Musings & Meta

Addendum to phone/tablet battery stuff

I had some good conversations with people coming out of my PSA on your phone/tablet battery and wanted to write some more.

Compromises

I wrote that your best bet is to keep your phone between 48%-58% at all times. At that rate, you’ll get 12 years of life from your phone. (not proven by experimentation, just statistics). This is stupid though, who has time for that?

There’s a parabolic graph that goes with the peak voltage numbers. It has 100% as the worst voltage you can keep your phone at*, and gets better the further away from 100% you get. But the curve is parabolic, so the further away from 100% you get, the less “health per percent” you get. Someone determined that 58% was the sweet spot, whereby you got the most use along with the most battery health. But 58% may not be the best spot for you. I know some people who wouldn’t get a full day use out of 58%!

I looked at my own usage and determined that for me, 80% would be best. On any given day I my phone would be used to ~45% and wouldn’t be stressed up to 100% every day. So that’s where I’d charge to, if I had the ability to choose. You would have to make your own choices.

*Actually, the article writes that 110% is the worst you can get. Did you know that your battery could contain more power than it current does? This is a corollary to what I said — manufacturers have made battery life decisions for you, and they’ve determined that your current “100%” is the sweet spot! You could charge more, and get more daily life from your battery, at the cost of increasingly poor lifespan! (Also, Li-Ion batteries explode when overcharged. >.>)

Games Suck

I touched briefly on games and how horrible they are for your battery. I mentioned that they put out a ton of heat, which is bad for your battery.

But there’s something else you may have noticed while playing — your battery is used at a much faster rate. You’ll remember that we were measuring battery lifespan in discharge cycles. If you play Hearthstone for 15 minutes and use 5% of your battery, you’ve discharged it a lot faster than if you were just playing Sudoku for 15 minutes and only use 2%. I mentioned the “really good CPUs” — Hearthstone has more graphics going on, so it spins those up and uses more power more quickly. Which means you’re running through your discharge cycle much quicker!

If you only get 300 discharge cycles, and you use those cycles over 300 days with Sudoku (because you just love Sudoku so much you have to run from 100%-0% every day), you’ll get far fewer days playing Hearthstone, because you’ll have to charge your phone in the middle of the day.

Unfortunately, if you want to play games on your mobile device your best bet is to buy a device that’s cheap to replace, or one with a replaceable battery. Because your game playing will destroy your battery and there’s nothing you can do about it.

Battery Meters

This was all new to me, I did the research. There was some fun stuff. Battery Meters are fucky.

Some of them will lie to you, first thing in the morning. I’ve never seen it, but apparently sometimes your meter will drop to 95% in the first 3 minutes of use!  I mentioned last time that chargers stop charging at 100%. But if your phone is on, it will still use power. Rather than keep the charge at 100% (bad for battery), your phone will stop charging. And then lie to you, claiming it’s at 100% even though it’s at 95%. This is a good thing! (from a battery health point of view).

This next bit is totally messed up. A co-worker asked why her old phone from went from 30% to 10% to shutting off over the course of a couple minutes. I had to buy an ACM article on the topic to find the complete answer, but I did it for you and I’d do it again.

At our present level of technology, your battery meter uses the current voltage level of your battery to determine how much capacity it has remaining. (Remember I didn’t define peak voltage in the last article? This is what I meant). There are a set of look-up tables built during development of your phone that equate certain voltages to certain remaining percent. There are different look-up tables for different points in your batteries life, and it chooses which table to use based on what the software finds while it is charging. Was the charge faster than the month before without getting more current? The software may decide to use a table that titled “getting on in life”.

With the 30-10-off scenario, what’s happened is that the software/look-up table set isn’t good enough. It thinks your battery is “pretty ok”, but it’s actually “not super great”. So it uses the wrong look-up table and the voltage drops faster than that table is expecting and whoops now your phone is suddenly off.

Given this, how can we re-calibrate the meter? You can’t. If you were able to, you’d also be able to set all sorts of other things about your battery meter, and you just aren’t smart enough to have that kind of power in your hands. Worse, I researched how to get this power…if you thought you were smart enough to use it right…and no one has looked into it. Lots of smart people on the XDA Developers forums (a place for phone developers), and this is a topic that no one is interested in.

Lastly – why does your phone shut off? You’d think it’s because it’s out of power, but that isn’t it. It’s because if it were to use any more power, the battery may not be usable anymore. If you run a Li-Ion battery to it’s actual 0%, it can’t be recharged! So again, the meter lies to us to save us – this one I have no problem with though!

TL;DR: If your battery meter is acting weird, you probably need a new battery.

The End?

Maybe. My articles on glue certainly went on longer than I expected. 🙂 But the take home message is that your battery will die, eventually, no matter what you do. Using your phone kills your battery.

You Might Also Like

No Comments

    Leave a Reply