Have a look at this thread : [Q] "Android OS" vs. "Android System". WTH is "Android OS"? - xda-developers

There are already plenty of battery threads on this forum... you should have done a search first and asked your question there!

Cheers

 

I haven't actually confirmed this, but my theory is such:

Android system is basically the battery usage that the ACTUAL OS uses, that meaning CPU, system processes, ROM reading, etc.

Android OS is basically the battery usage by PROGRAMS run by the OS. That means basically battery that has been drained from the RAM of the phone by keeping programs running.

As the Android OS is so high up, I can only guess you are not using a task-killer app, and this is also resulting in a low battery life cycle? Say your phone only lasts maybe less than 20 hours?

Read my post in the Battery Thread about the experiement I am currently doing:

 

Euhm task killers usually don't have anything to do with extending the battery life. Closing apps might even decrease the battery life... so please don't give any wrong information!

 

Right. I'm doing a experiement right now, and if you read my posts carefully, you will see what I'm writing is correct. If you have many applications open (i.e. haven't killed them), they will pull a lot of power in the RAM.

It's backed up by theory and by what I see when I'm using my phone without a task-killer app these days.

Can you please give any theory/links/experiement results to prove otherwise?

If not, please refrain from calling me a liar. I'm currently studying the field of electrical engineering, so I kinda know what I'm talking about.

PS! I do admit a BAD task-killer app can reduce battery life. But read my post, it says it's like a car. Chose the wrong car and your gasoline usage will be sky high. Chose a good car, gasoline consumption will be down ^^ The task-killer app I refer to is used by many, and is recommended many places.

 

First of all I didn't call you a liar!!!

Second of all this topic has been discussed plenty of times by android devs. Here is one example : AndroidSPIN Why you don’t need a task killer app with Android.

I might not be an "electrical engineer" but I've been a software developer for almost 10 years now. (Although not for smartphones, but I do have some knowledge about it)

If certain apps are draining your battery (when they're open in the background) it's either because they're a bit dodgy (like sending information when they shouldn't) or it's just shitty programming. So the use of the RAM is totally different here then it is on regular pc's so you can't compare those at all.

Sorry if I offended you, I didn't intended to.

cheers

 

No offense taken. I just misunderstood the part about "giving wrong information".

I see the validity of what you are writing from a programmers perspective, and I do admit, for efficiency and performance a task-killer app is maybe not the best. However, for battery life, it can make a difference.

Although I see what you claim about the programming structure, it is not about how the program behaves with the phone. It is the physical way that RAM works. RAM consists of thousands of small latches (a system made by electrical circuits). Still if the program doesn't do anything, if information from the program is stored in the RAM (as stated in many articles on this topic, the program needs to write to the RAM to keep the "last state of the program"), the RAM constantly needs to pull power to keep this memory there. Thus, the more memory in the RAM that is occupied, the more power the RAM needs to pull from the battery. ROM is different however, as the "latches" are magnetized to read the information.

Just to point out, these articles go into depth about how the programs are and how they behave. But do they consider what is best for the actual user of the phone, and take into account battery life?

Now, all these quotes are from comments on the article you linked.

And a good task-killer app, like the one by ReChild is programmed to ignore these system processes that run in the background and that should stay up. Thus, it does not drain unnecessary battery power

PS! Just to clarify, the RAM does not only require power to change itself (i.e. when receiving information from a program), but it requires power to keep its memory state. This is true for all RAM, as it is Random Access Memory. Thus, if something can be written "permanently" to RAM, a reboot would not automatically wipe it. Therefore RAM (per definition) works in a way of "temporarily storing" the information. To be sure these "latches" are temporary, they will always need to pull power to keep the information, so that when a power cut occurs, the RAM will automatically be wiped

If RAM overloads (like bluescreen of death, or a black screen followed by a reboot, etc.), it is important for the circuits not to be damaged, that the information can be quickly wiped, thus a "power cut" that dumps the RAM

 

I know I've read the comments But in the end they're just user opinions (just like my opinion is different) and not the devs.
All i can say is that I don't use an ATK and only use software from trustworthy devs and my battery seems to work longer then most people here claim... I get about 40hrs out of 1 charge and I'm a moderate user. I do turn off all my network connections when I don't need them though.

I guess in the end people just need to try and see what works best for them...

Cheers

 

My normal usage, I can get 50 hours out of my phone. I usually charge it at the 44-45 hour mark, so as to not drain it completely, but then it is still 15-20% battery left, AND I run with some form of data communication constantly on (mostly 3g/H). I use the phone moderately, and enjoy all the different functions the phone has, whenever I want to .

I see your point about it being just a user opinion, but when 95% of the comments I could bother read, stated that a ATK always increased their battery life, I can't help but wonder if there might be a good basis of truth to it

EDIT: Appologize, while writing this post my mind was on other things as well. My battery usually lasts for about 40 hours, and I charge it around the 34 hour mark with at least 15% juice left , thus reconing it could last around 40 hours =)

 

Unfortunately many people don't understand how Android works. This ad populi argument is made time and time again. It means nothing that many people use it. They are wrong.

