BQ E4.5 (r22): again battery down to zero

[Matthias Apitz]

I have again the situation wherein the battery was drained to zero in a bit
more than one hour;

my every-5-minutes cronjob about date+capacity shows:

Mon May 25 13:04:56 CEST 2015
battery capacity%: 72
Mon May 25 13:09:56 CEST 2015
battery capacity%: 71
Mon May 25 13:10:01 CEST 2015
battery capacity%: 71
Mon May 25 13:25:14 CEST 2015
battery capacity%: 68
Mon May 25 13:35:29 CEST 2015
battery capacity%: 66
Mon May 25 14:34:13 CEST 2015
battery capacity%: 12
Mon May 25 14:53:19 CEST 2015
battery capacity%: 1
Mon May 25 14:59:00 CEST 2015
battery capacity%: 0
Mon May 25 15:00:01 CEST 2015
battery capacity%: 0

i.e. between 13:35:29 CEST and 14:53:19 CEST the remaining 66% of the
capacity have been drained; it does not seem to be a non-suspend issue
because the cronjob was not run in this period every 5 minutes;

How this looks in the battery graphic is attached, as well the syslog of the time frame.

Thanks

 matthias
--
Matthias Apitz, <email address hidden>, http://www.unixarea.de/ [phone]

probably related to: LP: #1471913

[Matthias Apitz (gubu)]

the screen of the battery after reboot

[Matthias Apitz (gubu)]

Yesterday, between 5 pm and 6:45 pm my wife's BQ suffered again a total
drain. We recharged the device between 7pm and midnight and attached is
the picture what the battery says this morning. Interestingly, it shows
100% but the red line stays on zero.

Will try to see if changing the USB cable has some effect (thanks for the hint).

[Gabriele (gabrielemartelli)]

I think it's a hardware problem . Contact BQ . They are very professional and friendly .
Also my phone has had a hardware problem and was replaced in just 4 days , including the time required for transportation to and from Spain

g

[Matthias Apitz (gubu)]

I returned the device to BQ.com and they sent me within less than a week a new device (it is a new one with another IMEI and MAC, i.e. they did not repaired it). It runs now r23, the only additional apps installed is the terminal-app, and shows the same symtoms: while unused it drains the battery to zero within short time, see screen. This must be some software issue and the question is how to catch it.

[Jean-Baptiste Lallement (jibel)]

Thanks for your report.

- Could you please tell us if it happens randomly or systematically?
- When you say 'while unused' does it mean the phone is lying untouched and the display is black?
- Does it happen anywhere and any time or when you are in a specific environment (home, work, in transit, ...)
- Is the device particularly hot this battery drain occurs?
- When you reproduce can you please attach /var/log/syslog and the output of /android/system/bin/logcat
Can you please also open a terminal or connect with adb and check the output of top sorted by CPU consumption (press 'P' in interactive mode)

Thanks in advance.

[Matthias Apitz (gubu)]

answering the questions of comment #6:

- Could you please tell us if it happens randomly or systematically?

I do no have any pattern when it happens, i.e. until now it is not systematically; it happened 4-5 times, once now after changing the hardware;

- When you say 'while unused' does it mean the phone is lying untouched and the display is black?

Yes. It is the BQ 4.5 of my wife. She *very* seldom makes use of it and has the BQ in her bag. We always only detected the issue when the BQ did not wake up anymore and has to be charged. After charging we see the attached battery history (see attached screens).

- Does it happen anywhere and any time or when you are in a specific environment (home, work, in transit, ...)

It happened in different locations (in town, at work, at home). The BG runs without Wifi, only as a phone and a few data mobile (browser) usage.

- Is the device particularly hot this battery drain occurs?

Don't know because I detected it always hours after the issue occured.

[bzrudi (bzrudi)]

Just to confirm, the exactly same happened today with my BQ 4.5. I never had this before but today my E4.5 drained down to zero (from 50%) just within an hour or so. No apps opened, the only one I can think of is maybe Telegram, but not sure.

