The toggle you're talking about would still leave people with severely degraded batteries. How many people want to get the extra performance when it could come at the cost of random reboots? Phones with good batteries aren't slowed down. And the phones that are slowed down have poor batteries anyway so it makes sense that the people would like to change them. It's a good deal.
ddriver, you're assuming Apple's definition of "severely degraded" matches yours. Most companies state a 20% loss in total charge capacity is "severely degraded", which can happen in as little as a year with iphones (which are often recharged more than once a day from experience). Will a 20% loss in charge cause random reboots? Absolutely NOT, at least not with batteries designed to the maximum current specifications needed to run the damn phone.
Basically, Apple's been using underspecified (low margin, high internal resistance) batteries in their phones and "fixing" the over-current issue with software limits
20% degradation is the battery industry standard for "end of life". From experience I can tell battery performance at that point can be off from original specifications by some orders of magnitude.
Still, you are correct in that if they had been more conservative in their design choices, it wouldn't have been an issue.
"From experience I can tell battery performance at that point can be off from original specifications by some orders of magnitude."
I don't think you understand what an "order of magnitude" is. Even when a battery is basically dead, the voltage only drops ~25%, and maximum power output (i.e. current at that voltage) drops equivalently down by as much as 50%. That is LESS THAN ONE ORDER OF MAGNITUDE. And this is assuming internal resistance jumps ridiculously high too.
Apple seriously underestimated their margin or, much more likely, purposefully cut a corner in terms of battery specification to claim a much larger capacity than they actually had (since capacity is measured in energy output between a peak voltage and minimum voltage, they just set the min voltage lower than anyone else)
basroil, a severely degraded battery is one that cannot support the device functioning in standard operating conditions. It's not degraded capacity, it's degraded output voltage. Meaning it's not just a few hours less battery time, it's a reboot when power consumption spikes.
When a battery cannot deliver the necessary output in all standard operating conditions then I guess you can call it severely degraded. And I guess apple agrees since they eliminated the spikes by throttling the SoC.
I see most people here wrongly consider just the capacity. But this is meaningless in this case. The measure has nothing to do with the standard 80% capacity at 500 cycles. It's limiting instant power consumption, not extending battery life/autonomy.
@ddriver Voltage degradation without capacity degradation isn't a thing with Li-ion batteries. In fact, voltage degradation doesn't really exist either, just loss of capacity (which means lower voltage at any given time for the same discharge profile)
Like I said, the issue is high internal resistance and low margins, which means the same current draw drops voltage that much more. Since margins are razor thin to wrongfully state capacity, the battery ends up supplying less voltage than expected
Those are not 2 graph lines that go together linearly. Plenty of batteries lose a lot of their capacity but the voltage is still enough to support any peak. Meaning they don't exceed any recommended operating condition. Which is why I'll say it again: the theory has nothing to do with reality until we know what those peaks in SoC power consumption are. It's very likely that the tolerances are a lot lower for Apple's A SoCs so a battery that you'd normally consider works within parameters except for decreased capacity is not enough to power this SoC through peak usage.
It's not lack of capacity that kills the phone, it's the inability to provide the necessary voltage. I'm pretty sure there is no term for battery capacity in the power equation, which is what the SoC cares about. The lower capacity is the side effect of not being able to provide the necessary voltage, not the other way around. You're drawing an amateurish conclusion that if the battery capacity and voltage both go down in time then capacity is to blame. Hair turns white with age so white hair is the reason for the body deteriorating, no?
Your engine doesn't shut off even if your gas tank shrinks and is too small. It will just run for a shorter time. It turns off when your fuel pump is too weak regardless of capacity. So a battery might as well be a 10.000mAh, when the voltage drops below the required level to support peak power consumption the phone will reboot.
And that depends on the peak power consumption. Which is why I believe that Apple's SoCs trigger much higher spikes than other ARM SoCs. Capacity is irrelevant but the higher spikes trigger reboots earlier than in other phones due to the more powerful SoC exceeding operation conditions.
You're confusing everything I'm saying with nonsense right now.Here's the rundown: 1) Li-ion batteries have a set discharge profile, they go from a charge voltage, usually ~4.2V, to a discharge voltage, somewhere around 3V. That is a chemical voltage and CANNOT change without a change in chemistry (which does not happen during degradation) 2) Charge is the total energy stored between the charge voltage and discharge voltage for a certain use rate. The manufacturer determines the use rate, and depending on the battery model that may or may not be an optimal rate for maximum charge. 3) The EXTERNAL voltage can differ from the internal one due to internal resistance and current. A properly designed system will build in sufficient capacitance to smooth out any spikes. More importantly, a properly designed system will measure the battery charge based on the actual maximum use scenario. Remember that the internal voltage may be dropping to 3V but the external one is actually a bit lower! 4) Manufacturing processes induce variations, so if your first unit hit 3000mAh before 3V at half the actual power you're using, even the pilot run will likely have a half dozen or more that will get only 2400mAh at full power unless you built in enough margin during testing! Guess what? 2400mAh would be less than 80% max charge, and the battery would have to be replaced immediately from the factory!
Don't fixate on current spikes, any properly designed system will have enough capacitance to smooth it out. I happen to work with amp boards that drive nano-precision motors at 30A, and even at 90% use the total current used by all motors is nowhere near 30A peak at the wall because there's a few large capacitors providing the extra energy when necessary.
Let's have the provide a toggle with full performance mode and then compare. If it was due to battery degradation, then why doesn't the phone operate at max frequency when on charger? I am charging and have it still running at 911MHz.
I am defending Apple's response to a battery defect in the iPhone.
Since no other manufacturer uses Apple's A SoCs it's hard to tell if this behavior isn't caused by a design choice. Like the SoC triggering much higher spikes than SoCs used in other phones. The same design choices may even be the reason why Apple's SoCs outpace all others regardless of core count.
But most old phones actually get this behavior. They reboot when the battery drops below certain levels even if they shouldn't. I have plenty of old phones that reboot anywhere between 40%-60%. Some of them aren't even that old, and didn't go through that many charge cycles (like a ~3 y/o BB that was charged every 3-5 days and that now reboots consistently at ~50%).
If the battery in those iPhones was a dud or the users went through a lot of charge cycles due to the battery being small then it will just happen faster. To claim that this can't happen to others just suggests your parents buy you new phones every Christmas.
"Apple's A SoCs it's hard to tell if this behavior isn't caused by a design choice. Like the SoC triggering much higher spikes than SoCs used in other phones."
but that's not what happened, of course. a new OS decided that the battery was defective, once that OS was installed. if that's not a good definition of sabotage, I'll wait while someone finds a better one.
Basically the new OS was given the capability to 'throttle or not to throttle' depending on battery health. Whether or not it is 'sabotage' remains to be seen. It can also be seen as a device saving measure which ensures the user can still use the phone and not require an expensive (now rather cheap) new battery.
Bunny, you're not fun. The OS didn't decide "hey, let me do this and that". Apple engineers noticed the SoC exceeds battery operation conditions a lot sooner than in most other slower phones. So the software engineers introduced the throttling to compensate. It's not sabotage, it's very similar to every other protections put in place in every CPU since the dark ages of computing.
Your CPU slows down when it's hot. But what if you wanted it to be fast at ANY cost? And the damn capitalist pigs sabotaged you! No, the problem here is that they did the right thing but weren't honest with people and didn't tell them. Screaming sabotage is so idiotic it's not even funny.
Next you'll tell me Mercedes limiting their cars to 155mph is also sabotage. Forces you to buy the more expensive unlimited models.
"No, the problem here is that they did the right thing but weren't honest with people and didn't tell them."
I don't run iPhone, just to be clear, but tell that to all the folks here who've installed 11 and found that their battery is "degraded". the only change is the OS. battery was fine under previous OS, and so, apparently, was the phone. the only reasonable conclusion: iOS 11 (at least) tries to drive *all* iPhone as if it were the latest SoC, rather than the real one inside the case. it's not rocket science to configure an OS to the cpu it runs, rather than simply degrading the SoC. linux does that all the time.
