Comprehensive Battery Guide

[El Presidente]

Found this whilst perusing some of the links in this thread. hope it helps -

Do and don’t battery table

 

[bograma]

Just finished conducting another test.
Seems like even more juice can be saved (eventhough I didn't think so, as I have deactivated "always-on" communications), by shutting down the data communications when not needed.

That's really obvious. Disabling auto-sync saves a lot of battery life. And I mean a lot! Whenever I get down to 20%, I disable it and this keeps my phone from going empty in 15 minutes. You can still sync manually whatever you need.

 

[Gardium]

Hi. Just to clear off with CoolIce.

I do not disagree with most of what you have written here. However, I still can't see nor find the post where you claim I have mentioned battery memory effect in the extent you say. Yes, I have said we are teaching the battery how much charge/electrons/effect it can take up, and this has something to do with the fundamental chemistry of electrolysis. An alternating electrolysis system (which is basically what any battery is), will have a certain "memory" of how much electrons that can transfer between its components, before the reaction will slow down due to the saturation in the cell.

Also, the way a battery will "feedback" to the electronics and say "I'm full", is to send back a feedback signal that tells the rate of reaction in the cell (the electrolysis system). When this drops below a certain point, the phone (or electronics) will assume the battery is full, as the cells of the electrolysis will be very saturated. Due to constant discharge from any battery (yes, any battery will have a natural discharge, both as safety functions, but also because of the way an electrolysis is set up), the charging (the reaction in the cell) will also almost be "constant" when the battery is "full", but at a very low rate. Thus the phone will say the battery is fully charged, when the battery infact maybe only is 90-95% full, as the cells have become very saturated and are taking in electrons very slowly. Also, the amount of electrons, is also what decides how much current the battery can make, and thus also how much mAh the battery can produce. Thus by prolonging the first charges, the battery will learn that it can take in more electrons from the reaction before slowing down, increasing its cycle life. Even though the battery has been indexed to 1230 mAh, it does not mean it always will perform this, and this is also the "theoretical" value calculated by the phone manufacturer. There will always be slight variations, and also if the battery is not treated well, it might not take in the full 1230 mAh. So basically I am not saying that we are magically increasing the 1230 mAh, but we are making sure the full effect is being loaded into the batteries.

As for the NiMH battery, I see what you mean about the conditioning, as this battery suffers from the "dead" memory effect. However, the cells perform more optimally, if there have been few fully discharges. So about once a month, a NiMH battery should be completely drained to keep it really optimal.

As for the rest CoolIce writes, I completely agree, or agree to a large extent . All batteries will take some damage from a total discharge, but newer batteries have internal circuit protections against this happening. Also, a battery should NEVER be charged with a voltage/current/effect higher than what it has been regulated for.

As for the number of cells in various batteries (including the DHD), I will not try to speculate in. That is for others to determine . I only care about the battery working, and that it works well xD

Cheers

 

[Gardium]

That's really obvious. Disabling auto-sync saves a lot of battery life. And I mean a lot! Whenever I get down to 20%, I disable it and this keeps my phone from going empty in 15 minutes. You can still sync manually whatever you need.

I know some things have been obvious for some of us that really know this stuff, but for some people here it is not really always that obvious, and I am just conducting the tests and trying to help those people. Also there are many suggestions and other threads out there talking about things to do to improve battery cycle life, so I am just conveying what I read other places/see of suggestions, and what I have tried and their outcome here

 

[EarlyMon]

Second, Li-ions take no damage of being completely drained (at least not noticeable for a mobile phone, that probably will have its battery changed out after 12-18 months). That is the "middle" generation Nickel-Metal Hydride (NiMH) battery that will take damage of that (the kind of battery used in high performance units that require a lot of power fast, but for a short time).

Caution: When your phone is discharging rapidly, in event of heavy load, it will tend to run warm - in some cases, quite warm. Running to full discharge while very warm will deform structures in the battery and significantly reduce its shelf life. It also causes the battery itself to run warmer if that happens too often - a vicious cycle.

You are misunderstanding the term "overcharge". We are referring to "overcharging" the cell, as in forcing it to take in extra electrons to have more current to provide. The charging voltage is still the same. I have not told anyone to overload the charging circuit. This is potentially VERY dangerous, and OFC I am very aware of that.

