[PATCH Try#2] introduce-bq27000-battery-driver.patch

Andy Green andy at openmoko.com
Fri Feb 15 08:26:59 CET 2008

Hash: SHA1

Somebody in the thread at some point said:

Hi Tony -

>> Those time estimates are bogus because that battery was never charged at
>> that time.  The values are sensible once the charger patches are applied
>> (some of which are still coming), particularly "capacity" works nicely,
>> it's a percentage -- as you can see that battery was getting pretty
>> empty.
> Willie and our hardware people just meet battery vendor yesterday, and
> they required some parameter from us to make the coulomb work more
> _smart_ . I will update these parameters if hardware side finish the
> experiment today. Current parameter inside coulomb battery was some
> _default_ value, and these value might not reveal really battery
> performance. And wirting/re-program the battery need external 21v power
> and special fixture to erase eeprom inside the PCBA.

Yes the "bogus" figures got better from the battery started to make
correct results as soon as the battery was ever charged for a while from
the PMU, so it wasn't so bad after all.  But if we put sensible initial
guess in the battery EEPROM it will be more consistent for the user from
the start, so it is good.

> And the battery self update rate is about 2.5 sec, so the hdq
> application won't need visit battery so often.

Well the HDQ action via FIQ is only triggered to happen once each time
an application reads one of the /sys exposed "files"... and I guess a
GUI app will be doing that one time a minute or something to see if it
should change the battery state report.

- -Andy
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org


More information about the openmoko-kernel mailing list