cancel
Showing results for 
Search instead for 
Did you mean: 

[EVALSTDRIVE101] Hall Signal Noise Issue Causing Commutation and Speed Loop Instability

youhao
Associate

 

[PN]: EVALSTDRIVE101, NUCLEO-G474RE
[VERSION]: MCSDK 6.3.2
[TOOL]: Motor Control Workbench, Motor Pilot

[DETAILS] :


We are using the EVALSTDRIVE101 to drive a 250W, 1200RPM motor with Hall sensors. The Hall signals (H1, H2, H3) are filtered through the onboard RC low-pass filter before reaching the MCU on the NUCLEO-G474RE board.The control method uses 16kHz PWM to drive the UVW MOSFETs on the EVALSTDRIVE101.

When measuring the signals with an oscilloscope:

  • H1 shows minor jitter at high level.

  • H2 and H3 show significant jitter around 16Hz when high.

This noise seems to interfere with the motor commutation, resulting in audible "clicking" or "jerky" sounds and large speed fluctuations in the speed control loop during high-speed operation.

Interestingly, when we bypass the onboard low-pass filter on the EVALSTDRIVE101 and use an external RC filter instead, the noise is significantly reduced.

 

youhao_2-1751508892481.png

um2781-getting-started-with-the-evalstdrive101-stmicroelectronics.pdf   圖 4. EVALSTDRIVE101 –Hall sensors/BEMF 

 

oscilloscope motor h1 green,motor ch2 yellow 

 

youhao_14-1751509968322.png

 

16k noise

scope_11.png

 

oscilloscope mcu h1 green PA15, mcu h2 yellow  PB3 RC low-pass filtered signal

youhao_1-1751508776430.png

youhao_3-1751509060898.png

youhao_4-1751509065310.png

external RC low-pass filter 

youhao_9-1751509536546.png

 

oscilloscope motor h1 yellow ,motor ch2 green

youhao_10-1751509639672.png

youhao_11-1751509659835.png

oscilloscope mcu h1 yellow  PA15, mcu h2 green  yellow  PB3 RC low-pass filtered signa

 

youhao_15-1751510117704.png

 

youhao_16-1751510164025.png

To rule out any issue with the motor itself, we tested the same motor using a different commercial motor driver.
Oscilloscope measurements of the motor-side Hall signals (H1, H2, H3) showed no 16kHz noise.
Therefore, we preliminarily ruled out a fault in the motor's internal Hall sensors and suspect the interference originates from the EVALSTDRIVE101 board.

[EXPECTED BEHAVIOR] :


We would like to understand how to reduce Hall signal noise when using the onboard low-pass filter on the EVALSTDRIVE101, in order to achieve more stable commutation and speed loop behavior, especially at higher speeds.

[HOW TO REPRODUCE] :

 

  1. Connect the motor with Hall sensors to the EVALSTDRIVE101.

  2. Use an oscilloscope to probe the Hall signals:

    • CH1: Measure H1 and GND on the EVALSTDRIVE101 board (green trace)

    • CH2: Measure H2 and GND on the EVALSTDRIVE101 board (yellow trace)

  3. Additionally, measure the filtered signals at the MCU input pins:

    • H1 (PA15) and H2 (PB3) on the NUCLEO-G474RE

  4. For external RC low-pass filter testing:

    • Remove jumpers J10, J11, J12 on the EVALSTDRIVE101

    • Connect external RC low-pass filters between Hall signal lines and MCU pins:

      • H1 → RC filter → PA15, H2 → RC filter → PB3

    • Measure the filtered signals again at PA15 and PB3

0 REPLIES 0