Delta-wound FOC PMSM

Hello,

I am currently using the Three-Phase PMSM Drive example to model a Field-Oriented Control (FOC) PMSM. The model works as expected when the machine is configured as Wye-wound. However, when I change the winding configuration to Delta-wound, I notice that the measured torque is half of the reference torque.

Could you please help me understand where this discrepancy comes from? Is there a scaling factor or adjustment that needs to be applied when using a Delta-wound configuration?

Thank you in advance for your support.

Best regards,

Answer 1

[Orrin]

It helped me, Thank you so much.

Answer 2

[Yoshi Minagawa]

Hi Yacine,

As you mentioned, when switching PMSM setting between Y-connected and delta-connected windings, it’s necessary to rescale the values for winding resistance, inductance, and flux linkage.

In this demo model, the PMSM is controlled using the PMSM Field-Oriented Control block. While this block allows you to set the winding type to Delta-wound, it appears that this setting only adjusts the angle offset, and does not automatically rescale the winding resistance, inductance, or flux linkage values.

Therefore, to properly control a delta-connected PMSM, I think you should manually adjust the settings in the PMSM Field-Oriented Control block as follows:

Permanent magnet flux linkage (Wb)

＞PM/(3)^0.5

D-axis inductance for feed-forward pre-control (H)

＞Ld/3

Q-axis inductance for feed-forward pre-control (H)

＞Lq/3

[Yacine]

Hi Yoshi,

Thank you for your answer.

I tried adjusting the permanent magnet flux linkage (Wb) in the PMSM Field-Oriented Control block by setting it to PM/√3. However, I still observed a torque offset between the measured and reference values.

Interestingly, the torque aligns correctly when I set the flux linkage to PM/2 — though I’m not sure where this factor comes from. I would have expected √3 to be the correct scaling factor, as it matches the typical relationship for phase currents in Delta-wound systems.

Do you have any insight into why PM/2 works here ?

Best regards,

[Yoshi Minagawa]

Hi Yacine,

Thank you for checking so quickly.

In my environment, it looks like the actual torque is following the target torque as expected. I’ve attached my settings screen for your reference—I hope it helps.

Also, like you, I believe that √3 is the correct scaling factor. Unfortunately, I can’t think of a reason why using PM/2 would work.

　　

[Yacine]

Hi Yoshi,

All good, it worked correctly after I properly set all the parameters.

Thank you very much.