I've read this and your other posts on electro-chemical reactions within a battery and would like to offer a few things.

Once your battery is marked as fully charged by the system, it stops charging, and if your phone is on, it begins running off of the battery and depending upon where the limit is set (anywhere from 92~95% on HTC phones), will run down to that point before the charge cycle begins again. It never trickle charges, it never runs from the charging current - that would cause overcharging and fire, even when subjected to the correct charging voltage.

The phone will indicate a green-light full charge the whole time it's riding that cycle from 95%~100%.

So - depending upon luck, when you pull a phone off of the charger, it may be at 93, 95, or 100% - but it will always show 100% immediately. After a short time, the actual power available will show, and this often gives rise to the false presumption that the phone, its configuration, battery or charger isn't doing its job.

You seem to enjoy experimentation, I've not seen this posted around here, so please try this and report your findings:

Collective experience shows that sometimes, not always but often, the charging circuit may not be calibrated out of the box. Others have found that this procedure will calibrate:

1. with your phone on, charge while observing.
2. when it hits green, immediately remove from charger and power off.
3. plug back into charger. if not immediately green - your charging circuit is not calibrated. in such a case, charge until green, immediately remove from charger and power on.
4. plug back into charger. if not immediately green - repeat this process until off or on, if indicating full charge in one state, it immediately registers full charge in the other state.

I own 2 HTC phones, one took 15 minutes of this sort of thing, the other more than a half hour, but in both cases the outcome was beneficial.

Apologies if this forum has such a cal procedure, our site search is lousy (hence our Zero Tolerance Policy to not flame newbies for not searching), and I didn't see it - so I thought perhaps better to post twice than not at all.

I did note in your OP that you've heard of the "battery boost" stuff by

...that sounds like "charge the battery 8 hours, turn off, then charge one more hour, use it for 3-4 hours, charge it 8 hours again, then turn off and charge one more hour, then battery should last longer". However, I have seen posts that claim this only works for a certain period of time.

