Tracking two years of laptop battery use, sampled every minute | Hacker News Reader

Tracking two years of laptop battery use, sampled every minute

ifweassume.com

113 pointsdavidbarker35 comments

This doesn't involve the quantified self component but:

CoconutBattery Online (http://online.coconut-flavour.com/) has massively aggregated Coconut Battery data ("over 150,000" data points) from the last five or eight years of Apple laptop models.

It helps give a sense of laptop cohort's battery longevity.

The method he's using might influence the statistics negatively a bit too much, as there's periodical data that needs to be flushed to disk, plus an useless fork going on every minute.

This popped up on HN last year, uses a much cleaner implementation:

http://www.thregr.org/~wavexx/hacks/bcmon/

"A fun thing to notice: my computer apparently wakes up a few times every night... I wonder what it's dreaming about?"

The fix for this is disabling "Wake for network access" in the preferences.

Though note that "fixing" this will remove your mac's ability to handle network requests when asleep, which is a pretty cool feature that I wish the phones/pads had (they sleep the network cards to save power)

http://support.apple.com/kb/ht3774

This may also be due to the Power Nap feature: http://support.apple.com/kb/ht5394

Maybe I'm just missing it, but one statistic I would like to see is how long it takes your battery to fully recharge, and how that statistic has changed over time.

The definition of "full" decreases steadily over time, all the way to the point where "full" is no power at all.

Remember, two things degrade batteries - heat and load. The former more so than the latter, but they both go hand in hand. The heat is not a data point variable the author has I imagine, but he has a decent representation of state with the graph.

The first MBA seems to have more moments of "drained" battery, represented by the orange/red colours being more predominant on the left side of the graph. This may signify more load being placed on the first MBA than the second MBA.

The second MBA seems to have had an unusual beginning to its life - lack of daily data and perhaps being left in a discharged state rather than charged. Might have some play on the initial degradation of the battery's capacity, but probably not.

Would minute to minute logging effect the battery performance and thus make this data not entirely objective?

I would think that even if logging used a non-negligible amount of battery power to take measurements the data are still valid because such a load was applied consistently throughout the sample period. So, the measurements are valid within the set (you can accurately compare one day to another) but perhaps not transferable/generalizable (could not compare them to other laptops with differing monitoring periods).

note the temperature is an important consideration instead of just the number of charge cycles. for the whole story on a battery the log should include battery temp, even when the computer is asleep.

https://www.apple.com/batteries/notebooks.html

Is percent charged also a factor?

If I understand correctly, I've read that fully charging a hot battery will degrade the capacity.

If true, I'd like smart recharging to target some optimal charge, say 95% full. (I think one of the PC laptops did this.)

"Too much media/computer use isn't good for your brain. I worry computers are making people less creative in some ways, and too much time online is certainly bad for your soul. You need sunlight, air, dirt." Such statements abound but actually don't make much sense. It's just the progress our society is making and we should be proud and happy of it because it absolutely enhances our productivity. People just need to adapt to new things over time and overcome their resistance. I bet when pen and paper became popular similar statements were made of them! Which are of course quite nonsensical examined by time.

If you want to run this can of experimentation on your own, you can easily get these data points from the command line using iStats[1].

[1] https://github.com/Chris911/iStats

>This newer 2013 MacBook Air is holding up much better than the 2012 model, and I'm consistently still getting 6-8 hours of life out of the battery at least.

It may be just my usage patterns but I never seem to get more than 4 hours on the 2013 Air. I mostly have a fair number of Chrome tabs along with a terminal window open and that seems to be enough to drain the batter rather quickly.

Have you tried Safari? I hear it is more energy efficient: http://thetypist.com/482/chromefree/

Allowing Flash to autoplay is wasteful as well.

Chrome + many tabs is the biggest power draw on my 2013 Air, followed by my Virtualbox vms.

Chrome for MacOS seems to be quite badly behaved; you should see better results with Safari.

I'm confused, is he tracking computer usage, or % of battery drained? If the laptop is plugged in then battery drain won't show up. Also power draw will be different than duration of computer use, as some tasks are more CPU intensive.

He is tracking battery capacity (Current Capacity / Max Capacity), not battery drain. This is useful to track the lifetime of the battery itself, instead of its charge.

What tool did the author use for tracking? this seems like an interesting metric everyone should track.

edit: batlog a cronjob the author wrote

https://github.com/jradavenport/batlog

"Measure what is measurable, and make measurable what is not so." - Galileo

Playing Civilization V?

How does this show if one battery is better than the other? Doesn't it just show how often he charges his batteries?

>>Too much media/computer use isn't good for your brain

I was really interested in the article until this popped up. What a gross generalization.

This is obviously and objectively true for a sufficiently high level of "media/computer use", and I think many of us will say, anecdotally, that we have bumped up against this limit. What's so offensive about it?

People in denial (in this case about the health of endless computer usage) tend to react violently when you confront the issue.

All statements in the form "Too much [thing] is bad" are tautologically correct.

It's obviously true that all this computer use that I use to learn more about programming, math, and various other fields, and communicating with others, is rotting my brain, yes. I should go watch TV or something, right?

Reply on news.ycombinator.com