Develop Bugs

Cooked ROMs battery drain tests

Hi,
I'm making some tests on battery drain in WM6 ROMs: Meshle's Vanilla 3, Fracoon's, B&B and HTC WM6 original (all additional programs closed in all ROMs). It takes some time so I'll let you know later about results.
I have one question for Meshle's ROMs users: did you noticed differences between Vanilla 2 and 3 versions in battery drain?
Thanks and have nice weekend.

unfortunately I cannot tell you anything about the difference between vanilla 2 and 3, BUT am very interested if you already had some (preliminary) info
patrick

Results
Hi,
Finally I've made all the tests.
Procedure:
Every ROM was flashed to device with original HTC WM6, then I swiched all Today plugins, utilities (like pBar) off, so system was clean and not overclocked. Then battery was fully charged.
For every ROM I made following tests (3x every test): first - streaming Internet radio 128 kbit/s with WiFi connection. Backlight was off, sound - 1/4. Test was made with GSPlayer. In table you can see how much it drains battery in % per hour (lower is better). Second test - playing Divx movie (320x240, full screen, no sound, backlight set to maximum) with TCPMP.
Like I said all tests were made 3 times for each ROM so you can see averages in table. I think the results are very similar for ROMs - 32-40% battery drain with WiFi and 16-22% while playing AVI movies so it's not so important parameter when you chose ROM for yourself.
I noticed that changing WLAN power save mode to "Best Battery" lowered power drain while streaming over WiFi from 40% to 10-12% so if you have strong WiFi signal it's better to do this. I haven't noticed differences in power consumption during calls, but I have no time for additional tests.
I've checked also free RAM and Storage memory (clean system with no plugins etc...).
Last impressions - with Bepe ROM I didn't managed to obtain Edge connection, Meshle Vanillas have almost no memory leakages.
I hope it'll be useful for somebody.

varvocel said:
Every ROM was flashed to device with original HTC WM6
...
Then battery was fully charged.
Click to expand...
Click to collapse
Good idea to measure power consumption. Interesting to check WM5 for comparison, as many people think it was a lot better.
However, I'd like to point out several things, which might influence outcome.. or might not.
1. Probably it is absolutely irrelevantl, but still... was there exactly the same version of power meter program used? At least, we would get same errors
2. There were complaints that battery after flashing does not last as long as before. Some kind of recalibration helps: full discharge until device shuts down followed by full charge. It was adviced to flash fully charged device, but still after some flashings I thought that battery drains too fast, so did the trick too.
Suggestion is to measure after discharge/charge cycle.
3. Correct me if I'm wrong, but probably like most gauges, the best precision is in the middle of the working range.
Normal behaviour of my device (while it is on and just sleeps) with fully charged battery: one day number stays at 100%; then gets lower at constant rate; and last few percents almost fly out.
In my opinion it could be right first to discharge from 100% to maybe 75-80%, and only after that start measurements.
Purely hardware power meter could be 'the final answer' to all questions but you need at least to have one.

1. Battery level was checked in System -> Settings -> Power
2. Every ROM was tested 3 times. Because 1 WiFi test takes about 40% of battery power device was fully charged 2 times.
3. First measurement was done when battery was full, but second at about 50%. It's interesting that differences was only 2-3% in both ways, so I think they are not significant because WiFi drain was about 40% like I said.
By the way - I'm using Meshle Vanilla 3 now, device was recharged 5-6 times already and I didn't observed any better or worse battery drain.
I fully agree - the best way to measure power consumption is some kind of hardware, but concerning very similar results for all ROMs I don't think it's really necessary to do further tests.
Thanks for you suggestions.

Current battery drain measurment tools?

I tried Battery Status, but it gave me strange results - for example battery drain with full brightness was less then with screen off (tested many times).
Could you recommend any really working application?

Make a dummy battery with contacts that align with those in the phone, and connect this to the phones battery via lengths of wire about 12" long.
Use a current clamp (I use a Fluke i30 for low current measurement) then sample the current ten times a second for one minute and average the results out to derive the current consumption - its not difficult really, and this method is far more accurate than software running on the phone - Mike

[Q] Battery drain too fast

I have been using Htc Desire + Motorola Milestone + samsung galaxy s these few days. It seems that my Desire, although lowest usage used up battery faster than milestone and galaxy, even in sleep mode! ( As for background syncing, milestone and galaxy used instant while desire is set to sync every 4hrs)
Is there anyway to at least increase the battery life?

