Anybody notice one performing better than the other with regard to battery life? I have the paid version of SetCPU which I bought for my G1 a while back but I'm not sure I really need it for the CPU throttling on the Vibrant but it does seem to improve battery life and I was wondering if Juice Defender might be better.
JuiceDefender was fabulous for battery life but it did slow down my phone a lot. I ended up uninstalling it. If you want to try toggling your data to save battery, I recommend DCSwitch widget. Works like a charm for data control.
I just started using SetCPU on this phone so I can't say how it affects battery.
Tasker is better then Juice Defender. It does everything JD does and MUCH more stuff.....unlimited possibilities
You should probably understand that lowering the frequency of your processor doesn't save a bit of battery. The only way you can save battery is to lower the voltage which neither of those apps can do so whatever you're seeing as "saving battery" is a placebo effect. When the phone is in sleep mode the CPU is defaulted at 245mhz. Lowering the "Max" frequency does just that; it lowers the maximum frequency the processor is able to use.
bmj1086 said:
You should probably understand that lowering the frequency of your processor doesn't save a bit of battery. The only way you can save battery is to lower the voltage which neither of those apps can do so whatever you're seeing as "saving battery" is a placebo effect. When the phone is in sleep mode the CPU is defaulted at 245mhz. Lowering the "Max" frequency does just that; it lowers the maximum frequency the processor is able to use.
Click to expand...
Click to collapse
It's true that reducing voltage (which we can't do) would be more effective as power consumption at a given clock frequency is proportional to the square of the voltage, but you're wrong to dismiss frequency control as a power saving strategy because at any given voltage, power consumption is proportional to frequency. Hence Intel's "speed step" technology used in notebook CPU's.
Juice Defender and SetCPU are doing more or less the same thing as Intel speed step.
We CAN reduce voltages with the right kernel... Dragon TEST 5 kernel.
^ correct with the voltage control app from the market i believe.
Between setCPU and juice defender..
i say juice defender.
Set CPU is a hassle.
It shuts your phone off sometimes with profiles on, and freezes it etc.
If your phone experiences any of those problems, uninstall setcpu and bam, your phone is normal again.
SetCPU is only good for overclocking, and sometimes sucks at that too.
xriderx66 said:
^ correct with the voltage control app from the market i believe.
Between setCPU and juice defender..
i say juice defender.
Set CPU is a hassle.
It shuts your phone off sometimes with profiles on, and freezes it etc.
If your phone experiences any of those problems, uninstall setcpu and bam, your phone is normal again.
SetCPU is only good for overclocking, and sometimes sucks at that too.
Click to expand...
Click to collapse
Yeah, that's what happened to me with setcpu. Every time the screen turned off, the phone turned off as well.
With juice defender, I lost a lot more battery. About 15% in 3 hrs of standby vs. 3-4% in 3 hrs of standby without any power saving apps.
If people are having good results with both apps, could you please show how you set it up? I may have not set it up right.
burntrat said:
With juice defender, I lost a lot more battery. About 15% in 3 hrs of standby vs. 3-4% in 3 hrs of standby without any power saving apps.
If people are having good results with both apps, could you please show how you set it up? I may have not set it up right.
Click to expand...
Click to collapse
Same here.
I figured that people who got positive results from all the battery-saving apps are the ones that use their phones aggressively and switching between all types of connections (wifi/bluetooth/gps). I use my phone normal, so I'm better off without apps like that.
Yeah, that's what happened to me with setcpu. Every time the screen turned off, the phone turned off as well.
Click to expand...
Click to collapse
If I remembered correctly, it's because you've set the frequencies way too low for your phone to function properly.
I'm currently debating buying WidgetLocker, but I'm a bit skeptical since I'm not sure how bad it will lag with my current sleep profile. Can any WidgetLocker users possibly shed some light if they think it will lag on my phone? I'll put my profiles below.
Code:
Temp > 38.6C, Priority: 100, 480 max 160 min, ondemand
Charging/Full, Priority: 75, 710 max 245 min, ondemand
Battery < 25%, Priority: 50 ,480 max 245 min, ondemand
Screen Off, Priority: 25, 245 max 160 min, ondemand
If my sleep profile would make it lag, would bumping it up to 245 max 245 min help any?
