Humming Noise at High RPM Using Sensorless FOC

[Mani2]

Hi Everyone,

We are using an STM32F401RE Nucleo board with the IHM16M1 power board, running a sensorless FOC algorithm generated using ST Motor Control Workbench. The motor parameters were averaged over several runs using ST Motor Profiler, and the control is implemented in 3-shunt topology.

Observed Issues:

1. Humming noise becomes audible and increases as RPM increases.

   • This is more prominent at higher speeds.

   • At low RPM, noise is tolerable but increases as speed ramps up. 

     Speed graph

     Modifications done in Motor Control Workbench to mitigate issues:

     • Increased Kp divider to reduce aggressive control at low speeds (reduces noise at low RPM but not high RPM).

     • Reduced Ki to prevent instability.

     • Reduced Observer G2 gain to prevent speed feedback faults in Motor Pilot.

2. PWM Mode Testing:

   • Switching from center-aligned to discontinuous PWM seems to delay the onset of the humming noise slightly, but noise still eventually appears at high RPM.

Is there a recommended strategy to suppress audible noise in the high-speed region for sensorless FOC?

       Any suggestions or tuning tips from others who used the IHM16M1 + F401RE + sensorless FOC combination would be highly appreciated.

Best Regards,

Manikanta

[GMA]

Hello @Mani2,

Have you verified whether the PWM frequency affects the occurrence of this noise?

[Mani2]

Hello @GMA,
We checked with 20 kHz, 30 kHz, and 50 kHz, but the same humming noise was observed.

Best regards,
Manikanta

[LCE]

Have you tried another motor ?

Any gears involved ?

Measure the level and the frequency of the noise to find out how that is proportional to RPM.

Note: it might be that the noise only seems louder at higher RPM because human hearing is very insensitive at low frequencies.

[Mani2]

Hello @LCE,

I haven't tried another motor yet. There are no gears involved, it's a direct-drive fan. I can't measure the sound level or frequency since I don't have the proper equipment; I'm just using my ears. The noise is only noticeable between certain range of RPM, and it seems to disappear above or below that range.

[LCE]

Okay, it could be a motor issue still.

I'd check the PWM signals with a scope with FFT capability, or connect the probes somehow to some PC audio input (careful!) to get an even better FFT with some PC audio tools (eg audacity).

[Ozone]

> Humming noise becomes audible and increases as RPM increases.

I suspect this is the mechanical resonance of your motor. The closer you get, the more energy is tranferred into vibrations.

You could try to move the resonance point by mounting the motor to a solid (and heavy) frame, or dampen it (mount with rubber buffers).

Or try another motor.

[Mani2]

Hello @Ozone,

I don't think it's a mounting issue because I tried it with another controller, and I didn't observe any audible noise at that specific RPM range. That's the only reason I suspect something needs to be addressed in the code.

[Ozone]

> I don't think it's a mounting issue because I tried it with another controller, ...

And are we talking about the same PWM frequencies here ?

If yes, try to compare the waveforms.
"Humming noise" is mechanical oscillation. Try measuring alongside with a vibration gauge.

[LCE]

In case you don't even have a scope, try your smartphone with an app that can display an FFT.