In the first tow weeks the battery will not preform optimal, it needs a couple of charges before it operates as it should. Also if it seems that the battery drains to fast even in sleep mode, install current widget, set it to refresh every 5 minutes and to write a log. Normally the phone should use 5 to 10 mA on average during normal 3g standby (everything that allows push is set to push) and not more than 1 mA in flight mode. If your consumption is substantially higher than this you have either some very bad sync settings or there is a app running/syncing in the background when it should not be doing that. To check this dial *#*#4636#*#* and go to the battery usage and check the partial wake time.
Also are you using a task killer or does your Desire have worse mobile connection that the other phones (different network)?
To answer your question, yes there is something you can do to increase the battery life. There are different apps that help safe battery life, you can underclock and/or undervolt the cpu, swiching form 3g to 2g when no heavy internet is required, using as much a possible dark colors (only for amoled i believe) or lowering the background light.
But all this won't matter much if there is a background process which prevents the phone from sleeping.

TheGhost1233 said:
In the first tow weeks the battery will not preform optimal, it needs a couple of charges before it operates as it should. Also if it seems that the battery drains to fast even in sleep mode, install current widget, set it to refresh every 5 minutes and to write a log. Normally the phone should use 5 to 10 mA on average during normal 3g standby (everything that allows push is set to push) and not more than 1 mA in flight mode. If your consumption is substantially higher than this you have either some very bad sync settings or there is a app running/syncing in the background when it should not be doing that. To check this dial *#*#4636#*#* and go to the battery usage and check the partial wake time.
Also are you using a task killer or does your Desire have worse mobile connection that the other phones (different network)?
To answer your question, yes there is something you can do to increase the battery life. There are different apps that help safe battery life, you can underclock and/or undervolt the cpu, swiching form 3g to 2g when no heavy internet is required, using as much a possible dark colors (only for amoled i believe) or lowering the background light.
But all this won't matter much if there is a background process which prevents the phone from sleeping.
Click to expand...
Click to collapse
i have been using desire since june, its just that i use all 3 together recently and discover its's much battery draining. I seen the battery usage, and i saw radio cell is the highest standing at 80% while motorola milestone is around around 3%

magickiller9 said:
i have been using desire since june, its just that i use all 3 together recently and discover its's much battery draining. I seen the battery usage, and i saw radio cell is the highest standing at 80% while motorola milestone is around around 3%
Click to expand...
Click to collapse
Ofc it also depends on the roms you have on them and the radios, it'hard to tell...what have you got on your desire?

Thats not nice, my total cell standby and phone idle times total about 80% of the battery usage and i get 2 days out of one charge and if i don't play with it a lot and only use it as a phone( ) i got 3 days for one charge.
Oh one more thing, I got an S-LCD unit with a snow wallpaper.

vnvman said:
Ofc it also depends on the roms you have on them and the radios, it'hard to tell...what have you got on your desire?
Click to expand...
Click to collapse
I using the office Rom given by HTC
Sent from my HTC Desire using XDA App

[GUIDE] Galaxy Nexus Battery Life [12/27/11]