I very much suspect that procedure is actually attempting the calibration, but is not really well-considered (because leaving on charger for an hour past full charge indicated is likely to do nothing whatsoever other than give you a false reading on your actual charge level), so you may try this one and see if it matters.

~~~~

Android apps have a mechanism called the intent. This mechanism is used to set apps to start and then start other apps.

Once intents are widespread, the more you use a task killer, the more you need to use a task killer. Please see my post in an adjacent thread on a cure for using a task killer.

The controversy over task killers comes from the individual users use cases and how they vary - this leads to different intents being set, and leads to different behaviors from the same app for different users. So - all a task killer really accomplishes is chasing the symptoms.

Android is a Linux system with a thing called the Dalvik engine - with few exceptions, all apps are running within the Dalvik engine process. Android apps are compact because they rely on standard, typical Linux services below the Dalvik engine to do the actual heavy lifting. Killing an app only performs the kill within the Dalvik - the services are still there, ready to go work. So - a task killer is by no means equivalent to the unix "kill -9 proc_id" and that's yet another reason experienced Androiders hate task killers.

If the reset procedure I recommend in the parallel thread doesn't put things back into order because you have a bad app just wanting to run and setting intents for a lot of other apps, your superior alternative is to root your phone, and then use a root-only app called Titanium Backup to freeze the apps in question or those you know you'll never use. This will effectively disallow them from executing and they'll cease to bother you. Replace the offenders with better apps - they're out there.

If you want to see your system logs, use aLogcat, free in the Market. It will produce a endlessly-scrolling, very boring, but very detailed output of basically everything you could care about running under the hood. Use the menu key, pause it, and scroll back - entries in red with an initial E (for error) marking may be considered as areas to investigate for misbehaving apps or processes. (If you see one for LEDs not providing a proper termination, ignore that, it's a known HTC defect and is quite harmless.)

Hope some of this info is helpful in some way.

 

[Gardium]

Hi there EarlyMon.

Thanks for your input. I guess I can say I'm honored a moderator would join the thread

Caution: When your phone is discharging rapidly, in event of heavy load, it will tend to run warm - in some cases, quite warm. Running to full discharge while very warm will deform structures in the battery and significantly reduce its shelf life. It also causes the battery itself to run warmer if that happens too often - a vicious cycle.

This I am quite aware of. Rapid discharges will cause overheating. Any electrochemical reaction will of course produce heat. The amount of heat energy will be the same over time for the same amount of battery use. Thus, by rapidly discharging the phone (aka. using the phone heavily as to use battery power fast) will produce a lot of heat fast, and the phone's body cannot disappate this heat fast enough. Also, by completely discharing that fast, the internal structure of the electrolysis will be ruined, and thus the battery cell will not be able to function as it should any more (basically, the heat formed has "burned" the electrolysis component, rendering parts of its composition inoperable as a battery, thus the battery will not be able to charge, and is basically completely ruined). If this happens partially only, then the battery will cause more heat than under normal operation, as the battery will "struggle" to keep up its power delivery. Look upon the battery as a small mechanical motor. If the motor is running fine and only pulling the amount it is meant to, it won't overheat. If you put some strain to it for a short amount of time, it will heat a lot, but hopefully make it. However, prolong this strain, and the engine will in the end malfunction due to overheating. If the engine is not repaired for this damage (aka, this is the partial burn out of the battery), the engine will not function optimally, and overheat much more easily.

I've read this and your other posts on electro-chemical reactions within a battery and would like to offer a few things.

Once your battery is marked as fully charged by the system, it stops charging, and if your phone is on, it begins running off of the battery and depending upon where the limit is set (anywhere from 92~95% on HTC phones), will run down to that point before the charge cycle begins again. It never trickle charges, it never runs from the charging current - that would cause overcharging and fire, even when subjected to the correct charging voltage.

The phone will indicate a green-light full charge the whole time it's riding that cycle from 95%~100%.

So - depending upon luck, when you pull a phone off of the charger, it may be at 93, 95, or 100% - but it will always show 100% immediately. After a short time, the actual power available will show, and this often gives rise to the false presumption that the phone, its configuration, battery or charger isn't doing its job.

I see what you have written here, and I agree, after the batteries initial conditioning. The phone does not trickle charge or run of the current coming in, it will always run off the battery. However, during the initial conditioning, the phone will think it is fully charged, but the battery will not refuse to take in more charge before the battery itself is completely saturated. It is only when the battery itself refuses to take in charge, that the charger circuit is flipped off until it reaches the "point of recharge". The phone will measure its battery capacity from the voltages read over the cells, whilst the battery will cut the charger circuit only when its "electrolysis" reaction rate has reached such a low point that the cells are considered fully saturated with electrons. It is this difference, that I believe is the initial conditioning of the battery that make a huge difference. And as a alternating electrolysis system, the cell has a certain "memory" about how much charge/electrons it can take in, before it starts slowing down the reaction rate. Thus, by doing the earlier referred to "overcharging" as an initial conditioning, the cell learns how much it can take in before stopping to charge. I am not saying (as explained in an earlier post) that we are magically increasing the 1230 mAh, but we are making sure the battery has loaded that to its full extent. It was maybe not smart to call it overcharging before, but I thought that would be the easiest explanation for someone that maybe didn't understand what was happening, as the phone was already showing the battery was full. Basically the cells will after the initial conditioning "know" how much charge/electrons it can contain, and adjust its "voltage drop" over the cells accordingly. But this voltage drop is VERY VERY small. It is so small, it has no effect on the actual phone electronics. We are talking drops in the milivolts here.

