N900 battery duration

Frantisek Dufka wrote:
> I have posted my findings with one spare N810 to that thread. I got like
> 30 days with N810 sitting completely idle. That would roughly mean one
> reboot is 16 hours of standby (with no wi-fi).

30 days. That's quite nice. Any idea why n900 gets only to 4 days then?
What makes the difference so big? I didn't run it on offline mode though.

--
Tuomas

Aniello Del Sorbo wrote:
>
> The issue was (abstract) if a reboot would consume more energy than a
> night-long idle.

As I said - by Matan's equation in the forum post, one
reboot is 16 hours of standby. So at least with wi-fi off, keeping it
idle looks better. I was thinking about testing with wi-fi but my router
cannot keep connection for so long so it would be a bit mor complex than
just keeping it in the drawer.

> you should be able
> to squeeze up to 44 nights out of a charged battery.


I guess with 30 full days I could do 60 nights ;-) But I somehow need to
keep it powered off during daytime. Maybe I could shut it down in the
morning and boot it in the evening to save power ;-)

Frantisek

2009/11/11 Frantisek Dufka <dufkaf@seznam.cz>:
> Aniello Del Sorbo wrote:
>>
>> The issue was (abstract) if a reboot would consume more energy than a
>> night-long idle.
>
> As I said - by Matan's equation in the forum post, one
> reboot is 16 hours of standby. So at least with wi-fi off, keeping it
> idle looks better. I was thinking about testing with wi-fi but my router
> cannot keep connection for so long so it would be a bit mor complex than
> just keeping it in the drawer.
>
>> you should be able
>> to squeeze up to 44 nights out of a charged battery.
>
>
> I guess with 30 full days I could do 60 nights ;-) But I somehow need to
> keep it powered off during daytime. Maybe I could shut it down in the
> morning and boot it in the evening to save power ;-)
>
> Frantisek
>

Yeah, that's true. So far it looks like a reboot is mostly equal to a
night-long idle consumption.
To me at least. 44 reboots and 60 nights, considering a difference in
how old a battery is, could
very well be the case. I also take a bit into account battery heat
dissipation (in the case of 44 consecutive reboot).

So, I would guess that leaving it on in offline is better than turning
it off and than on during the night (it's faster to
switch between offline/online than actually rebooting).
But rebooting ensures and memory leak gets a workaround. oh well..
--
anidel
Sent from London, Eng, United Kingdom

On Wed, 11 Nov 2009, Frantisek Dufka wrote:

> Matan Ziv-Av wrote:
>> Here is some data:
>>
>> http://talk.maemo.org/showpost.php?p=361648&postcount=23
>>
>> Abstract: The often repeated claim that a reboot cycle wastes a lot of
>> battery is incorrect.
>>
>
> Hmm, you proved you can drain battery in 44 reboot cycles in 2 and half
> hours. To me it says reboot wastes a lot of battery :-) You can play video or
> surf the web longer than 2.5 hours. In fact there is not many things that
> could drain battery faster than reboots, is it?

I cannot play video for 2.5 hours on my almost 2 years old N810. I can
hardly read an ebook for this long.

In addition I have dual boot, including 30s delay. During this delay
screen is on maximum brightness, and I don't know which power saving
features are enabled, so without this delay, a few more reboots might
be possible.

> I have posted my findings with one spare N810 to that thread. I got like 30
> days with N810 sitting completely idle. That would roughly mean one reboot is
> 16 hours of standby (with no wi-fi).

This must be very atypical. In similar tests on a new N800 I never even
managed two weeks (in offline mode).

The main reason I suggest turning off when the device is not going to be
used for long periods (such as a night sleep) is the non-slim chance
that something might cause break power saving (metalayer-crawler, router
incompatible with power saving, forgetting to disconnect bluetooth,
etc.) and you wake up to see a dead device.


--
Matan.

Matan Ziv-Av wrote:
> I cannot play video for 2.5 hours on my almost 2 years old N810. I can
> hardly read an ebook for this long.
>
> In addition I have dual boot, including 30s delay. During this delay
> screen is on maximum brightness, and I don't know which power saving
> features are enabled, so without this delay, a few more reboots might
> be possible.

Good points.

>
>> I have posted my findings with one spare N810 to that thread. I got like 30
>> days with N810 sitting completely idle. That would roughly mean one reboot is
>> 16 hours of standby (with no wi-fi).
>
> This must be very atypical. In similar tests on a new N800 I never even
> managed two weeks (in offline mode).

N800 is worse, see Igor's post here
http://lists.maemo.org/pipermail/maemo-users/2008-February/009409.html

My numbers is from my own measurements I did later. The extra N810 was
new at that time.

>
> The main reason I suggest turning off when the device is not going to be
> used for long periods (such as a night sleep) is the non-slim chance
> that something might cause break power saving (metalayer-crawler, router
> incompatible with power saving, forgetting to disconnect bluetooth,
> etc.) and you wake up to see a dead device.

Yes, definitely.