This guide is a work in progress, please contribute to it so we can have a central place for battery info!
Table of Contents:
1. Quick Tips
2. Monitoring Battery Usage
3.The Nitty-Gritty Details
4.Frequently Asked Questions
5.References
6.Changelog
Quick Tips
NOTE: As with many devices in their early days, there seems to be a few bugs needing to be worked out. Some of which include:
High "Android OS" Battery % usage. Possible remedies include settings->Wifi->Advanced(menu)->Keep Wifi on ALWAYS when sleeping; Trying alternative kernels from the dev section; although 4.0.3 may have some improvements, early leaks are not yeilding substantial changes.
High "Media Server" battery % usage when using a media app with audio, video, etc.
High Preset voltage table in kernel can be fixed with UC Kernels, but proceed cautiously.
http://code.google.com/p/android/issues/detail?id=22878
Battery Capacities. NOTE: The batteries are NOT interchangeable between GSM and LTE!
GSM: 1750mah
LTE: 1850mah
LTE extended battery: 2100mah
GSM extended battery: 2000mah
Settings that use us battery:
Screen brightness: Using the lowest comfortable setting will save a good bit of battery, espescially on such a large screen. App LogGraph can be used to change auto-brightness levels
Haptic feedback: Switch this setting off (Settings->Sound->Vibrate on touch). You may need to turn this off in alternative keyboards. Vibrations on button press, unlock, etc will use power for the vibrating motor. This includes vibrating for a call or text.
Button Sounds: Switch this setting off (Settings->Sound->"Dial pad touch tones"; "Touch Sounds"; "Screen Lock Sounds"). Speakers are small electro-magnets and require energy to move, consuming a small amount of power.
Sync: The more frequently your apps sync the more energy will be consumed. Turn off unnecessary syncs and reduce excessive syncing (Settings->Accounts & Sync->...) Alternative apps such as Juice Defender and Tasker can be used to control sync settings. Some ROMs are now incorporating this feature as well.
Antennas: The short of it: Use Wifi as often as you can, it scans for a connection less often than mobile data. BT, and GPS antennas use very little power when idling. This is the most often confused aspect of battery life. Read more in the "Nitty Gritty" section.
Live wallpaper: Requires greater CPU usage to run, and currently ICS lags in this CPU processing, as demonstrated when switching an older phone over to ICS and finding lag in the wallpaper.
Widgets: Higher refresh rates, sync intervals and greater changes in the widgets will consume energy more rapidly.
AMOLED Screen consume less power when displaying black.
Auto Rotate uses greater CPU and changes screen.
Apps for reducing battery usage:
Juice Defender - Very easy interface for switching wifi/bt/3g/etc on and off to save battery. Great first start. WARNING: Some users report greater battery consumption when USING juice defender. This may be due to certain settings they have or the app itself.
Tasker -Much more advanced, and less user friendly application for automating processes on android. Can be used to closely control the behavior of your phone.

Table of Contents:
1. Quick Tips
2. Monitoring Battery Usage
3.The Nitty-Gritty Details
4.Frequently Asked Questions
5.References
6.Changelog
Battery Monitoring
Standard Android OS Battery Usage Utility
Settings -> Battery
When entering this section, you will see two distinct sections:
Battery % graph, which shows the amount of battery used since unplugged, as well as a time on battery.
Individual App/Process battery usage. The Percentage here corresponds to the amount of battery used by the App/Process of the already used battery. NOT the total battery capacity.
Clicking on a process or app will bring up more details on that process. Some definitions used are as follows:
CPU Total - Total CPU usage in time
CPU Foreground Total - Total CPU usage while app is visible to user
Keep Awake - Total time app prevented phone from going into “Deep Sleep”
Screen on Time - Time screen was active (duh...)
Time on - Time process has been running
Battery Monitor Widget
I highly recommend this app. It gives you a wealth of information.
Battery % usage over very long periods of time
Battery voltage recording
Battery drain (mah and watts)
Battery capacity estimates (both in time and mah)
Battery conditioning status
Support for multiple batteries
Estimates for various conditions (watching video, audio, games, etc)
CPU Spy
I use this for one purpose: to see how much time my phone spends in each CPU frequency. And more importantly, how much time it is in deep sleep. Deep Sleep is the condition in which the phone suspends as much activity as it can. This is the lowest consumption state your phone can be in... besides being off