Me Yo is right. Other phone designs, voltage regulators, power management, etc, seem to work around degraded voltage of a battery just fine. Basically Apple should have incorporated a capacitor or some sort of on the fly voltage upconverter to deal with spikes in power demand.
Other phone design are also a lot slower. By any benchmark. Apple simply decided to cram the performance in the first year of usage instead of spreading it nice and thin over 5 years when almost nobody uses the phone for that long.
So as I said, design choice not accident. And no amount of holding hands and blowing compliments in each other's ears about who's right will change this. The people who designed the fastest phone SoC in history probably know more about this than Me Yo and Samus with his "it's missing a capacitor".
ironargonaut, search term spikes explained by Apple releasing new version of OS typically around the time of new iPhone release. Apple is able to push new OS update to "every" iPhone applicable. New OS is optimized for newer phones (with more capable SOCs) and probably less so for 3+ year old phones. You don't see the same search term spike for Android because there is no spike in Android phones downloading new OS updates. Most Android users are not on latest release with in a year of release (maybe years?). Also note, far from all of the iPhone models apple admitted to throttling were actually throttled according to John Poole's data. If Apple's sole intention was to "sabotage" why would they not sabotage all iPhones. Also, the throttling occurred at several different bands supporting the notion the more worn a battery the more throttling.
Umm yes the very first cellular phone that used a battery had a power supply in it with energy storage elements and voltage regulation and every cellular phone since then has had that. They didn't disappear in the iphone 6s. This is a sham.
Sorry, but phones with good batteries are also slowed down. I have 5% battery wear and my SE sits at 911-1200MHz range. Sometimes at 100% it revs at 1.85GHz, but it lasts for a few minutes and drops to 1200 or 911MHz. It is bullshit that it is due to degraded batteries. Apple should provide a toggle and keep the 29$ dollar price for batteries.
"I bet they make a lot of money battery swaps, which was the plan all along."
I seriously doubt they make money on battery swaps - my guess almost 90% of battery upgrades are when they get trade in. Typically an iPhone user trades them in every other version. This is first time I didn't - because 1. the 7 did not look like much improvement but I like the 8 better - but X was even more desirable and the 6 has done me great.
But then there is labor and associated costs (buildings, lighting, tools etc). Of course if they would make the batteries replaceable in the first place, there would not be such a problem, but them a lot of people would not change their phones every year or two.
"But then there is labor and associated costs (buildings, lighting, tools etc)."
not really. the only marginal/variable cost to the replacement is the battery and labor. all the rest are fixed cost, not related to the replacement exercise. the total margin is cannibalized.
Yes battery cost is low, but installation is 20 minutes done properly on 6, and even more difficult on 7’s because of IP68 certification sealing. There are other costs associated with the battery replacement, such as pull strips, lcd gasket and seal, and the possibility of damaging a component during disassembly/reassembly. Since these repairs are happening in the United States where wages are high for technicians replacing these components, Apple may not even be breaking even after parts and labor. But it could be close to breaking even if they contract it out right...
What is a "scam" is that Apple hid the fact that they were throttling performance in an effort to mask their poor engineering choices. This has caused countless tens of millions of people (if not hundreds of millions) to purchase new phones when all that was needed was a new battery. $29 is not a good deal when you consider that this is Apple in damage control mode and they are just trying to gloss over their massively unethical behavior. When Nexus 6P users were hit with random reboots because of "degraded" batteries Google replaced them with new Pixel XL's at no charge.
SquarePeg: sorry but you're completely wrong. Rising output impedance of aging batteries (rechargeable or not) is just one of the constraints of developing high performance mobile devices. Apple engineers have addressed in the only possible way: to reduce current spikes by throttling the SoC. You can also look at it the other way around: they maximized performance of the SoC as much as possible on fresh batteries or when plugged in...
Other phone manufacturers do exactly the same to avoid random reboots and crashes. I agree that Apple have shown dubious business practices with batteries but mixing this up with an electronics engineering constraint is at best and uninformed opinion...
Apple engineers did not take into account these constraints. That's why they first recalled a batch of iphones and then issued a band-aid. Firs they admitted the flaw - the recall - then they hid it - the "fix".
It is not reasonable for the iphone 7 to start getting throttled after one year. This is flawed design, textbook planned obsolescence, especially in Europe where 2 years is the minimum mandatory warranty.
In abstract it's a correct decision, in reality it represents flawed design. Average users do not wear the battery to such an extent in one year.
The issue isn't caused by voltage drop. It caused by current drop. Thats what makes the phone reboot.
My iPhone 6 had this exact issue. The random reboots were annoying as hell. Having the phone be slightly slower would have definitely been preferred (The code for the throttling was not in place back then).
The cheapest battery replacement I've seen (WITHOUT labor) for an iPhone 6 Plus is $15. $15 for labor for Apple-official isn't exactly "raking in the dough." No way was the plan "to make a lot of money on battery swaps." I'd believe "convincing people to upgrade to the latest phone" much more than "make a lot of money on battery swaps."
How much should we bet that the "discount" is applied only if you sign a waiver stating you won't sue Apple for this issue? If they are really "crafty" they might even state you accept that discount as final compensation for any and all "features" of the phone!
My iPhone 6S has never, ever shut down on me. I never even noticed any drop in battery life before I updated to iOS 11. So why am I being throttled to half speed (according to geekbench) at 40% battery?
Can you just turn the throttling off you cheapskates? Or just give us the cruddy option in the settings app to disable this new throttling protocol. How hard is that? Settings has already become such a labyrinthine mess, what harm is there in giving us one more option?
My iPhone 6 shuts down a lot if I let the battery run down below 30%. It’s also slow as hell, especially launching apps like Weather Channel or Twitter.
I just installed a battery check app and it’s telling me my battery is “marginal” at 87%.
If your phone is shutting down below 30%, is slow, and the battery is old... replace the battery, seriously. You don't need an app to know that battery is shot. Jeebus. Also how many gens old is that phone now? I'm far from an Apple fan and I haven't broke down and bought one yet, but yeesh.
They probably measured the output voltage of your battery and noticed during power spikes it's dangerously close to rebooting the device. So they throttled just in case. I think they took the right decision but it should have been transparent.
They took the right decision in abstract but they still handicapped your phone. You are thus the owner of a phone that performs under its advertised power, possibly while under warranty. Maybe you are out of warranty but throttling started while under warranty and you didn't notice. In Europe, this is called a concealed defect.
This is a recall issue not a battery swap. Apple's iphones "burn too bright" and start suffering conveniently at around the 1 or 2 year mark when the warranty expires.
Finally, a guy who gets it. Apple concentrated performance in the first 1-2 years of operation because they know people change the phone after. Everybody else just spreads the performance and you get a watered down phone that lasts more than 2 years.
The only mistake Apple did was hiding this. They should have told people and they would have been able to sell the replacement for more than $29.
Ddriver, this was not an industry standard measure. This band-aid was issued on the background of an iphone recall. Apple thus admitted a flaw otherwise they wouldn't recall anything and just issue the patch. The "fix" just hid what could've been free(under warranty) battery swaps or worse, total recall.
This was not a degraded battery under normal operation. This throttling band aid was issued on the background of an iphone recall. Basically Apple admitted that there was a design flaw within the whole situation. After the recall they noticed the issue persisted and issued this patch. This was a sweeping under the rug of a design flaw. Better kneecap all iphones than issue free battery swaps.
It's not really that choice... the choice is do you want a normal operating phone that may reboot without warning or a slow phone that doesn't reboot. If it was a normal operating phone with a shorter battery capacity (like it's the case with every other phone) then people may take that choice and just charge it more often.
Your iPhone 6S has never, ever shut down on you BECAUSE it slows down to half speed at 40% battery. Before the software change my old iPhone 6 was shutting down randomly all the time, and the update stopped that. A couple months later I got a brand new iPhone 6 for the $79 cost of a battery replacement from Apple because my out of warranty battery started swelling.
