Permanent current drain from VBAT

[jvog.sen]

Hi,

we are using a coin cell to supply the internal RTC of an STM32MP157 when our system is unpowered.

This works as expected - as long as the external supply is off. As soon as the system is powered on, the battery is discharged with >20µA.

Observing the register ADC_CCR shows that VBATEN is permanently set, connecting VBAT to the ADC input and a resistor devider (as stated in Fig. 248 in the reference manual).

[Edit: removed misleading reference to ADC driver]

In my understanding VBATEN is supposed to be set only while (and shortly before) VBAT is actually sampled. Otherwise the switch as seen in Fig. 248 does not make much sense to me.

[PatrickF]

Hi @jvog.sen​ 

please confirm you are using ecosystem v4.x (as you refer to kernel 5.15).

Check your Device Tree wrt to https://wiki.st.com/stm32mpu/wiki/ADC_device_tree_configuration#ADC_properties

I cannot check immediately, but as far as I remember I did not notice this VBAT consumption on STM32MP157F-DK2 with latest Starter Package (which I think did not define the label = "vbat" in device tree).

As far as I know, with latest ecosystem v4.1.0, the VBATEN should be shortly enabled during ADC conversions, so there should be negligible consumption on VBAT.

Maybe you miss some patches (in uBoot and kernel).

Once clarified that you are using latest deliveries, I will double check internally with experts.

Regards.

[jvog.sen]

Hi Patrick,

We are still using kernel 5.10 - sorry, that was misleading.

I will test it with kirkstone as soon as I can and report again if the error persists.

Thanks for your support!

[Kevin HUBER]

Hello @jvog.sen​ ,

Do you still have the problem?

Best Regards,

Kevin

[jvog.sen]

Hi @Kevin HUBER​,

we did not make the switch to kirkstone yet, so I can not tell if it solves the problem.

So for now the problem persists, but has a pretty low priority on our side.

Regards

Jan