KevinOConnor

Interesting. Unless I'm missing something though, this change may break other users of hard_pwm. The current hard_pwm interface assumes that "cycle_time" is approximate while pwm duty times are accurate. This...

Thanks. It seems fine to me. The only thing I'm not sure on is the name "gpio_clock_output" - I know I suggested that name, but I'm unsure of my suggestion....

Another possibility might be "pwm_clock" (thus being similar to "pwm_tool" and "pwm_cycle_time"). Cheers, -Kevin

Thanks. In general it seems fine. Why does `PrinterClockOutputPin()` create a `ppins.setup_pin('pwm', config.get('pin'))` ? Wouldn't that cause the pin to be initialized twice - once as a pwm pin and...

Ah, yeah, I guess 'invert' doesn't make much sense if it's 50/50 duty cycle. Not a big deal either way. Probably best to remove the now unused `mcu_pwm.get_pin()` method. Otherwise,...

Thanks. I committed this along with a minor change in 7377da63 on top. -Kevin

I'm a little confused - the errata is described at https://www.st.com/content/ccc/resource/technical/document/errata_sheet/a9/f2/ac/eb/22/6b/4f/70/CD00260217.pdf/files/CD00260217.pdf/jcr:content/translations/en.CD00260217.pdf - section 2.8.7 "I2C analog filter may provide wrong value, locking BUSY flag and preventing master mode entry". The...

> Which from my reading of the errata ("This issue occurs randomly.") means that the errata condition can be observed at the start of any communication, not just at startup....

Okay, thanks for the information. However, I suspect there is something else going on beyond the errata. There are a lot of stm32f103 chips running Klipper and I can't see...

Oh, another useful test would be to see if you can reproduce the issue running the pristine Klipper code while using software_i2c. If it can be reproduced using that, then...