STM32 programmer failed PCB not working afterwards

[mt1]

I'm trying to flash my custom STM32L152RET6 board using the command-line version of CubeProgrammer. Initially, everything seemed fine, but I wanted to ensure the firmware couldn't be read out. So, I set the option bytes with -ob RDP=1 to enable read protection. However, after doing this, the flashing process failed, and now my program no longer runs.

I was able to flash it again using a Segger debugger, but when I try to run the program, it crashes at startup. The voltages all seem normal, so I don't suspect a hardware issue. Could this be due to some incorrectly set registers or option bytes in the CPU?

Here’s the command I used:

STM32_Programmer_CLI.exe -c port=usb1 -ob RDP=1 -d firmware.hex -v

Any advice would be greatly appreciated!

[Tesla DeLorean]

Your app running on the STM32 should be able to print out content/state of FLASH and/or OB

Check it it write protects the base sectors.

Usually you can clear these with a Mass Erase, or walked in via the UART connection to the System Boot Loader.

As you say Segger also has an un-brick app.

Can you use ST-LINK via SWDIO/SWCLK ? To connect/inspect?

[liaifat85]

Try reverting the RDP level back to RDP=0 (no read protection) and see if your program starts running properly. You can do this by issuing the following command:

STM32_Programmer_CLI.exe -c port=usb1 -ob RDP=0

Look for any unexpected settings, particularly those related to:

• Boot configuration (nBOOT0, nBOOT1)
• Write protection (WRP)
• Watchdog settings (IWDG_SW, WWDG_SW)
• Flash memory configuration
• If you want to redesign the board, you can see this for reference:

https://www.pcbway.com/project/shareproject/Getting_Started_with_STM32_Arduino_code_Programming_and_PCB_Design_b4b9891e.html

[mt1]

Yes I can just debug no problem. But when it inits the HAL driver it gives an exception and crashes

[mt1]

That is a good one I will try indeed the -ob RDP=0  option !

But what I do not understand is that the Segger SWD can just write new flash code. So its not write protected anymore. Or the Segger ignores the protection bits ?

[mt1 · Accepted solution]

I fixed the broken PCB, when I looked in the Cube programmer GUI the option byte  nBFB2 was not set so it would boot from bank2. When enable this option and also did a full chip erase everything was working just fine !

Why I could still flash it with the segger debugger (in bank 1 ?) and the program would still go to the main , but then  only crashed when it wanted to init the HAL is still a mystery to me.