Apple Addresses Alleged Throttling of iPhones With Degraded Batteries

[laz232]

As an engineer let me weigh in on this:


An ideal battery is a perfect voltage source and can deliver infinite current at a constant voltage.


In practice, batteries have a ESR (equivalent series resistance) in the order of 10’s of milliohms which increases with battery age.

The quality of the cells used in the battery is also a factor: cheaper manufacturer- higher ESR (Apple outsources battery manufacturing)

Ohm’s law tells us that the battery voltage drop is proportional to the current draw.


Additionally, LiIon batteries have an output voltage that is dependent on its state of charge.

Varying in a non-linear way from 4050 mV (Full) down to about 3300 mV (empty).


If the phone detects 3300mV or so it will issue a shutdown warning to prevent over-discharge of the LiIon battery (potentially destructive) and to ensure there is enough power to shutdown the file system safely.


And this is the problem: due to natural battery aging *or* use of lower quality batteries the ESR increases.

Increased ESR means bigger Voltage drop on current peaks.


Sufficient current drop will cause a temporary dip below low voltage cutoff threshold causing instant shutdown.


Worked example:

Battery voltage 3500mV (corresponds to, say 40% charged), ESR 12mOhm

Peak current draw 2A from CPU and rest of System at full speed.


Voltage detected by onboard controller: 3500mV - (12mOhm*2A) = 3476mV

Which is fine


Now with aging / bad battery, say the ESR increases to 120mOhm.

Voltage detected during peaks is now

3500mV - (120mOhm*2A) = 3260mV

Which is below threshold, say, causing instant shutdown.


Now, is Apple being consumer friendly, by having older phones work longer by reducing peak current consumption to prevent undervoltage shutdown or is it covering up battery manufacturing defects to avoid warranty replacements...?



Written on a mobile.

 

[ThaRuler]

Can this thread go past 1000?

 

[imronburgundy]

The problem here is transparency, not the fact that Apple wants everyone to have a good user experience.

 

[Aydy]

The internet states there are 90 million Apple customers. Should all 90 million have your point of view. Is it okay that people have different opinions of things? I’m not convinced this explains 100% of things anyway as there is a user population not on iOS 11 that still “complains”.

You do you and al do me. I couldn’t bear to be in agreement with you tbh. No offence! But so strongly against this do I feel. What you find is OK ethically and attitudes of the like simply infuriate me, and I can’t dress that up politely, sorry.

 

[Leomon84]

As an engineer let me weigh in on this:


An ideal battery is a perfect voltage source and can deliver infinite current at a constant voltage.


In practice, batteries have a ESR (equivalent series resistance) in the order of 10’s of milliohms which increases with battery age.

The quality of the cells used in the battery is also a factor: cheaper manufacturer- higher ESR (Apple outsources battery manufacturing)

Ohm’s law tells us that the battery voltage drop is proportional to the current draw.


Additionally, LiIon batteries have an output voltage that is dependent on its state of charge.

Varying in a non-linear way from 4050 mV (Full) down to about 3300 mV (empty).


If the phone detects 3300mV or so it will issue a shutdown warning to prevent over-discharge of the LiIon battery (potentially destructive) and to ensure there is enough power to shutdown the file system safely.


And this is the problem: due to natural battery aging *or* use of lower quality batteries the ESR increases.

Increased ESR means bigger Voltage drop on current peaks.


Sufficient current drop will cause a temporary dip below low voltage cutoff threshold causing instant shutdown.


Worked example:

Battery voltage 3500mV (corresponds to, say 40% charged), ESR 12mOhm

Peak current draw 2A from CPU and rest of System at full speed.


Voltage detected by onboard controller: 3500mV - (12mOhm*2A) = 3476mV

Which is fine


Now with aging / bad battery, say the ESR increases to 120mOhm.

Voltage detected during peaks is now

3500mV - (120mOhm*2A) = 3260mV

Which is below threshold, say, causing instant shutdown.


Now, is Apple being consumer friendly, by having older phones work longer by reducing peak current consumption to prevent undervoltage shutdown or is it covering up battery manufacturing defects to avoid warranty replacements...?



Written on a mobile.

I am an electrical engineer also.

Straight answer is that if what Apple says is true, they under sized their batteries. A 1000$ phone should have batteries that can handle peak current draw after 1 year or more of usage.

In the industry it is an IEEE standard to oversize size batteries to include battery degradation so that after 5 years you can still expect your UPS to have 4 or 8 hours of backup.

We do this all the time in heavy industrial designs. Its common knowledge..

 

[Blaze4G]

That's correct. If the choice is a dead phone or a slow phone. Slow wins. Unusable is subjective. Dead phone is not.

Yep, slow unusable all the time Vs dead temporarily.

 

[617660]

Well that is where we disagree. I believe they have every right to update power management, go through the eula and other stuff available online.

You of course can start a class action lawsuit, or be part of one, which I'm sure is being formulated at this very minute.

I think apple did the right thing, imo.

You can have any opinion you want.

This is mine.

You are using EULA as a defense mechanism now? If that thing ever means anything we wouldn't have thousands of product-related lawsuits and recalls every year.

