I have problem on my Xperia Play. I instal CM9 FXP112 and my battery was discharged in 5 hours. I used Darkforest CM9 kernel. So what can i do with it? (sorry for my english)
Developers are working on it! The next FXP/Darkforest releases should start to improve battery performance.
Disable Wi-FI if you are not using, that is the problem. Prevents the phone from going into Deep-sleep mode.
i had it disabled,killed apps by fast reboot but again, about 5 hours and battery was discharged. It is same problem with battery in CM7?
ojdon said:
Developers are working on it! The next FXP/Darkforest releases should start to improve battery performance.
Click to expand...
Click to collapse
Which version do you mean with next? FXP114?
Hello,
I search the best rom official or custom and the best kernel for the battery, only the battery.
Can you help me because i had only a utilisation intensive!!
Best regards
xamou711 said:
Hello,
I search the best rom official or custom and the best kernel for the battery, only the battery.
Can you help me because i had only a utilisation intensive!!
Best regards
Click to expand...
Click to collapse
at the moment i use the android revolution hd 6.0 rom with a perseus 29.1 kernel. the battery drain is 1,5%/h (betterbatterystats: 2d 42m battery 25%) and the cpu setting is stock. i will see if it change after the next charging.
i already downloaded the redpill rev1.00 kernel and i want to test it the next time.
unfortunately i've got some l2_hsic wakelocks with the perseus kernel (2h in 2 days) and will see what happens with the redpill kernel.
tenofnine
ROM and Kernel have only a small effect on battery life. The main culprits are processes that wake the phone up from deep sleep and engage the CPU or data transmittion.
When I lowered all auto-update intervals of apps from once-per-30-minutes to once-per-4 hours, my battery life nearly doubled itself.
i presume that you already disabled/lowered these "useful" things to improve your batterylife.
i don't play games on my cellphone, just some surfing and calling on viber (2-3h per day) however my battery lasts longer compared to the stock rom.
Hi all,
I am proud to be a small silent part of this community and to see how much the Galaxy W has grown due to the collective work of our members. I feel like the phone has damn near doubled it's value now with all the custom ROMs, kernels, tweaks, etc, etc that have been implemented in such a short time. Really, I can't see how this phone could be any faster! That being said I feel their is a void in resources dedicated to focusing on battery life while still retaining acceptable speed.
Moving on to Kernel 3.X has really brought great features and speed to our phone, but at the cost of extreme battery loss. It's early on in it's development, so I think there is room for improvement. I have noticed that members are starting to complain about battery life and I for one agree that this is a problem. I wanted to open a forum where we could collect ideas here so our precious phones could ideally make it through the day (or at least a half of a day!).
Anyway, I hope this will be a good jumping off point for discussing. I only ask if members could refrain from the old, "Just change the battery!" unless the user is reporting the battery is more pregnant then a Kardashian sporting twins.
Ideas
Some ideas to post about
ROM
Kernel
Governor
I/O SCHEDULERS
Tweaks
Apps
It would really help if contribute ideas on how to improve our battery life even for just a little.
Sent from my GT-I8150 using xda premium
I've had some success using conservative + row.
I'm really hoping to find a kernel that has more optimized conservative governors. Anyone have any tweaked conservative governor experience? I was looking at Scary or Lion heart, etc?
Btw I'm on RC 6 + Chronos kernel.
tebor8 said:
I've had some success using conservative + row.
I'm really hoping to find a kernel that has more optimized conservative governors. Anyone have any tweaked conservative governor experience? I was looking at Scary or Lion heart, etc?
Btw I'm on RC 6 + Chronos kernel.
Click to expand...
Click to collapse
I'm going to try Governor: Conservative. Is it better than SmartassV2? Anyway, I'm going to find out, just finishing the charge to 100%. I'm gonna test on CM10 Alpha 6 + Athena Kernel. Will report back after draining.
Okey bro, im using smartassv2 with athena on cm10. I'll post my observation here.
Sent from my GT-I8150 using xda premium
---------- Post added at 06:58 PM ---------- Previous post was at 06:57 PM ----------
Btw how much brightness are u guys using?