Table of Contents:
1. Quick Tips
2. Monitoring Battery Usage
3.The Nitty-Gritty Details
4.Frequently Asked Questions
5.References
6.Changelog
The Nitty Gritty
The key here is to think back to physics class. We have an energy storage device, and a few devices that consume energy and varying rates (power - NRG/Time)
Kernel Stuff
Adjusting Clock Speed - CPU and GPU frequenxcies will directly effect energy consumption when in use. That is to say, if your phone is just idling with the screen on, the CPU is using very little power. Much more is being eaten by the screen. There are two approaches to saving power in clock speed, both of which aim to reduce the time spent at higher frequencies.:
Reducing maximum clock speed.
Using cpu governer that keep the CPU at lower frequencies for longer periods of time. Power save, conservative, smartass, etc take different theories as how to achieve the best ballance between performance and energy efficiency.
[TIP] Use CPU spy to check much time the CPU spends in each frequency, and more importantly in deep sleep.
Adjustiing CPU Voltages - Using a kernel with an unlocked voltage table and setcpu to specify voltages for each individual frequency can lead to significant power savings from the CPU. When experimenting with voltages, do not check “set at boot” until you have established stable voltages.
Suggested Voltage testing: Coming soon...
Antennas
Wifi
Consumes more energy per time when trasmitting than mobile data, BUT...
Takes less time to transmit data, espescially large files, so it consumes overall less energy.
Wifi scans for a connection and pings the router less often than mobile data, thus consuming less energy at idle
Bluetooth
Consumes a small amount of power at idle to connect to new devices
Scanning for deveices consumes more energy
While connected, very little energy is consumed unless data is transmitted.
GPS
Unless an application specifically invokes the GPS antenna, it will not consume power
Using constant GPS updates (as in Navigation apps) will consume a large amount of power
Some apps will have bugs in them which will keep GPS on indefinitely. Try reinstalling those applications first, then remove if necessary.
Mobile Data
2G antennas consume the LEAST amount of energy at idle, but because they take so long to trasmit data, they can kill your battery.
3G antennas are Ok at idle, but take less time to transmit data, so are the best all around for traveling around metropilitan areas.
4G Antennas are FAST, but consume the most power. Best to leave these off when in transit, and used for stationary, stable connections
All Antennas consume much more power idling in areas of bad reception. So if you know you're going somewhere with poor reception, and don't need data, turn it off until needed (Settings->Wireless&Networks ->More ->Mobile Networks-> "Data Enabled"; "Data Roming" OFF; "Use Only 2G Networks"
Radios
Radios control how your phone connects to the mobile network. Specifically which proticols to use in which regions as well as how often to try to reconnect. This can lead to better or worse battery, signal, and network speeds.
Code:
NOTE: The Baseband is defined by 3 sets of codes. The first two are the Country Code, the next 2 denote the Year and Month (KK = 2011 November) and the last number denotes the Revision.
Country Code:
CE---> Belgium, Luxembourg, Netherlands
DC---> Thailand
DD---> India
DX---> Indonesia, Malaysia, Philippines, Singapore, Vietnam
DZ---> Malaysia, Singapore
JA---> South Africa
JC---> Algeria, Morocco, Nigeria, South Africa, Tunisia
JP---> Arabic, Algeria, Egypt, Iran, Iraq, Kuwait, Morocco, Nigeria, Oman, Pakistan, Saudi Arabia, Syria
JV---> Tunisia, Turkey
UG---> North America
UH---> Latin America, The Caribbean
XE---> Bulgaria, Estonia, Kazakhstan, Latvia, Lithuania, Russia, Ukraine
XX---> Austria, Belgium, France, Germany, Hungary, United Kingdom
XW---> Austria, Belgium, France, Germany, Hungary, Italy, Nordic, Spain, United Kingdom
ZC---> China, Hong Kong
ZH---> Hong Kong
ZS---> China, Hong Kong
ZT---> Taiwan

Table of Contents:
1. Quick Tips
2. Monitoring Battery Usage
3.The Nitty-Gritty Details
4.Frequently Asked Questions
5.References
6.Changelog
Frequently Asked Questions
Q: I just flashed a new rom or update, could that be the cause of battery issues?
A: Absolutely! A bad download, experimental features, or even settings that are not friendly with your device can greatly impact your battery life.
Q: What battery life should I expect out of my device?
A: That level can vary widely by usage. Generally, screen on time of 4+ hours is considered good.