In fact, let me summarize EULA for you:

"Apple can do no wrong. Any fault you find in our device is your fault and solely so. Apple has the freedom to change anything on the device without assuming any liability. You poor citizen, if you ever want to get your hands on our device, you have to agree to this and surrender your consumer right."

Hey...I did read it.

 

[DrBrush]

Sorry if this has been asked and answered but there are 39 pages of posts...

Is this true for iPads too? I have noticed my iPad Air 2 slower on iOS 11.

 

[cale508]

This "feature" should be called the "iThrottle"

 

[Aydy]

Sorry if this has been asked and answered but there are 39 pages of posts...

Is this true for iPads too? I have noticed my iPad Air 2 slower on iOS 11.

I would say Most likely. The code is written to any iOS device where battery has depleted Below a specific threshold. If it is coded into iOS software then all iOS devices are vulnerable. At least we now know when it’s ocurring and can consider battery replacement. Had we not been made aware we would have just assumed it was time to upgrade to newer. So I guess we can thank the guy who called this out

 

[I7guy]

Yep, slow unusable all the time Vs dead temporarily.

Unless you need slow at a point where you might dead permanently. Okay I'm having way too much fun arguing discussing this.
[doublepost=1513878038][/doublepost]

You do you and al do me. I couldn’t bear to be in agreement with you tbh. No offence! But so strongly against this do I feel. What you find is OK ethically and attitudes of the like simply infuriate me, and I can’t dress that up politely, sorry.

No worries. I on the other hand, don't understand what all of the hoopla is about in the myriad of things that come along with life.

 

[Analog Kid]

Yes, Apple should have designed a phone that will perform close to spec a year after it was purchased. Which should take into account higher internal resistance after a year. A phone should work as advertised for only a single year (or less)..

That is good design. Bad design is covering up the fact that you either under-specified the battery, or failed to account for the aging of the battery.

Proof of that failure is that the initial ios shipping with the phone did not have this behavior.

So you appreciate that your premium SMART phone turns into a dumb phone after a year?
Please show me these reference notes.

I'm not sure what reference notes you're referring to.

10.2.1 was released 16 months after the 6s, so repeating "a year" over and over again undermines your point. Nothing we're discussing here affects anything advertised by Apple, so while you might be upset that there's something happening under the hood, claiming that they aren't delivering what they advertised is undermining your point.

Also you keep suggesting that this is happening with certainty to everyone you're addressing. I don't know if you know how to read a statistical distribution, and I'm not going to bother carefully integrating the probability densities, but it looks like maybe 20% of phones that submitted themselves to Geekbench were affected at all. There is selection bias in these results. People running Geekbench may be more likely to be power users, and power users may be more likely to deep cycle their batteries. Apple spec for warranty battery repair is 80% health after 500 charge/discharge cycles. 10.2.1 was released 486 days or so after the 6s, so if you charge your phone more than once a day, as you would almost certainly need to if you're regularly running up against the performance limits of your device, then you're likely past the warrantied life of the battery.

We don't have multidimensional data so all we can do is look at broad generalizations. There's a lot of inference happening from what we have, but I don't see anything that can be called "proof of failure". What I see is that right around the time we'd expect to see the first heavily used batteries start to degrade, Apple started mitigating the risk of random restarts. I don't see anything insidious in that, I see incremental improvement in reliability via a point update which is exactly how the release notes identified it.

Good design is giving the user the most from what they pay for. This throttling mechanism is doing that. Your alternative of maintaining identical performance for some arbitrary time period (measured by you in years, not battery cycles, so it's only loosely correlated to the actual hardware constraints) means denying the user the full capability of their hardware for that same time period.

Better batteries won't change this truth. All batteries degrade, so setting the maximum performance at the worst case, high usage, 1 year, cold temp performance level will always mean you get less performance out of the box.

 

[whsbuss]

Question: If I get a battery replaced with a new one, does Apple's smart ware remove the throttle?

 

[Leomon84]

https://www.forbes.com/sites/gordon...phone-performance-ipad-problems/#2f58fa6db8e7 - Good Read

 

[KingslayerG5]

Look, at the end of the day people are still going to buy more iPhones. Nobody is going to leave their prison..uh..I meant their closed ecosystem in droves.

People are a little ticked off right now like Rian Johnson doing a character assassination to Luke Skywalker. But this heat of the moment will go away. People online just like being mad and complaining.

At the end of the day, I'm still going to watch Star Wars: Episode 9 - Rey's Anatomy (or Fifty Shades of Rey) and whatever SW drivel Disney tosses at me. MCU or SW, I'm there. And for Apple, same thing. I'm sorta forced to use iOS as a secondary for Apple Maps.

I don't even like iOS but I got to be like some who has to use both Android and iOS at the same time for options. We're slaves to them. We will be angry for a week but we will still continue to make Apple and Disney rich.

Present: We're angry like the Hulk. Waah waah.. Disney is ruining Star Wars. Waah waah.. Apple is deliberately slowing down my iPhone. I HATE THEM BOTH!

Future: Shut up and take my money!

 

[Leomon84]

I'm not sure what reference notes you're referring to.

