Skip to main content
Associate
July 18, 2025
Question

STM32F405RGT6 TIM Output Signal Becomes Unstable at 60°C While Room Temperature is Normal

  • July 18, 2025
  • 2 replies
  • 390 views

 

Tests Performed and Results:
1.A DC regulated power supply was used to provide 14V to the board directly. At room temperature, the CPU's TIM pin output is normal. However, at high temperatures, the output waveform becomes intermittent.

2.The on-board 3.3V and 5V power supplies remain stable and within specification at all times.

3.The external crystal oscillator connected to the microcontroller maintains a stable waveform under all temperature conditions.

4.The same issue occurs even after replacing the CPU with one from a previous production batch: the TIM pin output is normal at room temperature but becomes intermittent at high temperatures.

5.This TIM pin is used to simulate a frequency signal. When the output becomes intermittent, the total output duration becomes longer. The firmware uses a pulse count to determine when to stop the output, so it appears the missing pulses (due to missed interrupts) cause the output to last longer than expected.

Questions:
Given the above results, what should I do next?

Should I contact ST technical support, and if so, how can I submit a support ticket correctly?

2 replies

TDK
Super User
July 18, 2025

What is the waveform? Can you show the intermittent output on a scope plot?

"If you feel a post has answered your question, please click ""Accept as Solution""."
Associate
July 18, 2025

Thank you for your reply. The normal and abnormal waveforms are shown in the attached video.

Associate
July 18, 2025

The waveforms were captured from the TIM1 pin using an oscilloscope.

TDK
Super User
July 21, 2025

It may be hard to diagnose further without knowing what the firmware is doing. Looks like it's not a standard PWM output but rather something dynamic. Even the periods between pulse trains seems variable in your video. The firmware is certainly doing something here.

TDK_0-1753104151678.png

 

I would guess that if you programmed a chip to output a basic PWM signal, it would work fine at 60 C.

 

ST support request can be made here, but I don't think they're going to be able to assist without more information.

https://ols.st.com/s/

 

"If you feel a post has answered your question, please click ""Accept as Solution""."