fix(pa01): Boot failure at low temperature.

[sneone]

At low temperatures, the LDO voltage for core power supply stabilizes more than tens of microseconds slower, so there is a 48-millisecond delay before the initialization clock.

[raphaelcoeffic]

@sneone which LDO is too slow? Is that the internal one? Do you have any data / errata sheet on that? Which kind of "low" temperatures are we talking about?

Also, 48ms seem like a huge time frame! It's more than the complete boot at the moment. Isn't it a bit too much?

[richardclli]

@raphaelcoeffic PA01 uses external LDO for powering H7 MCU, at extreme cases in low temperature environment, it cannot bootup properly because the external LDO need some time to become stable.

[richardclli]

I am thinking, besides adding a delay in the beginning. Will it be better to use MCU internal LDO to boot up and switch to external LDO in a later stage? This will not slow down the boot up and solved the external LDO stability problem. @raphaelcoeffic What do you think? @sneone Maybe you can give it a try.

[richardclli]

@sneone You better make a 2.11 version of this PR for 2.11 branch fixing

[sneone]

I am thinking, besides adding a delay in the beginning. Will it be better to use MCU internal LDO to boot up and switch to external LDO in a later stage? This will not slow down the boot up and solved the external LDO stability problem. @raphaelcoeffic What do you think? @sneone Maybe you can give it a try.

We consulted STMicroelectronics’ FAE regarding this, and they do not recommend switching between internal and external power supplies after clock initialization is complete.