Skip to main content
JHSU.16
JHSU.16
Associate
December 23, 2021
Question

Weired output then OC, and controller damaged

  • December 23, 2021
  • 2 replies
  • 1634 views

Hi all,

​

I am using f303rb with ihm08-1sh to control a motor, motor control is MCSDK 5.3.2

Using HSI which has 64MHz clock speed.

I set the PWM frequency to 15K, and FOC execution rate = 2.

In most time, it works okay.

In a very low probability, it will OC when motor starts, and controller (MOS/driver IC) damaged.

Look at the waveform, the PWM duty looks very weired. It's not the same as normal situation.

The PWM duty looks full for couple of cycles and motor will rotate wildly.

I don't know if it's related to 1-shunt or any other else.

Does anyone have similar problems?

PS.

We can reproduce the problem with a run-4sec/stop-1sec without loading test in an hour.

Thanks,

Jungchu

0693W00000HpcxzQAB.bmp

This topic has been closed for replies.

2 replies

MK.1
MK.1
Associate III
December 25, 2021

Hi, Try to generate a project using the LL library. Describe what's on the oscillograms.

JHSU.16
JHSU.16Author
Associate
December 27, 2021

Hi MK.1,

​

Thanks for your reply.

Is that means the performance of F303 may not be enough?

​

I also tried in different PWM frequency and FOC execution rate, and I can only pass 1 hour test in 10K PWM and 10K FOC execution freq.

I cannot understand why 7.5K execution freq. fails and 10K passed.

​

I also checked the release notes, and I found ST fixed a bug for G0 series in 5.4.0.

--

Wrong PWM duty cycles generation for 1-Shunt configuration on STM32G0 devices. The issue was mostly visible with low PWM frequencies or with a Repetition Counter strictly higher than 1.

--​

I don't know if it's related, but the scenario is just like ours.

​

By the way, I will generate LL project first.

​

Thank you very much.

​

​

PS.

The oscillograms 1./2./3. shows voltage which is output A/B/C, and 4. is the current of output A.

​

​

Best Regards,

Jungchu

Terms & ConditionsAccessibility statement