For the last several weeks I've been having a battery drain issue with multiple ROMs, which is why I'm posting this (again) as a Android Development issue rather than in a ROM-specific thread.
Hardware: original 32B MyTouch 3G, from T-Mobile
Original ROM: Enomther's TheOfficial TMO
"Since flashing 2.6.1 and then 2.6.2, I've been having severe battery life issues, where when left alone the phone will drain completely within 7 or 8 hours. I've done much testing, and this is definitely not a battery meter calibration issue. I'm running Tmo, with RAM Hack. I've tried wiping and reflashing, picking out just about all third party apps, etc with no luck. I tried completely disabling location services via wifi/gps/cell, with no luck. Bluetooth/wifi are off, I tried both 2g and 3g only modes. The only hint I have is that under the battery usage screen, "Android OS" takes up anywhere from 60-80% so something is causing it to activate and not let the phone idle. Cyanogenized frequency scaling is on.
Running top -t shows "er.ServerThread" from process system_server usually leading the list, but the "top" process takes up the most CPU when I'm running it, so it appears to use 0% cpu."
Then I tried Dwayne's ROM, had the same problem. I then flashed TheOfficial AOSP, which seemed to fix the problem, but now it crops up again intermittently. Same battery drain but instead of 70% "Android OS usage I get about 30% "Android OS" and 30% "Cell Standby". I've tried uninstalling k9mail (the only background app), removing widgets, turning off sync, turning off location, and I can't track down what's causing this problem. It even seems to happen in airplane mode.
Some cursory searching on google shows that people are having the same types of issues on the Motorola Droid and on the Nexus One. Does *anyone* have suggestions on how I can track down what exactly is causing this increased power demand? I'm not afraid of the command line, but I couldn't find anything in addition to what I mentioned above.
Anyone?
***********************
**** I have been performing controlled experiments to the best of my ability. See below for results. ****
Experiment #1: Isolate if this is due to Wifi
Methodology (one test per night)
1. Keep wifi off. Perform a set usage pattern, then keep phone on sleep all night.
2. Perform the set usage pattern using wifi, then keep phone on sleep all night with wifi turned on.
3. Perform the set usage pattern using wifi, then keep phone on sleep all night with wifi turned off.
Results:
http://forum.xda-developers.com/showpost.php?p=50548879&postcount=27
Experiment #2: Isolate if this is due to GPS/Location Services (Ongoing)
Methodology (one test per night)
1. Keep GPS off. Perform a set usage pattern, then keep phone on sleep all night. [This is identical to the WiFi experiment, so there is no need to replicate this]
2. Perform the set usage pattern using WIFI, then keep phone on sleep all night with GPS turned on.
3. Perform the set usage pattern using GPS, then keep phone on sleep all night with wifi turned off.
Results:
http://forum.xda-developers.com/showpost.php?p=50685378&postcount=69
***********************
So, I've been battling idle drain battery problems for a while now . I was unable to make much progress because I did not have visibility into partial wakelocks on the system . So I did what any inquisitive person would, I rooted my phone. Bye bye warranty.
Now that I have root, I am able to see the partial wakelocks in better battery stats. To my partial surprise , I do not see any substantial partial wakelocks that would explain the significant battery drain at night. I have good cellular signal, and wifi signal where the phone is placed.
Any clues?
Almost 25 percent in 9 hours. Also notice the lack of activity between successive awake blips, and the constant battery drain that occurs between them. Before I even got root and was able to look at the partial wakelocks, I had already suspected that there were no wakelocks going on but yet the battery was draining.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Below are screenshots from the stock battery app showing the top battery drainers and the details of the major ones.
By the way, I would attribute approximately 6 percent of the battery drain due to the screens. I have quite a bit of data that suggests at the brightness levels I was using the screen for those 30 minutes, the drain rate should have been around 12 percent per hour. That still leaves 18 to 20 percent of battery to be accounted for. According to better battery stats, the phone was only awake for around 30 minutes total when the screen was off. And looking at the CPU Lodge, it's clear that those 30 minutes or so were spent in the lowest frequency mode , and that the phone was in deep sleep at all other times.
