Measuring inductive loads such as washing machines produces errors in power factor

I am measuring the consumption o different electrical devices such as resistive loads or capacitive and inductive loads.

when i measure resistive loads everything goes ok, but when it goes to complex loads it is not workin.

Can anybody help me. 

thank you.

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

See the FAQs for the information we need to be able to help you.

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

I have seen the FAQs but I cant find anything.

Maybe I have to be more clear: when I am measuring only a resistive loads everything is ok at this time when I plug in a washing machine everything changes the active power becomes greater than reactive one which is impossible. 

Can you help me because it is really important for me.

dBC's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Did you see the one entitled:

Q:   I'm having trouble with... Can you help me?

(http://openenergymonitor.org/emon/faq)

The answer to that one provides a list of details you need to provide if you want useful help, and the one immediately after it about electricity supply might be useful too.

Active Power is almost always greater than Reactive Power(*), and in any case I don't think the standard sketches report Reactive Power do they?  Are you sure you didn't mean...

the active power becomes greater than apparent one which is impossible

in which case I'd agree, that shouldn't happen.  

But without knowing which sketch you're running on which hardware, and seeing the actual output of that sketch, and knowing what sort of electricity supply you have, it's almost impossible to guess what the issue could be.

(*) I've got a few very low-powered devices where that's not the case, for example a small circulation pump on my thermal solar hotwater system that uses about 22W and 66VAR.  But in general Active Power is by far the bigger of the two.

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Thank you. yes I mean apparent power but I have done a mistake by saying reactive. I am runing it on emontx v3 with a arduino one incorporated. I am runing emonlib for energy monitoring. I detailed information could help me.

I think the problem is on phase shifting but I canot calculate it exactly.

dBC's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

emonlib is a library.  We need to see the sketch you're running, how it calls that library, how it prints out the results, and the actual output of those prints (pasted here).   Without some real hard numbers as output by the sketch it's difficult to interpret what you've reported.

I can see how  phase errors could make your Real Power under-read or even go negative, but I'm struggling to see how it could ever make it go larger than Apparent Power.  Is your electricity supply single phase, two phase, or three phase?

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

thank you. my electricity supply is single phase. later on I will give you my measurement. but can you send me the sketch that measures the power of the heating pump that you mentioned before it would help me.

dBC's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

I don't have a sketch that measures it. I made the quoted pump power measurements above with a power meter.  But I think the standard OEM sketches used in these parts should be able to measure it fine (subject to all the normal caveats about CT sensitivity, low-end readings, dynamic range etc.)

You've still not said which sketch you're running?

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Thanks dBC for helping here. You are right, no sketch measures the imaginary component. Power factor is calculated from the definition, i.e. the ratio of real power to apparent power. Real power may be reported as fractionally greater than apparent power due to rounding errors in the maths, and the same rounding errors can also give a power factor of fractionally greater than 1.

Shabani A:
Much as we would like to help you, until you tell us EXACTLY everything that you are using, we cannot. That is why I wrote very early on: "See the FAQs for the information we need to be able to help you."

If you think your problem is the phase correction, do the step-by-step procedure here:

http://openenergymonitor.org/emon/buildingblocks/calibration

You will find it very difficult or impossible to calculate, because the properties of the transformers change as voltage and current changes. You can only set it by experiment to be correct under your typical operating conditions.

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Thank you. I am using the sketch that i have sent attached at you and the results of the measurement are those.

thank you again.

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Did you have a resistive load?

The error in the power factor is less than ½% in every case but two. Those results look to me to be perfectly acceptable for a resistive load.

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Thank you for your comment but when the power factor increases the motor of washing machine is turned on, i think that the power factor should decrease. that is the problem that I have. because for active loads is going very well but when a reactive load takes part I should not have these results.

thank you again.

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

"when the power factor increases the motor of washing machine is turned on, i think that the power factor should decrease"

I think I know what you wanted to say there, but why do you think the power factor should change?
You are STILL not providing enough data for us to be able to help you. What is that table intended to show? All it tells me is you have a varying load that appears to be purely resistive.

Do you know exactly what happens inside the washing machine? Do you know what type of motor is in there? Do you know what type of controller is in there? Have you ignored the possibility that it is an electronic controller with power factor correction, and the controller is doing exactly as it is supposed to do?

Everybody else seems to get sensible results using the emonTx, so until you tell use what - exactly what - you are doing, so that somebody can see what you are doing wrong, nobody can help you.

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Thank you. I have calibrated the voltage and current measurement with a professional multi meter and everything is going well. the readings are very accurate. But I am using an old washing machine without controller or anything electronic. And that why I am making questions because I want to have the reactive power measured. 

Also I have add to my scheme a pure capacitor and it is going well.

But I think  that the problem should be in phase shifting and also I have read the documentation about phase correction but I think there should be the problem. Any more detailed information about that could help me. And I am going to try it again.

Also I have another question: Which is the largest frequency that the Emontx V3 can capture signals with high accuracy. 

Thank you.

calypso_rae's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

Which is the largest frequency that the Emontx V3 can capture signals with high accuracy. 

The Atmega 328P processor on the emonTx requires approximately 104 us to do each A->D conversion.  By sampling voltage and current alternately, 96 pairs of samples can be taken every 50Hz mains cycle.  So that's 4800 sample pairs per second which gives a Nyquist frequency of 2.4 kHz.

The above calculation is for when the processor has nothing else to to do.  When other tasks have to be done as well, it will be unable to perform in this way.  In practice, the highest frequency at which power can be measured will be very dependent on the software that is being used.

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

The standard calcVI( ) method runs at just over 50 sample pairs per 50 Hz cycle, so better than 2500 samples per second. Therefore, your input signal may contain components up to but not above 1.25 kHz.

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

I have now found the time to use an emonTx V3.4 to measure the power drawn by an unloaded transformer. This is what my (uncalibrated) emonTx running the standard discrete sampling sketch reported:

CT1 Irms = 0.259, CT1 Vrms = 247.09, CT1 Power = 12.22, CT1 VA = 64.09, CT1 pf = 0.191

I have not tried to measure the inductance and do the calculations, but that result looks reasonable to me. It certainly does not support your claim that the sketch (and by implication emonLib) reports that real power and apparent power are equal for inductive loads.

Therefore, until you say exactly how you are making your measurements, I am sorry but I cannot help you any more.

Shabani A's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

thank you for your answer. Can you send me tha sketch that you are running. But I think my problem is with phase shifting calibration. Maybe you can help me at this point. It is very important to me to have as much as possible to have an accurate reactive power measuremnt.

Thank you again.

Robert Wall's picture

Re: Measuring inductive loads such as washing machines produces errors in power factor

As I wrote, "... my (uncalibrated) emonTx [is] running the standard discrete sampling sketch..."

The only change was to add the print statement to report those values.

"It is very important to me to have as much as possible to have an accurate reactive power measuremnt."

Did you not understand what was written earlier? You CANNOT get a reactive power measurement from the emonTx. You can only have real power and apparent power.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.