stm32g474met3 keeps burning out for unknown reason

[Tom12]

i have MCU circuit with stm32g474met3 as the processor, designed such that it controls the main board charging circuit and it keeps burning randomly at random times could be 5 hours or 20 minuts or even few days I cant figure out why, it burns whether its connected to the main board or not. I tested pin configuration (i.e. inputs and outputs ) and they are all defined correctly. I cant find any issue. and the scheme looks fine. i also removed the 2 0ohms resistors on UART2, and permanently shorted the boot instead of the jumper,
i don't have access to the code of the MCU but how can I electrically find the issue?

[TDK]

VSSA should be shorted to VSS. It appears to be floating in your schematic.

If that's not it:

What does "burn out" mean in particular here? Is something visibly damaged or is it just not doing what you expect it to? What are you expecting it to do?

[Tesla DeLorean]

>>What does "burn out" mean in particular here?

Probably the IC / silicon acting as a fusible link..

Review the NET LIST, and confirm you don't have any oddly or inconsistently named nets or islands that don't connect to others as expected.

Test continutity on a PCB, or unpopulated solder-test item.

Physical damage frequently occurs due to very high potentials and unanticipated current paths, possibly through the device itself.

[Tom12]

I noticed it as well and connect it but it still burns. And by burn i dont see any physical damage however one or more of the pins i dont know which starts to sink current after fault happens. Under normal operation the mcu board consumes around 50-70mA and after the fault it consume over 170mA even after erasing and burning the chip again or just complete erase 

[TDK]

It may not be the only issue, but it's certainly one issue.

The schematic is quite blurry so it's hard to read. But I didn't find anything else. Certainly sounds like a protection diode got blown due to an overvoltage condition. Especially since it was working, then something happened to increase current consumption, yet is still functional. Note that VSS/VSSA should be hard shorted. Some "0 Ohm" resistors are a non-negligible resistance. Doubt that would cause this issue but maybe.

[TDK]

> I noticed it as well and connect it but it still burns.

Did you do this after replacing the chip? Because connecting VSS/VSSA if the chip is already dead won't fix the diode that got damaged.

[Tom12]

yes i replaced the chip it still burned out, also from my understanding VSSA is internality connected so it shouldn't

cause the issue even if its not entirely correct, i might be wrong thought.

[Tom12]

Update: 

i tested the burned chip and check which pins burn the processor and its randomly, meaning i check several chips and each of the chip pins sink the current at different pins unrelated to one another. i assume the damage is due to noise or other reason, if i cannot solve the issue how can i protect the pins from burning again.