You may be familiar with my previous threads testing the battery life of the Optimus V:
http://forum.androidcentral.com/optimus-v/71693-optimizing-battery-drain-optimus-v.html
http://forum.androidcentral.com/opt...attery-drain-optimization-tests-part-2-a.html
http://forum.androidcentral.com/opt...es-block-apps-using-3g-wifi-save-battery.html
I had originally planned to do a 4th round of testing pitting rom vs rom, kernel vs kernel, dev vs dev, in a no holds barred match of DEATH! That wasn't going to be that hard because when I started testing this phone, there were only two custom kernels to pick from and four roms. Then some inconsiderate jerks decided to develop a bunch of kernels and roms for the Optimus V. Now we have our choice of five kernels and 15 Roms! I would have to test over 50 kernel/rom combination to cover all of the possible ones. Combine that with the fact that these kernels and roms are changing on a daily basis, and it makes my previous testing techniques impractical. I even tried to collect user data to aide in my tests, but roms where being produced faster than people were reporting their data.
So for this round of testing, I took a different approach. I decided to test the capabilities of the hardware instead of focusing on any one piece of software. Instead of getting into the details of my tests, I am going to jump to the conclusions.
Kernel Effects on Battery Life
Before I changed my testing technique, I began running tests on the five kernels that were out. I installed a rom, configured it for "optimal" battery life while still running various practical apps like sipdroid and gmail, and then ran tests on how it behaved with different kernels. My results ended up being really boring. Aside from one of the kernels that I will not name, they all performed equally well. Even the "bad" kernel has since been fixed. Then I realized what I had done. By applying my optimizations, I had equalized the playing field.
You see, you can break a kernel down into its various parts, the most important being the hardware drivers. And while you may think those would be crucial for better battery life, they are actually fairly optimized from the factory. Even though all of the kernels were compiled differently and included slightly different drivers, and the devs may add modifications to them to make them more stable or have new features, the truth is that the hardware is going to consume a minimum amount of power no matter what you do. You can probably make the drivers less efficient, but that is probably the only way you'll have the effect the battery life directly. As long as all your hardware sensors and buttons work, the kernel i doing its job.
So now lets focus on the other parts of the kernel. There is the slash screen that can be included. This doesn't really effect anything, and it too can be "optimized" and disabled by the user without changing kernels. Lastly, there are the various startup settings and scripts that are run at every boot. And while this is a necessary evil to ensure stability since the kernel loads before the rom, it is here where kernels can go wrong, wifi scan rate to be too short, setting your CPU clock speed improperly, ...etc. But whatever the kernel sets, can be un-set by the rom or yourself, and this where I nuterlized the effect of having a different kernel.
So how critical is the kernel to battery life? Not very critical at all if you are willing to configure your rom to compensate for your kernel's shortcomings.
Then how critical is the rom to battery life? Again, if you are willing to uninstall battery hungry apps, and configure your phone, the rom because just a skin.
What does effect battery life? You do, you silly goose. Turning on the screen to check for messages, having your phone auto check for emails, auto downloading angry bird updates from the Market, this is what ruins your battery life. Anything the rom and kernel do can be changed. You, on the other hand, are who you are.
Optimus V Maximum Standby Battery Life
How long can the Optimus V be left on standby? I installed a stripped down version of Aphrodite (yeah, it is possible to remove even more things). I ran my optimizations which include increasing the memory read speeds and overclocking the CPU to 806.
Side not: Some may argue that I should have put the phone on airplane mode because even though I'm not making calls, the phone checks-in with nearby phone towers. Well here is a shocker I discovered by looking at the battery draw when in airplane mode, the phone still uses its radio to (I'm guessing) check-in with nearby towers. Airplane mode has the same periodic burst of current as when not in airplane mode. That can only be explained by the phone using its radio. Just a note for all the kidnappers that are reading this, no matter what you do, if your phone is on, it will check-in with nearby towers, and you can be tracked that way. I'm sure this is for 911 purposes, but for my tests, it means that airplane mode doesn't help me.
So with the phone stripped of apps, Wifi and 3G off, and everything else optimized for efficiency, we have a standby battery life of:
110 hrs plus or minus 5 hours
I never actually went the full amount of time because this is my primary phone and my stupid/inconsiderate friends and family would insist on calling me. Get a life, losers!