Attached are the better battery stats logs that shows the partial wakelocks as well. I don't know why the CPU's states are not included in the file but I can assure you I took a look look at it and the 30 or so minutes of awake time were perfectly captured by the 300mhz state.
http://pastebin.com/j685pa5q
Sent from my XT1034 using Tapatalk
I turned airplane mode on, and left location services on. Battery drain was 2.8%/hr (measured over a 5% battery drain interval) with the phone only being awake for 7min 28sec when the screen was off. There were around 5min of NLP wakelocks, and some other miscelaneous ones. I don't think that explains the battery drain fully.
Then, I turned off location services, and the battery drain has been 1.7%/hr (measured over an 8% battery drain interval) with a total awake time of 5 mins, and of that only 1 min of awake time when the screen was off. Very few to no wakelocks.
What is wrong? Why is the battery draining at ~2%/hr with airplane mode on, location services off, and the phone being in deep sleep pretty much the whole time!?
Hey I'm trying to find if there of a pattern with the battery drain and the battery voltage (mv) .I slightly have the felling the battery is a bit wonky . Could you check the battery voltage while charging just before 100% and the voltage when it has fully charged ? .. Try the code *#*#4636#*#*
I feel the my peak battery voltage has dropped significantly over the last 2 weeks.
Sent from my XT1033 using Tapatalk
I have the same problem with mine, no clue what the problem is, I turned off location services and I dont use 3g at all, only wifi while at home or other places sometimes, but the drain persists, still dont know what to do
Sent from my Moto G :3
Try a Cache wipe in recovery - http://forum.xda-developers.com/showthread.php?t=2634361
neu - smurph said:
Try a Cache wipe in recovery - http://forum.xda-developers.com/showthread.php?t=2634361
Click to expand...
Click to collapse
I don't think a cache wipe is the solution. For example, see the updated battery settings screenshot below. The flat line at the end is the result of just rebooting the device (no cache wipe). The battery is no longer draining constantly, even though the same services are activated.
I'm the OP of the thread that I think first reported this issue. Note that I found that the drain is noticeable in airplane mode as you expect there to be almost zero drain, however it still happens when not in airplane mode as well, you just notice it less as the device is doing other stuff which all adds up to battery drain.
I think there is a driver bug, most likely a WiFi driver bug or possibly GPS driver bug. For example maybe GPS is somehow constantly active but the bug manifests itself in such a way that this is not reported to the OS itself which is why you do not see it active in the battery screen .
I had the exact same issue as reported above, no wakelocks over and above the normal amount but the drain was 3% hr or so. A reboot fixed it but it would reoccur when i connected to another WiFi network.
A factory reset fixed it completely until a few weeks ago when I was on the road and used GPS. That night I had the battery drain issue again.
So whether its WiFi or GPS that's the cause I don't know, and why, for me, did it stay fixed for so long. I also can't remember if I had GPS switched on when I first had the issue before the factory reset so makes it harder for me to determine the cause.
The above was all on 4.3. I've now upgraded to 4.4.2.
I think the only way to debug will to be to pull off all the various logs from android, you may even need to turn on various debugging/logs from the OS to capture more logs. I'm pretty sure you wont diagnose from just looking at wakelocks and battery screens.
If you can recreate this problem, and sounds like some of you have it a lot, then you are in a good position to get the logs.
BTW does anyone use the BBC weather app, I noticed this caused a LOT of NLP wakelocks so don't think its related but I did notice on 4.4.2 that it drained battery fast so I uninstalled. There is a thread in xda that I started about it.
A poster above says to clear cache, this is normally pointless as /cache will usually contain nothing except a few empty folders and a couple of log files. Also clearing cache means a reboot which temporarily fixes the issue anyway.
scott_doyland said:
I'm the OP of the thread that I think first reported this issue. Note that I found that the drain is noticeable in airplane mode as you expect there to be almost zero drain, however it still happens when not in airplane mode as well, you just notice it less as the device is doing other stuff which all adds up to battery drain.
