Since the speculation of power throttling on iPhones with old batteries has been confirmed by Apple, I'm going to go ahead and change the battery on my 6S. I'm cheap, and I've swapped iPhone batteries before, so I'm doing it myself to save money. I figured I'd start a thread for science and perhaps to help those that might consider doing the same thing. As always, do your own repairs at your own risk. Just thought it would be cool to get some data on the thresholds.
The goal is post your device stats using Coconut Battery, how it benchmarks (reboot, let it rest, then post the best of 3 runs, allowing for cool downs), and then again how it does after the battery swap.
From Coconut Battery:
iPhone 6S (TSMC A9)
Age: 829 days old
Old Battery:
Cycle count: 255
Full charge capacity: 1533 mAh
Battery health: 89%
New Battery:
Cycle Count: 0
Full Charge capacity: 1688 mAh
Battery Health: 98.4%
3Dmark Icestorm:
Baseline 6S: 28051
Old Battery: 17686 (-37%)
New Battery: 28072 (+0.1%)
So even though my old battery was reporting 89% of design capacity, I saw a 37% drop in expected performance, which quite honestly, I could feel in day-to-day use already. The old battery dropped pretty quickly through the day, so that 89% health seems dubious. We'll see how the new battery holds up.
The new battery is a Yontex 0-cycle battery from Amazon. It came with everything I needed to replace the battery, including instructions (which seemed similar to iFixit's guide). They even threw in a new screen protector.
EDIT:
Conclusions:
New battery fixes all. Benchmarks aside, I could tell immediately upon startup that the phone was much more responsive. I'm surprised at how noticeable the improvement is--and to think I blamed iOS11 for my performance problems all this time!
The thing that really surprised me was the 89% health claim of Coconut Battery on the old battery. If this reading is accurate regarding what the phone reports, could this be why people have been getting turned away by Apple regarding replacements despite having issues? Speculation, I guess, but I could see that becoming a big concern if my phone was still under warranty and this report of 90% resulted in a denied repair. It makes me wonder what battery property the phone is using to determine when to throttle. If it's not net capacity--maybe cycle count? Does it drop once it goes below 90%?
The goal is post your device stats using Coconut Battery, how it benchmarks (reboot, let it rest, then post the best of 3 runs, allowing for cool downs), and then again how it does after the battery swap.
From Coconut Battery:
iPhone 6S (TSMC A9)
Age: 829 days old
Old Battery:
Cycle count: 255
Full charge capacity: 1533 mAh
Battery health: 89%
New Battery:
Cycle Count: 0
Full Charge capacity: 1688 mAh
Battery Health: 98.4%
3Dmark Icestorm:
Baseline 6S: 28051
Old Battery: 17686 (-37%)
New Battery: 28072 (+0.1%)
So even though my old battery was reporting 89% of design capacity, I saw a 37% drop in expected performance, which quite honestly, I could feel in day-to-day use already. The old battery dropped pretty quickly through the day, so that 89% health seems dubious. We'll see how the new battery holds up.
The new battery is a Yontex 0-cycle battery from Amazon. It came with everything I needed to replace the battery, including instructions (which seemed similar to iFixit's guide). They even threw in a new screen protector.
EDIT:
Conclusions:
New battery fixes all. Benchmarks aside, I could tell immediately upon startup that the phone was much more responsive. I'm surprised at how noticeable the improvement is--and to think I blamed iOS11 for my performance problems all this time!
The thing that really surprised me was the 89% health claim of Coconut Battery on the old battery. If this reading is accurate regarding what the phone reports, could this be why people have been getting turned away by Apple regarding replacements despite having issues? Speculation, I guess, but I could see that becoming a big concern if my phone was still under warranty and this report of 90% resulted in a denied repair. It makes me wonder what battery property the phone is using to determine when to throttle. If it's not net capacity--maybe cycle count? Does it drop once it goes below 90%?
Last edited:
