Back EMF Constant within the Simscape PMSM block

Hello guys,

in the Simscape PMSM block I want to type in the Back EMF constant (kE).

But I cant find if I should give the data as a peak, rms or mean value neither if it should me a Line-to-Line or Phase value?

Because every Motor-Designer gives these data in different ways.

Thank you in advance.

Best regards.

Answer 1

[youjarr]

Hello guys,

mathworks contacted me again and corrected their answer and provided me an example model.

The simscape PMSM block needs the BackEMF-Constant as peak and Phase value.

You will find the model attached.

Thank you all.

[Pat Gipper]

Thanks for posting this followup. The Mathworks example gives a nice implementation of a motor spin-test using Simscape. Specifying the flux linkage in Webers is always a bit of a question. Typically for a Wye connected motor I will have a measured phase to phase peak voltage which will be converted to either a torque constant in N*m/A or back EMF constant in V/(rad/s), which will have the same numerical value. But you can't just enter this value into the PMSM block parameters because this is asking for a "Phase value". So the value you enter needs to be divided by sqrt(3). It's a best practice to run this example to double check your selections.

Answer 2

[Tobias Schindler]

kE is line-to-line in rms by standard definition. Note that does not mean that motor designers supply the value according to the standard definition nor that matlab follows this standard.

[youjarr]

Hey Tobias, thanks for your fast response.

Yes I know that kE is by standard L2L and RMS, but what exactly does that mean for the Simscape PMSM?

And in PMSM with DELTA connection Phase is equal to L2L.

I mean the value has to fit in the equations behind the model?

But when I read the help it sounds like it needs the kE as peak Value per phase ?!

This is the only thing I can find in the block help:

Simscape - PMSM - Help

Alternative Flux Linkage Parameterization

You can parameterize the motor using the back EMF or torque constants which are more commonly given on motor datasheets by using the Permanent magnet flux linkage option.

The back EMF constant is defined as the peak voltage induced by the permanent magnet in each of the phases per unit rotational speed. It is related to peak permanent magnet flux linkage by:

ke=Nψm.

From this definition, it follows that the back EMF eph for one phase is given by:

eph=keω.

[Tobias Schindler]

https://de.mathworks.com/help/physmod/sps/ug/parameterize-a-permanent-magnet-synchronous-motor.html

Seems like Matworks takes the freestyle approach to defining Ke ;)

Personally, I would use psi_pm and check if the torque of the simulation makes sense.

[Pat Gipper]

I have double-checked the Simscape PMSM interpretation of Ke and Kt in the past by doing the following; 1) Connect each phase through a large resistor to ground 2) Make the motor inertia value large 3) Give the inertia a speed initial condition. Then I just capture the phase to ground voltage across the resistors as a function of time. It duplicates the motor free-spin test.

[youjarr]

@tobias it seems so... . I think I am going to ask the support because this couldn´t be the solution... .

@Pat can you share what you find out if u already have tested it? Can you share your model and the results?

Answer 3

[youjarr]

Hello,

I understand that we need a BackEMF constant peak and over one phase.

In a DELTA connection it doesn´t matter because phase and Line-to-Line is the same.

I tested the whole thing with a model and talked to a motor designer.

But what irritated me is the following text from Mathworks, which I asked for to do so just to make sure that it was there:

Quotation:

However, I checked the doc the doc for our internal database and I phase that the Back EMF constant means that the continuously magnet voltage induced by the permanently magnet each of the units per speed. This means you should use the peak line-to-line value. I am assuming this answer your question.

Ican therefore recommend to do a back EMF estimation according to the following example:

https://de.mathworks.com/help/physm....et-synchronous-motor.html

I find completely contradictory.. .

The attached model of the Mathworks people couldn't help me much.