Your car analogy is a false one rendering your argument fallacious.

Full RAM in Android is a good thing. Apps stored in RAM do not use CPU cycles and do not drain the battery. Only running apps use the battery.

Task Killers will show apps stored in RAM that are not using your battery life. Force closing apps with a Task Killer results in Android automatically opening some of them again - resulting in battery drain.

The constant force-closing of apps not only sucks battery life but it also compromises the smooth running of Android OS.

Task Killers are totally redundant on devices running 2.1 or higher.

Let Android do what it was meant to do.

 

All I can say is read the WHOLE thread again. You seem to be missing the point!

And how is my car analogy wrong? Can you please specify?

And *sigh* yet again, I'm an electrical engineer student. I THINK I might have a pretty good idea about how RAM works. It does drain battery, and the fuller the RAM the more battery is drained. Read my second last post again please, then maybe you will understand how ELECTRONICS works (not programming, programs or the OS).

This quote is quite interesting. Can you please refer to where you have this information from? My ATK always shows all my active processes and those in RAM. And btw, when information is in the RAM, it is still active by definition of RAM, and the same definition of RAM states that it must be pulling power. Again, read my second last thread, but in case you can't find it here (as you seem to have missed a lot of what's been said earlier), here is a link:

I have already stated, many times in these forums, that I admit, for performance and how Android would want to run, a task-killer might not be benefitial. However, we are talking battery life in this thread... ^^

PS! The ATK I have referred to is programmed to ignore standard processes that Android wants to keep up. If you have any costum systems, it is also very easy to add these to an ignore list, and they will no longer be "shut" when the ATK is run.

 

Sorry but I've been thinking about this all day and I couldn't let it go... Since you're studying electrical engineering could you tell me how much difference there is in power consumption for a total empty ram and a total full ram (I am talking about android-smartphones here, not pc's!) Because I've been doing a lot of research and pretty much all the sites say there is no difference or an insignificant difference.

Here is another good thread that discusses the need for (or not needing of) Task Managers.

Cheers

 

Hi there.

Your question is really really specific. There are so many RAM types out there, and they all require different set ups and power usage. I don't know what type of RAM is in the phone.

But, all RAM works essentially like a flip-flop in digital electronics.

Here is a link for flip-flops, and here is a link for digital electronic.

Basically, RAM stores information by "charging" on end of a latch with electricity. When information is accessed from the RAM, the "reader" sends an electrical signal through the allocated latches (the latches that the program has asked to be read). If one latch is "charged", it will let a signal go through it, and this will then act as a binary digit (depending on the RAM type either a 0 or 1).

Now this "charging" of the one end of the latch is what constanly pulls power. This means that the more information that is stored in the RAM, the more latches need to be powered to get the different binary digits.

RAM is constructed like this for various reason, but the most important is that, since programs write to it so fast, IF the RAM should be overloaded, the RAM must be able to dump all its memory fast so as to not get a oversurge of electricity, which can potentially damage the RAM memory. By simply "pulling the plug" (aka. dumping the phone's power), all the RAM will be cleared.

So as to your question about the difference in power consumption between a empty and full ram, when it comes to a difference of 700 MB, this can equate to quite a bit, as one latch is only 1 bit. And there are 8 bits in 1 byte. and you now have a difference of 700 000 000 bytes. So even if each latch only pulls a minor amount of power (again, dependent on specifications), it multiplies to quite a bit.

Hope this explanation helps

 

I know how ram works, but there is a big difference in architecture between regular pc's and smartphones... and since you keep preaching that full ram uses so much more power I want you to back up your theorie with some numbers since a lot of people say it doesn't use more or insignificantly more power.

 

Well, as I wrote, it is dependent on the type of RAM, and how efficient/powerfull it is. High speed RAMs will use more power, and so on.

I thought that the numbers I gave would give you an idea. If (say only using 500 MB) 500 000 000 * 8 latches are in use, that is 4 000 000 000 latches pulling power. Even if it is not a PC, or smartphone or whatever it is, RAM still works the same way. If each latch pulls 5 nanowatts (which I'd say is plausible) that is 20 watts for 500 MB. 20 W for only the RAM alone, each hour, on top of all other processes and things used. You figure the difference then, on such a "small" battery the DHD has

 

20 Watt ram alone??? So you're saying that my laptop which has 4gb of ram and a 1gb of video memory, 15inch screen and a big cpu to feed can actually keep running from a 90watt charger??

Kingston claims that their 2GB modules only uses a maximum of 1.9W and then again we're talking about pc ram which is totally different.

So you're saying i'm pretty much using a constant amount of 20watts (because I never close my apps) and in the end you only gain a few hours since I get 40+ hours out of a charge...

 

I did some calculating... 20watts for the ram running at about 1.8V (that's the average for pc's you can look that up, so probably even less for smartphones) That means you're saying it uses more than 10Amps (watts/Volts)... Although the battery of the DHD is only 1230mAh which means it could only run 1.23Amps for an hour. How do you explain this??