I'm far from an Apple fan, but really they only made two, maybe three mistakes:
1) Non-disclosure was clearly an error. It was, however, an error that's consistent with Apple, in that they tend to hide details from customers and try to "just work." 2) Their algorithms should be better tuned and provide more performance when the phone is charged and plugged in. Hopefully with the software update they'll provide some options, like "give me full throttle, I like to live dangerously," "Err on the side of more performance," and their current setting of "please don't let my phone ever shut down unexpectedly." 3) On the smaller screen phones they had battery issues on both the 6 and 6S (maybe 7 & 8?) that were likely the result of being overly aggressive with pushing design margins. In the end this caused these devices to have high battery failure rates and short battery lives. The $29 replacement program probably needs to be extended indefinitely for these devices and any other device with similarly razor-thin design margins.
Too me Apple designed their iphones wrong to burn too bright. The CPU is too strong for the battery output. It technically works but it is still a design flaw. Some would consider a defect.
Leaving a customer without a cell phone for a week (when for many it is their only phone), a solid response?
The price for the battery replacement is definitely a good deal. However, I customer should be able to make an appointment and have it done while they wait. If they can't do that, loaner phones should be available.
Now THAT much is true, it is seriously sucky to go without your phone. The real answer to this is for Apple to figure out how to pack at least SOME user replaceable battery capacity without compromising their design philosophy much. Even if it's like a slide-out battery that doesn't require a split-body design, and only accounts for maybe half the total battery capacity. The side benefit would be the ability to hotswap a battery. Pretty unlikely to happen, though.
As for the "How is Apple measuring battery impedance?" question in the article..
Around 8-10 years ago me and others were experimenting with this on certain Nokia Linux devices. The basic method is simple: you measure voltage and current, as current increases, voltage increases. Yes the voltage drops naturally with declining state of charge, but keep your time window short and you can ignore that. Calculate impedance from the voltage and current data points.
The challenge was that the hardware updated the voltage once per second and current once every 5 seconds. So, we had to look for two periods of 5 seconds where voltage remained stable, two periods where current use was very different in order to get useful data.
In the end, the only practical approach was to wait until the device was otherwise idle, and periodically peg the CPU with a busyloop to create a synthetic and consistent power draw.
Perhaps today 10 years later, hardware battery fuel gauges are able to report data more frequently, or even report impedance directly.
But I still wonder if they're not simply doing something like: If system shut down unexpectedly, then bump power limit down one notch.
Given the choice between a phone that mysteriously turns itself off when you need to make an emergency call and a phone that does not, I'll take the phone that works.
The problem here was that if it is necessary to reduce the voltage draw of the device because the battery is reaching it's end of life, then the user should be notified.
This is what happens when you continue to provide updates and security patches to phones that are more than two years old. The phone remains useful past the lifetime of a single battery.
Here is a big question, I have an iPhone 6 and so far no shutdowns, but if the new iOS detects that I need new battery - and I replace it for $29. Does iOS still limited my performance with new battery.
If so the problem is not in iPhone 6 but in the new iOS.
I have no problem with my iPhone 6 ( in fact I have another one for work ). I just glad I didn't upgrade to 7 or even the 8 - planning on new version that will be like the iPhone X but size of the iPhone 6 ( not 6 plus or X size )
TLDR: Good luck getting Apple to replace your 'degraded' battery!!
I have an iPhone 6+, bought directly from Apple at release, noticed super slow-downs with iOS 10.2+ (and also iOS 11) Did the requisite Geekbench tests for sanity, noticed the throttling at lower battery levels (50%), charged it up 100%, results miraculously back up to par for the phone. Used a few battery test apps, which indicate degraded performance. Gave Apple a call today after seeing this news, they wanted to do a remote diagnostics, so humored them and did it - here is the exact wording from that effort - "That is normal behavior for the device to slow performance when the battery gets a lot lower. We can do some steps to try and help with the performance, but the battery does appear to be working fine." and then later "Alright, but I’m just letting you know it’s not battery related or related to needing a battery replacement. Your battery passed the diagnostic with flying colors!" Then they wanted me to do a complete wipe/reset, etc. Which I didn't have time for today, so I'll give them the benefit of the doubt (however slim that is), and try that tomorrow. But... I don't expect much from them anymore.
What is it about Apple’s devices that necessitates this throttling? I suppose we would have heard by now if it were an issue with Android or Windows devices. And anecdotally, my wife’s iPhone has always shut down quite quickly if she takes it out in the cold, whereas I’ve never seen another phone do that. Do they have a different (well, inferior) power management approach in the SOC, or are their batteries spec’d differently (well, with less margin) than the competition?
It is an issue with the lithium batteries in Android and Windows Phones.
Lithium batteries are only good for so many recharge cycles. Eventually, they stop being able to supply the necessary voltage even though the phone still says it has some charge left.
When the device demands a higher voltage than the battery can provide, the phone suddenly turns off without warning.
Perhaps the difference is that Android devices only get two years of software updates, so they tend to be replaced before the battery has time to degrade as much.
Apple chose to throttle down devices whose batteries are failing instead of just letting them turn off without warning, which is good, but they failed to warn the user that their battery was hitting it's end of life when they did so which was bad.
"It is an issue with the lithium batteries in Android and Windows Phones."
I don't think so. at least, it's not a construct of *a* Li-ion battery. my Braun toothbrush is at least a decade old, has a Li-ion battery, spends most of the time charging (of course), so it's been through thousands of charge cycles. let it run out once a month or so, and it recharges and runs like new. so, it's not a Li-ion issue, per se.
You might want to read up on it. Lithium batteries have a limited lifespan.
>Lithium-ion batteries age. They only last two to three years, even if they are sitting on a shelf unused. So do not "avoid using" the battery with the thought that the battery pack will last five years. It won't. Also, if you are buying a new battery pack, you want to make sure it really is new. If it has been sitting on a shelf in the store for a year, it won't last very long. Manufacturing dates are important.
You might want to experience it. I have a 20 year old ThinkPad, 10 year old Toughbook and 9 year old Palm Centro whose batteries perform 90% or higher.
The reason why the battery only lasts 2 years is because they're continually kept at 100% which burns them up faster than using them.
Your tooth brush also only has one spike, when it's turned on and jut two modes of operation: on or off.
It's nice to see your level of understanding of a complex SoC power usage profile by comparing it with a tooth brush. I have a LED light in my remote and the batteries are 4 years old, Apple could learn from those LED manufacturers and alkaline battery makers.
Thatr's a dumb example: your toothbrush doesn't spike at a dozen amperes for a few milliseconds and expects the battery voltage to hold steady... Search for Ohm's law and see what happens to your voltage output when the impedance increases and you try to suddenly pull 10A...
Those same guys can't realize that the capacity is not relevant to the problem. It's the voltage/impedance that's the problem. Both degradations may occur at the same time but it's the second one that caused this whole debacle.
no, it isn't. the problem is that iOS 11 made the problem by being one-size-fits-newest-iPhone, thus sabotaging existing units. having an OS that switches functions based on the hardware it runs on isn't rocket science or unprecedented. as I said on one of these posts, Gates had DOS and Windows written to the latest x86 cpu, and devil take the rest. nothing new. Cook faces a slightly different problem: battery progress has always been glacial, and there's some reason to believe that energy density has reached the asymptote of density increase. since we're nearing the quantum problem of smaller node size, another asymptote, dealing with being on the leveling part of the curve is something humans haven't had to deal with before. previous Dark Ages of anti-progress were the result of decisions to ignore science. here, even though some of us are anti-intellectual (guess who?), the issue is reaching the limits of physics.
an interesting observation, but still doesn't offer a rational explanation for why: older SoC with older batteries and older iOS run fine, but tank when iOS 11 is installed. well, other than Apple thought it a neato way to get folks to toss an old iPhone for a new one. just read the multitude of comments describing that behaviour. again, Apple has the smarts to configure iOS 11 to meet the needs, and just the needs, of older SoC installed. they chose not to do that.
