STM32H563 Revision Z Creates a HardFault in the UCPD Library.

[mathieuM]

With revision Z of the STM32H563VIT6, which looks to be an older revision of the MCU, we get a HardFault issue in the UCPD library. We noticed the issue with revision Z, but revision X is working as expected (see errata in attachments)!

 

When connecting a device through USB-C, the MCU goes in a HardFault when setting Rp resistors to SinkTxNG.

In the STM UCPD library, I see that there is some added code to manage older revision of the component, and the line with the red arrow creates the HardFault.

It seems that the MCU is not able to access "pUCPD_TRIM_1A5_CC1" value in non-volatile memory at address 0x08FFF844, even though there is a value at this address if I manually look for it in the flash.

What could create this HardFault?

Is there something we can do to fix this HardFault or should we discard all revisions other than the X?

[STOne-32]

Dear @mathieuM ,

Thank you for the question, I do not have a board now to check, but Indeed I confirm that RevX is the latest revision in production and fixes few major limitations - you can see in the first table between A, Z and X.   May be you hit this Knowledge Article case as that @Address is impacted by RO data if you can check and let me know : How to avoid a hard fault when ICACHE is enabled o... - STMicroelectronics Community

Hope it helps You.

STOne-32

[mathieuM]

Hi,

I tested what is recommended in the link you sent, and it fixed the issue.

The HardFault was indead created because of the ICACHE and access to RO memory areas.

Thank you so much for your quick response!

Best,

MM