[Robie Basak (racb)]

I've seen similar issues on my Aquaris 4.5 across multiple OTAs but came to the conclusion that the software was overestimating the battery percentage - I think I was getting a fair amount of battery use out of the phone, but a poor indication of percentage remaining, rather than having a battery drainage problem.

I focused on making sure that the OS saw a number of full discharge and charge cycles in the hope of fixing any internal calibration, and I think the issue is much better for me now. While doing this I allowed the phone to die from low battery and then turned it on as soon as it reached 2% (the minimum it'll allow me to turn the phone on at) so the main OS saw both full discharge and charge cycles. I don't know if this helped, but that's what I did.

For others seeing the same sort of symptom, it might be worth considering how much battery life you're getting out of the phone in total in order to determine if this is a battery drain problem or remaining percentage indication problem.

[Matthias Apitz (gubu)]

re/ the Comment #9 of Robie Basak:

Please note, the issue here is not about how much energy can one get from the battery (i.e. how much hours of usage w/o charging), but about that the device drains all remaining energy, sometimes from 90%, to zero within a relatively short time (of an hour or so) and without that the user makes use of the phone. He/she detects the device as a brick after this in his pocket without any alarm before.

Of course this could either be a drain of energy for something running in the device, or as well a false alarm of the firmware of the battery or hardware of the BQ. At least after recharging the device the presented graph in the battery window shows this drain, maybe even this could be wrong and we are hunting a phantom.

[Robie Basak (racb)]

On Wed, Jul 22, 2015 at 05:28:19PM -0000, Matthias Apitz wrote:
> Please note, the issue here is not about how much energy can one get
> from the battery (i.e. how much hours of usage w/o charging), but about
> that the device drains all remaining energy, sometimes from 90%, to zero
> within a relatively short time (of an hour or so) and without that the
> user makes use of the phone. He/she detects the device as a brick after
> this in his pocket without any alarm before.

Sure, but note that what I'm saying is that is that although this is
what you might be observing (and it is certainly a problem), it could be
because the phone drains the last 1% but falsely reads 90% charge
remaining before it happens. So it appears to drop 90% but it actually
just drops from 1% to 0%.

Of course that's still a problem because you might expect the claimed
90% to last the day, and it won't if it's really 1%, and that needs to
be fixed. But since my observation suggests that it might be a false
reading rather than an actual drain, so this might be worth looking at
when trying to pin the cause down.

It might also be worth other reporters reporting their estimate of the
real battery remaining before a failure event without relying on the
reported percentage remaining, as I'm not sure that is reliable. This
might help developers pin it down more easily too, by figuring out if my
observation is consistent with the other reporters or if I'm observing a
different issue.

> At least after recharging the device the
> presented graph in the battery window shows this drain, maybe even this
> could be wrong and we are hunting a phantom.

I think that if the battery percentage remaining report is wrong, I'd
expect the graph to also be wrong since it is presumably based on that.

[Matthias Apitz (gubu)]

The cronologiy in the original bug description does not fit with the idea of a transition from 90% to zero in one step:

Mon May 25 13:04:56 CEST 2015
battery capacity%: 72
Mon May 25 13:09:56 CEST 2015
battery capacity%: 71
Mon May 25 13:10:01 CEST 2015
battery capacity%: 71
Mon May 25 13:25:14 CEST 2015
battery capacity%: 68
Mon May 25 13:35:29 CEST 2015
battery capacity%: 66
Mon May 25 14:34:13 CEST 2015
battery capacity%: 12
Mon May 25 14:53:19 CEST 2015
battery capacity%: 1
Mon May 25 14:59:00 CEST 2015
battery capacity%: 0
Mon May 25 15:00:01 CEST 2015
battery capacity%: 0

[Robie Basak (racb)]

I'm not claiming it goes down in one step, just that it goes down unusually fast for which one explanation is that the percentage readout is wrong rather than the battery is actually draining that quickly (which you'd probably notice from the alarming heat). AFAICT, this is still consistent with what you quoted.