Table of Contents:
1. Quick Tips
2. Monitoring Battery Usage
3.The Nitty-Gritty Details
4.Frequently Asked Questions
5.References
6.Changelog
References
References:
Better Battery Stats - http://forum.xda-developers.com/showthread.php?t=1179809&highlight=battery
Battery Calibration The easy way - http://forum.xda-developers.com/showthread.php?t=1024867&highlight=battery
Tasker Sync Control - http://forum.xda-developers.com/showthread.php?t=1031743&highlight=battery
Tips Battery Usage - http://forum.xda-developers.com/showthread.php?t=1374133&highlight=battery
SetCPU - https://market.android.com/details?id=com.mhuang.overclocking
Autostarts - https://market.android.com/details?id=com.elsdoerfer.android.autostarts
List of Battery wasting apps - http://forum.xda-developers.com/showthread.php?t=1251897&highlight=battery
Build.prop & init.d stuff - http://forum.xda-developers.com/showthread.php?t=1289554&highlight=battery
Battery monitor Widget - https://market.android.com/details?...d_apps#?t=W251bGwsMSwxLDEwOSwiY2NjNzEuYm13Il0.
System Tuner - https://market.android.com/details?...eloper#?t=W251bGwsMSwxLDEwMiwiY2NjNzEucG13Il0.
CPU Spy - https://market.android.com/details?id=com.bvalosek.cpuspy&hl=en
Juice Defender - https://market.android.com/details?...xLDEsImNvbS5sYXRlZHJvaWQuanVpY2VkZWZlbmRlciJd

Changelog
Code:
12/19/11 Basic Guide structure, info, etc
12/27/11 Added notes for current GNex issues; updated Nitty Gritty w/ antennas; added guides for settings.

Very nice thread. You might want to mention JuiceDefender as well, as that's a really useful utility too.

wanderfowl said:
Very nice thread. You might want to mention JuiceDefender as well, as that's a really useful utility too.
Click to expand...
Click to collapse
Thanks, added that and tasker to the first post as apps that can help save battery

why many posts...
Hey instaed of reposting everything can you try putting all together in one post..???

I think this should be mentioned:
USE WIFI!!!
Using 3G uses about double the battery of wifi. Using 4G uses about triple the battery of wifi. If there is wifi near you, use it! Use wifi at work and at home if at all possible.
Sent from my Galaxy Nexus using XDA App

You should also mention this on-going issue with battery life:
http://code.google.com/p/android/issues/detail?id=22878

For what its worth, my preliminary experiments seem to be showing that on the Galaxy Nexus, JuiceDefender uses more battery than it saves. FYI

is there any way to turn off the media server?

my battery is very inconsistent and at times I have horrendous drain

Not that it'll mean much, but Wi-Fi is the battery saver on this phone for me.
Light use all day, screen at 100% (which I turn off manually every time before putting it down)
E
verything on but bluetooth
Did some MP3 playback (2 hours), Lots of Tweetdeck, some maps - all while actually NOT on wifi (about 3 hours total) - the rest of the time I was at home.
I got about 18 hours before the phone shut off from a dead battery. The same without Wifi with 3G only would net me 8'ish. With 4G about 6ish - all with screen set to Auto.
So my casual and un-scienitific observation for my phone shows that there is something to be said about using WiFI instead of cellular service. I'll try and narrow it down further this week since I'm on vacation.

matt2053 said:
I think this should be mentioned:
USE WIFI!!!
Using 3G uses about double the battery of wifi. Using 4G uses about triple the battery of wifi. If there is wifi near you, use it! Use wifi at work and at home if at all possible.
Sent from my Galaxy Nexus using XDA App
Click to expand...
Click to collapse
Absolutely! Added a note for it, and a section in "Nitty Gritty" about all the antennas.
gogol said:
You should also mention this on-going issue with battery life:
http://code.google.com/p/android/issues/detail?id=22878
Click to expand...
Click to collapse
Added, thank you!
wanderfowl said:
For what its worth, my preliminary experiments seem to be showing that on the Galaxy Nexus, JuiceDefender uses more battery than it saves. FYI
Click to expand...
Click to collapse
I had the same experience, but I like Tasker for finer grain control anyway. The main thing to be weary of is Location-based settings. Tasker has a guide called "Location Without Tears" to explain it.
jnyce87 said:
is there any way to turn off the media server?
Click to expand...
Click to collapse
Not currently. You could freeze it through titanium backup, but that would break a bunch of apps. Some devs are working on fixes for this and Android OS
jnyce87 said:
my battery is very inconsistent and at times I have horrendous drain
Click to expand...
Click to collapse
All Three attached photos show pretty good battery life. Keep in mind: for a device early in its development, 12 hours is great!
Sal Khan said:
Not that it'll mean much, but Wi-Fi is the battery saver on this phone for me.
Light use all day, screen at 100% (which I turn off manually every time before putting it down)
Everything on but bluetooth
Did some MP3 playback (2 hours), Lots of Tweetdeck, some maps - all while actually NOT on wifi (about 3 hours total) - the rest of the time I was at home.
I got about 18 hours before the phone shut off from a dead battery. The same without Wifi with 3G only would net me 8'ish. With 4G about 6ish - all with screen set to Auto.
So my casual and un-scienitific observation for my phone shows that there is something to be said about using WiFI instead of cellular service. I'll try and narrow it down further this week since I'm on vacation.
Click to expand...
Click to collapse
Thank you for the perspective. I've added some sections referencing these issues.