The closest I got was 80 hours, but using the remaining battery to calculate the total, I averaged 110 hrs. Pretty amazing considering that my first MP3 player was the size of an alarm clock and it would play songs for 2 hours at most.
Practical Standby Time
Same setup as before, except that I left 3G or Wifi on. I also used a firewall to block all internet traffic, and I left the default rom apps installed. In this state, the phone was completely usable as a web browser, GPS device, phone...etc the whole time during the test. All you had to do was lower the firewall and you were good to go.
Standby time on 3G:
95 hours plus or minus 5 hours
Standby time on Wifi:
85 hours plus or minus 5 hours
Again, these are estimated projections, but they were fairly consistent during the testing.
Real World Use
Here I used my phone with the setup I personally use. I left all the apps the way I have them. Everything was stock mmarz. I made no calls. I left Gtalk logged in as well as a popular GV VOIP app and the actual GV app. I also had my firewall setup to block all internet traffic except: VOIP, Gtalk, GV.
I actually managed to get through the entire test and my phone, after being on a mix of 3G and Wifi, shutoff at:
74 hours!
How to Get the Most Out of Your Phone
What does this all mean? Well, for starters, this smartphone has a pretty good battery life. I think these numbers could even be improved upon, not by a new kernel or rom, but by manually configuring the settings differently from what I used. So far, I have not found the magic formula that gives you the best battery life. What I did find is that the solution does not lie with the kernel or the rom. Throughout the testing I switched the rom that I was using. That did not effect the averages one bit!
Go ahead and pick the prettiest rom you want with the features you are looking for. Then pick a compatible bug free kernel. Once you do that, only install apps that you really need. Avoid apps that have location based features as those will wake up your phone and check the GPS location on a regular basis. Also avoid apps that have advertising (just buy the pay version you cheapskate). That is wasted bandwidth and therefore wasted battery life on advertising, and who knows how often the app will update itself in the background. Finally, when you have your phone setup and configured the way you enjoy, forget everything that I wrote here. You bought the darn thing. Use it the way that you want! If you are already getting +2 days of battery life, you probably won't do any better unless you are willing to turn your phone off. If you have less than 1 day of use on a single charge, you are either a heavy user of your phone, or you have something configured incorrectly. Don't blame your rom or kernel though. In the end of day, everything can be fixed by you.
http://forum.androidcentral.com/optimus-v/71693-optimizing-battery-drain-optimus-v.html
http://forum.androidcentral.com/opt...attery-drain-optimization-tests-part-2-a.html
http://forum.androidcentral.com/opt...es-block-apps-using-3g-wifi-save-battery.html
I had originally planned to do a 4th round of testing pitting rom vs rom, kernel vs kernel, dev vs dev, in a no holds barred match of DEATH! That wasn't going to be that hard because when I started testing this phone, there were only two custom kernels to pick from and four roms. Then some inconsiderate jerks decided to develop a bunch of kernels and roms for the Optimus V. Now we have our choice of five kernels and 15 Roms! I would have to test over 50 kernel/rom combination to cover all of the possible ones. Combine that with the fact that these kernels and roms are changing on a daily basis, and it makes my previous testing techniques impractical. I even tried to collect user data to aide in my tests, but roms where being produced faster than people were reporting their data.
So for this round of testing, I took a different approach. I decided to test the capabilities of the hardware instead of focusing on any one piece of software. Instead of getting into the details of my tests, I am going to jump to the conclusions.
Kernel Effects on Battery Life
Before I changed my testing technique, I began running tests on the five kernels that were out. I installed a rom, configured it for "optimal" battery life while still running various practical apps like sipdroid and gmail, and then ran tests on how it behaved with different kernels. My results ended up being really boring. Aside from one of the kernels that I will not name, they all performed equally well. Even the "bad" kernel has since been fixed. Then I realized what I had done. By applying my optimizations, I had equalized the playing field.
You see, you can break a kernel down into its various parts, the most important being the hardware drivers. And while you may think those would be crucial for better battery life, they are actually fairly optimized from the factory. Even though all of the kernels were compiled differently and included slightly different drivers, and the devs may add modifications to them to make them more stable or have new features, the truth is that the hardware is going to consume a minimum amount of power no matter what you do. You can probably make the drivers less efficient, but that is probably the only way you'll have the effect the battery life directly. As long as all your hardware sensors and buttons work, the kernel i doing its job.
