![]() |
400% battery performance improvement?
This is freaking me out!:eek:
My N900 used to have terrible battery life. 3-4 hours max. really crappy. I talk all about it in this thread Suddenly, yesterday I noticed it had gone ~8 hours and still had juice. Today, 14 hours after I took it off the charger, it still reports 35% battery. This is after 9 months of crappy battery life. WhhaawwtThaFaaarrk? 400% increase?!? Why is this? I'm afraid to reboot it else the magic will escape! Anyone care to speculate on this? I'm really interested in folks ideas on this. Here's some stats & observation.:
I started poking around and found lots of this in the syslog: Code:
[39710.458679] bq27x00-battery 2-0055: battery is not calibrated! ignoring capacity values And for comparison (see my other thread), here's typical top output: Code:
Mem: 142680K used, 92580K free, 0K shrd, 396K buff, 28824K cached Code:
[1|root@Nokia-N900|~]powertop Here's lshal for the battery: Code:
[1|root@Nokia-N900|~]lshal -s -u '/org/freedesktop/Hal/devices/computer_power_supply_battery_bq27200_0' what do you think is going on with this? lemme know if you'd like any other info. |
Re: 400% battery performance improvement?
There are many possible causes of crappy battery life that might invisibly go away, and they are discussed in the many battery life threads. One of my favorites is the cell phone switching from one type of signal to another.
|
Re: 400% battery performance improvement?
i have similar experience with NITdroid...
more then 5hrs of On-screen(continual use till battery cry for recharge) time and 4hrs on wifi hotspot... more then 15hrs on normal use... and it was cold too as you mentioned the reason i felt for it was 'heapsize' which was not set by me so it ws '20m' only & no swap... after changing both of them i.e heapsize to '36m' & enabling swap.. NITDROID BECAME A MAMMAL i.e always hot & hungry ;) this is what i observed, it might be wrong so is there a similar concept of heapsize in maemo too |
Re: 400% battery performance improvement?
@knoxy420,
What kernel are you using? Are you loading the bq27x00_battery module? Can you remember if you did anything in these last days that might have affected the kernel? |
Re: 400% battery performance improvement?
Every BL-5J has a chip inside which keeps track of the amount of charge a battery can hold. It seems like yours was off by a large amount or simply defect.
You can re-calibrate it yourself or purchase a new battery if the one you're currently using does turn out to be defect. http://talk.maemo.org/showpost.php?p...&postcount=846 http://talk.maemo.org/showpost.php?p...postcount=1176 |
Re: 400% battery performance improvement?
you're aware that the speed patch is generally not considered good?
|
Re: 400% battery performance improvement?
Quote:
|
Re: 400% battery performance improvement?
Quote:
Code:
Linux Nokia-N900 2.6.28.10-power50 #1 PREEMPT Sun Mar 18 20:10:56 EET 2012 armv7l GNU/Linux Code:
[1|root@Nokia-N900|~]lsmod One day it would die quickly. Now it lasts all day. |
Re: 400% battery performance improvement?
Quote:
|
Re: 400% battery performance improvement?
Quote:
Whatever was eating batteries was generating a large amount of heat so that the phone was always noticibly warm. I figured there was a short somewhere or something. |
Re: 400% battery performance improvement?
Quote:
|
Re: 400% battery performance improvement?
It seems that something was draining Your battery life (some script that stopped working? Cell signal quality was improved, so Your device isn't constantly switching between 2G and 3G?), and now it's gone. It's hard/impossible to tell what it was, without powertop readings from situation, when the problem actually *occurred*.
/Estel |
Re: 400% battery performance improvement?
@freemangordon and @estel,
Check the thread linked in the first post. The poor guy suffered from this problem even with the N900 being *turned off* (whether completely or ACT_DEAD is unclear, I vote for ACT_DEAD). Powertop outputs are also on that thread... The problem apparently only happened when the phone was *not* connected to the charger, hence my question about bq module. I don't know if things have changed with KP50, but the bq27.. module was not supposed to be loaded when bme is also running. At least I haven't loaded it since KP46 days. Now AFAIK nobody has (yet?) had a real problem when combining bq27 and bme. Perhaps we have a case here?! |
Re: 400% battery performance improvement?
Quote:
The only thing I've noticed that different now, is the messages in dmesg about bad battery calibration. That was not present before. |
Re: 400% battery performance improvement?
Quote:
It would drain if the battery was in it at all. Even if I removed it for awhile and then put it back in, without powering on, it would get warm and drain batteries. |
Re: 400% battery performance improvement?
@knoxy420,
I don't know what to think anymore. It might have been some bad contact, making the battery actually drain (letting the current flow) even when nothing was "using" it, hence the warming. |
Re: 400% battery performance improvement?
1 Attachment(s)
What will you say about this friends ? :confused:
From the graph you can see my battery is not fully charged still it runs for 49 hours with lots of cpu usage.:D :D and good thing is....:cool: This is my regular battery back up time. :eek::eek: |
Re: 400% battery performance improvement?
Quote:
Quote:
|
Re: 400% battery performance improvement?
Quote:
Is this low-normal use ?:confused: almost 65 % of that time cpu usage is showing more that means it's not a low-normal use. |
Re: 400% battery performance improvement?
Hey all, just an update on my mysterious/miraculous n900 improvement.
The previous issues were not battery dependent. i can use either of my batteries and get 12+ hours of usage (vs 3-4 hours battery previously). so something in the phone's hardware "cured" itself. I can also now over clock up to 1000MHz and it stays stable. Previously, if I went over 850MHz it would reboot immediately. I really wish I knew what happened for this improvement. |
Re: 400% battery performance improvement?
Quote:
is it (was it?) turned on for you? |
Re: 400% battery performance improvement?
Quote:
|
Re: 400% battery performance improvement?
Quote:
Cranked the cpu up to 1150, left Smartreflex off. Started installing stuff and checked conky to see it actually hitting full speed. Seems stable so far. Sweet! |
Re: 400% battery performance improvement?
Quote:
Quote:
Not to mention the fact that the SoC has a severely reduced guaranteed operational lifespan at this speed. |
Re: 400% battery performance improvement?
Yeah, Just wanted to see if it would go there. I scaled it back to 900Mhz. Good to know I can use smartreflex at that speed. I usually get ~24hr now with my usage.
|
All times are GMT. The time now is 15:21. |
vBulletin® Version 3.8.8