NUCLEO-G431 does not communicate on M5 Mac with a direct USB-C port connection

[CTapp.1]

CubeProgrammer 2.22.0 under MacOS 26.4.1 is not able to connect to my NUCLEO-G431, with the log window showing:

ST-LINK error (DEV_CONNECT_ERR)

Strange thing is this only happens when I have a direct cable connection between the USB-C port on the Mac and the ST-LINK V3 USB on the board - it works as expected if I connect via a USB-C hub.

VS Code is also not able to connect to the board to run a debug session without the hub, so I suspect this is an ST-LINK USB driver issue on the Mac.

[Pavel A.]

>  it works as expected if I connect via a USB-C hub.

Is the hub powered?

[Xiaojie]

It also works abnormally on my Mac Studio (M4 Max) with STM32WBA65I-DK1

[CTapp.1]

Whilst it does support PD, it does not need to be connected to external power for things to work (one wire to Mac, on to the board).

[Pavel A.]

These devices work well with many Windows, Linux and Intel Mac hosts. The drivers are very thin - no dedicated kernel modules. So it looks unlikely that ST could "fix" anything in their drivers. Using a hub is a good workaround.

Of course, people that are motivated and have a type-C analyzer can debug and find the real cause of the connection failures.

[CTapp.1]

I'm using the latest MacBook Pro M5, so it will soon be "one of the many" ;)

If there is a potential issue here that can't be fixed, it would be useful if that could be documented somewhere - it took me a couple of hours to find the workaround, and the diagnostic messages that were given didn't give any clues.

CubeProgrammer also appeared to connect to the probe at times (though in a non-functional way - just the serial number and target voltage shown), and even allowed the "firmware update" for the ST-LINK to be activated - though that failed with something like "bad firmware checksum". Sounds like USB packet loss, but that's just a guess.

[Kouthair · Accepted solution]

Hello there @CTapp.1 ,

This behavior appears to match the known issue reported in this FAQ: Possible communication failure between STLINK-V3 and some recent computers. The FAQ might help explain the root cause of this behavior and how to mitigate it, including one of the workarounds which you have already found.

Kind regards,
Kouthair

[Andrew Neil]

---

@CTapp.1 wrote:

it works as expected if I connect via a USB-C hub.

---

Checking with/without a hub is, indeed, one of the recommended steps in dealing with ST-Link comms issues:

https://community.st.com/t5/stm32-mcus/how-to-solve-debugger-connection-issues/ta-p/49693#:~:text=If%20connecting%20via%20a%20hub%2C%20try%20without%20it%3B%20if%20connecting%20without%20a%20hub%2C%20try%20with%20one

[Pavel A.]

Type-C connectors on these thin laptops are too overloaded: they carry USB and Thunderbolt video and smart power functions, maybe also something Apple proprietary. Hubs help to untangle this complexity. Without a good analyzer it's hard to tell more. Better treat this as Somebody Else's problem.