Possible reasons for STM32F4 to generate a reset.

[Maor Avni]

Posted on February 22, 2017 at 12:26

We have a strange problem on one of our boards. One cabinet has 3 identical boards, running the same binary, (the application is configured remotely from the main server so the software behavior is different), but only one of the boards has intermittent resets.

I tried disabling the watchdog, but the resets still occur.

While trying to debug the problem by attaching to the board with a JTAG (Jlink Pro), the resets frequency rises dramatically, but the JTAG loses connection to the board, as though the STM32 lost power, so I can't witness the place where the software resets.

We connected a scope to see the reset signal. We deducted that the reset signal source is internal to the STM32, and not external.

The only place where a forced reset is used by the software is at the end of bootloading, and it is performed with NVIC_SystemReset(). 

We use STM32F427ZG/I.

Any other reasons that the STM32F4 will reset itself, other than NVIC_SystemReset and a power outage?

[Tesla DeLorean]

Posted on February 23, 2017 at 01:15

The STM32 has a reason register for the Reset, start by reviewing that.

Internally a power-on-reset circuit might fire if the supply browns out. It is connected to VDDA, it is going to be sensitive to glitches, so make sure your regulator circuit has adequate bulk capacitance, and there is sufficient bulk capacitance placed close to the STM32. Check the voltage and capacitors on the VCAP pins. Make sure values placed on board are as specified in the schematic/design, and meet recommendations in the data sheet. Sorry, no salient details on board design or powering to work with here.