I have two different STM32WB55 Nucleo boards in a state where the debugger cannot connect to the target. Is there an errata I'm missing?

[WMcKi.1924]

I am using the ST32CubeIDE and the imported HeartRate project from the STM32CubeWB package. For both boards the project compiles and is loaded into the board, and then the boards just stop working during debugging. I am not probing the boards or touching them in any way. This is the oddest thing I've ever seen. Has anyone else experienced this?

[WMcKi.1924]

I'm using the latest jlink firmware, etc. by the way. Ok, if I hold down the reset button, plug a board into USB, click, "erase chip" in jlink and then let go of reset, the board seems to come back.

However -- the full chip erase always fails. So more questions:

1) Why in the world would your example code redefine the SDIO debug pins (which I assume you're doing having not looked at the code yet)? I've never seen ANY other device manufacturer do that in long experience.

2) Is the fact that I can't fully erase a STM32WB55 Nucleo board a bug in the jlink firmware?

[WMcKi.1924]

Yep, app_entry.c, Init_Debug(). Why in the world isn't CFG_DEBUGGER_SUPPORTED turned on by default? It's just silly to expect people new to this processor to be aware that you're pulling this stunt. Maybe I've missed a write up somewhere. If not, this should be in very large font somewhere very early in the documentation.

[Christophe Arnal]

Hello,

The BLE radio protocol is mostly used in IoT product where power consumption is a key driver, Keeping the debugger enabled would drive significant power in the product.

All BLE applications provided do have their debugger enabled except the BLE_HeartRate example which is configure by default to provide the best power consumption ( Debugger disabled).

Regards.

[WMcKi.1924]

The question about full chip erase failing for the STM32WB55 Nucleo still stands.

[Christophe Arnal]

Hello,

The full chip erase is not supported on that product. This is due to the fact the secure flash area where CPU2 is running shall not be erased except when using the dedicated services.

In order to erase all public flash area, this shall be done sector by sector.

This is documented in the reference manual (chapter 3.3.7 Flash main memory erase sequences)

Regards

[WMcKi.1924]

And, lo and behold, the process for flashing the wireless coprocessor binary, described in UM2550 Rev 2, p23/24 ALSO RELIABLY FAILS TOO. This is pretty lame guys. What gives?

[WMcKi.1924]

I can't emphasize how lame this is. I managed to get my bluetooth stack back bu upgrading the FUS, which also reported that it failed. However 0x20030030 reported a different value, 0x10000200. After that the update finally reported success. I'm trying to evaluate this device for a medical product, but these tools are leaving a lot to be desired.