NUCLEO-H7A3ZI-Q are not working after few hours (no ST-link connection possible)

[JPodh.1]

Hello all,

I would like to ask for your experience with NUCLEO-H7A3ZI-Q board. It is new board which started to be produced. I ordered 3 same boards. They were used at two workbenches independent.

All products have same behavior:

First 2-3 hours are working without any issue (during whole time, no external components connected, just USB port for debugging and power)

There was used just example from st.com supplier ...NUCLEO-H7A3ZI-Q\Examples\UART_TwoBoards_ComDMA, there was used Virtual Serial port....

I would like to note. There were not activated/deactivated any lock bits in processor just flashing example and debugging software + small modification in code for serial port

• There were used just debugging and flashing procedures. Board will stop to communicate by integrated ST-Link interface (after 2-3 hours). There is same behavior if external ST-Link v3 debugger is connected to the board which is not working with integrated ST-Link
• The same behavior was observed at two workbenches independently.

I expect that some of you can have similar experience. Maybe, there is some solution or procedure how to exchange boards for good one from the supplier (www.mouser.de). From my perspective, there is some production issue at least in the same production batch.

thank you in advance

Best regards

Julius

[TDK]

Does the ST-Link chip still communicate?

Could be ESD. Are you grounded while you're working on these?

Did you ever have the board powered from the user USB port and not ST-Link? This can kill the board.

It's easy to jump on the "must be a production issue" train, but there's usually another explanation.

[Tesla DeLorean]

Download the standalone firmware update tool, and apply the most current firmware.

DELL computer?

Make sure you don't have any background tools running, like diagnostic app, AV, or power down for MSC

You'd presumably need to get an RMA from Mouser if you want to return.​

[PMath.4]

It has been discovered that the clock source to the STM32H7A3 chip from the ST-link is based on an RC oscillator in the ST-link chip and not the crystal.

You state that your code is running a USB example. Depending on how you are clocking this it may be that the Nucleo warming up is changing the USB frequency and causing the problem.

You can try removing SB72 and install SB71. Then use an external crystal oscillator on OSC1 and see if the problem goes away.

I was seeing huge jitter on the clock input to the STM32H7A3 chip from the STlink which was making LTDC unusable but have fixed this using an external oscillator.

If the ST-link connection issue seems permanent set Boot1 to run the bootstrap loader and then connect a USB cable from the computer to CN13. Then use STM32CubeProgrammer in USB DFU mode to connect directly to the STM32H7A3. Erase the STM32H7A3 memory completely. Remove the BOOT1 link and restart and magically ST-Link starts working again.

Just don't ask me why!!!!!