Nucleo ST-Link V3 holds board in reset when connected

[musicmindmachine]

Been working on the Nucleo H5 dev board for about a month.  Everything with the built-in ST-Link/V3 was fine until randomly yesterday, the board would no longer get out of reset when the ST-Link is powered.

Board powers up and boots if getting power from the USB-OTG port AND the ST-Link is unplugged.  Tried with STProgrammer and GDB in VSCode debugger, neither can connect to the target chip.  To be clear, the STLink enumerates and works OK, but the MCU is held in reset state.

Board is held in reset state when ST-Link gets power/host connection.  Attempting to flash the board always gives "Reason: Target held under reset.".

All solder jumpers and physical jumpers are in the default positions

Fixes I've tried:

• Any set of settings via the STProgrammer, including Soft Reset, the "under reset" connection type, etc.  Any time STLink is powered from host, the board is held reset and will not connect still.
• Updating the STLink/V3 drivers/firmware multiple times.
• Connecting the STLink in hot--plug mode, which still holds board in reset state when powered.
• Shorting, holding the STLink reset jumpers during power up, etc.

Issue randomly started after seemingly no specific cause.

[Imen.D]

Hello @musicmindmachine ,

Check your power VDDA​ and don't pull NRST low.

Have a look at this FAQ: Possible communication failure between STLINK-V3 and some recent computers.

[musicmindmachine]

Thanks for the reply.

I have no issue enumerating and connecting with the STLink V3 from the computer.

When the STLink is powered, connected, etc... it holds the MCU in a reset state.  I can see in GDB/Programmer that it attempts to reconnect several times at different clock frequencies, but cannot connect.

I have been able to connect and program the MCU many times before, until suddenly it began holding the MCU in a reset state.

If I power the board via the USB FS Type-C port on the MCU and do not connect a host to the STLink, then the MCU boots and runs code properly.

If hold the STLink in Reset, it also holds the board in reset. 

However, through my attempts to find the fault, if I removed the crystal oscillator clocking the STLink, then the board can run even when the STLink is powered via the host.

This seems to be an issue between the STLink and the MCU, not with the Host to STLink over USB.

[musicmindmachine · Accepted solution]

Ok here's an update...

Out of desperation, I tried shorting some capacitors/resistors to see if there was a voltage or something that was causing the reset issue.

Shorting out R15, R14, R91 magically forced the STLink to go into a reset state.  This somehow kept the STLink Powered and let the MCU boot up properly.  I was then able to connect in hot-plug mode and gracefully disconnect.

This seemed to resolve the issue and I can now re-flash and debug the MCU via the ST Link.