Stlink-V3Set programmer burned and MCU H7 destroyed

[PAdam.11]

Hi, I was using brand new Stlink V3 Set to program custom device with H7 onboard for 4-5 months. The programmer was connected with 3 lines: GND, SWCLK, SWDIO, and was working fine.

Today, after weekend, I wanted to test new code. I had programmer connected as always, not ejected since last work day.  USB Device was detected fine by windows 10. Programmer was connected as always to MCU with goldpin connector.

I turned on mains power supply delivering 12V to the board. I used to use lab supply, burt after testing hardware I switched to target supply.

As I toggled switch HDMI monitor connected to laptop turned off, something cracked inside programmer. Some parts of the device started workign (gpio, I2C, LEDs) but LCD not started. MCU started getting really hot.

It turned out that something burned in the programmer. I havent found schematic for it :\ And MCU is broken, It is getting 100*C hot after several seconds - see thermo photos.

Here programmer:

Here MCU, programmer connector marked as SWD. Coil on the left side is part of step down regulator - the regulator IC is visible in thermo fotos when broken MCU draws too much current.

Here 115*C on broken MCU - 12V, 5V and 3.3V power sources are fine.

Here other working board programmed afterwards with Nucleo programmer.

What could couse this fault? If it was USB overcurrent windows will poup message about it. There is one direct way to mains power - the device is controlling chamber heater with triac and opto isolation. But if it was 230V leak then the damage should be more severe to power ICs.

  

[hashem]

I have experienced something similar with the STLINK-V3SET and the STM32F302RE, the STLINK still works, but the MCU is broken. It is still possible to flash the MCU and flashing succeeds, however the software that used to work doesn't work anymore. I used the STLINK-V2 for many months, however only seeing these issues with the V3. This happens intermittently, because I have flashed with the V3 successfully many times without burning the MCU, but then suddenly a MCU gets burned.

[Tesla DeLorean]

Not sure if this is the ST-LINK/V3's fault or that it just became a conduction path.

In different systems / powering, I'd probably lean to a ground bonding issue, and very large potentials. Boards on a conductive surface

Now something is vaporized in the IC, and it's shorted across the supplies now.