You call suddenly turning off without warning "running fine"?
I'm kind of waiting for someone to file a class action lawsuit against Android device makers because they can't make emergency calls when their device suddenly shuts down without warning.
"You call suddenly turning off without warning "running fine"?"
go through the comments. the number of shutdown when running old iOS on old iPhone with old battery is a small fraction of the iOS 11 on such an iPhone tanking. FWIW. again, with vigour: Cook could have ordered the pointy headed guys to build iOS such that it configured to the SoC/battery combination it found, and didn't implement the "new" iOS functions that cause problems with said old SoC/battery combos. that would have been putting the customer first. make it transparent that older iPhones wouldn't get all the new goodies of iOS 11, so the owner could choose. and, for the sake of argument, how long do you think old iPhones will take to eat that new battery running iOS 11 functions that still clobber older SoC???? a few months, is my guess.
the problem is iOS 11 only works as designed on 8/X phones. but Cook won't tell you that.
Wow...For only $30.00 it's problem solved! I've got two iPhone 6+ phones that will get new batteries and keep on trucking. People will still complain, but for me, this is a cost effective solution.
If after a year your smartphone gets kneecapped that constitutes a defect. It is not reasonable for your phone to lose performance because the battery is undersized by design. This is a design flaw either way you take it. Basically textbook planned obsolescence.
No, if after your battery reaches it's end of life and your device starts turning off without warning, this is an avoidable problem.
How long until somebody sues Samsung because they were unable to make an emergency call because their device powers off unexpectedly, despite showing that it still has a charge?
Batteries do not reach EOL after 1 year. This is a design flaw by underspeccing the battery in relation to the SoC's requirements. It's planned obsolescence and the hidden "fix" robbed you and millions others from free warranty service or replacements.
I've never used an iPhone but I've tinkered with enough Lithium Polymer batteries over the years to understand the following:
*The temperature charging range for Li-Po is much narrower than the discharge range. The higher the temperature the faster it degrades. It's safe to charge up to 45C but not usually advisable. Your phone will last longer if you charge it while not in use. *The higher the voltage is kept the faster the battery degrades. So in theory if the phone's battery is kept at 100% for 8 hours a day one would lose about 7% of the battery's lifespan over the course of a year. *There is no standard as to how many charges a battery is designed to go through. Some say 300 others say 500. *I've understood that if never fully charged or discharged the stresses upon the battery are halved. *I don't know about the iPhone but some devices will continually cycle the top 5% of the battery under external power in a vain effort to prolong battery life. This practice is highly destructive. *I've read that some iPhone and Samsung Galaxy battery malfunctions are due to third party adapters and power banks. I know on micro-USB devices the voltage regulation is on the external side rather than within the phone itself. *The health of a Li-Po pack is correlated to how much swelling it has. Something that can be seen with my Galaxy S3 and S4 that are still perfectly fine with the former being 5 years old. *Lithium Ion prefers to always be in a constant dis/charging state. So with laptops even cycling them once a week is better than leaving them on the wall for months at a time.
I tend to be rather delicate with my batteries so I get well above the 'two years' expected out of Lithium Polymer. I can't compare Li-Ion cells because most of what I use in laptops are Panasonic and those things are durable.
I forgot to add, with cell phones in order to be able to be charged via USB quickly the batteries are on the primary circuit rather than independent like a laptop (i.e. I can turn on the laptop without a battery). So once the battery is charged it cannot 'relax' as it's always being kept at voltage despite not being charged...At least assuming the voltage readouts are not on the controllers in my phones.
Battery throttling is one of the most speculated outcome of the year. Apple played it correctly and fooled people for over more than a decade. Finally, the accepted what they were doing and came back with another program to make more out of their products - https://ios12guide.com
Here is how you can download iOS 12 beta IPSW for iPhone or iPad devices. Our download page contains direct links of iOS 12 firmware IPSW. You can use our website to get the latest update for your device.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
95 Comments
Back to Article
Sarah Terra - Saturday, December 30, 2017 - link
What a scam, seems to me a simple toggle in settings to enable or disable the throttling would cost customers exactly $0.I bet they make a lot of money battery swaps, which was the plan all along.
ddrіver - Saturday, December 30, 2017 - link
The toggle you're talking about would still leave people with severely degraded batteries. How many people want to get the extra performance when it could come at the cost of random reboots? Phones with good batteries aren't slowed down. And the phones that are slowed down have poor batteries anyway so it makes sense that the people would like to change them. It's a good deal.basroil - Saturday, December 30, 2017 - link
ddriver, you're assuming Apple's definition of "severely degraded" matches yours. Most companies state a 20% loss in total charge capacity is "severely degraded", which can happen in as little as a year with iphones (which are often recharged more than once a day from experience). Will a 20% loss in charge cause random reboots? Absolutely NOT, at least not with batteries designed to the maximum current specifications needed to run the damn phone.Basically, Apple's been using underspecified (low margin, high internal resistance) batteries in their phones and "fixing" the over-current issue with software limits
shadowjk - Saturday, December 30, 2017 - link
20% degradation is the battery industry standard for "end of life". From experience I can tell battery performance at that point can be off from original specifications by some orders of magnitude.Still, you are correct in that if they had been more conservative in their design choices, it wouldn't have been an issue.
basroil - Saturday, December 30, 2017 - link
"From experience I can tell battery performance at that point can be off from original specifications by some orders of magnitude."I don't think you understand what an "order of magnitude" is. Even when a battery is basically dead, the voltage only drops ~25%, and maximum power output (i.e. current at that voltage) drops equivalently down by as much as 50%. That is LESS THAN ONE ORDER OF MAGNITUDE. And this is assuming internal resistance jumps ridiculously high too.
Apple seriously underestimated their margin or, much more likely, purposefully cut a corner in terms of battery specification to claim a much larger capacity than they actually had (since capacity is measured in energy output between a peak voltage and minimum voltage, they just set the min voltage lower than anyone else)
Jaaap - Saturday, December 30, 2017 - link
Research Li-ion battery degradation.A degraded cold battery will have a much higher internal resistance.
ddrіver - Sunday, December 31, 2017 - link
basroil, a severely degraded battery is one that cannot support the device functioning in standard operating conditions. It's not degraded capacity, it's degraded output voltage. Meaning it's not just a few hours less battery time, it's a reboot when power consumption spikes.When a battery cannot deliver the necessary output in all standard operating conditions then I guess you can call it severely degraded. And I guess apple agrees since they eliminated the spikes by throttling the SoC.
I see most people here wrongly consider just the capacity. But this is meaningless in this case. The measure has nothing to do with the standard 80% capacity at 500 cycles. It's limiting instant power consumption, not extending battery life/autonomy.
basroil - Sunday, December 31, 2017 - link
@ddriver Voltage degradation without capacity degradation isn't a thing with Li-ion batteries. In fact, voltage degradation doesn't really exist either, just loss of capacity (which means lower voltage at any given time for the same discharge profile)Like I said, the issue is high internal resistance and low margins, which means the same current draw drops voltage that much more. Since margins are razor thin to wrongfully state capacity, the battery ends up supplying less voltage than expected
ddrіver - Monday, January 1, 2018 - link
Those are not 2 graph lines that go together linearly. Plenty of batteries lose a lot of their capacity but the voltage is still enough to support any peak. Meaning they don't exceed any recommended operating condition. Which is why I'll say it again: the theory has nothing to do with reality until we know what those peaks in SoC power consumption are. It's very likely that the tolerances are a lot lower for Apple's A SoCs so a battery that you'd normally consider works within parameters except for decreased capacity is not enough to power this SoC through peak usage.ddrіver - Monday, January 1, 2018 - link
It's not lack of capacity that kills the phone, it's the inability to provide the necessary voltage. I'm pretty sure there is no term for battery capacity in the power equation, which is what the SoC cares about. The lower capacity is the side effect of not being able to provide the necessary voltage, not the other way around.You're drawing an amateurish conclusion that if the battery capacity and voltage both go down in time then capacity is to blame. Hair turns white with age so white hair is the reason for the body deteriorating, no?