I think there is a driver bug, most likely a WiFi driver bug or possibly GPS driver bug. For example maybe GPS is somehow constantly active but the bug manifests itself in such a way that this is not reported to the OS itself which is why you do not see it active in the battery screen .
I had the exact same issue as reported above, no wakelocks over and above the normal amount but the drain was 3% hr or so. A reboot fixed it but it would reoccur when i connected to another WiFi network.
A factory reset fixed it completely until a few weeks ago when I was on the road and used GPS. That night I had the battery drain issue again.
So whether its WiFi or GPS that's the cause I don't know, and why, for me, did it stay fixed for so long. I also can't remember if I had GPS switched on when I first had the issue before the factory reset so makes it harder for me to determine the cause.
The above was all on 4.3. I've now upgraded to 4.4.2.
I think the only way to debug will to be to pull off all the various logs from android, you may even need to turn on various debugging/logs from the OS to capture more logs. I'm pretty sure you wont diagnose from just looking at wakelocks and battery screens.
If you can recreate this problem, and sounds like some of you have it a lot, then you are in a good position to get the logs.
BTW does anyone use the BBC weather app, I noticed this caused a LOT of NLP wakelocks so don't think its related but I did notice on 4.4.2 that it drained battery fast so I uninstalled. There is a thread in xda that I started about it.
A poster above says to clear cache, this is normally pointless as /cache will usually contain nothing except a few empty folders and a couple of log files. Also clearing cache means a reboot which temporarily fixes the issue anyway.
Click to expand...
Click to collapse
Thanks you very much for your reply Scott.
As you can see in my troubleshooting, I eventually turned location services and wifi off, but still measured a considerable battery drain. Do you suppose that once you use wifi, or location services after bootup, the effect may remain afterward even if they are switched off? If so, do you agree that if I turn off location services and wifi, then reboot so that the phone boots up with neither - that should be able to verify your hypothesis?
I'll go ahead and try that for a few days and see how it goes. Will report back.
Others have mentioned something to do with the cellular radio power management, and how rebooting and clearing the cache helps with that (apparently this came from a motorola rep...). I've included a data point in this thread where I merely rebooted without a cache clear and the problem went away. So, perhaps the clearing of the cache really does nothing at all ...
If this is a software issue, why aren't all phone with the same software version effected? How are people getting 6-7 days of standby time consistently? Maybe they never use wifi or location services? Well, I intend to test the wifi/gps hypothesis soon .
google+, google now etc caused me a massive battery drain until i greenified them. now i get a solid 2 days and about ~6 hours screen on time.
twist3d0n3 said:
google+, google now etc caused me a massive battery drain until i greenified them. now i get a solid 2 days and about ~6 hours screen on time.
Click to expand...
Click to collapse
Do you need the donate version to hibernate them?
twist3d0n3 said:
google+, google now etc caused me a massive battery drain until i greenified them. now i get a solid 2 days and about ~6 hours screen on time.
Click to expand...
Click to collapse
So that´s exactly what I did, too. Turns out my Moto G is a battery monster, 2 days with average screen time without any problems.
tictac0566 said:
So that´s exactly what I did, too. Turns out my Moto G is a battery monster, 2 days with average screen time without any problems.
Click to expand...
Click to collapse
This thread is specifically about battery drain caused by something unknown, not an app or wakelock, I think we should try to keep it clean by not cluttering with app related issues as that can then be confusing for people who are reading through the thread. Thanks.
---------- Post added at 11:40 AM ---------- Previous post was at 11:16 AM ----------
if0rg0t said:
Thanks you very much for your reply Scott.
As you can see in my troubleshooting, I eventually turned location services and wifi off, but still measured a considerable battery drain. Do you suppose that once you use wifi, or location services after bootup, the effect may remain afterward even if they are switched off? If so, do you agree that if I turn off location services and wifi, then reboot so that the phone boots up with neither - that should be able to verify your hypothesis?
I'll go ahead and try that for a few days and see how it goes. Will report back.
