Help to find why NUCLEO boards are getting burnt
First of all the board that we use is "NUCLEO-L552ZE". The upper section of the board is the ST-Link which we use to program and debug the chip. The ST-Link also forwards one of the UARTs (LPUART1) over USB. So we conveniently use this feature to test the software.
However on the final product there is no ST-Link available to us (it's custom designed board). We're going to use ST programmer v3 (which is a ST-Link) for the same purpose as above.
For simulating what we have on final board I changed the JP6 on the Nucleo board and set it to 5V_EXT. I connected the 5V (current limited to 500 mA on power supply and SB1 is disconnected) to the 6th pin of CN11 and GND to 8th pin of CN11. I couldn't send and receive anything to UART (over USB). So then I decided to connect the USB cable directly to the ST-Link on the Nucleo board to check the UART on that. Although I hadn't changed anything on the board (specially JP6 jumper and the 5V_EXT pin) the board appeared to get burnt. According to the schematic the current shouldn't damage anything (unless I overlook something).
Here are two thermal pictures of the NUCLEO-L552ZE board:
with ST-Link:
with 5V_EXT:
After this experiment I tried to do the same with another board. This time the board was "NUCLEO-F401RE". However the board got burnt again right after I connected the rx and tx cable of ST programmer v3 to UART.
Here is a thermal picture of NUCLEO-F401RE:
Has anybody an idea of what's causing damage to the boards?