Problems with the F103 hardware I2C

[风正豪.1]

A year ago I heard from a friend that there was a problem with the STM32F103's hardware I2C. All the ones I found on the internet said there was no problem with the hardware I2C of the ST, but not the F103 series. So, I have two questions. First, I want to know if the hardware I2C of the F103 series still has problems now. Second, what is the problem with the hardware I2C of the F103 that was once circulated and how to deal with it.

Translated with DeepL

[Foued_KH]

Hello @风 正豪​ ,

If you are facing some erratic behavior on STM32, do not forget to check Errata sheet of the device. There are few points related to I2C interface. 

STM32F101xC/D/E, STM32F103xC/D/E device errata - Errata sheet

Foued

[Tesla DeLorean]

As I recall it was very fussy at initialization, and would latch busy.

Unnecessarily complex implementation. ​

[AScha.3]

btw - i got this "busy forever" also, on H743 .

according to the recommendations of the more experienced i tried writing a very simple I2C in software,

--- until now no more errors on I2C . (although I have a rather problematical timing : driving it with 600..700 kHz clock :) )

[Tesla DeLorean]

On more recent STM32 I've used HAL base code to manage OLED displays continuously for months on end without issues.

[AScha.3]

yes, on H7A3 board with I2C temp.+humudity sensor and an eeprom onboard, about > 500 boards going to customers, most run 24/7 , 2 or 3 boards make curious things on I2C, but only sometimes. (dont know why.) so: usually working correct. :)