Others have mentioned something to do with the cellular radio power management, and how rebooting and clearing the cache helps with that (apparently this came from a motorola rep...). I've included a data point in this thread where I merely rebooted without a cache clear and the problem went away. So, perhaps the clearing of the cache really does nothing at all ...
If this is a software issue, why aren't all phone with the same software version effected? How are people getting 6-7 days of standby time consistently? Maybe they never use wifi or location services? Well, I intend to test the wifi/gps hypothesis soon .
Click to expand...
Click to collapse
Well I cant answer all your questions but I will tell you about my test last night. I hardly use GPS at all (and my battery is normal) and last time I did use GPS I had the battery drain as per my post above, so last night I:
1. Drove 15mins from my house.
2. Switched on GPS (high accuracy).
3. Started google maps and drove for 2mins with GPS in use.
4. Shutdown maps app and drove home (GPS still on - ie high accuracy).
5. At home used phone for bit of XDA on wifi for a bit.
6. Switched on airplane mode but location service still on (note airplane mode does NOT switch off GPS), woke up at 02:00AM and noted I had the annoying battery drain - arghhh.
7. Switched off GPS - woke up 2 hrs later and drain still happening.
8. Switch off location altogether - woke up in morning and drain still happening.
9. Rebooted phone and now battery back to normal, Im back on wifi/mobile location (battery saving).
So to me it looks like its maybe related to GPS in some way but whether its purely GPS or a GPS/wigi/location combo I have no idea.
However I have in the past switced on GPS in my house just to play about and havent has the drain, so is it due to a larger change of location when GPS is on (ie 15mins drive from my house) that causes it to go bad.
So annoying, I took a CWM backup before I booted up my phone so I will pull that to my laptop and see if there are any obvious error logs to look at but I doubt it.
Im still leaning towards a driver issue or maybe some component, eg GPS chip, staying alive/awake even when not in use but m sure others have reported they have it with location/GPS off completely. Maybe its even a few different problems.
You said not everyone has the error, but I bet quite a few non-xda people do but do not realise. A lot of people will just charge their phones overnight so wont even realise, during the day they will think ~2% drop per hour is normal. Even if they do notice the drop in airplane mode they probably think thats normal and ignore it.
I think getting Motorola to accept there is an error is going to be almost impossible especially if its hardware related. Also as we have unlocked bootloader I doubt they will even accept a call once they know that.
I think it would be interesting to get a list of phones that are affected, ie model, serial no, location brought, ROM etc. maybe its a certain batch of phones, if we can determine that it would be some help, maybe we could buy one that isnt affected
So to start it off my phone is
XT1032 UK retail brought from Basildon, Essex, UK from phones4u.
Problem happens on 4.3 stock and 4.4.2 stock.
Baseband version is MSM8626BP_1032.340.63.00R.
Im tempted to flash GPE and/or new basenband radio but I bet the same thing happens.
I may see if there is a new GPS driver but no idea how to flash it etc.
Such a shame as this phone is excellent for me otherwise.
Just seen this thread as well, it blames GPS
http://forum.xda-developers.com/showthread.php?t=2607641
if0rg0t said:
Do you need the donate version to hibernate them?
Click to expand...
Click to collapse
if they're system apps, yeah. totally worth it tho cos you can do maps/fb etc too, also you can allow apps that get push messages to still give you notifications whilst greenified.
e: just to note, i have gps (high accuracy) enabled constantly and still manage 2 days.
twist3d0n3 said:
if they're system apps, yeah. totally worth it tho cos you can do maps/fb etc too, also you can allow apps that get push messages to still give you notifications whilst greenified.
e: just to note, i have gps (high accuracy) enabled constantly and still manage 2 days.
Click to expand...
Click to collapse
Maybe you can help test if you don't mind.
1. Use Google maps with high accuracy when you go out for 5mins or so.
2. Once home use phone as normal on WiFi for a bit, just 10mins will be enough.
3. Take note of battery percent.
4. Switch on airplane mode for 2hrs - DO NOT USE PHONE during this time.
5. After 2hrs take note of battery percent.
6. Report the battery percents recorded in this thread.
If you can do this we would be grateful.