You seem to enjoy experimentation, I've not seen this posted around here, so please try this and report your findings:

Collective experience shows that sometimes, not always but often, the charging circuit may not be calibrated out of the box. Others have found that this procedure will calibrate:

1. with your phone on, charge while observing.
2. when it hits green, immediately remove from charger and power off.
3. plug back into charger. if not immediately green - your charging circuit is not calibrated. in such a case, charge until green, immediately remove from charger and power on.
4. plug back into charger. if not immediately green - repeat this process until off or on, if indicating full charge in one state, it immediately registers full charge in the other state.

I own 2 HTC phones, one took 15 minutes of this sort of thing, the other more than a half hour, but in both cases the outcome was beneficial.

Apologies if this forum has such a cal procedure, our site search is lousy (hence our Zero Tolerance Policy to not flame newbies for not searching), and I didn't see it - so I thought perhaps better to post twice than not at all.

I did note in your OP that you've heard of the "battery boost" stuff by I very much suspect that procedure is actually attempting the calibration, but is not really well-considered (because leaving on charger for an hour past full charge indicated is likely to do nothing whatsoever other than give you a false reading on your actual charge level), so you may try this one and see if it matters.

Yes, I do enjoy experiementing, and will try this. However, I am not going to buy a new phone. So I will try this, but if it turns out that my phone was correctly configured, then I cannot really determine if this set up will help. As for the other method, I have basically roamed a bit on the net, and found that this helps about 50/50. So it is no sure method to increase battery life permanently. But it could be worth a shot. However, again, I have not tried it, so I don't know. But based upon the theory just served by you there, I would tend to agree, that it is not well thought through. But it came officially from HTC as a help.

Android apps have a mechanism called the intent. This mechanism is used to set apps to start and then start other apps.

Once intents are widespread, the more you use a task killer, the more you need to use a task killer. Please see my post in an adjacent thread on a cure for using a task killer.

The controversy over task killers comes from the individual users use cases and how they vary - this leads to different intents being set, and leads to different behaviors from the same app for different users. So - all a task killer really accomplishes is chasing the symptoms.

Android is a Linux system with a thing called the Dalvik engine - with few exceptions, all apps are running within the Dalvik engine process. Android apps are compact because they rely on standard, typical Linux services below the Dalvik engine to do the actual heavy lifting. Killing an app only performs the kill within the Dalvik - the services are still there, ready to go work. So - a task killer is by no means equivalent to the unix "kill -9 proc_id" and that's yet another reason experienced Androiders hate task killers.

If the reset procedure I recommend in the parallel thread doesn't put things back into order because you have a bad app just wanting to run and setting intents for a lot of other apps, your superior alternative is to root your phone, and then use a root-only app called Titanium Backup to freeze the apps in question or those you know you'll never use. This will effectively disallow them from executing and they'll cease to bother you. Replace the offenders with better apps - they're out there.

If you want to see your system logs, use aLogcat, free in the Market. It will produce a endlessly-scrolling, very boring, but very detailed output of basically everything you could care about running under the hood. Use the menu key, pause it, and scroll back - entries in red with an initial E (for error) marking may be considered as areas to investigate for misbehaving apps or processes. (If you see one for LEDs not providing a proper termination, ignore that, it's a known HTC defect and is quite harmless.)

Hope some of this info is helpful in some way.

I see what you mean about task-killers, and I do agree to it mostly, as task-killers are probably not the best solution for the OS (as I also have stated before). However, there are a few points regarding the battery life.

In terms of a "better" task-killer, I mean basically a task-killer that will not "monitor" the systems, so as to keep shutting down an app that restarts, and a task-killer that does not run all the time itself. These things will be very system demanding, and there are task-killers like this out there. These are the task-killers NOT to be used. But as the for actual "killing", I agree, there is no "better" task-killer in that sense. They all execute the basic "kill this app" function.