So now lets focus on the other parts of the kernel. There is the slash screen that can be included. This doesn't really effect anything, and it too can be "optimized" and disabled by the user without changing kernels. Lastly, there are the various startup settings and scripts that are run at every boot. And while this is a necessary evil to ensure stability since the kernel loads before the rom, it is here where kernels can go wrong, wifi scan rate to be too short, setting your CPU clock speed improperly, ...etc. But whatever the kernel sets, can be un-set by the rom or yourself, and this where I nuterlized the effect of having a different kernel.
So how critical is the kernel to battery life? Not very critical at all if you are willing to configure your rom to compensate for your kernel's shortcomings.
Then how critical is the rom to battery life? Again, if you are willing to uninstall battery hungry apps, and configure your phone, the rom because just a skin.
What does effect battery life? You do, you silly goose. Turning on the screen to check for messages, having your phone auto check for emails, auto downloading angry bird updates from the Market, this is what ruins your battery life. Anything the rom and kernel do can be changed. You, on the other hand, are who you are.
Optimus V Maximum Standby Battery Life
How long can the Optimus V be left on standby? I installed a stripped down version of Aphrodite (yeah, it is possible to remove even more things). I ran my optimizations which include increasing the memory read speeds and overclocking the CPU to 806.
Side not: Some may argue that I should have put the phone on airplane mode because even though I'm not making calls, the phone checks-in with nearby phone towers. Well here is a shocker I discovered by looking at the battery draw when in airplane mode, the phone still uses its radio to (I'm guessing) check-in with nearby towers. Airplane mode has the same periodic burst of current as when not in airplane mode. That can only be explained by the phone using its radio. Just a note for all the kidnappers that are reading this, no matter what you do, if your phone is on, it will check-in with nearby towers, and you can be tracked that way. I'm sure this is for 911 purposes, but for my tests, it means that airplane mode doesn't help me.
So with the phone stripped of apps, Wifi and 3G off, and everything else optimized for efficiency, we have a standby battery life of:
110 hrs plus or minus 5 hours
I never actually went the full amount of time because this is my primary phone and my stupid/inconsiderate friends and family would insist on calling me. Get a life, losers!
The closest I got was 80 hours, but using the remaining battery to calculate the total, I averaged 110 hrs. Pretty amazing considering that my first MP3 player was the size of an alarm clock and it would play songs for 2 hours at most.
Practical Standby Time
Same setup as before, except that I left 3G or Wifi on. I also used a firewall to block all internet traffic, and I left the default rom apps installed. In this state, the phone was completely usable as a web browser, GPS device, phone...etc the whole time during the test. All you had to do was lower the firewall and you were good to go.
Standby time on 3G:
95 hours plus or minus 5 hours
Standby time on Wifi:
85 hours plus or minus 5 hours
Again, these are estimated projections, but they were fairly consistent during the testing.
Real World Use
Here I used my phone with the setup I personally use. I left all the apps the way I have them. Everything was stock mmarz. I made no calls. I left Gtalk logged in as well as a popular GV VOIP app and the actual GV app. I also had my firewall setup to block all internet traffic except: VOIP, Gtalk, GV.
I actually managed to get through the entire test and my phone, after being on a mix of 3G and Wifi, shutoff at:
74 hours!
How to Get the Most Out of Your Phone
What does this all mean? Well, for starters, this smartphone has a pretty good battery life. I think these numbers could even be improved upon, not by a new kernel or rom, but by manually configuring the settings differently from what I used. So far, I have not found the magic formula that gives you the best battery life. What I did find is that the solution does not lie with the kernel or the rom. Throughout the testing I switched the rom that I was using. That did not effect the averages one bit!
Go ahead and pick the prettiest rom you want with the features you are looking for. Then pick a compatible bug free kernel. Once you do that, only install apps that you really need. Avoid apps that have location based features as those will wake up your phone and check the GPS location on a regular basis. Also avoid apps that have advertising (just buy the pay version you cheapskate). That is wasted bandwidth and therefore wasted battery life on advertising, and who knows how often the app will update itself in the background. Finally, when you have your phone setup and configured the way you enjoy, forget everything that I wrote here. You bought the darn thing. Use it the way that you want! If you are already getting +2 days of battery life, you probably won't do any better unless you are willing to turn your phone off. If you have less than 1 day of use on a single charge, you are either a heavy user of your phone, or you have something configured incorrectly. Don't blame your rom or kernel though. In the end of day, everything can be fixed by you.
Last edited:
Facebook
Twitter
Instagram