cancel
Showing results for 
Search instead for 
Did you mean: 

Flash write give different results when MCU is programed with ST-Link or PE-Micro Cyclone.

rcard
Associate II

I wrote a lib to handle different routine that read/write the flash. Everything works fine.

The linker file is customized with some page declarations later use to affect some variables to some flash area.

I have done all the debug using the embedded STLink of the NUCLEO-G071RB. No error when writing the flash, everything is writen where it should be, so great no problems.

Then i decide to use a PE-Micro Cyclone, because i find it's working better than the ST-Link. The cyclone never lose the connection while debugging, and it's also faster and more responsive.

I used the Cyclone a lot with the STM32F030, and it worked really well, never had any trouble.

But now i have a really strange issue. When i use the cyclone to program the MCU, everything works fine, except that when i write something in the flash, it trigger a PROGERR.

Nothing changed in the code between the STLink and the Cyclone. The disassembly show the same instructions.

The adress writen are the same, the data writen are the same, and yes, the address where i want to write is in erase state (0xFFFFFFFFFFFFFFFF) so there should be no PROGERR.

Any idea why i have different results like that between STLink and Cyclone?

2 REPLIES 2
SMcGr
Associate II

Hello cardetti,

Could you elaborate on the exact error you are receiving?

Since this post, PEmicro has updated the STM32G071RB algorithm and the problem you are having may have been resolved with the release of the new Cyclone firmware incorporating these new algorithms. You can download the new Cyclone installation software on the PEmicro website by navigating to the Support Page -> Documentation & Downloads -> Software Update -> “Cyclone LC and Cyclone FX Programmer Installation Software�?.

Hope this helps!

Steve

rcard
Associate II

I contacted PEMicro support after i realized that asking for PEMicro related issue directly to PEMicro was probably a better idea than asking here.

They sent me the new algorithms and it solved the issue.

Thanks for your answer to my question.

Have a good day.

Romain