Limiting charging to 80% to improve battery lifetime
Affected hardware version
Bangle 2
Your firmware version
Any
The bug
Bangle seems to charge to 100%. That will reduce battery lifespan from maybe 8 years to maybe 4.
Given that battery can last for few weeks, limiting charge to 80% might be good idea (at least for times when you don't plan to use GPS).
So far I have this: it attempts to vibrate when we are at 80% to make me disconnect the charger. If there's hw trick to stop the charging, that would be better. It probably needs more experiments / calibration to vibrate at right time.
https://gitlab.com/tui/tui/-/commit/b3780e1d169fdd09fc1eb0db94b2143dfb1b68d8
Installed apps
No response
Sounds like a cool idea - I do find that if, while charging, the watch reports 100%, it's actually more like 80% when I unplug, since the voltage read is that of the battery + what the charger is putting in, so I figure if the watch is at 100% for a little while, it's ok since it's more like being at 80% anyway.
But halting changing via hardware sounds good - I do the same with my phone. @gfwilliams do you think this is possible with what the firmware can do?
So for me, if I unplug around 95% indicated while charging, I get to ~80%, but it is not too consistent. (Which is what the code does).
I'm experimenting with turning on backlight & gps at ~90% to slow charging down (and maybe getting more consistent results from the voltmeter). But if there's GPIO to control charging, that would be even nicer :-).
I'm not aware of anything you can do in hardware to stop charging - that's not to say it's not there, but I never found it. You could try fiddling around with some of the unused pins while charging in case one of them does affect it.
One thing to consider is I know a full charge will reduce life in a phone, but I believe that can be to 4.3v or so. Bangle.js only charges to 4.2v absolute max, so you may find that it's already charging to what a phone would consider 80%
... You could of course have some external hardware if you want to do it. Like you could power the charge cable from a Jolt.js, then advertise on Bluetooth when there's enough charge and get the jolt to turn it off
Yeah, so the "external hardware" that cuts the charge will be "me" in this case :-).
And yes, 4.35V charging is done in some new phones, but I believe those have special cell chemistry. I don't believe Bangle.js2 has one of those cells.
Not sure how applicable this is to smartwatches, but someone analyzed different ways of managing battery charging for phones in this youtube video: Is Fast Charging Killing the Battery? A 2-Year Test on 40 Phones.
They basically come to the conclusion different alleged battery health conserving efforts (e.g. keep within 30%-80%) are not super effective and you may as well just charge away without thinking too much about it.
Dunno. Clearly the author spent lot of time doing the experiments, but... he clearly does not have data for his conclusions.
He clearly demonstrated that 30%..80% charging makes sense. Whether the battery degrades by 12% or by 8% in half a year is not insignificant. Note that control only lost .5% capacity during that time. And he failed to demo the "bad idea" strategies. Leaving the thing plugged in at 100% for a year vs. plugged in at 60% for a year, for example. Always charging from 80% to 100% is not going to be good for the battery, either.
....plus he did experiments on high-end phones with reasonably advanced charge controllers. We have very simple/basic CC-CV charger AFAICT.
I believe the difference is easy to demonstrate between brands. Nokia N900 was pretty good keeping their batteries healthy. Droid 4... not so much. And that difference is from charging strategies, whether it is human doing the control or charging circuit.
(And he basically would have demonstrated significant difference between brands, if his results could be trusted. Androids lost 8% capacity during testing, iPhones lost 12%. Please also note aging is not linear).