Sent from my GT-I8150 using xda premium
I'd love to get a CM9 or CM10 RC7 with 2.6 kernel along with a kernel 3.0 version for people who loves kernel 3.0
I am not asking to develop 2.6 kernel further. Previous 2.6 kernel is very good, but we need a ROM (base updated/sources updated) which can support 2.6 kernel fully. 2.6 Kernel was giving me much better battery life compared to kernel 3.
i tried flashing respective 2.6 kernels on CM9 and CM10 RC6. It seems to work fine but if i plug the usb cable to pc then all memory cards get unmounted. ie.. i can't copy files to phone or vice verse. I hope arco gives us a updated source for 2.6 kernel with RC7 release..
My suggestions for better battery life:
1. First thing first: make sure your battery is not dying; if it's bulging, replace it immediately.
2. Use a kernel that's undervolt-able. I use ICS, and Yoon.Kernel is perfect for me.
2a. Then, undervolt your phone. The Incredicontrol app is usually compatible with the method used by your kernel. Or, read the kernel's thread to find out which app to use.
3. Use Greenify (app) to 'hibernate' apps you don't need to run in the background at all times.
4. If you have the money (and can purchase apps via the Play), purchase "Autostarter" and disable apps' starting points. You should focus on the "After Startup" and "Connection Changed" events.
5. Every now and then, while the phone is on, turn off all data connection (wifi or 2G/3G), turn off sync, and let your phone charge to 100% *and* keep charging for awhile (trickle charging, 1-2 hours should be enough), then let the battery drain until your phone shuts down (by using it normally). Let the phone cool down, plug it into the charger, and power it on, and let it charge until 100% again.
6. Use BetterBatteryStats regularly to find apps consuming too much power and/or holding wakelocks too often; complain to the app maker, and replace it if the maker won't/can't change it.
7. Remember to turn off wifi and/or use 2G only mode when you're no longer using them.
Sent from my GT-I8150 using xda app-developers app
Pep, with full respect to your suggestions that are legit, but this can´t fix the battery drain issue CM9 is having with it´s new 3.X kernel.
They all will reduce the battery drain of course,but wont bring it down to the level you will reach with 2.6 kernel.
I´m still not sure what exactly is causing this, but in my case it´s for sure not my batteries nor other apps causing this.
As I already posted in the Bug report threat, I now flashed RC6 by going totally pristine. Formated both SD Cards and whiped everything, just installed BatteryMonitorWidget to control the battery drain. It´s twice as hight in DeepSleep compared to RC2 and RC3 with it´s old 2.6 kernel.
So first of all we need to narrow down what exactly is causing this.
Therefore I suggest to install first the BatteryMonitorWidget to comapare our results and to share the settings whe have changed after installing RC6 the pristine way.
First my configuration:
1. SDCards swapped
2. Dithering disabled
3. 16-bit Transparency enabled
4. Freeing up resurces enabled
5. Kernel sample Merging enabled
6. Zram disabled
7. default governor and scheduler
8. Syncing Accounts disabled
9. Updates and notifications in GooglePlay disabled
10. Wlan and MobileData disabled
11. Baseband XXLMB
My results with 2.6 Kernel are:
~16mA at daytime in DeepSleep
~ 5-8mA at night in DeepSleep
with kernel 3.x
constantly ~35 - 50mA in DeepSleep
and I checked it´s really in DeepSleep Mode.
So it looks like the radio is consuming more power while keeping the phone connected with the radiostation while the phone is in DeepSleep.
but....
I once also tested hadi´s 1.6 yoon Kernel with RC2, which is basing on Arcos RC2 kernel. By using the Ondemand govarnor in his kernel to compare the battery consumprion against Arcos Ondemand governor, hadis Ondemand governor also comsumed more than Arcos Ondemand Governor. With his OndemandX givernor the consumption was pretty much the same like with Arcos Ondemand.
So maybe it´s also something that changed in Arcos ondemand governor in 3.x kernel.
honeyx said:
Pep, with full respect to your suggestions that are legit, but this can´t fix the battery drain issue CM9 is having with it´s new 3.X kernel.
They all will reduce the battery drain of course,but wont bring it down to the level you will reach with 2.6 kernel.
I´m still not sure what exactly is causing this, but in my case it´s for sure not my batteries nor other apps causing this.
As I already posted in the Bug report threat, I now flashed RC6 by going totally pristine. Formated both SD Cards and whiped everything, just installed BatteryMonitorWidget to control the battery drain. It´s twice as hight in DeepSleep compared to RC2 and RC3 with it´s old 2.6 kernel.
So first of all we need to narrow down what exactly is causing this.
Therefore I suggest to install first the BatteryMonitorWidget to comapare our results and to share the settings whe have changed after installing RC6 the pristine way.
