Bug 736599

Summary: battery power drains quickly and takes long to charge; gnome-power-manager reports incorrect battery lifetime
Product: [Fedora] Fedora Reporter: Pekka Savola <pekkas>
Component: gnome-power-managerAssignee: Richard Hughes <hughsient>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: medium Docs Contact:
Priority: unspecified    
Version: 15CC: hughsient, rhughes
Target Milestone: ---   
Target Release: ---   
Hardware: i686   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-11-24 16:13:04 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
upower -d output when almost full and discharging.
none
upower -d output when almost empty none

Description Pekka Savola 2011-09-08 08:10:50 UTC
Created attachment 522055 [details]
upower -d output when almost full and discharging.

Description of problem:

1) For some time now when using Fedora 15, X61s battery has drained quickly. Regular idle use (with 5 firefox windows), the battery goes to zero in 1h15min - 1h30min.  A while back with Fedora 14, this was 4+ hours. This may have been due to kernel or other component changes, but I haven't been able to isolate it.

Charging the battery also takes a very long time. Yesterday, after 3.5 hours of charging, it was still at 60% and the estimate for 100% completion was 13 hours. Right now at 31% the estimate to completion is somewhat better, just 1.7 hours.

2) gnome-power-manager and upower report incorrect battery lifetime estimates.
Gnome-power-manager, when full, reports battery lifetime usually as 2h30min. It decreases slowly to about 1h45min. Then it jumps down to 4% lifetime and starts to complain about critical lifetime.  Battery stays at 4% or 0% for 15-30 minutes, even when /proc/acpi/battery/BAT0/state report "remaining capacity" as 0.

The jump down to 4% seems strange, as well as staying for half an hour in 0-4% state.

Version-Release number of selected component (if applicable):
kernel-2.6.40.4-5.fc15.i686
gnome-power-manager-3.0.2-2.fc15.i686
upower-0.9.12-1.fc15.i686

How reproducible:
Charge the battery to full. Take off the power cord and observe the time and behaviour to exhaust the battery. I need to kill gnome-power-manager and upowerd to make sure that the system runs out of power and doesn't hibernate/shutdown prematurely.

Additional info:
Attached are two outputs from upower -d.

Comment 1 Pekka Savola 2011-09-08 08:11:27 UTC
Created attachment 522058 [details]
upower -d output when almost empty

Comment 2 Richard Hughes 2011-09-08 10:24:49 UTC
(In reply to comment #0)

> Charging the battery also takes a very long time. Yesterday, after 3.5 hours of
> charging, it was still at 60% and the estimate for 100% completion was 13
> hours. Right now at 31% the estimate to completion is somewhat better, just 1.7
> hours.

Sounds like you need a new battery. How old is the battery and how many cells does it have?

Richard.

Comment 3 Pekka Savola 2011-09-08 12:31:57 UTC
I'm not sure if a new battery would fix the issues. This is 3 years old now. The model is (according to upower) Sony 42T4571, which appears to have 8 cells.  I do have the original 4-cell battery somewhere at another location (3yrs old, unused), but I suspect that would also exhibit weird behaviour.

Comment 4 Richard Hughes 2011-09-08 12:42:21 UTC
Can you test with another battery please. 3 years for a battery is quite old. Thanks.

Comment 5 Pekka Savola 2011-09-13 10:17:40 UTC
I've not been able test with the new battery yet, but I'll observe that after downgrading upower, the time calculation and battery lifetime is much more reliable now.

I suppose this change might be the culprit:

Description : - Fix how we estimate the device rate for batteries that do not
            :   provide this data.

I don't know how to verify if my battery does provide this data or not.

Comment 6 Richard Hughes 2011-11-15 13:54:08 UTC
(In reply to comment #5)
> I've not been able test with the new battery yet, but I'll observe that after
> downgrading upower, the time calculation and battery lifetime is much more
> reliable now.

Cool. Can we close this bug? You'll likely find the power hog by using either gnome-power-statistics or powertop.