Your engine doesn't shut off even if your gas tank shrinks and is too small. It will just run for a shorter time. It turns off when your fuel pump is too weak regardless of capacity. So a battery might as well be a 10.000mAh, when the voltage drops below the required level to support peak power consumption the phone will reboot.
And that depends on the peak power consumption. Which is why I believe that Apple's SoCs trigger much higher spikes than other ARM SoCs. Capacity is irrelevant but the higher spikes trigger reboots earlier than in other phones due to the more powerful SoC exceeding operation conditions.
basroil - Monday, January 1, 2018 - link
You're confusing everything I'm saying with nonsense right now.Here's the rundown:1) Li-ion batteries have a set discharge profile, they go from a charge voltage, usually ~4.2V, to a discharge voltage, somewhere around 3V. That is a chemical voltage and CANNOT change without a change in chemistry (which does not happen during degradation)
2) Charge is the total energy stored between the charge voltage and discharge voltage for a certain use rate. The manufacturer determines the use rate, and depending on the battery model that may or may not be an optimal rate for maximum charge.
3) The EXTERNAL voltage can differ from the internal one due to internal resistance and current. A properly designed system will build in sufficient capacitance to smooth out any spikes. More importantly, a properly designed system will measure the battery charge based on the actual maximum use scenario. Remember that the internal voltage may be dropping to 3V but the external one is actually a bit lower!
4) Manufacturing processes induce variations, so if your first unit hit 3000mAh before 3V at half the actual power you're using, even the pilot run will likely have a half dozen or more that will get only 2400mAh at full power unless you built in enough margin during testing! Guess what? 2400mAh would be less than 80% max charge, and the battery would have to be replaced immediately from the factory!
Don't fixate on current spikes, any properly designed system will have enough capacitance to smooth it out. I happen to work with amp boards that drive nano-precision motors at 30A, and even at 90% use the total current used by all motors is nowhere near 30A peak at the wall because there's a few large capacitors providing the extra energy when necessary.
picka - Tuesday, January 2, 2018 - link
Let's have the provide a toggle with full performance mode and then compare. If it was due to battery degradation, then why doesn't the phone operate at max frequency when on charger? I am charging and have it still running at 911MHz.Samus - Sunday, December 31, 2017 - link
Hopefully this opens up the possibility Apple will change their industrial design to incorporate user replaceable batteries...Hurr Durr - Monday, January 1, 2018 - link
One thing that certainly will never happen. If the goy can change his battery at will, how to make him buy a new phone?Me Yo - Sunday, December 31, 2017 - link
ddriverYou're defending a hardware defect on iPhones.
Every battery degrades over time and doesn't cause the issues Apple is describing.
ddrіver - Sunday, December 31, 2017 - link
I am defending Apple's response to a battery defect in the iPhone.Since no other manufacturer uses Apple's A SoCs it's hard to tell if this behavior isn't caused by a design choice. Like the SoC triggering much higher spikes than SoCs used in other phones. The same design choices may even be the reason why Apple's SoCs outpace all others regardless of core count.
But most old phones actually get this behavior. They reboot when the battery drops below certain levels even if they shouldn't. I have plenty of old phones that reboot anywhere between 40%-60%. Some of them aren't even that old, and didn't go through that many charge cycles (like a ~3 y/o BB that was charged every 3-5 days and that now reboots consistently at ~50%).
If the battery in those iPhones was a dud or the users went through a lot of charge cycles due to the battery being small then it will just happen faster. To claim that this can't happen to others just suggests your parents buy you new phones every Christmas.
FunBunny2 - Sunday, December 31, 2017 - link
"Apple's A SoCs it's hard to tell if this behavior isn't caused by a design choice. Like the SoC triggering much higher spikes than SoCs used in other phones."but that's not what happened, of course. a new OS decided that the battery was defective, once that OS was installed. if that's not a good definition of sabotage, I'll wait while someone finds a better one.
RanDum72 - Monday, January 1, 2018 - link
Basically the new OS was given the capability to 'throttle or not to throttle' depending on battery health. Whether or not it is 'sabotage' remains to be seen. It can also be seen as a device saving measure which ensures the user can still use the phone and not require an expensive (now rather cheap) new battery.ddrіver - Monday, January 1, 2018 - link
Bunny, you're not fun. The OS didn't decide "hey, let me do this and that". Apple engineers noticed the SoC exceeds battery operation conditions a lot sooner than in most other slower phones. So the software engineers introduced the throttling to compensate. It's not sabotage, it's very similar to every other protections put in place in every CPU since the dark ages of computing.Your CPU slows down when it's hot. But what if you wanted it to be fast at ANY cost? And the damn capitalist pigs sabotaged you! No, the problem here is that they did the right thing but weren't honest with people and didn't tell them. Screaming sabotage is so idiotic it's not even funny.
Next you'll tell me Mercedes limiting their cars to 155mph is also sabotage. Forces you to buy the more expensive unlimited models.
FunBunny2 - Monday, January 1, 2018 - link
"No, the problem here is that they did the right thing but weren't honest with people and didn't tell them."I don't run iPhone, just to be clear, but tell that to all the folks here who've installed 11 and found that their battery is "degraded". the only change is the OS. battery was fine under previous OS, and so, apparently, was the phone. the only reasonable conclusion: iOS 11 (at least) tries to drive *all* iPhone as if it were the latest SoC, rather than the real one inside the case. it's not rocket science to configure an OS to the cpu it runs, rather than simply degrading the SoC. linux does that all the time.
ironargonaut - Tuesday, January 2, 2018 - link
https://trends.google.com/trends/explore?date=2006...Sorry but they only slow them down when new phone is released, if it was important they would do it when necessary not when phones released
ironargonaut - Tuesday, January 2, 2018 - link
Should have added this link.https://wattsupwiththat.files.wordpress.com/2017/1...
Clearly shows slow downs correspond to release dates not to when battery levels drop.
Samus - Sunday, December 31, 2017 - link
Me Yo is right. Other phone designs, voltage regulators, power management, etc, seem to work around degraded voltage of a battery just fine. Basically Apple should have incorporated a capacitor or some sort of on the fly voltage upconverter to deal with spikes in power demand.ddrіver - Monday, January 1, 2018 - link
Other phone design are also a lot slower. By any benchmark. Apple simply decided to cram the performance in the first year of usage instead of spreading it nice and thin over 5 years when almost nobody uses the phone for that long.So as I said, design choice not accident. And no amount of holding hands and blowing compliments in each other's ears about who's right will change this. The people who designed the fastest phone SoC in history probably know more about this than Me Yo and Samus with his "it's missing a capacitor".
Full of geniuses around here it is. LOL =))
ironargonaut - Tuesday, January 2, 2018 - link
Please explain this correlation,https://wattsupwiththat.files.wordpress.com/2017/1...
Argosy - Tuesday, January 9, 2018 - link
ironargonaut, search term spikes explained by Apple releasing new version of OS typically around the time of new iPhone release. Apple is able to push new OS update to "every" iPhone applicable. New OS is optimized for newer phones (with more capable SOCs) and probably less so for 3+ year old phones. You don't see the same search term spike for Android because there is no spike in Android phones downloading new OS updates. Most Android users are not on latest release with in a year of release (maybe years?). Also note, far from all of the iPhone models apple admitted to throttling were actually throttled according to John Poole's data. If Apple's sole intention was to "sabotage" why would they not sabotage all iPhones. Also, the throttling occurred at several different bands supporting the notion the more worn a battery the more throttling.willis936 - Tuesday, January 2, 2018 - link
Umm yes the very first cellular phone that used a battery had a power supply in it with energy storage elements and voltage regulation and every cellular phone since then has had that. They didn't disappear in the iphone 6s. This is a sham.picka - Tuesday, January 2, 2018 - link
Sorry, but phones with good batteries are also slowed down. I have 5% battery wear and my SE sits at 911-1200MHz range. Sometimes at 100% it revs at 1.85GHz, but it lasts for a few minutes and drops to 1200 or 911MHz. It is bullshit that it is due to degraded batteries. Apple should provide a toggle and keep the 29$ dollar price for batteries.HStewart - Saturday, December 30, 2017 - link
"I bet they make a lot of money battery swaps, which was the plan all along."I seriously doubt they make money on battery swaps - my guess almost 90% of battery upgrades are when they get trade in. Typically an iPhone user trades them in every other version. This is first time I didn't - because 1. the 7 did not look like much improvement but I like the 8 better - but X was even more desirable and the 6 has done me great.
Alexvrb - Saturday, December 30, 2017 - link
At $30 they won't be making money off the battery swaps themselves. They will be retaining customers, however.Albertc - Sunday, December 31, 2017 - link
The battery cost is less than $10. They're not giving the farm away.peevee - Sunday, December 31, 2017 - link
But then there is labor and associated costs (buildings, lighting, tools etc).Of course if they would make the batteries replaceable in the first place, there would not be such a problem, but them a lot of people would not change their phones every year or two.
FunBunny2 - Sunday, December 31, 2017 - link
"But then there is labor and associated costs (buildings, lighting, tools etc)."not really. the only marginal/variable cost to the replacement is the battery and labor. all the rest are fixed cost, not related to the replacement exercise. the total margin is cannibalized.
Samus - Sunday, December 31, 2017 - link
Yes battery cost is low, but installation is 20 minutes done properly on 6, and even more difficult on 7’s because of IP68 certification sealing. There are other costs associated with the battery replacement, such as pull strips, lcd gasket and seal, and the possibility of damaging a component during disassembly/reassembly. Since these repairs are happening in the United States where wages are high for technicians replacing these components, Apple may not even be breaking even after parts and labor. But it could be close to breaking even if they contract it out right...piroroadkill - Sunday, December 31, 2017 - link
At the end of the day, the batteries degrade. That's not a scam. An original, OEM new battery, professional replaced, for $29, is a good deal.SquarePeg - Sunday, December 31, 2017 - link
What is a "scam" is that Apple hid the fact that they were throttling performance in an effort to mask their poor engineering choices. This has caused countless tens of millions of people (if not hundreds of millions) to purchase new phones when all that was needed was a new battery. $29 is not a good deal when you consider that this is Apple in damage control mode and they are just trying to gloss over their massively unethical behavior. When Nexus 6P users were hit with random reboots because of "degraded" batteries Google replaced them with new Pixel XL's at no charge.rsandru - Monday, January 1, 2018 - link
SquarePeg: sorry but you're completely wrong. Rising output impedance of aging batteries (rechargeable or not) is just one of the constraints of developing high performance mobile devices. Apple engineers have addressed in the only possible way: to reduce current spikes by throttling the SoC. You can also look at it the other way around: they maximized performance of the SoC as much as possible on fresh batteries or when plugged in...Other phone manufacturers do exactly the same to avoid random reboots and crashes. I agree that Apple have shown dubious business practices with batteries but mixing this up with an electronics engineering constraint is at best and uninformed opinion...
id4andrei - Monday, January 1, 2018 - link
Apple engineers did not take into account these constraints. That's why they first recalled a batch of iphones and then issued a band-aid. Firs they admitted the flaw - the recall - then they hid it - the "fix".It is not reasonable for the iphone 7 to start getting throttled after one year. This is flawed design, textbook planned obsolescence, especially in Europe where 2 years is the minimum mandatory warranty.
In abstract it's a correct decision, in reality it represents flawed design. Average users do not wear the battery to such an extent in one year.
Stuka87 - Tuesday, January 2, 2018 - link
The issue isn't caused by voltage drop. It caused by current drop. Thats what makes the phone reboot.My iPhone 6 had this exact issue. The random reboots were annoying as hell. Having the phone be slightly slower would have definitely been preferred (The code for the throttling was not in place back then).
CharonPDX - Wednesday, January 3, 2018 - link
The cheapest battery replacement I've seen (WITHOUT labor) for an iPhone 6 Plus is $15. $15 for labor for Apple-official isn't exactly "raking in the dough." No way was the plan "to make a lot of money on battery swaps." I'd believe "convincing people to upgrade to the latest phone" much more than "make a lot of money on battery swaps."basroil - Saturday, December 30, 2017 - link
How much should we bet that the "discount" is applied only if you sign a waiver stating you won't sue Apple for this issue? If they are really "crafty" they might even state you accept that discount as final compensation for any and all "features" of the phone!shabby - Saturday, December 30, 2017 - link
So far that's a myth.basroil - Sunday, December 31, 2017 - link
Have you read the service agreement? It actually states pretty much thatsonicmerlin - Saturday, December 30, 2017 - link
My iPhone 6S has never, ever shut down on me. I never even noticed any drop in battery life before I updated to iOS 11. So why am I being throttled to half speed (according to geekbench) at 40% battery?Can you just turn the throttling off you cheapskates? Or just give us the cruddy option in the settings app to disable this new throttling protocol. How hard is that? Settings has already become such a labyrinthine mess, what harm is there in giving us one more option?
SavedByTechnology - Saturday, December 30, 2017 - link
My iPhone 6 shuts down a lot if I let the battery run down below 30%. It’s also slow as hell, especially launching apps like Weather Channel or Twitter.I just installed a battery check app and it’s telling me my battery is “marginal” at 87%.
Alexvrb - Saturday, December 30, 2017 - link
If your phone is shutting down below 30%, is slow, and the battery is old... replace the battery, seriously. You don't need an app to know that battery is shot. Jeebus. Also how many gens old is that phone now? I'm far from an Apple fan and I haven't broke down and bought one yet, but yeesh.ddrіver - Sunday, December 31, 2017 - link
They probably measured the output voltage of your battery and noticed during power spikes it's dangerously close to rebooting the device. So they throttled just in case. I think they took the right decision but it should have been transparent.id4andrei - Sunday, December 31, 2017 - link
They took the right decision in abstract but they still handicapped your phone. You are thus the owner of a phone that performs under its advertised power, possibly while under warranty. Maybe you are out of warranty but throttling started while under warranty and you didn't notice. In Europe, this is called a concealed defect.This is a recall issue not a battery swap. Apple's iphones "burn too bright" and start suffering conveniently at around the 1 or 2 year mark when the warranty expires.
RanDum72 - Monday, January 1, 2018 - link
A degraded battery will handicap your phone regardless. So which one do you want, slow but longer use or blazingly fast but shorter run-time?ddrіver - Monday, January 1, 2018 - link
Finally, a guy who gets it. Apple concentrated performance in the first 1-2 years of operation because they know people change the phone after. Everybody else just spreads the performance and you get a watered down phone that lasts more than 2 years.The only mistake Apple did was hiding this. They should have told people and they would have been able to sell the replacement for more than $29.
id4andrei - Monday, January 1, 2018 - link
Ddriver, this was not an industry standard measure. This band-aid was issued on the background of an iphone recall. Apple thus admitted a flaw otherwise they wouldn't recall anything and just issue the patch. The "fix" just hid what could've been free(under warranty) battery swaps or worse, total recall.id4andrei - Monday, January 1, 2018 - link
This was not a degraded battery under normal operation. This throttling band aid was issued on the background of an iphone recall. Basically Apple admitted that there was a design flaw within the whole situation. After the recall they noticed the issue persisted and issued this patch. This was a sweeping under the rug of a design flaw. Better kneecap all iphones than issue free battery swaps.Strunf - Tuesday, January 2, 2018 - link
It's not really that choice... the choice is do you want a normal operating phone that may reboot without warning or a slow phone that doesn't reboot.If it was a normal operating phone with a shorter battery capacity (like it's the case with every other phone) then people may take that choice and just charge it more often.