[Lennart Jern (lennart-jern)]

My observations agree with Robie Basaks comments. A few times now my phone has claimed that the battery charge has not dropped a single percent during the day, just to suddenly drop later on. This seems to indicate that the charge is not reported correctly.

[Pat McGowan (pat-mcgowan)]

@YC is this something ou can look at

[Robie Basak (racb)]

I just by chance caught my phone with virtually no battery after a few hours of no use (it was just on standby with the screen off) when I turned the screen on to send a text message. As I typed my text I received multiple critical battery notifications (two or three I think) as the apparent battery percentage remaining dropped rapidly. Worried that the battery would die before I had a chance to finish my text and take a screenshot, I set it to charge and took a screenshot within a minute of doing this. Screenshot attached. The phone was not hot so definitely did not burn 50% of battery capacity in the space of one hour.

I have previously monitored /sys/class/power_supply/battery/capacity and I believe that the graph is accurately reporting this value, so perhaps this is a driver problem?

[Cosmin Humeniuc (cos64)]

This also happened to me a few times in a row.
Battery level is over 50% for a few days, then all of a sudden I get a phone call and when the screen turns on to show the incoming call it also displays a warning that the battery level is 1% or 2%. The graph looks similar to the one attached by Matthias.

[Simon Fels (morphis)]

Btw. I saw something very similar with having Android installed on a BQ Aquaris E4.5. When I left home the battery level was still shown as 30% in the UI but when leaving the card 15min later it dropped down to 1% and power off a bit later.

[Yuan-Chen Cheng (ycheng-twn)]

guess this is related: LP: #1471913

[Cosmin Humeniuc (cos64)]

I also think the battery level is overestimated.
Yesterday, 24 hours ago, I had 53%. Now I have 51%.
In the mean time, I made 2 calls (2-3 minutes each), and wifi and data have been permanently on.
Last charge was 105 hours ago.
I am almost sure that my battery is not 51% full, but rather almost empty.

[Matthias Apitz (gubu)]

I have had the same issue again two or three days ago with mi wifes BQ E4.5 running OTA-8; i.e. it is not fixed, if it is a software issue and not the firmware of the battery reporting nonsense.

One should prepare a special USB cable to measure the charging rate and voltage to see if a battery reported as zero is really charged from zero to full;

[Alexander Pitsch (alexander-pitsch)]

Being affected by the same bug on my bq Aquaris E5 running OTA 9, I've had the experience that I use the phone for a long time without any significant changes in the battery charge level - followed by a huge loss of power within minutes (please see attached file, looking very similar to other already uploaded images). Therefore, I think that the sudden rapid decay in the battery level must be a correction of a previous overestimation.

[Vincas Dargis (talkless)]

I have same issue with Aquaris E5 HD with OTA-10.1 (was the same with 8 and 9). I have same graphs like Alexander Pitsch has attached.

I achieve >160 hours without charning. WiFi, Mobile data is off all the time, _except_ for few minutes in the morning and some random short browsing through the day. Around 20 calls and messages per month.

I believe that graphs is overestimated, "freezes" around 60-50% and then drops rapidly after few minute of usage. 160 hours is a lot, so it's more like estimation problem, not actual drainage.

[Roman Shchekin (mrqtros)]

BQ E4.5 running OTA-10, my issue: https://yadi.sk/i/WjgrWanqrDYia
Phone was inactive, without SIM, with WiFi.

[Konrad Hofer (konrad.hofer)]

BQ E4.5 running OTA-11, yesterday I left phone for a couple of hours unused, when I wanted to make a call, battery was nearly 0%.
No Wifi, Bluetooth, only 2G was on. Battery Graph: http://imgur.com/a/DPve5

[Alexander Kinne (alexanderkinne)]

As suggested on the mailing list i am attaching a screenshot of the latest overnight power graph and /var/log/syslog

[Alexander Kinne (alexanderkinne)]

The corresponding syslog.

[Matthias Apitz (gubu)]

this 18 months old bug is at least related, if not even the same as https://bugs.launchpad.net/canonical-devices-system-image/+bug/1540062

the bug description in #1540062 is even more precisely then here; maybe we should close this one here;