Slashdot Mirror
Studying the Slow Decay of a Laptop Battery For an Entire Year
First time accepted submitter jradavenport writes "I've been keeping a log of the health of my MacBook Air battery for the past year, taking samples every minute I use the computer (152,411 readings so far!). This has allowed me to study both my own computing/work habits, but also the fascinating rapid decay of battery capacity. Comparing it to my previous 2009 MacBook Pro, the battery in this 2012 Air is degrading much faster."
You're discharging it wrong, don't do that.
Let's see the comparative graph where you did identical tracking over time for both, instead of detailed now against casual before, which seems a bit weak. I'd also like to see how you factor out the constant logging's effect as well.
Bad batteries something Apple is famous for, RAM fixed to the logic board, insecure and buggy OS, and a host of other complaints makes me wonder why anyone pays the premium for Apple any longer.
"If any question why we died, Tell them because our fathers lied."
I got 10 hours of battery life on my 2011 macbook air when I first got it. I don't just mean 10 hours of it sitting idle either. I could get 7 hours of continuous play of movies. Then Mountain Lion came out and I was lucky to get 3 hours tops. That lasted 6 months until they "fixed" it and I was able to get 5 again. Now in I can consistently get 4 hours with it sitting mostly idle.
I love the machine but I hate that I cant change the battery myself. I'll have to pay the Apple tax to get this fixed. I am holding out hope for Mavericks though, hopefully the power saving features can breathe some new life into this thing.
I seem to be constantly fighting the battle with battery life, and it is a topic that I am acutely worried about, thanks to the newest generation of phones which seem to have settled on non-replaceable batteries. I found this very interesting. Glad someone took the time to take these measurements and write it up.
"Here Lies Philip J. Fry, named for his uncle, to carry on his spirit"
1) It's "without further ado," not adieu.
2) Is it possible that querying the battery health once per minute actually invokes some mechanism which causes it to degrade at a higher rate?
you should have brushed it at least twice a day.
He finds the failure of a product he paid good money for fascinating, rather than infuriating.
Nice to have if the power goes out unexpectedly, but shut down ASAP.
Why aren't we filling our laptops, tablets and phones with compressed H2 that we make in our own homes with our new appliances (electrolyzers) that take distilled water? Why isn't this happening yet! Storage that does *not* degrade. The power density isn't great enough for cars but is awesome for portable electronics.
Duh...
Your software is querying the battery, which in turn takes battery power. If you keep asking it to check on itself every minute, you are decreasing its life expectancy in so doing. While it wouldn't produce as rich of a data set, if you really want to know how long your battery lasts at idle, you need to track it with pen and paper.
Damn_registrars has no butt-hole. Damn_registrars has no use for a butt-hole.
So far as I can tell, you're looking at the reported capacity from the intelligent charging circuitry. From my experience, and the reports I've heard, this has only a tenuous connection to reality.
From what I can see, the author has not actually ever measured the unit's battery life.
What the study shows is that logging the battery life every 1 minute prevents the laptop from staying in a low power state for long periods of time. This is like flipping your house lights on and off every 1 minute for a whole year and wondering why your power bill is rising through the roof.
Most modern lithium batteries should *not* be cycled or discharged "fully"--such a practice degrades the battery capacity quite rapidly. I think the practice of fully discharging the battery comes from the NiMH-type rechargeable AA(A) batteries.
Yeah, sometimes people recommend fully discharging a lithium battery during operation so that the monitoring software can recalibrate it's battery power meter to adjust for the decline in total capacity, but I'm not sure it's worth it.
As mentioned earlier, temperature is a big factor as well. Maybe Haswell will save the day...
They are considered disposable and if you can make it until the next upgrade cycle/treadmill they don't care.
For those of us who actually buy our phone outright, it sucks.
---- Booth was a patriot ----
Maybe polling the battery every minute reduces battery life?
You should replace the battery or have a spare waiting.
Oh, that's right. Apple doesn't allow you to do that. They expect you to upgrade to the new model every year and everyone obeys gladly.
I love studies with a sample size of one. No statistics, no variability. Definitive.
Almost every iPhone on record has been more likely to shatter its screen than the one before it. Apple went from 1st in lowest malfunction rates to 6th from 2007 to 2011. The new ipad is heavier, runs hotter, and gets worse battery life than the one before it. This isn't exactly a new pattern that the battery in the new air is inferior. Everything Apple is going downhill.
Calculated battery capacity is an estimate, nothing more, used by power management to decide when the computer should be force-slept, then suspended to disk to keep from damaging the battery (ie, it's not useful to wake up too late from sleep to do the suspend-to-disk.)
The SMC's estimate is just that: an estimate. Errors build up over time, and certain things fake it out a bit. For example, note the capacity, unplug the laptop, use it for 30 minutes, plug it in. Immediately the value will be different. It'll change again when fully charged. Your battery capacity didn't actually change. Even in a perfect world, since batteries have internal resistance, capacity gauges can never be perfect(if you draw at X you'll get less power out than if you draw out at X*0.8), and the battery's capacity varies with temperature. Battery degradation is impacted by temperature as well, so unless you're controlling for temperature of the pack, this was a completely useless endeavor. The only way this would have been useful would've been to cycle several (probably a dozen or more) batteries on lab-grade equipment in a temperature-controlled environment.
The noise and big upward swings alone should tell you that using the SMC's estimate for the purposes of statistical analysis or trending is virtually useless.
The stupid shit I see "enthusiasts" of any product obsess over is absurd. The time wasted on such an exercise far outweighs the impact it possibly could have had on the author (and probably even 9-10 other people combined.) The batteries last for well over 6 hours. Most people using a ultrabook with the battery life of a Macbook Air have plenty of opportunities to charge their machines during the course of a day.
Please help metamoderate.
I'm afraid you've been conned by several digital phantoms.
(1) There is no way to actually know the remaining charge in a battery. The little microprocessor in the battery or the laptop can take a good guess, by integrating the current draw over time, it can measure how many watt-hours have been drawn out, and from integrating how much charge has gone into the battery, and how the battery temperature went up, which signals that it's fully charged (or dying due to high ESR), and from previous discharge cycles, it can roughly compute what the battery capacity used to be, very roughly, but only if you discharged it until the voltage was significantly drooping. And so then the microprocessor can very roughly tell you that there is xx% of charge remaining. Very roughly. Very roughly.
(2) You probably did not measure the laptop's current draw over time. Perhaps you've been using the CPU more intensively, or you're using apps or web sites that need more processing power. Watching video or playing video games use a whole lot more power than light web site browsing. The Macbook air takes very aggressive steps to minimize power draw, that's good, but that also makes for a big jump when you go from light-duty to heavier-duty power-hungry apps.
(3) You are just one data point. And people with bad experiences tend to stand out, so we really can't generalize from your possibly bad experience to anything more general.
While your conclusion might be true, there are also all those other confounding factors that cast substantial doubt.
My laptop is rarely off AC power. When I had the charger set to stop charging at 100% (and to recharge when 90%), my battery life greatly improved. OId battery dropped 60% in reported capacity in less than 2 yrs; new battery is barely down 30% in the following 4 years.
I call it Chinese electron torture for your battery -- drip, drip, drip.
I don't know how OS X controls battery charging, but all OS's should provide an option to stop charging at 100%.
o wait my car is built on laptop batteries
Tesla Powered
Without knowing how deep the cycle was. Furthermore, the rate of discharge and temperature during the discharge will also have a fairly significant affect.
Basically, you will get about 1/10th the charge cycles out of a battery that is nearly completely discharged vs one that is only discharged to 10%.
Its better to really think of LI as providing a fixed number of watt hours. You can consume them in small bites, or you can consume them in big chunks but once you consume them they are gone.
The best rule for laptops is carry the charger and use it when at all possible.
Oh, and don't buy machines without replaceable batteries. If the "average" user can get 3 years out of it, and you decide your going to use all 8 hours of battery life every day, and charge it at night, you will probably be lucky if it lasts a year.
Plus, if your laptop gets hot (either by being left somewhere hot, or because of poor cooling) then the battery life is going to nose dive even if your plugged in. If your going to be doing a lot of intense gaming, your probably better taking the battery off and placing it somewhere cool (just don't let it discharge).
See this link http://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries
Maybe I'm a bit exaggerating here but, man... Anyone can measure how long the battery of a laptop lasts, but this guy actually put some effort to patiently capture the values long-term and, making the graphs of the battery decay and his computer use times. Good stuff, interesting report!
can be found at http://batteryuniversity.com/learn/
People misunderstand rechargeable batteries. Your laptop isn't psychic. It can't really know the current state of its battery. Instead, it has to use a series of tests, measurements, statistics and guesses to try to know what charge is left, and how best to bring the battery back to full charge. THIS IS AN IMPRECISE SCIENCE. The so-called memory effect is frequently nothing more than your computer device losing track of the actual status of its battery.