Colin1497 - Sunday, December 31, 2017 - link
Your iPhone 6S has never, ever shut down on you BECAUSE it slows down to half speed at 40% battery. Before the software change my old iPhone 6 was shutting down randomly all the time, and the update stopped that. A couple months later I got a brand new iPhone 6 for the $79 cost of a battery replacement from Apple because my out of warranty battery started swelling.I'm far from an Apple fan, but really they only made two, maybe three mistakes:
1) Non-disclosure was clearly an error. It was, however, an error that's consistent with Apple, in that they tend to hide details from customers and try to "just work."
2) Their algorithms should be better tuned and provide more performance when the phone is charged and plugged in. Hopefully with the software update they'll provide some options, like "give me full throttle, I like to live dangerously," "Err on the side of more performance," and their current setting of "please don't let my phone ever shut down unexpectedly."
3) On the smaller screen phones they had battery issues on both the 6 and 6S (maybe 7 & 8?) that were likely the result of being overly aggressive with pushing design margins. In the end this caused these devices to have high battery failure rates and short battery lives. The $29 replacement program probably needs to be extended indefinitely for these devices and any other device with similarly razor-thin design margins.
id4andrei - Sunday, December 31, 2017 - link
Too me Apple designed their iphones wrong to burn too bright. The CPU is too strong for the battery output. It technically works but it is still a design flaw. Some would consider a defect.Notmyusualid - Saturday, December 30, 2017 - link
Seems like a solid response to the issue from Apple.$29 for a battery, replaced?
I don't even think I could buy an S8+ battery for that amount of money where I live...
smartthanyou - Saturday, December 30, 2017 - link
Leaving a customer without a cell phone for a week (when for many it is their only phone), a solid response?The price for the battery replacement is definitely a good deal. However, I customer should be able to make an appointment and have it done while they wait. If they can't do that, loaner phones should be available.
Alexvrb - Saturday, December 30, 2017 - link
Now THAT much is true, it is seriously sucky to go without your phone. The real answer to this is for Apple to figure out how to pack at least SOME user replaceable battery capacity without compromising their design philosophy much. Even if it's like a slide-out battery that doesn't require a split-body design, and only accounts for maybe half the total battery capacity. The side benefit would be the ability to hotswap a battery. Pretty unlikely to happen, though.beisat - Saturday, December 30, 2017 - link
So what about non-US customers?shadowjk - Saturday, December 30, 2017 - link
As for the "How is Apple measuring battery impedance?" question in the article..Around 8-10 years ago me and others were experimenting with this on certain Nokia Linux devices. The basic method is simple: you measure voltage and current, as current increases, voltage increases. Yes the voltage drops naturally with declining state of charge, but keep your time window short and you can ignore that. Calculate impedance from the voltage and current data points.
The challenge was that the hardware updated the voltage once per second and current once every 5 seconds. So, we had to look for two periods of 5 seconds where voltage remained stable, two periods where current use was very different in order to get useful data.
In the end, the only practical approach was to wait until the device was otherwise idle, and periodically peg the CPU with a busyloop to create a synthetic and consistent power draw.
Perhaps today 10 years later, hardware battery fuel gauges are able to report data more frequently, or even report impedance directly.
But I still wonder if they're not simply doing something like: If system shut down unexpectedly, then bump power limit down one notch.
Colin1497 - Sunday, December 31, 2017 - link
My battery charger will measure impedance, I'm sure Apple's charging circuit can do the same. It's really not rocket surgery.BillBear - Saturday, December 30, 2017 - link
Given the choice between a phone that mysteriously turns itself off when you need to make an emergency call and a phone that does not, I'll take the phone that works.The problem here was that if it is necessary to reduce the voltage draw of the device because the battery is reaching it's end of life, then the user should be notified.
This is what happens when you continue to provide updates and security patches to phones that are more than two years old. The phone remains useful past the lifetime of a single battery.
HStewart - Saturday, December 30, 2017 - link
Here is a big question, I have an iPhone 6 and so far no shutdowns, but if the new iOS detects that I need new battery - and I replace it for $29. Does iOS still limited my performance with new battery.If so the problem is not in iPhone 6 but in the new iOS.
I have no problem with my iPhone 6 ( in fact I have another one for work ). I just glad I didn't upgrade to 7 or even the 8 - planning on new version that will be like the iPhone X but size of the iPhone 6 ( not 6 plus or X size )
shabby - Saturday, December 30, 2017 - link
Performance goes back up when a new battery is put in.ddrіver - Sunday, December 31, 2017 - link
Yes, it detects the battery is ok and disables the throttling.HilbertSpace - Saturday, December 30, 2017 - link
TLDR: Good luck getting Apple to replace your 'degraded' battery!!I have an iPhone 6+, bought directly from Apple at release, noticed super slow-downs with iOS 10.2+ (and also iOS 11) Did the requisite Geekbench tests for sanity, noticed the throttling at lower battery levels (50%), charged it up 100%, results miraculously back up to par for the phone. Used a few battery test apps, which indicate degraded performance. Gave Apple a call today after seeing this news, they wanted to do a remote diagnostics, so humored them and did it - here is the exact wording from that effort - "That is normal behavior for the device to slow performance when the battery gets a lot lower. We can do some steps to try and help with the performance, but the battery does appear to be working fine." and then later "Alright, but I’m just letting you know it’s not battery related or related to needing a battery replacement. Your battery passed the diagnostic with flying colors!" Then they wanted me to do a complete wipe/reset, etc. Which I didn't have time for today, so I'll give them the benefit of the doubt (however slim that is), and try that tomorrow. But... I don't expect much from them anymore.
ABR - Sunday, December 31, 2017 - link
What is it about Apple’s devices that necessitates this throttling? I suppose we would have heard by now if it were an issue with Android or Windows devices. And anecdotally, my wife’s iPhone has always shut down quite quickly if she takes it out in the cold, whereas I’ve never seen another phone do that. Do they have a different (well, inferior) power management approach in the SOC, or are their batteries spec’d differently (well, with less margin) than the competition?BillBear - Sunday, December 31, 2017 - link
It is an issue with the lithium batteries in Android and Windows Phones.Lithium batteries are only good for so many recharge cycles. Eventually, they stop being able to supply the necessary voltage even though the phone still says it has some charge left.
When the device demands a higher voltage than the battery can provide, the phone suddenly turns off without warning.
Perhaps the difference is that Android devices only get two years of software updates, so they tend to be replaced before the battery has time to degrade as much.
Apple chose to throttle down devices whose batteries are failing instead of just letting them turn off without warning, which is good, but they failed to warn the user that their battery was hitting it's end of life when they did so which was bad.
FunBunny2 - Sunday, December 31, 2017 - link
"It is an issue with the lithium batteries in Android and Windows Phones."I don't think so. at least, it's not a construct of *a* Li-ion battery. my Braun toothbrush is at least a decade old, has a Li-ion battery, spends most of the time charging (of course), so it's been through thousands of charge cycles. let it run out once a month or so, and it recharges and runs like new. so, it's not a Li-ion issue, per se.
BillBear - Sunday, December 31, 2017 - link
"I don't think so"You might want to read up on it. Lithium batteries have a limited lifespan.
>Lithium-ion batteries age. They only last two to three years, even if they are sitting on a shelf unused. So do not "avoid using" the battery with the thought that the battery pack will last five years. It won't. Also, if you are buying a new battery pack, you want to make sure it really is new. If it has been sitting on a shelf in the store for a year, it won't last very long. Manufacturing dates are important.
https://electronics.howstuffworks.com/everyday-tec...