 

Look. I said, it was a specific question, and I don't know the details. I tried looking up how much one latch uses, but couldn't seem to find that information currently. So I said a number that was plausible. But I guess for smaller things (like laptops and smartphones) it is much smaller when I think about it.

But that is not the point, the point is that RAM does pull power, and as I said, with such a small battery as the DHD, just small amount of power drainage constantly, does take quite a bit of power from the battery overall. Thus having many apps running on Android does take battery, either you are willing to accept this or not. I'm not gonna discuss this any further, because unless the Technical University of Denmark is wrongly informing their students, I know that RAM pulls power constantly, thus by not shutting programs off they will use power, no matter if the program is "inactive" in the CPU or not

PS! You also said you shut off communications when you don't use them. I have communcations on all the time, and I still get same number of hours more. And it's a known fact that the data communications of the phone take up most battery. Thus, end of discussion Thank you for your time

 

Hi first of all, I just registered to reply to you!

First of all, why do you mention always that you are an engineer student? That means actually nothing....!!!
I am actually an engineer (ECU programmer/ Dipl. Ing. and not only a student...)and yes I do know as well how a RAM works but I am not an mobile-phone specialist. However, see it from the point that the 'flag' setting/holding needs less power as to open and to use a task killer!

"CPU power overshadows RAM by
a factor of two or more. Audio displayed a largely static
power consumption in the range of 28

 

Why You Shouldn’t Be Using a Task Killer with Android

http://forum.xda-developers.com/showthread.php?p=6426198#post6426198

 

Hi there. I must say I'm honored that you signed up just for me . However, none of my posts were meant in a negative way, so my appologize if it was taken in another way, tones can be difficult to convey online.

However, I would advice you to read the full paper you linked to. There are a few points that catch my eye. The phone they tested was first of all not a commercially used phone. Thus it cannot apply to this phone under this thread, as the hardwares may function very differently (it is the electonics we are discussing now, NOT the OS). Secondly, the device under test had "only" 128 MB of RAM. Thirdly the paper does state this:

Now please note that the specific phone under this thread has over 700 MB of RAM, meaning the power needed to use this RAM is much more, significantly more. Meaning its outcome will be different.

Also, please understand this: I'm not saying Android enjoys apps get shut, and I'm not saying RAM will pull the most power. I know CPU or the Mobile Communications highly likely will. However, those are parts/systems that need to run for the phone to function. Thus we are discussing here how to optimize battery life in another way. My suggestion is using a task-killer app. If this will prolong the battery life significantly for everyone, I do not know. All I know (and as I have said in other battery threads the latest days) is that I'm conducting an experiement about using, or not using a task-killer app. Now on my third day of trying without a task-killer app (meaning no reboot in these days), I have unplugged my phone for almost 4 hours, and 12% battery is already used, with a screen "on-time" of less than 15 min. Android OS has used over 50% of the battery used, and this is just from having normal programs running, like facebook, messages, calendar, google search, browser, gmail, camera, Spotify, a few widgets, etc., basically all the common programs that most people would use. Normally, when I did use a task-killer app before, I would get basically 50 hours out of the phone with ease, still having data communications on constantly and using the phone whenever I wanted to.

Now the only difference from my normal usage in this experiement, is that I'm not using a task-killer app to clear my memory ever 4-6 hours.

So this tells me that RAM in the DHD is quite power hungry, compared to how much this "little" battery can deliver

 

Ok, sorry I guess I misunderstood your tone.

However I have no idea why a task killer should add anything to your battery life and I am be honest you are the first person who claims to get increased battery life from a task killer (I am talking here about the well known one, i.e. xda, android-hilfe or android central).

Best Wishes and a nice weekend,
Tiago

 

No problem my friend =). If I'm angry, its whole sentences in caps usually =P.

Yea, I see what you mean, however, this is my experience with the phone I am having right now .

And a nice weekend to you to sir.

If you wonder, I will be posting a comprehensive guide here about how I use my phone and what I do to keep battery up and stuff... so check in soon again

PS! There are some quoted comments from article in one of my earlier posts on this thread . They claim the same about ATKs

 

Sorry but for an electrical engineer saying 20W on a 1230mAh is funny to me. Because that number isn't plausible at all... and I haven't had electricity courses for more then 10 years now. You claim a lot of stuff here but can never actually back anything up with evidence.

I actually just realize that you wrote 50 hours with some sort of data-communication constantly on... I personally call that bullshit, or otherwise you're saying that google is lying about the architecture of their OS (which wouldn't make any sense!)

So please make a guide for us to use ATK in a good way and even I will test it (and I advice others to test it too) but I can already tell you now that NOONE
will get 50 hours with data-network constantly on. As far as I know noone has ever got that much out of 1 charge (not even close) with moderate use and data on.

Anyway so if your guide is online and we find 3 people here who get the same results I'll keep my mouth shut and apologize, till then I call it BULLSHIT.

Cheers