Hi ST Team,
We have an issue concerning a PWM generation.
We have a MCU : STM32F303 driving directly the LM25101 driver from Texas instrument with no limiting resistor or other protection on it.
The PWM drives two mosfet legs with 2 mosfet configured as high side and 2 mosfet as low side. (Buck and boost legs).
The TIM 1 inside the uC drive both legs through PA8, PA9 PA11 and PA12 thanks to TIM1_CH1 and TIM_CH2.
This is working great most of the time.
However, it happens that the stm32F303 doesnt deliver any PWM any longer on TIM1_CH2 but deliver clear and correct signals to TIM_CH1.
We have tried to toogle the pin of TIM_CH2 "manually" through HAL command and they are working great meaning the output capability is not the cause of this non functionning.
The stress case (where this does happen) seems to be some power load start up . Sometimes, we need several tries to brake it (over 500 happens)
We have analysed signals, tried to put more protection diode and so on to prevent both overvoltages and overcurrents and the GPIO pins, however, the stm32 seems to have the same default on several boards (same issue). We have also tried to measure those stresses with our lab but we never managed to catch something
We are currently beleving something is breaking the TIM_CH2 only. The rest os the function seems to be fine (ADC measurements are fine, TIM1_CH1 are fine, led driving are ok, UART communication works and so on....)
Would you have any clue why this specific part of the driver could burn ?
Thnaks for your advice.
Edgar