Profile recommendations are welcome as well if theres anything that could be done to improve performance without sapping an excessive amount of battery. I have noticed my phone lags waking up sometimes and occasionally just when it's being used.
I would put your sleep profile at either the 2nd or 3rd highest priority - either right after the overtemp profile, or after the "on charger/full" profile.
my profiles usually look something like this (ordered from highest to lowest priority) :
- temperature (can have more than one; highest (over-)temp has highest priority)
- charging/full
- screen off
- battery (can have more than one; lowest battery % has highest priority)
- battery < 101% (this is the sort of like the "off charger, not overtemp, screen on state)
If the "screen off" profile has a higher priority than all the battery conditions, then all the "battery" conditions certainly can only be effective when the screen is on. (Note that the way you have it now, your < 25% battery condition always "wins" over the screen-off profile)
Also, note that your profiles don't contain anything matching the "off charger, 25% < battery < 101%" condition as they stand - what is controlling that, the default profile?
I'm not sure that there are big gains to be had by ratcheting down the min/max cpu frequency for low battery conditions: when you have the display on, it could well be drawing more power than the CPU - so, if your goal is to save battery, it would make sense to have whatever it is you want "computed" to show up on the screen as quickly as possible - so that you can shut the screen off that much quicker.
For the same reason, if most of the time the phone is on you are just "staring at the screen", and nothing is going on in the background, allowing the rate governor to drop to a low frequency saves you only a little bit (if the screen is really the dominant juice-user in that scenario.)
My GSB v1.6 Eris reports using only 2-3 % overnight - it's screen off profile is set to 160/245 - and I seem to be using the "interactive" governor on all profiles ... but I also use Toggle2G with a 10-minute sleep onset delay.
( Probably that means that I don't get email notifications as soon as I could, but that doesn't bother me much - I get them as soon as I unsleep the phone. If you are the type of person that needs to look at their phone every time your facebook updates, then Toggle2G perhaps isn't for you.)
I can't prove it, but so long as you don't peg your CPU minimum frequency too high, and rely on the rate governor to do it's job, most power savings are likely to come from managing the radio state and the screen brightness.
bftb0
What percentages do u recommend bftb0
Sent from my Ginger Tazz using Tapatalk
bftb0 said:
I would put your sleep profile at either the 2nd or 3rd highest priority - either right after the overtemp profile, or after the "on charger/full" profile.
my profiles usually look something like this (ordered from highest to lowest priority) :
- temperature (can have more than one; highest (over-)temp has highest priority)
- charging/full
- screen off
- battery (can have more than one; lowest battery % has highest priority)
- battery < 101% (this is the sort of like the "off charger, not overtemp, screen on state)
If the "screen off" profile has a higher priority than all the battery conditions, then all the "battery" conditions certainly can only be effective when the screen is on. (Note that the way you have it now, your < 25% battery condition always "wins" over the screen-off profile)
Also, note that your profiles don't contain anything matching the "off charger, 25% < battery < 101%" condition as they stand - what is controlling that, the default profile?
I'm not sure that there are big gains to be had by ratcheting down the min/max cpu frequency for low battery conditions: when you have the display on, it could well be drawing more power than the CPU - so, if your goal is to save battery, it would make sense to have whatever it is you want "computed" to show up on the screen as quickly as possible - so that you can shut the screen off that much quicker.
For the same reason, if most of the time the phone is on you are just "staring at the screen", and nothing is going on in the background, allowing the rate governor to drop to a low frequency saves you only a little bit (if the screen is really the dominant juice-user in that scenario.)
My GSB v1.6 Eris reports using only 2-3 % overnight - it's screen off profile is set to 160/245 - and I seem to be using the "interactive" governor on all profiles ... but I also use Toggle2G with a 10-minute sleep onset delay.
( Probably that means that I don't get email notifications as soon as I could, but that doesn't bother me much - I get them as soon as I unsleep the phone. If you are the type of person that needs to look at their phone every time your facebook updates, then Toggle2G perhaps isn't for you.)
I can't prove it, but so long as you don't peg your CPU minimum frequency too high, and rely on the rate governor to do it's job, most power savings are likely to come from managing the radio state and the screen brightness.
bftb0
Click to expand...
Click to collapse
Never a post that isn't enlightening from you.
