My recent experience with my battery failing has me wondering could the phone detect & report this?
I am assuming that the slow wakeups I reported were due to the battery being unable to hit the right voltages or respond to increased demand fast enough which I would expect to be detected & flagged by the hardware though I saw no sign of this in journalctl.
If my assumption is correct is this something that the hardware can/does expose to the OS and if so wouldn’t it be nice to have some form of warning in the battery page of settings and maybe a notification if whatever algorithm is more than X% sure that there is an issue?
Information about the battery, e.g. capacity and health, is stored in /sys/class/power_supply/battery, I imagine Battery Buddy reads the information from there. My Xperia 10 II has apparently run through 854 cycles so far, is at 77 percent capacity and in ‘good’ health.
No, “Battery Health” does not indicate long-term status or effects.
If it goes from “Good” to something else it’s because of too high or low temperature, or overvoltage, or things like that.
Details are driver-dependent, see the Linux Kernel Power Supply Class documentation for details.
However both Battery Buddy and the /sys/ system report my very sick battery as healthy.
@nthn It seems there is some luck of the draw here, my battery has had significantly less cycles than yours and yet is at a lower capacity, what have you been doing with your phone that I may learn?
I still suspect that the very slow wakeups and their reason [which I suspect to be battery related] are probably registering somewhere in the hardware but it doesn’t seem to bubble up to the OS.
It’s my daily driver and since 2G/3G has been shutdown where I am falling back to my xperia x or JP1 is not an option.
I’m still waiting for my battery in the mean time the current battery is making my life when I want to replace it easier since it has detached half the front (screen) and half the back of the phone.
This is why I don’t trust general SMART reports either; the calculation of usage cycles/ age etc. says something about the hardware, but nothing about its actual health.
Things break for other reasons than old age.
Nobody seems to have found an automated way to check the actual health of hardware…? Either that means the industry is in cahoots to force us to buy more, or there are hurdles that make it impossible.
It can be only estimated statistically, but not detected somehow. But for the statistic you may need huge number of tests fir beginn of life, mid of life and end of life at least. Then based on voltage and max capacity you can extrapolate the numbers and estimate the actual state. Not easy
It’s kind of like the philosophical debate on existence, are we just a simulation, meaning etc. it is impossible to detect that from the inside…
I think in this case if the PMIC is reporting back to the OS it can report events where the battery is unable to perform and that is probably a good indicator that something is wrong.
My new battery finally came today and added a twist to the story -
During the installation I realised that the phone still had an SD card in it which I didn’t use most of the time and wasn’t showing in the storage overview, after installing the new battery I still had the yellow light issue ( [bug?][10ii][5.0.0.73] Slow wakeup with yellow LED ) which resolved after removing the SD card, I haven’t yet checked the card in a computer but it looks like another card killed by a phone.
