Battery drain in recent builds

[ozjeremy]

It seems to me that the battery is draining suddenly faster in recent builds. I gather this is the case for both Norbusan's and Kosh's builds, which implies that it's a CVS issue.

Who else is noticing the battery appears to be draining faster?

Does anyone have any idea what is causing it?

[Llorean]

I notice that you haven't tried it with CVS Rockbox, which implies that you have no idea what you're talking about by saying it implies a CVS build.

Norbusan and Kosh's builds both share a few of the same patches, so all that it implies until you've verified it happens in CVS is that it's a problem with code they both share, rather than specifically code in CVS.

It's entirely possible that it's CVS Rockbox that has the problem, but the evidence you've provided does not actually imply it.

[ozjeremy]

That was unnecessarily hostile. I didn't say it proves it, I said it "implied" it. You're right, it could be a patch in both builds. That's why I'm asking here - I'm wondering who else is experiencing it, and with what builds.

[Llorean]

The term "implies" in the context of A implies B actually means that A points directly to B (in general context it means that B is most likely, in actual logic it means that B is a logical necessity as a result of A).

The problem I have with your statement is that you're apparently not bothering with actual testing, and stating that CVS Rockbox is the most likely cause.

At the same time, you could've been very slightly ethical and actually *tested* it before suggesting it was the cause, or said that it merely "suggests that it MAY be the cause" rather than saying it implies it (which it in no way does).

[iriveronian]

Norbusan's build was oe of the first I used when I converted to tockbox. I've also tried drippy donut's half way builds. However battery drain seems to be an issue on the last few builds.

Due to regular "codec failure" and "freezing" issues with norbusan's builds (still enjoyed them by the way so not a criticism of the work I have started using the daily builds from the rockbox website. However, these too seem to be leading to a faster than before draining of the battery.

Don;t know what might be the problem

[Llorean]

Now, the second step, since you've got it happening on an official build, is to try and see if you can figure out a "last good build" and maybe post a battery log of a quickly draining official build attached to a proper bug report in the right place.

[musichound]

The term "implies" in the context of A implies B actually means that A points directly to B (in general context it means that B is most likely, in actual logic it means that B is a logical necessity as a result of A).

OK Cut it out now you propeller head.

[ozjeremy]

For f*cks. It wasn't a scientific claim. I was simply asking a question. I don't have time to test it myself - I was simply seeing if others were experiencing the same thing or if I was imagining it.

Jeeez-us.

[Llorean]

Out of curiosity, you had time to test two patched builds but miraculously ran out when it came time to actually test the CVS build before saying you felt it was to blame?

[ozjeremy]

No, I've only used Norbusan's. Someone else said they'd noticed it with Kosh's build.

[musichound]

I for one can attest that my battery life is much shorter now. I was using the next to the latest Koch build but I am going to try out the latest one today. The problem could also be an overused battery because my player is almost two years old. If my battery life is still short I will next try a CVS build.
I am glad that this was brought up so it can be discussed further.

[Llorean]

Well before you know what's going on, you should also try it in the retail firmware of course, to know whether your hardware is just getting old.

The most important thing though, is to try to track down which build it started with, if it is something in CVS.

[musichound]

I did notice my battery drains very quickly when recording in native firmware, even on a full charge. I believe the best way to determine if it's a battery issue or not is to use a fresh copy of native firmware without using Rockbox.
Maybe over the weekend I will try this.

[fizze]

well, all I can say is that those Li-whatever batteries, once they near the end of their life-cycle, drain and drop voltage very fast.
So the curve goes down steeply.

I've had my battery life decreasing from ~10h in rockbox to about 2h within 1-2 charge cycles. same game in the retail firmware.
so try the iRiver firmware first to see its NOT a battery issue.

@ Llorean: nice nitpicking, couldnt have done it better myself *scnr* *g*

[AcN]

i would say : if you're not happy with last builds, use an older one ^^ and before changing my battery, rockbox sucked the battery more than iriver fw, BUT when i changed the battery, it was like "whoaaaa ! i got 22 hours out of it on rockbox"

so if your battery is dying, rockbox is not recommended, but recommended anyway for the better sound... hahah ^^

[ratykat]

hey,
compared to the original iriver firmware, i have noticed a cosiderable drain from the battery using kosh's build, but alas, with this i did use doom for about 30 mins, so that will have some affect, but i wasnt quite prepared for a 45%battery drop in 30 mins :/ oh well, still love kosh's build.

and by the record, my battery is fine, i get about 18-20 hours on a full charge using standard iriver firmware, just like new :P

raty

[sanjay]

i have noticed this too, and its REALLY frustrating, to the point where i dont use rockbox anymore (using the latest build from the rockbox website)

i would be lucky to get 2 hours playback tops from rockbox whereas on the iriver firmware im getting at least 10. what gives? i used to get LONGER battery life on rockbox when i first started using it.

is there any way we can go back to getting some reasonable battery life without going back to really old rockbox builds (which have their own problems)?

[Llorean]

Well, the first step is to actually file a bug report.

The best way to do this is to actually run a proper benchmark and attach the logfile to the bug report, as well as running a test with the same settings (no equalizer, similar volume and backlight settings, same playlist) in the retail firmware and posting an approximated time that ran, so you can show the significance of the difference.

Something like this: http://img142.imageshack.us/my.php?i...gnuplotkq2.png which clearly shows that at least for that user there's no difference in the battery life between the Sep 20 build and the Oct 14th build.

[sanjay]

the first rockbox build i used was back in february. in fact, i still have that file. might try to run some battery tests actually