Prismatic - Wednesday, January 3, 2018 - link
You might want to experience it. I have a 20 year old ThinkPad, 10 year old Toughbook and 9 year old Palm Centro whose batteries perform 90% or higher.The reason why the battery only lasts 2 years is because they're continually kept at 100% which burns them up faster than using them.
ddrіver - Monday, January 1, 2018 - link
Your tooth brush also only has one spike, when it's turned on and jut two modes of operation: on or off.It's nice to see your level of understanding of a complex SoC power usage profile by comparing it with a tooth brush. I have a LED light in my remote and the batteries are 4 years old, Apple could learn from those LED manufacturers and alkaline battery makers.
rsandru - Monday, January 1, 2018 - link
Thatr's a dumb example: your toothbrush doesn't spike at a dozen amperes for a few milliseconds and expects the battery voltage to hold steady... Search for Ohm's law and see what happens to your voltage output when the impedance increases and you try to suddenly pull 10A...cpy - Sunday, December 31, 2017 - link
Capacity down to 10% of original = order of magnitude lowerCapacity down to 1% of original = orders of magnitude lower
Apparenlty we're dealing with idiots here.
ddrіver - Sunday, December 31, 2017 - link
Those same guys can't realize that the capacity is not relevant to the problem. It's the voltage/impedance that's the problem. Both degradations may occur at the same time but it's the second one that caused this whole debacle.FunBunny2 - Sunday, December 31, 2017 - link
"It's the voltage/impedance that's the problem. "no, it isn't. the problem is that iOS 11 made the problem by being one-size-fits-newest-iPhone, thus sabotaging existing units. having an OS that switches functions based on the hardware it runs on isn't rocket science or unprecedented. as I said on one of these posts, Gates had DOS and Windows written to the latest x86 cpu, and devil take the rest. nothing new. Cook faces a slightly different problem: battery progress has always been glacial, and there's some reason to believe that energy density has reached the asymptote of density increase. since we're nearing the quantum problem of smaller node size, another asymptote, dealing with being on the leveling part of the curve is something humans haven't had to deal with before. previous Dark Ages of anti-progress were the result of decisions to ignore science. here, even though some of us are anti-intellectual (guess who?), the issue is reaching the limits of physics.
ddrіver - Monday, January 1, 2018 - link
You'd better stick to toothbrush comparisons. Your other rants are boring and look like the ravings of a drunk.FunBunny2 - Monday, January 1, 2018 - link
an interesting observation, but still doesn't offer a rational explanation for why: older SoC with older batteries and older iOS run fine, but tank when iOS 11 is installed. well, other than Apple thought it a neato way to get folks to toss an old iPhone for a new one. just read the multitude of comments describing that behaviour. again, Apple has the smarts to configure iOS 11 to meet the needs, and just the needs, of older SoC installed. they chose not to do that.BillBear - Monday, January 1, 2018 - link
You call suddenly turning off without warning "running fine"?I'm kind of waiting for someone to file a class action lawsuit against Android device makers because they can't make emergency calls when their device suddenly shuts down without warning.
FunBunny2 - Tuesday, January 2, 2018 - link
"You call suddenly turning off without warning "running fine"?"go through the comments. the number of shutdown when running old iOS on old iPhone with old battery is a small fraction of the iOS 11 on such an iPhone tanking. FWIW. again, with vigour: Cook could have ordered the pointy headed guys to build iOS such that it configured to the SoC/battery combination it found, and didn't implement the "new" iOS functions that cause problems with said old SoC/battery combos. that would have been putting the customer first. make it transparent that older iPhones wouldn't get all the new goodies of iOS 11, so the owner could choose. and, for the sake of argument, how long do you think old iPhones will take to eat that new battery running iOS 11 functions that still clobber older SoC???? a few months, is my guess.
the problem is iOS 11 only works as designed on 8/X phones. but Cook won't tell you that.
Pork@III - Monday, January 1, 2018 - link
https://image.prntscr.com/image/sEFoujk-TKuIwO3LrP...TEAMSWITCHER - Monday, January 1, 2018 - link
Wow...For only $30.00 it's problem solved! I've got two iPhone 6+ phones that will get new batteries and keep on trucking. People will still complain, but for me, this is a cost effective solution.FunBunny2 - Tuesday, January 2, 2018 - link
"for me, this is a cost effective solution."update this comment in 6 months: is that "new" battery now considered "degraded" by iOS 11??? enquiring minds will want to know.
Tams80 - Monday, January 1, 2018 - link
The throttling itself is not an issue. It's what's best for the phones. However, not telling people this is most certainly not acceptable.That this battery replacement price is only going to be this low for a limited time is irresponsible, insulting, and greedy.
id4andrei - Monday, January 1, 2018 - link
If after a year your smartphone gets kneecapped that constitutes a defect. It is not reasonable for your phone to lose performance because the battery is undersized by design. This is a design flaw either way you take it. Basically textbook planned obsolescence.BillBear - Tuesday, January 2, 2018 - link
No, if after your battery reaches it's end of life and your device starts turning off without warning, this is an avoidable problem.How long until somebody sues Samsung because they were unable to make an emergency call because their device powers off unexpectedly, despite showing that it still has a charge?
id4andrei - Wednesday, January 3, 2018 - link
Batteries do not reach EOL after 1 year. This is a design flaw by underspeccing the battery in relation to the SoC's requirements. It's planned obsolescence and the hidden "fix" robbed you and millions others from free warranty service or replacements.Prismatic - Wednesday, January 3, 2018 - link
If you charge the battery once per day, it will reach End of Life in 10 months.FunBunny2 - Wednesday, January 3, 2018 - link
my antique LG phone is on the charger every night for at least 4 years. runs fine.Prismatic - Wednesday, January 3, 2018 - link
I've never used an iPhone but I've tinkered with enough Lithium Polymer batteries over the years to understand the following:*The temperature charging range for Li-Po is much narrower than the discharge range. The higher the temperature the faster it degrades. It's safe to charge up to 45C but not usually advisable. Your phone will last longer if you charge it while not in use.
*The higher the voltage is kept the faster the battery degrades. So in theory if the phone's battery is kept at 100% for 8 hours a day one would lose about 7% of the battery's lifespan over the course of a year.
*There is no standard as to how many charges a battery is designed to go through. Some say 300 others say 500.
*I've understood that if never fully charged or discharged the stresses upon the battery are halved.
*I don't know about the iPhone but some devices will continually cycle the top 5% of the battery under external power in a vain effort to prolong battery life. This practice is highly destructive.
*I've read that some iPhone and Samsung Galaxy battery malfunctions are due to third party adapters and power banks. I know on micro-USB devices the voltage regulation is on the external side rather than within the phone itself.
*The health of a Li-Po pack is correlated to how much swelling it has. Something that can be seen with my Galaxy S3 and S4 that are still perfectly fine with the former being 5 years old.
*Lithium Ion prefers to always be in a constant dis/charging state. So with laptops even cycling them once a week is better than leaving them on the wall for months at a time.
I tend to be rather delicate with my batteries so I get well above the 'two years' expected out of Lithium Polymer. I can't compare Li-Ion cells because most of what I use in laptops are Panasonic and those things are durable.
Prismatic - Wednesday, January 3, 2018 - link
I forgot to add, with cell phones in order to be able to be charged via USB quickly the batteries are on the primary circuit rather than independent like a laptop (i.e. I can turn on the laptop without a battery). So once the battery is charged it cannot 'relax' as it's always being kept at voltage despite not being charged...At least assuming the voltage readouts are not on the controllers in my phones.Kaly - Thursday, January 4, 2018 - link
Hi @Team ANANDTECH kindly review iphone X.harryseth - Wednesday, April 25, 2018 - link
Battery throttling is one of the most speculated outcome of the year. Apple played it correctly and fooled people for over more than a decade. Finally, the accepted what they were doing and came back with another program to make more out of their products - https://ios12guide.comharryseth - Monday, May 28, 2018 - link
Here is how you can download iOS 12 beta IPSW for iPhone or iPad devices. Our download page contains direct links of iOS 12 firmware IPSW. You can use our website to get the latest update for your device.https://iostrends.com/ios-12-beta-download/