So far 2 boards have died and I am uncertain how they failed, is there a chance they could be fixed through software?

[AZeli.1]

I was using a Nucleo F746zg to develop firmware and in the middle of testing after consistently functioning for over an hour, it stopped communicating with the st link and started showing an overcurrent error. After further testing, the voltage regulator now outputs only 1 volt when it should be 3.3v and I got a "The interface firmware FAILED to reset/halt the target MCU" when i was able to access the board.

After this I bought a Nucleo F401RE and had the same exact thing happen, the board worked for many hours and then suddenly in the middle of outputting to stlink, suddenly stopped outputting data and I got a overcurrent error and a "The interface firmware FAILED to reset/halt the target MCU" error.

I heard that a way to fix this was to use stm32programmer to do a full flash wipe on the chip but I was unable to do that for either board since every time I tried to connect to the board it crashes the software or makes it go unresponsive.

I hope I can get these boards to work soon.

[Karl Yamashita]

Do you have anything connected to the development board?

[AZeli.1]

When they died I was communicating with another component via i2c and was using 2 adc channels, but while I have been trying to fix the boards I have had nothing connected to them.

[MM..1]

Primary seems more as your USB port failure.

Secondary what is connected to I2C and ADC and how is powered ?

Work you on ESD controlled environment?

[AZeli.1]

I was connected to a bms dev board, which was externally powered, I am not working in a very esd controlled environment but I made sure to isolate the higher powered components, and the power supply I used was limited to only 100ma for the bms dev board and it never went above 20ma during regular use.

[MM..1]

MCU can be killed with overvoltage. Current isnt here relevant.

Too you can kill STLink STM and aplication MCU is OK, but powering stop work.

Designs with more as one power source is critical ...

[Karl Yamashita]

I assume you were using jumper wires to connect the I2C device to the Development board? Could it be possible the I2C device draws more current than the Development board can handle? I am assuming you're using the built-in STLINK? I've never seen and over current error ever happen

[AZeli.1]

the i2c device was powered externally with a seperate powersupply, all of the cables that were connected were purely for data, having current draws in the micro amps.

[Karl Yamashita]

You say the I2C device was powered by a seperate PS, but do you have a ground reference from the I2C device to the development board?

[Tesla DeLorean]

There are some issues where you can get H7 boards into odd states, this doesn't sound like that.

Sounds like you've damaged them electrically / physically

There is no protection on the ADC inputs and they don't expect to see voltages above AVDD, or negative voltages.

You need to make sure the grounds are well bonded.

Having various supplies and poor connections can result in significant differences in potential.

[AZeli.1]

The maximum voltage the adc should have seen was around 3.06 volts which is the maximum the bms is supposed to output, but I had tested the adc on a bench previously up to 6 volts without any issues, thought that probably was not the best of ideas. The grounds were secure, but the negative voltage might be a cause. One of the adcs was getting really low voltages which might have sometimes gone negative, would that be bad enough to completely short out the mcu though?

[Tesla DeLorean]

You can create current conduction paths, and if excessive or prolonged will cause the silicon to act as a fuse. Could fail open, or closed, with the latter something is likely to get hot

Excessive currents on the regulators can cause them to overheat and fail, you'd expect multiple parts to get hot.

In situations where you think your code is configuring pins that might conflict with other circuitry or voltages on the board, you can pull BOOT0 to VDD (on the NUCLEO-144 boards you can jumper this as the pins are next to each other on the headers), so that user code doesn't run and you can erase the part.

[AZeli.1]

So I just tried externally powering the F746ZG and giving it more than 500ma and it now works again, it just requires a really large current when starting, likely due to the broken voltage regulator.

it now uses around 300-400ma consistently when powered with 5v externally, and I still am unable to connected it to stm32cubeprogrammer to wipe the flash on it, which is worrying, whenever I hit connect to debugger, the program goes unresponsive.

[Karl Yamashita]

I've had an test engineer testing a board that I have given him. He took it to his desk and could not program the MCU. So I took the board back and was able to program it with no issues. Then we used his power supply and it failed. Turns out his current output was turned way down, lol

[MM..1]

Check thermal state on board components, somethink will be hot... and ofcourse disconnect all external

[AZeli.1]

I have been seeing if anything has been getting hot, the mcu was getting warmer than I remember it getting before when i powered it externally but not sure if this all that bad.