i.MX28EVK charger "breaks" after a few hours

[ReinhardMeyer]

I connected a Li-Pol battery to the EVK which is mainly working as one would expect: charging, running from battery or external power etc. Only one problem repeatedly persists:

After a few hours (can be a day or more) the charger is reported broken.

Restarting the system resumes the Charging/Not Charging. One can catch the system either charging or not charging, however the charging is quite often.

root@freescale ~$ uname -a
Linux freescale 2.6.31-855-ge067785 #1 Fri Sep 10 09:22:06 MDT 2010 armv5tejl GNU/Linux

root@freescale ~$ cat /sys/class/power_supply/battery/uevent
POWER_SUPPLY_NAME=battery
POWER_SUPPLY_TYPE=Battery
POWER_SUPPLY_STATUS=Charging
POWER_SUPPLY_PRESENT=1
POWER_SUPPLY_HEALTH=Good
POWER_SUPPLY_TECHNOLOGY=Li-ion
POWER_SUPPLY_VOLTAGE_NOW=3920000
POWER_SUPPLY_CURRENT_NOW=600000
POWER_SUPPLY_TEMP=58
root@freescale ~$ charger------ ddi_bc_gBrokenReason=1

root@freescale ~$ cat /sys/class/power_supply/battery/uevent
POWER_SUPPLY_NAME=battery
POWER_SUPPLY_TYPE=Battery
POWER_SUPPLY_STATUS=Not charging
POWER_SUPPLY_PRESENT=1
POWER_SUPPLY_HEALTH=Dead
POWER_SUPPLY_TECHNOLOGY=Li-ion
POWER_SUPPLY_VOLTAGE_NOW=3744000
POWER_SUPPLY_CURRENT_NOW=0
POWER_SUPPLY_TEMP=42
root@freescale ~$

Can anyone reproduce this behaviour, or has seen it?

Reinhard

[ReinhardMeyer · Accepted solution]

I want to add, that this problem still persists, even with the latest kernel:

Linux version 2.6.35.3-670-g914558e (reinhard@ubuntu) (gcc version 4.4.4 (4.4.4_09.06.2010) ) #1 PREEMPT Mon Mar 25 08:23:40 PDT 2013

root@freescale ~$ cat /sys/class/power_supply/battery/uevent

POWER_SUPPLY_NAME=battery

POWER_SUPPLY_STATUS=Charging

POWER_SUPPLY_PRESENT=1

POWER_SUPPLY_HEALTH=Good

POWER_SUPPLY_TECHNOLOGY=Li-ion

POWER_SUPPLY_VOLTAGE_NOW=4224000

POWER_SUPPLY_CURRENT_NOW=600000

POWER_SUPPLY_TEMP=55

root@freescale ~$ charger------ ddi_bc_gBrokenReason=1

What I think is suspicious, is that a current of 600mA is reported, at a battery voltage of over 4.2V !

Does nobody else observe a similar problem?

Addendum: it is probably a Jumper Problem: JP4 should be set to drain power from the 5V supply, NOT the Battery.

Which, of course, means that the EVK will not be useable on battery power alone. In that regard, the EVK design is a bit "broken".

I will later remove R230 and connect REG_3V3 to VDDIO_3V3.

I will add the results of this experiment later.

Reinhard

As promised: here the result:

Any larger load on the DCDC_BATT line will prevent the current from dropping below the 10% margin and thus prevent detecting the charge end. This will result in a charge timeout (BrokenReason=1).

That also means that a GSM Module, which usually runs off a Lithium cell, cannot be directly connected to the battery.