Battery Lifetime

Andy Green andy at openmoko.com
Sun Jul 20 21:02:31 CEST 2008


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:

| I've noticed significantly reduced runtime when using one of the
| latest daily builds (2008-07-16 I think).  I have a theory.  I suspect
| bluetooth isn't really being turned off when I think it is any more.
| The factory image ran with a dimmed screen (no suspend) for a good 6
| hours and only drained to 64% according to asm.  The latest daily
| build completely drained the battery in 6 hours (even after asm -s).
| How can I check to ensure that the BT module isn't drawing power?

There's two levels of supicion you can apply to check what's really
going on.  First there is a logical enable signal that is reported by

cat /sys/bus/platform/devices/neo1973-pm-bt.0/power_on

This would report 0 if the BT stuff was logically "off".

If it still doesn't satisfy the suspicion, you can quite directly check
if it is being given power by the PMU.

cat /sys/devices/platform/s3c2440-i2c/i2c-adapter/i2c-0/0-0073/dump_regs

will (after a little pause while it grabs them) dump the whole PMU
register state.  BT is powered off LDO4, it means registers 0x33 and
0x34 are interesting.  The LDO is powered if b0 of 0x34 is '1', it
should be disabled if it claims BT is off.

However, WLAN is much hungrier than BT.  Maybe it can be that?

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkiDi8EACgkQOjLpvpq7dMoOhQCeN07Vy/RbA5sY+Lg5vmkuaROv
QA8An0pzeJYXvmtYsIT0bV01eViu08Uu
=9EYb
-----END PGP SIGNATURE-----




More information about the community mailing list