First my configuration:
1. SDCards swapped
2. Dithering disabled
3. 16-bit Transparency enabled
4. Freeing up resurces enabled
5. Kernel sample Merging enabled
6. Zram disabled
7. default governor and scheduler
8. Syncing Accounts disabled
9. Updates and notifications in GooglePlay disabled
10. Wlan and MobileData disabled
11. Baseband XXLMB
My results with 2.6 Kernel are:
~16mA at daytime in DeepSleep
~ 5-8mA at night in DeepSleep
with kernel 3.x
constantly ~35 - 50mA in DeepSleep
and I checked it´s really in DeepSleep Mode.
So it looks like the radio is consuming more power while keeping the phone connected with the radiostation while the phone is in DeepSleep.
but....
I once also tested hadi´s 1.6 yoon Kernel with RC2, which is basing on Arcos RC2 kernel. By using the Ondemand govarnor in his kernel to compare the battery consumprion against Arcos Ondemand governor, hadis Ondemand governor also comsumed more than Arcos Ondemand Governor. With his OndemandX givernor the consumption was pretty much the same like with Arcos Ondemand.
So maybe it´s also something that changed in Arcos ondemand governor in 3.x kernel.
Click to expand...
Click to collapse
Ah, I misunderstood the intent of this thread, then, my bad :banghead:
Hmmm... I'm willing to do some experiment... but there are several apps called "Better Battery Widgets"... which one?
Sent from my GT-I8150 using xda app-developers app
i don't know how will you classify my battery life because for me if the battery is enough to last at least for 8-9hours(hours spent in college) then i call it good enough
i use athena with no UV.
gov lulzactive 245-1024 scheduler row io
1 think usually done wrong in forum i'm following is they're using task killer or app to manage their free ram. this app is THE REAL BATTERY KILLER. they stop needed service and make your phone awake . not to mention the choppyness because the os needs to restart services and many bad effect it causes
Sent from my GT-I8150
pepoluan said:
Ah, I misunderstood the intent of this thread, then, my bad :banghead:
Hmmm... I'm willing to do some experiment... but there are several apps called "Better Battery Widgets"... which one?
Sent from my GT-I8150 using xda app-developers app
Click to expand...
Click to collapse
Oh I see there are several apps out now using the same name.
I´m using this BatteryMonitor Widget from 3C.
So when using it:
1. Do not recalibrate the battery as it will otherwise need much more time to show you accurate readings.
2. The App needs at least one or two charging cycles to recalibrate itself (not the battery) as it needs to determinate the exact capacity of your battery.
hadidjapri said:
i don't know how will you classify my battery life because for me if the battery is enough to last at least for 8-9hours(hours spent in college) then i call it good enough
i use athena with no UV.
gov lulzactive 245-1024 scheduler row io
1 think usually done wrong in forum i'm following is they're using task killer or app to manage their free ram. this app is THE REAL BATTERY KILLER. they stop needed service and make your phone awake . not to mention the choppyness because the os needs to restart services and many bad effect it causes
Sent from my GT-I8150
Click to expand...
Click to collapse
Hm well I´m used to recharge my battery every 3 to 5 days because it´s most of time in DeepSleep Mode. I´m just doing calls and surf the net in my breaks at work. So only turning on MobileData if it´s needed. At home I´m having my PC for surfing the net, so MobileData and WLAN is disabled as well.
With RC4 my battery drain was so high it even wouldn´t last as long till I´m back home from work. With RC6 i think it would last at least 12 to 18 hours, but it´s still a huge difference compared to the RC´s with it´s 2.6 Kernel.
Agree with you about using such apps like Task Killer or scripts like V6 Supercharger. They all need extra CPU time to do their jobs and often are causing more problems than they are solving.The Android OS can manage such things better itself and terminate unneeded services itself too.
---------- Post added at 08:23 AM ---------- Previous post was at 07:51 AM ----------
Additionally you can use BetterBatteryStats in case your phone is keeping a long time active to see what is causing this. It´s free for XDA members.
honeyx said:
My results with 2.6 Kernel are:
~16mA at daytime in DeepSleep
~ 5-8mA at night in DeepSleep
with kernel 3.x
constantly ~35 - 50mA in DeepSleep
and I checked it´s really in DeepSleep Mode.
So it looks like the radio is consuming more power while keeping the phone connected with the radiostation while the phone is in DeepSleep.
Click to expand...