I've been really surprised with the reviews about the battery life. So far I got the phone for a month ( in UK) and battery life is pretty good.
I have sync accounts, use the phone moderately (about 3 to 4 hours with screen on per day), no battery saving app or anything.
My daily routine is to unplug my phone from it's charger a 7am and every day I end up at night around 11pm with a battery around 40%... Some day I will forgot to plug it at night and be able to have it last almost for the entire next day (being careful though)
first week wasn't so good but after a few days battery improved significantly.
to me the key to save tons of battery life was to set the screen brightness to the second lowest level. It's really bright enough to my taste and literally since then I never got the phone to drain off within 20 hours after a full charge.
Also I have the wifi always on as mentioned earlier on the thread.

I have the GSM Nexus. I have had the phone for 1 week. My phone for phone calls, which drains the battery pretty quickly has not been real heavy as of late, but I have been typically seeing a 3% drop per hour of use. So right now I am at 78%, and have been up for 6h 48m.
Overall I am very happy with my battery life.
JOHN

Tubes6al4v said:
The key here is to think back to physics class. We have an energy storage device, and a few devices that consume energy and varying rates (power - NRG/Time)
...
Reducing maximum clock speed.
Click to expand...
Click to collapse
Is this really a good idea? Consider the following:
Higher clock speed means getting things done quicker. If the phone is running at 1400MHz with a voltage of 1225mV for one second, instead of running two seconds at 700MHz with a voltage of 900mV, what will demand the most energy?
(Voltages are the ones I'm using on my OC/UV'd Nexus, not the stock ones which are *way* high)
I'd rather focus on lowering the CPU voltage as low as it goes without instability than waste time with a slow phone