10.2.1 was released 16 months after the 6s, so repeating "a year" over and over again undermines your point. Nothing we're discussing here affects anything advertised by Apple, so while you might be upset that there's something happening under the hood, claiming that they aren't delivering what they advertised is undermining your point.

Also you keep suggesting that this is happening with certainty to everyone you're addressing. I don't know if you know how to read a statistical distribution, and I'm not going to bother carefully integrating the probability densities, but it looks like maybe 20% of phones that submitted themselves to Geekbench were affected at all. There is selection bias in these results. People running Geekbench may be more likely to be power users, and power users may be more likely to deep cycle their batteries. Apple spec for warranty battery repair is 80% health after 500 charge/discharge cycles. 10.2.1 was released 486 days or so after the 6s, so if you charge your phone more than once a day, as you would almost certainly need to if you're regularly running up against the performance limits of your device, then you're likely past the warrantied life of the battery.

We don't have multidimensional data so all we can do is look at broad generalizations. There's a lot of inference happening from what we have, but I don't see anything that can be called "proof of failure". What I see is that right around the time we'd expect to see the first heavily used batteries start to degrade, Apple started mitigating the risk of random restarts. I don't see anything insidious in that, I see incremental improvement in reliability via a point update which is exactly how the release notes identified it.

Good design is giving the user the most from what they pay for. This throttling mechanism is doing that. Your alternative of maintaining identical performance for some arbitrary time period (measured by you in years, not battery cycles, so it's only loosely correlated to the actual hardware constraints) means denying the user the full capability of their hardware for that same time period.

Better batteries won't change this truth. All batteries degrade, so setting the maximum performance at the worst case, high usage, 1 year, cold temp performance level will always mean you get less performance out of the box.

I have already responded to this.

Including Battery degredation into the design will lead to an overall better experience and less throttling required over the lifetime of the phone. All you need to do is oversize the battery slightly. Apple engineers are not stupid, they make good products. They willingly knew this was going to happen.

 

[mtneer]

I am an electrical engineer also.

Straight answer is that if what Apple says is true, they under sized their batteries. A 1000$ phone should have batteries that can handle peak current draw after 1 year or more of usage.

In the industry it is an IEEE standard to oversize size batteries to include battery degradation so that after 5 years you can still expect your UPS to have 4 or 8 hours of backup.

We do this all the time in heavy industrial designs. Its common knowledge..

That is if you have real engineers running the show. When you have half-wits like Apple's bald guy calling the shots.....

 

[jjjoseph]

The part about "Apple offers battery replacement in stores" is not really true. I've gone twice to an Apple store to have them assess my battery. They refused to let me pay them to replace the battery because their diagnostics passed.

I'm in the situation where Apple won't even let me pay to replace my battery even though I have random shut downs and blatant CPU throttling when below 50%.

I have dealt with this same exact thing. Is their a way to purposely degrade Lithium-ion battery so it will show as a bad battery to apple so they will replace it? My battery is at like 75/80 some percent and my iPhone6S is definitely throttled. It is also only 6 months old because of a refurbished AppleCare replacement. My iPhone should not be this slow.. The guts are still decent..

 

[trifid]

Just curious, does anyone know how Samsung or other Android manufacturers deal with this battery issue?

 

[Blaze4G]

Unless you need slow at a point where you might dead permanently. Okay I'm having way too much fun arguing discussing this.
[doublepost=1513878038][/doublepost]
No worries. I on the other hand, don't understand what all of the hoopla is about in the myriad of things that come along with life.

Me too! Or unless you need speed at a point where you might be dead permanently! *someone is chasing you* can you stop chasing me so fast! I need time for my phone to load the phone app!

 

[opfreak]

That’s correct if the choice is a slightly slower phone or dead phone.. slow wins.

Or apple could make a phone that doesn't slow down after a year. Kind of like every phone before the 6/s. I know my phone didn't come with a warning on it that said phone will run 20-50% slower after a year
[doublepost=1513879799][/doublepost]

I have already responded to this.

Including Battery degredation into the design will lead to an overall better experience and less throttling required over the lifetime of the phone. All you need to do is oversize the battery slightly. Apple engineers are not stupid, they make good products. They willingly knew this was going to happen.

I don' know if they did. After all is was patched in after phone were on the maket

 

[jgdeschamps]

Question: If I get a battery replaced with a new one, does Apple's smart ware remove the throttle?

According to the article, yes.

 

[AdonisSMU]

Or apple could make a phone that doesn't slow down after a year. Kind of like every phone before the 6/s. I know my phone didn't come with a warning on it that said phone will run 20-50% slower after a year
[doublepost=1513879799][/doublepost]
I don' know if they did. After all is was patched in after phone were on the maket

The phone does slow down if there is significant wear and tear on the battery. Its not about if its a year or not. Every phone manufacturer has this problem currently.

 

[Phonephreak]

I get how you feel, but for all you know this is done on Androids as well.

We aren’t talking about android phone. This is all iPhone.

 

[Williesleg]

I knew it!

That's the end of Apple. I'm loading Ubuntu on my Mac Mini.