Yes my default clock profile (710 max 160 min, might bump it up to 245 min) is the one managing the clock rate when the phone is off the charger. As for Toggle2G that might not be a bad idea, I keep my phone on Airplane Mode most of the time I'm at school (horrible signal = battery draining out the wazoo) but it might not be a bad idea to have for when I'm not at school. Of course I'm not sure I would need it though since I have an extended battery, maybe something to experiment with.
As for profiles, something like this?
Code:
Temp > 38.6C, Priority: 100, 480 max 160 min, ondemand
Screen Off, Priority: 75, 245 max 160 min, ondemand
Charging/Full, Priority: 50 ,710 max 245 min, ondemand
Battery < 25%, Priority: 25, 480 max 245 min, ondemand
If you are already using Airplane Mode because of poor signal conditions, Toggle2G won't help much - the battery will drain just looking for voice/1xRTT service (not to mention 3G). I have a pretty good signal where my phone spends most of it's time, including overnight ( -70 dBm to -80 dBm )
ToastPwnz said:
As for profiles, something like this?
Code:
Temp > 38.6C, Priority: 100, 480 max 160 min, ondemand
Screen Off, Priority: 75, 245 max 160 min, ondemand
Charging/Full, Priority: 50 ,710 max 245 min, ondemand
Battery < 25%, Priority: 25, 480 max 245 min, ondemand
Click to expand...
Click to collapse
Seems OK. I have a couple of battery profiles where I slowly drop the max frequency, and make the default condition explicit by using a "battery < 101%" profile, as in:
Pr Min/Max Governor Condition
90 245/480 interactive Temp > 50.0 C *audible alarm
80 245/604 interactive Temp > 45.2 C
70 528/710 interactive Charging
60 160/245 interactive Screen Off
50 160/480 interactive Battery < 15%
40 160/528 interactive Battery < 25%
30 245/604 interactive Battery < 40%
20 480/710 interactive Battery < 67%
10 480/729 interactive Battery < 101%
I'm not sure this really produces much effect in terms of battery savings, though - for instance, it runs counter to my suggestion (above) that it might actually hurt to slow things down with reduced battery reserve because you just end up with the screen on longer.
I'll also mention that for a period of time I was actually convinced that using setCPU was causing random freeze-ups when used with certain combinations of the CFS kernel, governor choice, and min frequency - and so I didn't even have it installed on my phone. I only recently put setCPU back on my phone because I was noticing that CPUboost settings did not seem to be sticking even within the same session (perhaps I was misinterpreting something, I'm not sure). No random freeze-ups lately.
Because some of the things which actually drain the battery - cell radio, screen time on & brightness, 3G activity level, etc., are so highly variable, I'm not sure that any of the reports on this forum (including mine) can be regarded as particularly meaningful. It takes real dedication to the task (creating reproducible signal conditions and compute/activity workloads) to be able to produce quality data about battery usage. Most folks need to use their phones throughout the day (moving continuously from place to place with widely varying signal levels), so they can't set their phones aside to do such things; their usage is not repeatable, so it is hard to believe them, whether they are saying "great battery life", or "battery life sucks!".
Heck, even at night, when my phone isn't "doing anything", (which you would think would be sort of "reproducible") if I move it 8 inches one way or another on the nightstand, my signal level can change by 10 dBm. That means that if I don't use Airplane Mode overnight, it is really not too meaningful to compare "configuration X" to "configuration Y" on successive nights - because the battery usage by the radio is not reproducible.
I guess the bottom line is - if you can get a full day out of the phone, be happy with that; it's just about what the phone was designed to do.
bftb0
bftb0 said:
If you are already using Airplane Mode because of poor signal conditions, Toggle2G won't help much - the battery will drain just looking for voice/1xRTT service (not to mention 3G). I have a pretty good signal where my phone spends most of it's time, including overnight ( -70 dBm to -80 dBm )
Seems OK. I have a couple of battery profiles where I slowly drop the max frequency, and make the default condition explicit by using a "battery < 101%" profile, as in:
Pr Min/Max Governor Condition
90 245/480 interactive Temp > 50.0 C *audible alarm
80 245/604 interactive Temp > 45.2 C
70 528/710 interactive Charging
60 160/245 interactive Screen Off
50 160/480 interactive Battery < 15%
40 160/528 interactive Battery < 25%
30 245/604 interactive Battery < 40%
20 480/710 interactive Battery < 67%
10 480/729 interactive Battery < 101%
I'm not sure this really produces much effect in terms of battery savings, though - for instance, it runs counter to my suggestion (above) that it might actually hurt to slow things down with reduced battery reserve because you just end up with the screen on longer.