If i use my GNex normal i charge every two days, i like maximum display

[INFO] SGS2 T989 current consumption tests

I'm not sure if this would be of an interest for general public but those who like numbers can skip "OMG this/that SuperBattery/ROM/app let my phone run forever" statements.
I had read somewhere about similar testing for other phones (such posts are very rare). So I was curious to do such tests for my T989. And here we go:
SGS2 T989, ICS stock rooted/debloated ROM.
No Juice/CPU keepers/defenders/sleepers etc.
All syncs are manual.
All tests are done with dark background at 20% brightness if not otherwise stated.
Most of the time current jumps up and down so some results are with Low/High mA range.
Cell signal was -84dBm (pretty good 4 bars).
CPU was at default range 384MHz -1.5Ghz, ondemand.
Anker 2200mAh battery was at 80% - around 4V. When voltage goes down, the current would go up but I didn't bother to calculate in Watts instead because knowing ballpark in Amperes let you calc budget easier.
Deep sleep test.
In deep sleep,current stays below 10mA some of the time but every 5 sec jumps to 40/50mA, every 15 sec to 150mA. So, something is running on background giving an average of 11mA (somewhere around 0.5% of charge per hour).
If I put phone in airplane mode then there would be no jumps and average current will be 3/4mA. Over 10 hours (overnight) phone would loose only 1-2% vs. 5% when not in airplane mode.
Interesting, my previous phone SG S 4G had low current 2/3mA even when not in airplane mode. Well, anything running/looping on background (like sleeper script) would only consume more energy.
So, if you don't need to pickup calls at night - your best battery savior is an airplane mode... otherwise just put phone on a charger.
AMOLED Screen brightness test was done in airplane mode, idle:
Dark background 20% - 80/90mA, 50% - 110mA, 100% - 130mA
White background 20% - 120/130mA, 50% - 170mA, 100% - 250/270mA
This should give you around 10 hours of screen time if you say just reading white text on black background during boring intercontinental flight.
GSM voice call screen off - 280mA, screen on (speakerphone) - 350mA
(So basically, I can talk non stop for 2200/280 = 7 hours straight, not saying that I would do it )
Market apps update on WiFi - 250/300mA
SpeedTest WiFi - 450/500mA,
4G - 600/900mA (900 most of the time when uploading) (actual 4G speed was around 18Mb down and 5Mb up)
Everybody saying "too much 4G data will kill your battery" - but how much is to much? We can see that 2200mA budget enough for 2200/900 = 2.4 hours of intensive 4G usage. (There is some CPU part in it too)
And my cell signal was pretty strong, so wicker signal would require even more current to get your bits back and forth to the cell tower.
Anyway, in my case 4G requires double vs. WiFi.
CPU stress test 1.5GHz (364 Native Benchmark) - 570/600mA,
1.3GHz (402 Native Benchmark) - 460/480mA,
1.0GHz (506 Native Benchmark) - 350/390mA
You can see with performance goes power consumption... lowering from 1.5 down to 1.3GHz can save you around 120mA...
NetFlix playing through WiFi - 200/300mA, 4G - 240/600mA (600mA when loading buffer)
So, you can expect about 2200/300 = 7 hours of NetFlix from one charge.
Pandora, screen off, 4G - 70/350mA (350 when loading buffer 1/5 of a time)
Google Music screen on, local - 150/200mA, WiFi - 300/340mA, 4G - 500/700mA
I think Google Music buffering more then Pandora and I was not waited long enough... GM should consume about the same as Pandora.
Dice MP4, HW decoder, local video - 150/200mA
Google Map idle WiFi - 250mA, with GPS on - 350/400mA
Game Quell Reflection - 400mA (all the time)
Angry Birds - 250/350mA
Well, here are some of my screen shots as an illustration. Enjoy.

So you are saying all these battery savers are making things worse?

Here's mine with AOKP ICS Eugene CPU sleeper and TDJ kernel
Truth is battery life varies hugely, with usage and app setups.
Sent from the pink Unicorn from the Darkside.

I am trying to show that with the given battery budget it's up to user how to spend it.
Knowing how stuff works (at least in general ) allow to plan actions.
Speaking about CPU sleeper - it loops in background (already not good), it's trying to put CPU1 offline regardless what kernel scheduler think about it meaning next moment kernel would wake up CPU1 back if needed, so in next 2 sec sleeper script would try to put it off again and so on and so on....
Do you really need those intensive background processes running? Like Google++ etc. - if answer is Yes then CPU script is counterproductive, it's fighting against your wishes, spending even more energy... if answer is No then you better make sure your phone is clean of such rough apps, so it would go to deep sleep on it's own...
Deep sleep is not exactly up to Samsung specs on this phone - 0.5% per hour - meaning 200 hours in standby with excellent signal and with 2200mA Anker battery... only in airplane mode T989 got 3/4mA. So, could sleeper scrip help in this situation? Ask the author. Because when I ask him he just congratulated myself and that's it.
Band aids are always less effective then the cure of the root problem. Band aids could/would produce side effects (like something stopped working).
If you don't need dual CPU power - get yourself a single core phone... or lower CPUs frequency. According to my tests decreasing CPU frequency by 15% will save you 25%.
If you don't want to flip airplane mode by yourself then get apps which will do it for you but there might be some other gotchas...
Just try to get to the bottom of the problem and avoid the idiocracy "when the plants crave for... you know what".

LoopDoGG79 said:
Here's mine with AOKP ICS Eugene CPU sleeper and TDJ kernel
Truth is battery life varies hugely, with usage and app setups.
Sent from the pink Unicorn from the Darkside.
Click to expand...
Click to collapse
Looks fine. 20% in 12 hours with light usage. Test idle overnight with and without CPU sleeper to see if this is placebo effect.