Now people want a battery that is light, lasts for a long time, and charges rapidly. They want a battery that doesn't get too hot in use, and is unlikely to explode or catch fire. Each of these aims tends to contradict another. Take the 'explode' one. Rapid placing of electrical energy into a chemical battery would seem to follow a Bell curve, where an actual percentage (hopefully low) of batteries would be destroyed by the rapid charging process. However, we know that if such an eventuality is higher than one in a million, the company is going to have big PR issues. Even below one in a million, such events hit the world wide press.
Given that phone and tablet companies are thus clearly between a rock and a hard place, they clearly to a wonderful job hitting a technological compromise. Consumers have to accept that the trade-off is their battery will appear to 'fade'. Some devices may offer a battery 'reset' monitor that attempts to recalibrate the computer that monitors the current battery charge, but usually this mitigates the problem only a little bit.
A modern device uses a dedicated computer to always integrate (monitor) the total power use from the battery across time. Then, because each battery should have the same essential nature (not quite true is one takes into account ambient temperature of different use environments), the computer can estimate with high accuracy the current charge ability of the battery. After a time (n recharge cycles) the effective battery life will be so low, the device or battery should be replaced. However, if a battery, say, gives 10 hours for two years, and five hours for five years after, the device probably remains useful to many. But more will have been broken or thrown away in the meantime.
Anyway, again, the battery business is complex and ever evolving. One may, for instance, measure minute changes in voltage for loads that vary, and use this to estimate battery capacity far more accurately. The temperature of the battery causes by the chemical reactions within also gives valuable data. Battery science and battery monitoring inside the best mobile devices is infinitely more sophisticated than most people here ever realise. However, companies have to compete on price as well, even Apple, so sometimes crappy cheaper solutions will replace better ones, because market research suggests the more cheaply made product will sell just as well. If you care, you should always demand a user replaceable battery.
Hey, we can't help if if Apple makes their power adapters too small to use on a continuous basis.
Is it just my observation, or are there way too many stupid people in the world?
I learned this with my white MacBook.
Keep it plugged in.
You will use it sometimes when it's not plugged in, but for the most part, make sure it's plugged in when you're using it.
I tested my old white MB and the battery only lasted 1 year with one year of daily a "down to 20%" discharge and then a full charge.
I've been using my early 2011 MBP 17 with the "keep it charging, stupid" approach and the battery is still good for 2+ hours when unplugged, as long as I make sure to put Safari on standby when it's not in the foreground. Safari is a memory and processor pig to the extreme. Make sure to 1) not use it, or 2) use a script to set it and its child processes to be paused when in the background, or at least disable JavaScript when Safari is in the background.
- Zav - Imagine a Beowulf cluster of insensitive clods...
there are multiple factors, that dermine the decay of battery:
1. temperature, the higher it gets, the more decay you have. at 40C you get noticable decay - 10% per year, when you go to 50C you get exponential decay
2. higher temp can be caused by multiple factors: overloaded battery, dying cell, or just hot cpu too close to it
3. keeping battery above 80% of charge or bellow 20% of charge for too long (2 weeks) will cause it to decay, that's why when you order a battery or a laptop it's usualy at around 50%
4. overdischargning is bad, always bad, but packs have protection against it so the theory about full cycling will cause no harm
5. if you have battery lying at 50% for long time - 1 year - if you cycle it around 2 times you should get it to the nominal range, which can be very often lower than you can expect
6. lithium batteries from my experience (RC racing, laptops) is about 2 years. On my buggy they last way less due to abuse they take
Stop checking your battery so often. It's a well known fact. The more you worry about shit, the worse it gets.
It shows that you don't know the details of Apple's power delivery architecture. Magsafe-equipped Apple laptops are intentionally crippled in that the charger is artificially disabled if you use an unauthorized one. There's a chip in the magsafe plug that connects to the middle pin and is interrogated by the system management on the mainboard. If the interrogation fails, you can still use the power source, but the charger is disabled.
All it takes not to charge the battery is to cover the middle magsafe pin. I've done it by keeping in use one charger with a bad magsafe plug where the chip had died. Died how? Ah, exposure to the saliva of a 1 year old, he liked to lick those plugs, they admittedly taste "funny" since they are energized :)
That way you have the best of both worlds: you don't lose your work if the magsafe plug is kicked loose, and you don't charge the battery if you don't want to. Win-Win.
A successful API design takes a mixture of software design and pedagogy.
A commercial life span of a car is 10 years or 200000 kms in Europe. In the USA, that's still 10 years or 125000 miles. Your 10 to 15 year old car is worth nothing and if it were to fall apart, that is considered normal according to commercial standards.