I'll also mention that for a period of time I was actually convinced that using setCPU was causing random freeze-ups when used with certain combinations of the CFS kernel, governor choice, and min frequency - and so I didn't even have it installed on my phone. I only recently put setCPU back on my phone because I was noticing that CPUboost settings did not seem to be sticking even within the same session (perhaps I was misinterpreting something, I'm not sure). No random freeze-ups lately.
Because some of the things which actually drain the battery - cell radio, screen time on & brightness, 3G activity level, etc., are so highly variable, I'm not sure that any of the reports on this forum (including mine) can be regarded as particularly meaningful. It takes real dedication to the task (creating reproducible signal conditions and compute/activity workloads) to be able to produce quality data about battery usage. Most folks need to use their phones throughout the day (moving continuously from place to place with widely varying signal levels), so they can't set their phones aside to do such things; their usage is not repeatable, so it is hard to believe them, whether they are saying "great battery life", or "battery life sucks!".
Heck, even at night, when my phone isn't "doing anything", (which you would think would be sort of "reproducible") if I move it 8 inches one way or another on the nightstand, my signal level can change by 10 dBm. That means that if I don't use Airplane Mode overnight, it is really not too meaningful to compare "configuration X" to "configuration Y" on successive nights - because the battery usage by the radio is not reproducible.
I guess the bottom line is - if you can get a full day out of the phone, be happy with that; it's just about what the phone was designed to do.
bftb0
Click to expand...
Click to collapse
Hm, I might try those profiles with a few tweaks, or I might just end up bumping up my screen off profile in terms of priority, not sure yet.
Another question, what do you suggest as far as governors go? I seem to have the most success with ondemand, but I've never tried interactive. I learned the hard way that smartass lags my phone horribly, not sure why.
With my extended battery, I've gone roughly 65 hours total uptime with about 45% battery left with moderate use. I guess I should be proud of that.
ToastPwnz said:
I'm currently debating buying WidgetLocker, but I'm a bit skeptical since I'm not sure how bad it will lag with my current sleep profile. Can any WidgetLocker users possibly shed some light if they think it will lag on my phone? I'll put my profiles below.
Code:
Temp > 38.6C, Priority: 100, 480 max 160 min, ondemand
Charging/Full, Priority: 75, 710 max 245 min, ondemand
Battery < 25%, Priority: 50 ,480 max 245 min, ondemand
Screen Off, Priority: 25, 245 max 160 min, ondemand
If my sleep profile would make it lag, would bumping it up to 245 max 245 min help any?
Profile recommendations are welcome as well if theres anything that could be done to improve performance without sapping an excessive amount of battery. I have noticed my phone lags waking up sometimes and occasionally just when it's being used.
Click to expand...
Click to collapse
Wow you actually set a max of 480 for merely being over 38.6C??? I think that's a bit extreme. Other than that your profiles aren't that different than what I use (using Conap's CPUBoost tool, what are you using?).
I love WidgetLocker, and yes it's one of the only two apps I've actually purchased. Lag or not I wouldn't trade it.
The only "issue" (not necessarily WL's fault) is that using the Theme Chooser under GB ROMs, at least with the Speedometer Battery theme, definitely causes WL to lose what widgets are on it between reboots, although not always every reboot, but most of them.
It's confirmed that the solution is to not use Theme Chooser, at least with the Speedometer Battery theme. I don't know for sure if other themes cause it as well, but it's easily reversed (go back to the stock theme) if it does.
roirraW "edor" ehT said:
Wow you actually set a max of 480 for merely being over 38.6C??? I think that's a bit extreme. Other than that your profiles aren't that different than what I use (using Conap's CPUBoost tool, what are you using?).
I love WidgetLocker, and yes it's one of the only two apps I've actually purchased. Lag or not I wouldn't trade it.