I was trying to solve this by send SIGSTOP to offending processes
(browser in particular) when keyboard and screen is locked and it
somehow worked. It is relatively easy to do it as powerlaunch script.
With nice UI that would allow to picks up offenders by monitoring CPU
usage this could be useful tool for poor man's suspend. Basically by
locking screen and keys every process except few critical onces could be
stopped and later resumed (SIGSTOP,SIGCONT) when device is unlocked.
Would not solve keeping audio open etc.

Frantisek

On Wed, Nov 11, 2009 at 5:58 AM, Frantisek Dufka <dufkaf@seznam.cz> wrote:
> Aniello Del Sorbo wrote:
>>
>> The issue was (abstract) if a reboot would consume more energy than a
>> night-long idle.
>
> As I said - by Matan's equation in the forum post, one
> reboot is 16 hours of standby. So at least with wi-fi off, keeping it
> idle looks better. I was thinking about testing with wi-fi but my router
> cannot keep connection for so long so it would be a bit mor complex than
> just keeping it in the drawer.
>
>> you should be able
>> to squeeze up to 44 nights out of a charged battery.
>
>
> I guess with 30 full days I could do 60 nights ;-) But I somehow need to
> keep it powered off during daytime. Maybe I could shut it down in the
> morning and boot it in the evening to save power ;-)
>
> Frantisek
>

The absurdity knows know bounds. You're saying a battery will last
longer with a device turned on in sleep mode than it is even capable
keeping a charge sitting all by itself out of the device on a counter.
Give me a break. We're talking real-world usage here. If you actually
*use* the device, rather than just letting it sit idle doing nothing
(in which case why do you even own the thing?!?!?!), you're not going
to get anything like that kind of battery duration. I've tried putting
my N800 in offline & sleep mode at night, and never got more than 3
days out of it, even with very little usage during the day. If I turn
it off when I'm not using it and turn it on and back off 2-4 times a
day, I can get 1 to 2 weeks out of it. Even if the N810s are better,
there are too many variables to defend such outrageous claims,
especially if you (again) actually use the thing and install any apps
that run in the background.

Let's leave theory to the theorists and take a dose of reality, okay?

Mark

Tuomas Kulve wrote:
> I repeated three runs with ffvorbis and got a bit over three hours
> more battery life with the headset connected compared to my earlier
> ffvorbis tests without the headset.

And now I repeated the test runs with Flac. The battery lasted roughly
1.5 hours longer than while playing MP3.


--
Tuomas

On Wed, Nov 25, 2009 at 6:54 PM, Tuomas Kulve <tuomas@kulve.fi> wrote:
> Tuomas Kulve wrote:
>> I repeated three runs with ffvorbis and got a bit over three hours
>> more battery life with the headset connected compared to my earlier
>> ffvorbis tests without the headset.
>
> And now I repeated the test runs with Flac. The battery lasted roughly
> 1.5 hours longer than while playing MP3.

Xiph's FLAC decoder?

--
Felipe Contreras

Felipe Contreras wrote:
> On Wed, Nov 25, 2009 at 6:54 PM, Tuomas Kulve <tuomas@kulve.fi> wrote:
>> Tuomas Kulve wrote:
>>> I repeated three runs with ffvorbis and got a bit over three hours
>>> more battery life with the headset connected compared to my earlier
>>> ffvorbis tests without the headset.
>> And now I repeated the test runs with Flac. The battery lasted roughly
>> 1.5 hours longer than while playing MP3.
>
> Xiph's FLAC decoder?

Yes.



--
Tuomas

All,

for what it is worth. battery charge duration is appearing as an issue
with the Motorola Droid as well. I have started using a Droid and I
notice that I need to charge it every night, and, if I fail to do that,
then the next day it runs low on battery charge. Here is the url to
some advice posted on the Verizon Wireless www site concerning battery
duration related issues:

http://search.vzw.com/?do=viewdoc&id=27662

Of course the Verizon Wireless Droid uses EVDO for Internet access which
may be more or less power efficient than HSDPA as implemented on the
N900, I am not sure. One think I like about the Droid is that there is a
way for the user to be informed of WHAT is consuming the power on the
device in each of 9 categories including

Display

Voice Calls

Phone Idle

Wi-Fi

Cell Standby

Browser

Android System

Bluetooth

Android OS



Best Regards,



John Holmblad



Acadia Secure Networks, LLC

* *




Alberto Garcia wrote:
> On Fri, Oct 16, 2009 at 10:48:27AM +0200, Andrea Grandi wrote:
>
>
>> after some days using the N900 (with already lot of charge/discharge
>> cicles) I'd like to compare my battery duration with other people.
>>
>
> During the first days I used the N900 a lot, including browsing the
> web, instant messaging, e-mail, etc. The battery didn't last much more
> than a day, so I had to recharge it everyday at night.
>
> But then I wanted to compare its battery life with that of my old
> phone, so I started using it in a more conservative way: phone calls,
> text messages, calendar, alarm clock and occasionally as a media
> player. I also used the camera to take some pictures here and there
> and I connected to the Internet a few times to read my e-mail or check
> a couple of websites, but only for a 2-3 minutes each time.
>
> I charged the N900 on Wednesday night, and it's almost exhausted now,
> so it lasted for ~4 days, which is quite good, I think, to get an
> idea.
>
> Berto
>
>
>