Now that we have established that, I do agree that if a car that was still made when cars weren't as expendable as they are made currently, it should be reasonably expectable that you won't need major overhaul on most of them after 10 years or 125000 miles, if it has been maintained decently. However, most dealerships only do the minimal required to drag the car to it's 10 year commercial life end and hardly will do proper preventative maintenance. They may charge you for it, but that doesn't mean they will actually do what's required to keep your car in the best condition, regardless of it's commercial value or life span left.
I maintain my own cars, I have been doing this for the last 10 years and my youngest car is 15 years old. All my cars are close to or way over the 200000 km mark. All of them are made by Alfa Romeo. I work on several other brands as well and I can tell you that given to a certain degree, they all are made expendable. I've done practically everything you can do to repair a car on my own cars and/or other peoples cars. once a car is 10 years old, nobody in the professional motor industry is going to consider it economically relevant any more and most cars are made to not last longer than that. Most gasoline engines will have serious wear to piston rings and valve guides, requiring a rebuild to get the power back and the oil consumption reduced. Wheel bearings will have worn. Suspension bearings/bushings will have worn. With the paint technology they had 10 to 15 years ago, your car will most likely be rusting. They had rust proofing pretty well under control, but then the environmental problems of those processes made them illegal and they had to resort to new "cleaner" processes that didn't work very well. There are plenty of other parts that wear out as well. Fans in the car ventilation, door locks and latches, even the car body itself will start showing more flex and metal fatigue. Plastics and such that are responsible for crash absorption will lose strength, seat belt tensioners, the list goes on. It takes a lot of maintenance to keep a car in a "young" condition, especially once you reach the moment that the designed life span for most parts is reached. Making them last longer costs more in production and will result in your vehicle being more expensive than the competition, without a lot of direct benefits that the potential buyer will be aware of. Most car companies choose to produce cheaper and just let the car fall apart after a little over 10 years or 125000 miles as a result.
I was promised a flying car. Where is my flying car?
It's not the charging that is the issue, it's heat from the laptop. If you removed the battery and just stuck it in an oven at say 40C it would degrade. Batteries are chemical devices, they react to temperature.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
My Macbook Pro permanently sits on my desk at work. I always thought the battery was just for the built in UPS that cuts in whenever the power goes out.
Only last week I sat there peacefully working when I heard desperate screams of 'failing to save' reverberating over my partition wall. At first I assumed they'd experienced a bad throw on their d20. But, out of curiosity I prairie dogged up to see what was going on, and saw that some of my co-workers had succumbed to the dreaded power outage. They mostly just stared blankly at their partition walls murmuring incoherently, but at least one grown man was reduced to tears. It brought back memories of the crash fest that was Windows 95/98.
I've owned various Apple portables (as well as PCs) since the Powerbook 170. Until I owned a G4 iBook, Apple's battery longevity was nothing different than the PC counterparts. I bought a used G4 powerbook when that model was 1.5-2 years old. When I got it, the battery was good for about 6 hours of solid average use. I had it for five years (only replacing it with this Macbook I have now last year) and in all that time, the battery only ever lost about 30% of it's capacity. I still saw a solid 4 hours of use out of it per charge. It was by far the longest I've ever had a laptop battery last (even keeping the comparison to Li-Ions). Comparatively, this Macbook never got great life (maybe 3hrs tops) but in over a year, I've yet to notice it losing any capacity. And in the past, a year is about all the time it ever took for previous laptops' batteries to hold zero charge at all. I had a Sony Vaio that killed it's battery every nine months, and before I bought the iBook, I was just used to a $50 replacement cost annually or so.
From this article (as well as what I've heard from others) at some point between this Macbook and the current models, the batteries or charging method Apple's used have significantly shortened the usable lifespans. Which will probably prove annoying when it eventually comes time to replace this machine in another 3-5 years.
I don't suffer from insanity, I enjoy every minute of it! --Longbottle
This is one of the reasons I've avoided iDevices (or others with hardwired batteries).
Yes, with some work you can change the battery, but being able to drop in a fresh battery on-the-fly is a very useful thing - especially if you travel - and any product which requires disassembly to do so is crippled IMHO.
much?
Fucking hell mate!
There's people with a lot of free time or no live at all!
-- 29A the number of the Beast
This is not a computer question. This is a battery question. Lithium-Ion batteries should NEVER be completely discharged. I don't care if your battery meter doesn't know how to measure it, that's the computer's problem. NEVER let your laptop battery die and it will last a hell of a lot longer.