19

when I look at my Nexus 5 with android 5.1. battery screen, I can see something like this:

Screenshot Screenshot
Screenshots (click images for larger variants)

Sum of all per-app percentages in list is 10. However as you can see, there is 60% battery consumed (40% remaining). So, what is the deal here? Who consumed 50% battery? Or what do the percentages next to apps mean?

As you can see, screen was on just for few short times, the same for awake, and yet battery life about 18 hours? I dont get it.

Also, I found this issue reported on AOSP, but noone from google seems to care: https://code.google.com/p/android/issues/detail?id=80131

  • Find a raw answer below. Some more clues might turn up when you tap the graph on the top of your screenshot – so if you can make a screenshot of that screen as well, we might be able to tell more (and I might update my answer then ;) – Izzy Apr 28 '15 at 16:17
  • Thanks for the second screenshot! I've updated my answer accordingly, as promised :) – Izzy Apr 28 '15 at 19:48
  • I was frustrated by this, so I installed GSam (best non-root app I could find), and the battery usage stats that it reports are radically different than what the built-in app says. One or the other (or both) is just plain wrong, and since GSam's stats are at least self-consistent, unlike the Battery settings page, I'm guessing that the built-in app is broken. – Josh Aug 9 '15 at 17:36
14

There's usually a wrong conception on how these "numbers" should be interpreted. For example, your screenshot shows "Screen: 3%". So what does that mean?

  • a. the screen ate 3% of your battery, thus reducing it e.g. from 100% to 97%
  • b. the screen was responsible for the overall discharging by 3% – in this case, it "ate 1.8% juice"

Hint: it's not a) – so your numbers just account for 6% drain (10% of 60% drain).

Second, not all participants of the party were named – there's a threshold. I'm not sure where exactly that threshold lies, but it's safe to assume that somewhere below 1% of participation, they simply get "uninteresting".

Which still leaves the question open: what was responsible for the remaining 90% of the 60% "battery loss"? To find out about that, you will need a helper to sum up everything. I recommend you taking a look at my list of apps to Identify Battery Consumers. Personally, I'm using BetterBatteryStats – which at least in the "tech scene" (especially on XDA) is "the reference product". Less technical oriented users often hold to GSam Battery Monitor.


Now lets go for your second screenshot. Focus on the bars below the graph:

  • Cellular Network Signal: You had pretty strong cell signal – except for several minutes before ~10 am (where you've been out of coverage – red) and after ~10 am (weak signal – yellow). This bar looks fine.
  • GPS On: Not a big deal here, just two tiny stripes. If you see big colored areas in this bar, there might be apps flooding your GPS modul with requests, which quickly drains the battery.
  • Wi-Fi: Your WiFi was turned on all the time. Unless you've used it all that time – and especially when you could not use it as there were no network/AP you wanted to use), you might consider turning it off when not used. There are WiFi automatics which can take care for that, see e.g. Auto disconnect wifi when out of wifi zone? – otherwise it uses power to permanently check for available hotspots. Not a big issue in your case it seems – as according to the first screenshot, WiFi accounts for only 1% of overall consumption.
  • Charging: In all the time, you've never connected a charger. Good to know :)

Ah – you've noticed I've skipped two bars? Yes, and two important ones at that. These two must be seen together: Your device was "awake" more often than your screen was turned on – so some apps get active while you're not actively using your device. That might be OK (and doesn't really look excessive here) – think of background-syncs to check for new mail, updates for your wheather widgets, Spotify streaming music to your ears (no screen-on needed for that), and the like. Though, "big blocks" in the awake-bar with no corresponding part in the screen-on-bar should draw your attention. Using the Spotify-example, you might be able to easily account for them (did you stream music at the time? Oh, OK: then that's it, probably).

Again, you cannot always correlate those with the first screen. For a really thorough analysis, you would need a specific app. And the really good ones when you want to "dig deep" are not coming for free. My first look would go to the paid version of SystemPanel and comparable candidates, which can to good background monitoring (check out the screenshots; you will also notice this app doesn't forget about stuff having participated with 0.2% only). You can give the free version a look if you're curious, it can do most of the stuff as well.

  • How can be build in tool so inaccurate? not accounting for 90%? As for BetterBatteryStats: "Due to major changes by Google in Kitkat root access is required from 4.4 and upward" - well, I am not gonna root my phone for this... – rouen Apr 28 '15 at 17:25
  • You had to scroll until your fingers are wound if you wanted to see through them. If, as I wrote, the "other processes" consume less than 1% each, and you want make up 90% of usage – simple math explains why :) But I agree with you: that built-in tool is far from being perfect. Still it is sufficient in most "normal cases". // BBS: forgot about that root issue indeed (newest Android version in use here is 4.3 – and all my devices are rooted). Pick another one from the list then: how about GSam? – Izzy Apr 28 '15 at 19:24
  • Thanks for update... as for wifi, I am not considering turning it off and on again 4 times a day (home wifi, work wifi).. and I certainly didnt buy Nexus5 smartphone to be offline whole day to save battery :) And I am not sure about that saving - if I turn off wifi, 3G/LTE will turn on and that can be more consuming that strong local wifi – rouen Apr 29 '15 at 10:24
  • That's often depending on your device (my LG O4X e.g. uses much more power when on WiFi) and availability (if there is no WiFi your device can log into, it will fall back to 2G/3G/LTE anyway – but in addition to that keep up its permanent scanning for some WiFi becoming available). So if you're sure there's no "usable" WiFi in reach, it definitely costs you extra power when turned on – without any benefit. – Izzy Apr 29 '15 at 11:03
  • SystemPanel (I have the paid version) has its own weird quirks... for instance, on my phone it shows average CPU use running close to 10% all of the time and yet battery use frequently dips below 3%/hour; furthermore, if I (mentally) add all the CPU time reported by every app, it doesn't add up to 10% either! – Michael Dec 7 '15 at 23:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.