I also see what you mean about intent. However, if Android (and I do not believe it is made this way from my experience) is constructed such that the apps will just start themselves, without having any actual function (like say an updating, or a command that does so), then I am very shocked and will consider Android a very bad OS. An OS should only start up apps that its user wants to run. I do however see the point, that if you have an app running that needs or wants another app to run, that the OS will start up this system. However, if you then use a proper task-killer, it should shut down all apps that the user has started (of course the task-killer should not even see or touch OS systems), and the intent chain should be temporarily deactivated, meaning the apps should not restart by themselves and waste unnecessary resources. However, if the user is a very heavy user of the phone, and constantly has updates running, this could then trigger a app to start again immediately, initating a chain that will then lead to more consumed system power in reopening these apps. I will admit to this. However, for my set up (as in this guide) and my use, as I only have gmail and weather (once a day) syncing, I do not have many intent chains that start up by themselves. Thus a task-killer is saving me a lot of battery. In fact, over the last couple of charges, I've been having an estimated battery life of almost 80-90 hours+. And this is with moderate usage (meaning 50-60 sms in that period, on screen time of 2 hours+ and calling for 1.2 hours, data communications always on and syncing gmail and weather). For me that is more than enough usage on the phone, as I prefer a computer for browsing and writting emails.

As for the parallel thread about the task-killers as well, when you say that you can prove that apps don't use battery even if they are "running". I am guessing this proof takes into account main system resouces, like seeing how much CPU and battery the actual app is taking from running. But does this take into consideration raw energy used by the actually RAM system that stores the information about the program that is on? When you have about 500-600 MB of information in such a small RAM system, with a 1230mAH battery, that battery can be used up very fast just to power the RAM alone (granted, you still should be able to get about 20-30 hours from the battery). Just a thought, as explained in my guide how RAM uses battery

But thanks again for your input, it has been very appreciated, and I will at some day soon take a look at your tip and try out the experiement

 

[EarlyMon]

Yes, Android apps will start seemingly by themselves, not by the user and that does not make Android a bad OS - it makes it a mobile one.

There are a great many apps that the system needs to run.

The name intent is a software mechanism here - it's defined and documented so its theory of operation of is not mysterious.

Example - you start a music player, it can set an intent to start an app like an art downloader, for example, and then that will hybernate but be seemingly instantly available when desired by the user.

This doesn't prevent poorly-written apps from over-setting intents, tho, and that's where your own configuation management comes in. And some are downright pernicious. The US online retailer, Amazon, worked a deal with my carrier to install its MP3 store app on my phone. For most people, this app was harmless - would wake up, do nothing, go to sleep. In my case - I visited their online store via my web browser once and ever after, that app decided it needed to be often-awake, probing other processes and using the net to phone home. My carrier cleverly marked it as a system app, so I couldn't delete it until I gained root access. That's a single example.

Note that the SoC Snapdragon in your phone features a variable clock speed - it actually cycles down, conserving battery when running few tasks, then cycles up as loads increase so that your experience is such that the phone won't seem to bog down under load.

Apps in hibernation do not consume processor power for management, nor does their image need to be retained in memory.

Your concept of RAM sucking more power is tangentially correct, but not in the way you're thinking at present, so let's see if I can change your mind.

On a poorly-behaving phone, as bloat creeps in and memory use triggers system actions, the system _does_ need to make more decisions about what's been sleeping long enough to simply swap out. Too much of that is the power waster.

When your task manager wakes up and deletes things per its own schedule, you've just interrupted that control process and have now forced it to re-calculate everything - at a processor speed (hence battery) premium.

Task killers are not the way to go - there's no third way that makes them so. Configuring correctly is the way to go.

Here's a snapshot of a running Android system - if you're not already intimately familiar with Linux, please ask, don't assume - what you're seeing is that the apps are sleeping, but services remain and task killers do nothing about that -

View attachment sample_of_stock_evo_processes.pdf

In addition, I'm under the impression that you're not running any system monitors to help you understand your system - is this true? If so, you might try the free-in-Market apps like System Panel Lite (pro version is better) and Android System Info by Electric Sheep. The latter is itself a power hog, so it's really good for learning your system, but not for everyday use - word to the wise.

If my credentials may serve to help influence you away from task killers and into consideration of re-configuring, I work in semiconductor test in R&D (per my profile here) and have been doing systems programming since before Linux existed.

As for mods joining theads - we're just very interested and motivated Android users just like you and we're chatty by nature.

 

[Bedroom Eyes]

Random thing that worked for me, I deleted my Barcode Scanner app. Now I get more than 1.5 days with medium to heavy use.

 

[djdillpickle]

Would just like to preach the wonders of manual screen brightness. Keeping it as low as practical manually has a big difference.

Auto adjust fails to attenuate the brightness consistently.

 

[stuarta]

Random thing that worked for me, I deleted my Barcode Scanner app. Now I get more than 1.5 days with medium to heavy use.

Bizarre if it's that that's helped you

 

[EarlyMon]

Yep - sometimes a bad app install or weird app conflict happens.

 

[Xyro]

Yep - sometimes a bad app install or weird app conflict happens.

Yup. After I installed that Camera360 app recommended on the phandroid podcast I saw it was contributing to 50% of my battery usage, if the readings are to be believed.

 

[EarlyMon]

Wow!