Click to collapse
I can testify to this, my BBS shows the phone going into deep sleep normally but the drain with data on is immense even in deep sleep. Your initial findings already point to the most likely source of the problem. Given the varying opinions on this(some say it's fine I others don't agree), is it possible we have the same device but different hardware revisions, some of which are much more power-friendly than others? Because for the life of me I couldn't NEVER anything more than a day from kernel 3.x(CM10) but some people claim they could get upto 3-4 days on light use!
Sent from my Wonder powered by CM10
I don´t think it has something to do with different hardware revisions, else it wouldn´t matter which kernel is in use. But it does.
There is for sure something strange going on. I´ve additionally installed BetterBatteryStats and now the phone started keeping very often in awake mode. Not sure this has something to do with this app, because it didn´t happen when I had it installed on an older CM9 release.
However I can´t find any suspicious processes that are causing this. But to be sure I will have to go back to RC2 to compare with.
So here my results:
Within of 5 hours it keept Awake for 25 minutes.
BBS also says NO Data Connection and No or Unknown SIgnal during that time
But I´m sure it was connected to the radiostation and made a short testcall.
About wakelocks:
KernelWakelocks:
"mmc0_detect" 4m 32s
"DPRAM" 3m 11s
"alarm_rtc" 2m 21s
"PowerManagerService" 1m 36s
"radio-interface" 1m 10s
"alarm" 36s
"mmc1_detect" 13s
Partial Wakelocks:
AlarmManager 44s 997 Counts
GSM 20s 4 Counts
ActivityManager-Launch 18s 104 Counts
sleep_broadcast 4s 23 Counts
RILJ 2s 146 Counts
*backup* 1s 109 Counts
NotificationService 1s 113 Counts
Ok I´m back to RC2 and am having some results from BBS to compare with.
After fully charging the battery I unplugged the phone and left it over night at the same place I´m always leaving. So the conditions were the same like with RC6, though in RC6 there were no apps exept BatteryMonitorWidget and BetterBattery Stats installed.
To make it short the results are nearly identical. Even the Awake time is almost identical in a period of 5 hours.
There is also the No or unknown Signal entry for the whole time. So I guess this is due to the MobileData being disabled.
But what I´m missing in RC2 is "mmc0_detect" and "mmc1_detect" in the Kernel Wakelocks. They are not showing up there.
So now we need some other results to compare with my results.
Been using this phone for 1 and half year now..is it normal to have a battery life of 5 hours with 3G on? Currently using cm10
I think the 3.0 kernel really makes a big difference in a negative way in regards to the battery life.
To fight against this, I have used the following settings for some success:
Conservative + Row... still very fluid with Yoon Chronos
245 - 1024 MHz
Screen off after 30 secs
Under volt
Its much better than before, but still I would like more battery life. Probably 8 hours with heavy use up to 16 hours lite.
Sent from my GT-I8150 using xda app-developers app
Agree on that. I tried now RC6 together with hadis chronos kernel and the consumption is the same like with Arcos 3.x kernel.
The "mmc0_detect" and "mmc1_detect" entries also appear with hadis chronos kernel. Not sure they are responsible for it, but they didn´t appear with 2.6 kernel.
So what I can say for sure it´s not the governor but something else in the 3.x kernel and so far these mmc detect entries are under suspicion.
First of all, sorry for not testing against the two kernels; there's a major project going on in my office, and I totally can't risk to be offline... plus with the random times I'll receive calls, the battery usage might be skewed...
honeyx said:
Agree on that. I tried now RC6 together with hadis chronos kernel and the consumption is the same like with Arcos 3.x kernel.
The "mmc0_detect" and "mmc1_detect" entries also appear with hadis chronos kernel. Not sure they are responsible for it, but they didn´t appear with 2.6 kernel.
So what I can say for sure it´s not the governor but something else in the 3.x kernel and so far these mmc detect entries are under suspicion.
Click to expand...
Click to collapse
I did some Google search and arrived at the following thread:
http://forum.xda-developers.com/showthread.php?t=1827676
scroll down to the second post. The OP of that thread noticed that mmcX_detect will utilize a lot of wakelocks if the SD Card is defective.
Since the battery drain issue with kernel 3.0 does not seem to happen to everyone, could people experiencing such drain try the experiments without an SD Card being installed?
Sent from my GT-I8150 using xda app-developers app
No problem pep. I need my phone too during the week, so only can do some tests at the weekend.
So to narrow down this problem we need to know how it works without a SD Card installed and if all who are having this issue are having the SD Cards swapped.