The only "issue" (not necessarily WL's fault) is that using the Theme Chooser under GB ROMs, at least with the Speedometer Battery theme, definitely causes WL to lose what widgets are on it between reboots, although not always every reboot, but most of them.
It's confirmed that the solution is to not use Theme Chooser, at least with the Speedometer Battery theme. I don't know for sure if other themes cause it as well, but it's easily reversed (go back to the stock theme) if it does.
Click to expand...
Click to collapse
I'm still using SetCPU.
Yes I do knock it down that far, my phone heats up like crazy while I'm at school (terrible signal) and once it reaches 100F I could use the phone as a hand warmer and if I take off the back I can smell burning plastic. Maybe thats normal for the Eris, but I use it just as a precautionary measure.
ToastPwnz said:
Another question, what do you suggest as far as governors go? I seem to have the most success with ondemand, but I've never tried interactive. I learned the hard way that smartass lags my phone horribly, not sure why.
Click to expand...
Click to collapse
I haven't spent much time trying to figure out the "best" governor - I tend to use either "ondemand" or "interactive" without giving it too much thought. I tried smartass for a while, and remember being not happy with it for some reason or another; but I don't remember what those reasons were. ( It might have been nothing more than "choppiness" of scrolling behaviors. )
bftb0 said:
I haven't spent much time trying to figure out the "best" governor - I tend to use either "ondemand" or "interactive" without giving it too much thought. I tried smartass for a while, and remember being not happy with it for some reason or another; but I don't remember what those reasons were. ( It might have been nothing more than "choppiness" of scrolling behaviors. )
Click to expand...
Click to collapse
Now my curiosity is getting the best of me. Whats the difference between ondemand and interactive? (Sorry for all my questions <.<)
ToastPwnz said:
Now my curiosity is getting the best of me. Whats the difference between ondemand and interactive? (Sorry for all my questions <.<)
Click to expand...
Click to collapse
I don't know for sure...
... but I can tell you that the exact answer to your question is right here.
bftb0 said:
I don't know for sure...
... but I can tell you that the exact answer to your question is right here.
Click to expand...
Click to collapse
Hidden within that garbled mess.
I kid, I think I found the explanation you had in mind, and from what I can find the only major difference is that interactive is more aggressive then ondemand. I'm not sure if I should take that governor for a test run or not, I do notice that my phone lags when waking up occasionally. It kind of sounds like switching to interactive might help that.
ToastPwnz said:
I'm still using SetCPU.
Yes I do knock it down that far, my phone heats up like crazy while I'm at school (terrible signal) and once it reaches 100F I could use the phone as a hand warmer and if I take off the back I can smell burning plastic. Maybe thats normal for the Eris, but I use it just as a precautionary measure.
Click to expand...
Click to collapse
You probably have your reasons, but when you're at school, why don't you put it in Airplane mode and just take it off it between classes to check emails and such? Save your battery and the heat of your Eris a bunch.
Or do you and you just use the temperature profile as a fallback in case you forget to put it back in Airplane mode?
roirraW "edor" ehT said:
You probably have your reasons, but when you're at school, why don't you put it in Airplane mode and just take it off it between classes to check emails and such? Save your battery and the heat of your Eris a bunch.
Or do you and you just use the temperature profile as a fallback in case you forget to put it back in Airplane mode?
Click to expand...
Click to collapse
You are quite correct sir. That and my I bump my phone occasionally, and I think two days ago it turned on and I didn't notice, and when I checked it I noticed over the course of about 2-3 hours it lost 14% and was at around 95ish F.
I do keep it in Airplane Mode when I'm not using it (and it's off as well), but the profile is pretty much just a precautionary measure.
Does anyone underclock? I only do it in certain situations. Do you see it as beneficial? Why or why not?
Sent from my Galaxy S2
EDIT: My settings are attached as a screenshot. My device is underclocked 50% of the time.
In my opinion and from personal use, I don't find underclocking to really be beneficial. I never really saw any better battery life. On the same note, I never really found overclocking and undervolting to be extremely beneficial either. It's like the gains aren't worth the time to tweak everything.
Undervolting, yes. Limiting clock speeds, and setting governer to conservative, yes. I say 1ghz is the best spot without losing much performance, and I get 2 days of battery life (sometimes) compared to the one day before... Stock kernel, you aren't going to see much of a difference because you can't undervolt...
