electricpete
Android Expert
I have downloaded the program PowerTutuor and have been trying it out on my Infuse phone.
First some terms / assumptions related to battery life:
Battery Life-time = 1750mA-hr = 100%
17.5 milliamps corresponds to 1% per hour.
If we want to say it in milliwats, assume averageit voltage ~ 3.8v.
1% per hour corresponds to 17.5 milliAmps = 17.5*3.8 =66 milliwatts
Our battery lifetime (still assuming 3.8V average) is 1.75*3.8=6.5 watt-hours. (in other words, or battery can provide energy to run a hundred watt light for only 60minutes * 6.5/100 = 3.9 minutes..... We should not be surprised at how fast the battery goes down but instead simply amazed at how much our phones can do in this tiny package... I certainly am).
Since there are 3600 seconds in an hour, that 6.5 watt-hours lifetime is the same as 23,400 watt-seconds = 23.4 kJ
The program can break power usage into categories: Display, cpu, wifi, data network, gps, audio. The program is developed for another phone but states it should work for most Android phones. On my Infuse it certainly responds qualitatively in the right direction for all inputs except data network. It consistently states zero power usage for this function regarldess of whether my network is of or on and transferring data. Initially I assumed it must have some problem accessing this particular data on this particular phone and the rest are probably (?) accurate. But some of the data are strange, I’m not sure.
Here are some results
Display:
1% screen - 300mW ~ 4.5%/hr
12% screen - 350mW ~5%/hr (20 hr life running from 100% to 0% batt with no other loads)
30% screen 430mW ~ 6%/hr
100% screen 900mW = 13%/hr (7.5hr life running from 100 to 0 with no other loads)
* Display power usage does not change when I load a black screen or a white screen or a blue screen. Have we been taught wrong all this time? Beats me.
CPU
At the time I was doing real-time monitoring, my cpu tended to use 75milliwatts with static display. If I scrolled fiercely back and forth among screens, the cpu usage went up to 200milliwatts, display power did not change.
Audio
I used relax and sleep program to test audio usage.
It can it in various white noise sounds, overlapped (I call it multple channels).
The audio category usage was 375 milliwatts with the program running.
Oddly, nothing changes it. Plugging in headphones, removing headphones, adding channels.
Changing the volume with headphones removed! Does not result in any change in power usage I can see on this program. I suspect something isn’t 100% right there since it's got to use more power to put out more noise.
.
Changing the number of channels played simultaneously on my audio program (“relax and sleep) doesn’t change the usage reported in the audio category, but results in about 15milliwatts increase per channel in the cpu category, tested all the way up to 8 channels running at the same time and added 15 milliwatts for each one.
Wifi – seems to use about 20 milliwatts while on/connected, but not using any internet apps. Jumps up to 300-600 milliwatts periodically even though I didn’t do anything...background programs connecting to internet on their own.
My total energy usage for the past week divided by category:
Display 7.7 kilohoules
Wifi ~ 3.1 kilijoules
Audio ~ 3 kilojoules
CPU ~ 2.7 killojoules
There is a lot more data available. For each application we can see the energy usage, either instantaneous or total over a period we specify. And within each application, we can also see if the consumed power went to LCD or CPU or Wifi or audio etc. Quite detailed and perhaps helps us identify the power-hungry things if they’re not already obvious. Also in the instantaneous display with no programs running that I’ve launched, it’s interesting to see what programs pop up and use the various system resources. If it’s a program that I have installed but hardly ever use, I may uninstall it.
All in all an interesting toy. Brings some insights which may or may not be 100% reliable.
You guys might want to download it and try it out. It’s free. Together maybe we can figure the best strategies for saving our infuse battery life.
First some terms / assumptions related to battery life:
Battery Life-time = 1750mA-hr = 100%
17.5 milliamps corresponds to 1% per hour.
If we want to say it in milliwats, assume averageit voltage ~ 3.8v.
1% per hour corresponds to 17.5 milliAmps = 17.5*3.8 =66 milliwatts
Our battery lifetime (still assuming 3.8V average) is 1.75*3.8=6.5 watt-hours. (in other words, or battery can provide energy to run a hundred watt light for only 60minutes * 6.5/100 = 3.9 minutes..... We should not be surprised at how fast the battery goes down but instead simply amazed at how much our phones can do in this tiny package... I certainly am).
Since there are 3600 seconds in an hour, that 6.5 watt-hours lifetime is the same as 23,400 watt-seconds = 23.4 kJ
The program can break power usage into categories: Display, cpu, wifi, data network, gps, audio. The program is developed for another phone but states it should work for most Android phones. On my Infuse it certainly responds qualitatively in the right direction for all inputs except data network. It consistently states zero power usage for this function regarldess of whether my network is of or on and transferring data. Initially I assumed it must have some problem accessing this particular data on this particular phone and the rest are probably (?) accurate. But some of the data are strange, I’m not sure.
Here are some results
Display:
1% screen - 300mW ~ 4.5%/hr
12% screen - 350mW ~5%/hr (20 hr life running from 100% to 0% batt with no other loads)
30% screen 430mW ~ 6%/hr
100% screen 900mW = 13%/hr (7.5hr life running from 100 to 0 with no other loads)
* Display power usage does not change when I load a black screen or a white screen or a blue screen. Have we been taught wrong all this time? Beats me.
CPU
At the time I was doing real-time monitoring, my cpu tended to use 75milliwatts with static display. If I scrolled fiercely back and forth among screens, the cpu usage went up to 200milliwatts, display power did not change.
Audio
I used relax and sleep program to test audio usage.
It can it in various white noise sounds, overlapped (I call it multple channels).
The audio category usage was 375 milliwatts with the program running.
Oddly, nothing changes it. Plugging in headphones, removing headphones, adding channels.
Changing the volume with headphones removed! Does not result in any change in power usage I can see on this program. I suspect something isn’t 100% right there since it's got to use more power to put out more noise.
.
Changing the number of channels played simultaneously on my audio program (“relax and sleep) doesn’t change the usage reported in the audio category, but results in about 15milliwatts increase per channel in the cpu category, tested all the way up to 8 channels running at the same time and added 15 milliwatts for each one.
Wifi – seems to use about 20 milliwatts while on/connected, but not using any internet apps. Jumps up to 300-600 milliwatts periodically even though I didn’t do anything...background programs connecting to internet on their own.
My total energy usage for the past week divided by category:
Display 7.7 kilohoules
Wifi ~ 3.1 kilijoules
Audio ~ 3 kilojoules
CPU ~ 2.7 killojoules
There is a lot more data available. For each application we can see the energy usage, either instantaneous or total over a period we specify. And within each application, we can also see if the consumed power went to LCD or CPU or Wifi or audio etc. Quite detailed and perhaps helps us identify the power-hungry things if they’re not already obvious. Also in the instantaneous display with no programs running that I’ve launched, it’s interesting to see what programs pop up and use the various system resources. If it’s a program that I have installed but hardly ever use, I may uninstall it.
All in all an interesting toy. Brings some insights which may or may not be 100% reliable.
You guys might want to download it and try it out. It’s free. Together maybe we can figure the best strategies for saving our infuse battery life.