• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 83 subscribers
  • Views 415 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

constant torque

Bene
Prodigy 20 points
Other Parts Discussed in Thread: DRV8301-69M-KIT

Hello,

i´ve got the DRV8301-69M-KIT and i am trying to test the performance of the board.

Í blocked the rotor, so the motor can´t turn around. With the blocked motor i´ve measured the torque of the Motor, but the torque isn´t constant if the motor is blocked. The graph of the torque shows a sinus signal.

Is there any possibility to get a constant torque with a blocked rotor?

Thanks a lot

Benjamin

  • 0
    Intellectual 310 points

    Bene said:

    Í blocked the rotor, so the motor can´t turn around. With the blocked motor i´ve measured the torque of the Motor, but the torque isn´t constant if the motor is blocked. The graph of the torque shows a sinus signal.

    Is there any possibility to get a constant torque with a blocked rotor?

    I've got the same problem. Perhaps this thread answers your question: http://e2e.ti.com/support/microcontrollers/c2000/f/902/p/335596/1172684.aspx#1172684

  • 0 in reply to MKo
    Prodigy 20 points

    Thanks a lot.

    So the forced angle didn´t use any feedback from the motor at low frequencys, so the controller can´t see that the rotor is blocked and doesn´t move, right?

    So the PWM-Signal at low frequencys with a blocked rotor is the same as with an open rotor?

    Please tell me if i´m right

    And one last question: Does the controller know the startposition of the rotor?

     

    Benjamin

  • 0 in reply to Bene
    Intellectual 310 points

    Bene said:

    So the forced angle didn´t use any feedback from the motor at low frequencys, 

    Right. Somewhere below a certain motor frequency the measured signals vanish to zero and can't be used to reliably feed the motor model.
    Bene said:

    so the controller can´t see that the rotor is blocked and doesn´t move, right?

    Not quite. It assumes that the rotor is blocked or rotating slowly below a certain threshold  and therefore generates a rotating angle for the FAST observer to get it up and running again (ForceAngle feature) since currently there is no other way to get a "valid" rotor angle for this case. (Maybe the HF-Injection library which has been announced for later this year might help.)
    If ForceAngle has been disabled, however, you get whatever angle FAST provides...

    Bene said:

    So the PWM-Signal at low frequencys with a blocked rotor is the same as with an open rotor?

    No, not really.

    Bene said:

    And one last question: Does the controller know the startposition of the rotor?

    No. It has to move the rotor in order to generate enough feedback for the FAST estimator.

    Perhaps you should read chapter 15 of SPRUHJ1C which is about "Managing full load at Startup, low-speed and speed-reversal" things.

  • 0 in reply to MKo
    TI__Guru** 119340 points

    Thanks for helping to answer and by linking to another thread where I responded. Very helpful!

     

    The only thing I would add is

    - to produce Torque you need to know the rotor flux alignment so you can produce a stator flux with an orientation of Theta which maximizes the expression T = Flux * polepairs * 1.5 * Current * sin(Theta)

    If you don't know the rotor flux angle you have no chance of producing constant torque.

    FAST will not conitnuously track the rotor flux angle if the rotor isn't moving above some speed (motor and inverter dependent). It can track through zero speed, but not at zero speed. You will also note that very small movements allow FAST to re-lock on (assuming ForceAngle is disabled) but if you don't/can't command the rotor to a higher speed the estimations will drift again.

    For some applications the use of an initial position detection algorithm will help. This allows you to get the initial rotor flux angle and start generating your stator flux angle properly aligned for maximum torque.  We are testing and developing these solutions now.  In some cases we will still use ForceAngle for the initial start-up anbut in most cases we don't.

     

    For an advanced solution to your problem, assuming you have been moving at a speed where FAST is tracking well, as you come to zero speed or into a stall condition, bit may be possible to get a known angle value, and if you know you are stopped/stalled, continue to write this angle value (or this angle + a bit of an increment depending on deceleration rate) into the control system.  This would essentially emulate the feedback of a rotor that isn't moving.  I think this is possible in theory, but probably hard to implement across all possible use cases.