I keep my phone at 800max 200min on conservative with a 85% up and 20% down threshold. I don't even see the phone slow down at 800 plus it saves some battery!
I force underclock when I'm in a situation where I KNOW the clock should never ramp up (Screen-off I set to 500) as a "safety measure" to prevent surprises.
I also may force underclock to 800 when running Navigation once my vehicle dock comes in - since we can't crank our charge current up.
Entropy512 said:
I force underclock when I'm in a situation where I KNOW the clock should never ramp up (Screen-off I set to 500) as a "safety measure" to prevent surprises.
I also may force underclock to 800 when running Navigation once my vehicle dock comes in - since we can't crank our charge current up.
Click to expand...
Click to collapse
My screen off is set to this also (200-500, conservative). You can use Tasker to change frequencies for specific apps automatically.
I've install Rom Toolbox,
and i saw there is a "CPU slider" where i control the clock speed.
i've put it to 1000MHz instead of 1200MHz and tested it for several days
i really dont feel any difference in performance.
browsing seems same, games like asphalt is equally smooth.
heating is similar, equally warm.
the only difference is quadrant benchmark.
1200MHz scores 3200-3400
1000MHz scores 2600-2900
frankly speaking, i'm not sure if there's any difference in battery life.
is there any way to accurately test whether the clock speed affects the battery life?
i've seen other threads, where there are very different opinions.
some say it will improve battery life, and some say its worst.
http://forum.xda-developers.com/showthread.php?t=726019
Quote: (SetCPU doesn't make a difference in battery life, it can only shorten it. The kernal already has the best settings for CPU speed built in.)
http://forum.xda-developers.com/showthread.php?t=1305465
Quote: (if you are able to stand the side effects of underclocking, it will surely boost your batery life.)
On my SGS2 program called CpuSpy shows that 1200MHz is about 1% of total cpu time (remember that governor is ondemand and CPU is at 1200 only when need it). If power consumption is directly proportional to clock speed by limiting it to 1000MHz you will get about 20% less power usage by 1% of time... looks like 0.2% power saved ? Soo if Your phone works for about 48h on one charging this way You can get about 6 extra minutes. It's just my guess...
Also have to consider if slower cpu causes screen to eat power for longer time... (because You have to wait longer for operation to complete)
slig said:
On my SGS2 program called CpuSpy shows that 1200MHz is about 1% of total cpu time (remember that governor is ondemand and CPU is at 1200 only when need it).
If power consumption is directly proportional to clock speed by limiting it to 1000MHz you will get about 20% less power usage by 1% of time... looks like 0.2% power saved ? Soo if Your phone works for about 48h on one charging this way You can get about 6 extra minutes. It's just my guess...
Also have to consider if slower cpu causes screen to eat power for longer time... (because You have to wait longer for operation to complete)
Click to expand...
Click to collapse
HI, thanks for replying. I understand what you mean. the phone dont operate at 1200MHz all the time. but when using browser, and playing games, such as asphalt, it runs at max CPU usage almost the entire gaming duration.
Anyway.....
the real question is whether the clock speed is directly proportional to the battery consumption.
while reading your post, i thought of a brilliant ideal how to verify this.
the CPU slider not only allows you to set the max CPU speed,
you can set the min CPU speed as well.
So, i thought of an experiment, lets set the min & max CPU to 1200MHz,
this way, the phone will be running constantly at max CPU even when its idle.
let the phone be turn on till it run out of battery, record the time, T1.
then repeat again with max and min CPU set to 1000MHz.
record the time it is turn on till it run out of battery, record time as T2,
then compare T1 & T2, this could certainly work.
it would be nice if any member here happens to have 2 sgs2, and tried them ;-)
There are two more things to consider
1. CPU is not the only element that consumes power.
2. SGS2's Exynos is always clocked at 200MHz when the screen is off - check if this minimum slider affects that too.
Please let know how your experiment goes.
Regards
when the screen is off, the phone will be in "deep sleep" state. i think thats less than 200MHz.
anyway, i wont be doin this experiment any time soon.
you see, this is my only phone, i need to use it.
i dont have much oportunity to leave it and wait for it to run out of juice.
still, i'll try it when i have the chance.