• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 84 subscribers
  • Views 455 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TMS320F28027F: Interior PMSM motor in torque loop loaded by regenerative dyno is causing estimator to get stuck in Idle

Palanivel Subramaniam
Prodigy 10 points
Part Number: TMS320F28027F

I am trying to run Interior PMSM motor in torque loop.

My estimator becomes stuck in Idle suddenly during torque control less than 2 seconds after load ramps up in a regenerative dyno.

I am running a modified proj11b on 28027f in HVkit connected to 1hp interior PMSM.

Scope current waveforms show no overcurrent or transient.

The average leakage is around 50 mH.

No other faults such as dc bus voltage out of range or hardware or software over current or external TZ inputs are observed.

Same code works fine for surface mount PMSM.

Idref is set to 0 during the run.

 

Please help find out what could cause estimator to suddenly get stuck in idle from online state.

When not loaded by a regen dyno, the drive runs fine.

___________________________

  • 0
    TI__Guru** 112790 points

    Any results if you just add a slight load or implement speed-close loop? And what's the Ld and Lq values of the motor? Are you using the Ld and Lq value according to the spec. of the motor, or using the identified Ls value for the IPM motor?

    Could you please monitor and capture the current during increase the load?

  • 0 in reply to Yanming Luo
    Prodigy 10 points

    Hi Yanming Luo,

    the load is being added by regen dyno

    torque loop is required to work

    Ld=18mH, Lq=25mH are used; values are from nameplate;

    current waveforms look clean on scope. No spikes or glitches.

  • 0 in reply to Palanivel Subramaniam
    TI__Guru** 112790 points

    You might try to identify the motor as SPMSM, and use the identified parameters to run the motor, and see what happens.

    Is the torque current enough to overcome the adding load? And you might try to run the motor with speed-close loop first for checking if the issue is from the estimator or control parameters.

  • +1
    Prodigy 10 points

    The motor and drive runs fine in torque loop as long as it is not loaded by regen dyno. Other types of loading are not causing estimator issues. 

    will the estimator work with interior pmsm in all four quadrants?

  • +1 in reply to Palanivel Subramaniam
    TI__Guru** 112790 points

    Yes, the FAST estimator can work with this IPM motor since the salient ratio is not too high. We don't test a motor with four quadrants, just verify spin the motor with forward and reverse direction.

  • +1 in reply to Yanming Luo
    TI__Genius 10435 points

    Latest from Pal:

    Just to recap the discussion:

    1. Motor control works well in speed and torque loop when loaded by loads such as a blower wheel or a braking dyno.
    2. When connected to regenerative dyno, (dyno is set to load dynamically to maintain a certain speed while the motor and drive configured to run in torque mode) the observer enters idle mode within 2 seconds of delivering torque.
    3. Same behavior is observed under following conditions:
      1. The delivered torque can be anywhere from 5 oz-ft to 85 oz-ft
      2. The speed is at 200 rpm regulated by the dyno
      3. The motor control starts regulating its generated torque after dyno has been running the shaft at 200 rpm for some time.
      4. No observed spikes or glitches in actual motor currents in any phase
      5. No other faults are detected
      6. All other operating conditions are nominal

    We are still experiencing the issue with no suggestions on how to proceed. Any ideas on how to proceed?

  • +1 in reply to Jason Reeder
    TI__Guru** 112790 points
    Jason Reeder said:
    Motor control works well in speed and torque loop when loaded by loads such as a blower wheel or a braking dyno.

    So that means the FAST estimator is good for this motor. So the issue should be the incorrect gains of the speed/current controller or other control logic codes.

    Jason Reeder said:
    When connected to regenerative dyno, (dyno is set to load dynamically to maintain a certain speed while the motor and drive configured to run in torque mode) the observer enters idle mode within 2 seconds of delivering torque.

    Please make sure that the adding load is less than the output torque of the motor, and the motor will be not stall for there is no enough torque output the torque control mode is using in this case.

    Please check if there is any faults to stop the motor like over current, stall, even the control program/device is reset. The FAST estimator will not be changed from ONLINE to IDLE mode automatically if there are no any errors, The issue might be from the control codes they added if the motor works well as mentioned above.

    Since lab11b is using in this case, you might try to use the speed closed-loop control with FAST first to check if the motor works well at constant speed and load, and set the learning rate value (gAlpha) to zero. And try to run the motor at different speeds to see what happens.

    Also, you have to tune the gains of the speed/current controller to ensure that the output torque current can response the load changes.