Hi
I have a custom gateway using the i.MX28 MCU with both 5V and battery power supply.
The board is running a custom uBoot+Linux solution based on buildroot with our own kernel module to control charging and 5V<->battery transitions.
Also we have a user space application to monitor battery voltage, battery voltage maintenance and shutdown on low battery. The user space application will regularly check the battery voltage from the LRADC and keep it between 3.8 and 4.1V according to battery and application specific specs.
It all seems to works fine but I've received reports of situations where boards with 5V power supply attached where the battery voltage drops like the battery is not being charged.The battery charging works fine again if the gateway is power cycled.
I've validated the state of the relevant registers without finding any clues of what has happened.
I've made a script to dump all the PMU related registers on a gateway with and without the issue - both running the same software.There are very little differences except for the BATT_BO bit being set.
If you can think of anything that could explain the behaviour I will be grateful.
Is there anything that can shut of the charger except the PWD_BATTCHRG bit?
Regards
Kim
Hi
Thanks for the reply. If I understand it correctly the BATT_BO bit indicates that charging has been turned of because of lack of power to the battery charger.
Reading the iMX28 reference manual and AN4199 this behaviour is not mentioned - only that the bit is the output of the brownout comparator (which can then be used to trigger and interrupt).
So what will be the process for recovering from the situation where BATT_BO bit is set? The HW_POWER_CHARGE register seems unchanged in this situation and toggling the PWD_BATTCHRG bit does not seem to restart the charger.
Regards
Kim
one can try with bit PWD_CHARGE_4P2 register HW_POWER_5VCTRL
Best regards
igor
Hi gain
Thanks for the reply!
I'll experiment with toggling PWD_CHARGE_4P2 to recover from the situation although there's a risk of an MCU reset if the battery has already been discharged to a degree where it cannot take over from the 4P2 LinReg.
As it's not documented I still need to have the functionality of BATT_BO clarified.
The bit is set when there's not enough power for the charger to run. But do you think this could be improved by lowering the charging current used or is it a total brownout that would prevent the charger from functioning even with the lowest charging current?
Regards
Kim
yes, I think this could be improved by lowering the charging current
Best regards
igor
Hi Kim
BATT_BO bit indicates that there was brownout condition, so charger was
turned off. This may be due to insufficient current for charger, if for example chip
is powered from 5V and some current also is consumed by LDOs, as depicted on
Figure 1. i.MX28 PMU AN4199 Using the i.MX28 Power Management Unit and Battery Charger
may be useful to check sect.5.2.3 Managing Charger and 4P2 Rail attached AN4199.
Best regards
igor
-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------