I just did some tests on endurance. 4 days with 1 hour of talk and 1 hour of screen time and still have 28% to go. Or previously I did 7:40 of screen time with 40% battery still to go...
Someone asked what is my settings, well here we go:
1. I don't use automatic sync anywhere. When I need something I just hit manual sync.
2. Most of the time I use fixed brightness about 20%. (sometimes I set more when I need it)
3. Observe system behavior periodically, especially when installing new app, using OS monitor or just simple terminal (top 10 command). Are there any processes running without any reason? If so, deal with them...
4. My "Screen On" CPU frequency range is limited to 1.3GHz. (usually I don't play games on a phone, I've got a tablet for it). This will cap max current and will save around 120mA.
5. "Screen Off" CPU is limited to 0.9GHz - for same reason, no need to use 1.5GHz CPU for background tasks.
6. If there is spotty 4G - better switch it Off. Sure this will delay reports on you going from your phone to Google servers but I don't care about it.
7. Use WiFi instead of 4G - it costs at least two times less for you battery.
8. Switch WiFi Off when non needed. Idle WiFi will not consume much more mAmps but will set internal connection status = DISCONNECTED which will prevent different processes wakening up and transmitting.
9. Overnight, I prefer use airplane mode... two reasons: I don't like to wake up due to some "wrong number" calls. Second, deep sleep mode is still consuming around 11mA and airplane mode is just 3mA. So, instead of 4% of charge per 8 hours my phone uses just 1%.
(FYI: Eugene told in other post that his best overnight idling was about 10%)
10.Last but not least by using dark background and reading white fonts on black screen can save another 50-70mA. Very handy if you like read a lot. That's how I've got 7:40 hours of screen time with 40% still left in battery.
When you look at your time and compare what you were running/doing - keep in mind that there will be about 80% of battery capacity available. (Fuel Gauge chip will show 0% when there is still 3.45V) Manufacturers rate batteries discharging them down to 3.3V.
If you want to post your results, please include at least screen time along with type of usage and other basic settings. Otherwise such posts would be useless in terms of objective comparison different ROMs and settings.

i got my telus refubrished phone few days ago, i am getting around 12 hours on factory settings, i didnt change them. i use the phone for calling emails and news.
is that normal, if not, could it be the battery

omarnajat said:
i got my telus refubrished phone few days ago, i am getting around 12 hours on factory settings, i didnt change them. i use the phone for calling emails and news.
is that normal, if not, could it be the battery
Click to expand...
Click to collapse
Well, to say if your battery needs to be replaced you either have to test it on capacity tester or indirectly in your phone with some static (and well known load). Read the first post, choose benchmark/load, run it and then calculate your battery capacity approximation.
If your calculated capacity much less then 80% of rated one then probably you want to replace battery.
You may try to reset Fuel Gauge chip by removing full charged battery for several minutes from phone and try those tests again.
You can check for calibration errors by comparing voltage and what % Android reports.
100% - you should see 4.2 - 4.1V
50% - you should see 3.8-3.7V
1% - you should see around 3.45V
To answer on your other question "if this normal?" - one would need more information. Or you can analyze your usage pattern yourself based on metrics I provided at first post.

Really interesting post, thanks for taking the time to document and write this up. This is the sorta thing that is really great on XDA.

I went on a one week campout a few weeks ago. Put phone in airplane mode and unplugged around noon Monday, didn't plug it in again until around noon Saturday. Airplane mode the whole time, but I was also using Eugene's CPU Sleeper app and AOKP milestone 6 (July 15th release). Used phone for camera throughout the week, some very light reading and more than a few hours of gps on a long hike. Didn't think to take a screenshot though, so I can't back my claim. My battery was flirting with 0% by the time I plugged it in though, but that's still ~5 days. If I didn't use gps I could have easily gone much longer. Gps was the biggest battery killer.
Airplane mode is the best battery saver there is, but I like to be able to receive phone calls so I just let AOKP turn data off after about 10 minutes of screen off. Saves a lot of battery right there.

I have some crappy metro PCS
huawei ascend that I don't use at all. I turned it on airplane mode and left it for about a month and it was at 27 percent when I picked it up yesterday. I was running cm6 and I think I under clocked the prosessor running at 400mkz max
It has a 1500mah batt
my s2 has a extended 3650 mAh bat
Sent from my SGH-T989 on Jelly Bean goodness!
Edit: I just carry around an portable charger with me. I get about a week on my phone but you never know whats gonna happen

Updated to ICS 4.0.4 and looks like deep sleep got improvement.
Last night battery lost just 2% instead of 4% as expected for 8 hours on previous rom.