I assume that if the period is 1 second the sampling rate is probably going to get us close enough? I would assume if it is a 1 second period there's no problem, but if it is some fraction of a second calibrated to a specific frequency (like 50 Hz) there could be a misreading -- but possibly negligible. I could be way off track here. My knowledge of electrical theory is trade level.

Actually, to be able to sense phase and distortion power factor properly a digital watt meter would have to sense a large number of data points per half cycle, not once per second.
Do not confuse sampling rate with reporting rate.

And if the sampling is adequate but the meter thinks that one second has gone by after only 50 cycles of voltage it could overstate the power by a factor of 6/5.

 

[ike5547]

Actually, to be able to sense phase and distortion power factor properly a digital watt meter would have to sense a large number of data points per half cycle, not once per second.
Do not confuse sampling rate with reporting rate.

And if the sampling is adequate but the meter thinks that one second has gone by after only 50 cycles of voltage it could overstate the power by a factor of 6/5.

Thanks, man. I think I get it. The meter might report when it senses 50 peak to peak values and decide that was 1 second when in fact it was less than a second. Or it might report more frequently (per half cycle or maybe even less) based on an expectation of 50 Hz.

This forum is great.

 

[JFletcher]

Actually, to be able to sense phase and distortion power factor properly a digital watt meter would have to sense a large number of data points per half cycle, not once per second.
Do not confuse sampling rate with reporting rate.

And if the sampling is adequate but the meter thinks that one second has gone by after only 50 cycles of voltage it could overstate the power by a factor of 6/5.

That would be my admitted layman-in-this-area's concern.

Do clocks designed to run off of 50Hz run fast on 60Hz? Different animal than a kW meter?

 

[ike5547]

Or something like that. Close enough.

 

[ike5547]

And if the sampling is adequate but the meter thinks that one second has gone by after only 50 cycles of voltage it could overstate the power by a factor of 6/5.

And this makes sense, as well. Watt-hours are time dependent for an accurate measurement. The last horse is finally crossing the finish line here.

 

[mbrooke]

So wait, KWH meters are frequency dependent?

 

[GoldDigger]

So wait, KWH meters are frequency dependent?

Only if they do not have their own clock, NOR are mechanically independent of any need for an external clock input, like an old mechanical meter.

Modern crystal oscillator time sources probably used in smart meters are easily accurate to within a few seconds per day and can also be resynchronized every time they check in if necessary.
But I have no idea how the OP's non-revenue-grade meter was designed.

 

[mbrooke]

Only if they do not have their own clock, NOR are mechanically independent of any need for an external clock input, like an old mechanical meter.

Modern crystal oscillator time sources probably used in smart meters are easily accurate to within a few seconds per day and can also be resynchronized every time they check in if necessary.
But I have no idea how the OP's non-revenue-grade meter was designed.

So an older dial watt hour meter relies on frequency for accuracy?

 

[GoldDigger]

So an older dial watt hour meter relies on frequency for accuracy?

Nope. If you spend a few minutes analyzing the logic of my admittedly tortured sentence you will, I hope, see that I am saying that the old spinning disk units are not strongly frequency dependent. They will definitely not work on DC since they involve induced eddy currents in the conductive disk.
As I think about it a bit more, I am not absolutely convinced that they are not frequency sensitive between 50 and 60 Hz.

The question remaining is whether the difference is solid ration of 6 to 5 or just a small percentage difference caused by practical design factors.

 

[iwire]

I know nothing of the science behind metering but I can infer like a pro.

I just looked at a bunch of pictures of old style meters and they seemed to be marked with 60hz. From that I infer it matters.

 

[mbrooke]

Nope. If you spend a few minutes analyzing the logic of my admittedly tortured sentence you will, I hope, see that I am saying that the old spinning disk units are not strongly frequency dependent. They will definitely not work on DC since they involve induced eddy currents in the conductive disk.
As I think about it a bit more, I am not absolutely convinced that they are not frequency sensitive between 50 and 60 Hz.

The question remaining is whether the difference is solid ration of 6 to 5 or just a small percentage difference caused by practical design factors.

I disagree for the same reason as Iwire- they are marked 60hz.

 

[mbrooke]

I know nothing of the science behind metering but I can infer like a pro.

I just looked at a bunch of pictures of old style meters and they seemed to be marked with 60hz. From that I infer it matters.

And those in 50Hz land are specifically marked 50Hz:

 

[gar]

1612117-1030 EST

The spinning disk KWH meter is not inherently frequency dependent except at high and low frequencies. However, these are designed and calibrated at a specific frequency. There are a number of factors in a KWH meter (friction, electrical losses, phase shift of the coil curent in the voltage coil, temperature, etc.) that influence the meter accuracy. To get the best overall accuracy a single frequency is selected for the meter.

How an electronic KWH meter performs vs frequency is dependent upon its design. Could be very dependent, or relatively independent.

As I previously mentioned run a test on your meter at 60 Hz.

Kill-A-Watt EZ meters from Home Depot at about $30 are quite good, but long term or even out-of-the-box reliability and quality may not be good. Of the 6 or so that I have some are now good and some not. When working correctly accuracy is good and performance with varying types of loads is good. On a correctly working unit I go from a purely resistive to capacitive load and inbetween and get the results I expect. Not so with a TED energy monitor.

.

 

[Besoeker]

And this makes sense, as well. Watt-hours are time dependent for an accurate measurement. The last horse is finally crossing the finish line here.

I can think of two things you could do.

1. Contact the supplier and get him to confirm that it is or isn't suitable for 60hZ.
2. Test it with a known load. Say, a 1kW fire bar (heater element). Run it for a measured period of time then you can calculate the kWh it has consumed and compare that with your unit reading.
   

 

[gar]

161220-2114 EST

ike5547:

Can you run a test on the meter so that we gets some feedback on how it works?

If you need suggestions on how to do the test we can help.

.