• State
  • Replies 2 replies
  • Subscribers 82 subscribers
  • Views 59 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TMS320F280037C: PMSM torque ripple vibration compensation

Genatco
Guru 56398 points

Part Number: TMS320F280037C

Hello,

Firmware: UMCSDK v5.03 20kHz modulation

Busvoltage: +151v DC under full speed load

  Is it possible to use vibration compenstaion module, VIB_COMP_A to reduce noise and torque ripple vibration for PMSM under FAST estimator control? If so is there an example universal motor control module with setup parameters for automatic mode in v5.03? Does vib comp require hardware changes to monitor FAST sensor inputs like TIDI-02010 dual motor fan and compressor control? 

Already reduced commutation rough noise setting; USER_M1_EST_FLUX_HF_SF  0.125f  but rotor vibration 365hz -> 495Hz range is more than trapezoidal FOC produces. Obviously less efficient FOC yet less rotor vibration and bearing noise results.

Should VIB_COMP_A not be possible Is there any other user_motor1.h define that might further reduce noise and similar torque compensation other than feild weakening? What is being perceived due to very rapid harmonic vibration, an odd noise coming from the radial ball bearing cage.

  • 0
    TI__Expert 6170 points

    Hi,

    Yes, VIB_COMP_A is designed to work directly with the FAST estimator and it doesn't need new sensor hardware. You can refer to the TIDA-02010, which implements automatic field weakening + torque compensation for the compressor PMSM under FAST estimator control, and documents vibration-compensation algorithms.

    Thanks,

    Jiaxin

  • 0 in reply to Jiaxin Teng
    Guru 56398 points
    Jiaxin Teng said:
    which implements automatic field weakening + torque compensation for the compressor PMSM under FAST estimator contro

    H Jiaxin,

    Does it require FW to work properly? This odd rotor noise sounds like rock tumbler or random screeching in the stator smooths out at higher rotor speeds.

    Seems as if negative Id is being injected Kp/Ki_speed loop because I can see the torque value randomly change ± sign. Not in CCS debug but in the user HMI widget reporting the torque output.

    Seemingly for AC phase current motor should be running much quieter 1600 RPM with no load other than the rotor mass. It is quieter than older Trap FOC but not by much and the torque ripple is about equal in both commutation modes. Though FAST estimator wins hands down for efficiency!