Let us know your phone model, where purchased, ROM etc.
Thanks.
Hi,
Following this MotoG thread for some time I have to admit that my MotoG also has the same issue from time to time.
This is the reason now that I joined XDA because this issue is a real pain.
Having my MotoG for three weeks now I had to restart it four or five times to get rid of the excessive battery drain.
My MotoG (XT1032) is bought in the Netherlands, unlocked, rooted, CWM and stock ROM.
I did remove some Motorola apps and disabled some.
Because of the last remarks that it seems to be GPS related, I want to let you know that it also happened to me without using the GPS.
Yesterday it happened at school, at first the drain was not present but suddenly it started.
In the between the two situation I only checked my mail through Wifi and checked some news.
My MotoG has Wifi always on, bluetooth off, GPS off.
There must be something more than just GPS which causes this drain.
Hopefully we can find the root cause soon.
Mickeyvdb said:
Hi,
Following this MotoG thread for some time I have to admit that my MotoG also has the same issue from time to time.
This is the reason now that I joined XDA because this issue is a real pain.
Having my MotoG for three weeks now I had to restart it four or five times to get rid of the excessive battery drain.
My MotoG (XT1032) is bought in the Netherlands, unlocked, rooted, CWM and stock ROM.
I did remove some Motorola apps and disabled some.
Because of the last remarks that it seems to be GPS related, I want to let you know that it also happened to me without using the GPS.
Yesterday it happened at school, at first the drain was not present but suddenly it started.
In the between the two situation I only checked my mail through Wifi and checked some news.
My MotoG has Wifi always on, bluetooth off, GPS off.
There must be something more than just GPS which causes this drain.
Hopefully we can find the root cause soon.
Click to expand...
Click to collapse
Thanks for your reply, have you confirmed this drain happens when in airplane mode as well and that no wakelocks occur, no unusual statistics in battery screen etc. I ask as some people say they have the issue but then say they have greenified an app etc to stop it which basically means they dont have the issue at all. Im not saying I dont believe you, far from it, I just want to confirm its exactly the same issue as a few of us are seeing. Next time can you NOT reboot but leave phone in airplane mode for a few hours and DONT use it at all, then see if battery has gone down by 2-3% per hour.
Thanks.
Have not used airplane mode to check it, but will sure do this when it happens next time.
What I observed in battery stats was a high percentage of android-os.
I have not used greenify so far.
Sent from my GT-P3110 using xda app-developers app
Mickeyvdb said:
Have not used airplane mode to check it, but will sure do this when it happens next time.
What I observed in battery stats was a high percentage of android-os.
I have not used greenify so far.
Sent from my GT-P3110 using xda app-developers app
Click to expand...
Click to collapse
Doesn't sound like the same issue if android OS shows high in battery stats. That sounds more like a wakelock issue of some kind which still means you may get drain in airplane mode but for a different reason.
This thread is specifically for the problem of ALL battery stats looking normal, so android OS doesn't show up as high.
I think your best get is to root and install 'wakelock detector' app and you will probably see some high wakelock for android OS.
Thanks for your post on the Motorola forums Scott. It was well written, and to the point. I hope this issue gets the attention it deserves. I'll be reporting in the next few days with my experiments.
scott_doyland said:
Doesn't sound like the same issue if android OS shows high in battery stats. That sounds more like a wakelock issue of some kind which still means you may get drain in airplane mode but for a different reason.
This thread is specifically for the problem of ALL battery stats looking normal, so android OS doesn't show up as high.
I think your best get is to root and install 'wakelock detector' app and you will probably see some high wakelock for android OS.
Click to expand...
Click to collapse
Sorry to post this wrong condition.
I had the impression that I was suffering the same problem.
Will have a look with a 'wakelock detector' to see if I can tackle the high android-os consumption since I rooted my device already.
Although the issue seems different I will have a good look this morning since I have to use the GPS.
Will keep you informed if I notice something weird.