[BUG] Full Charge Oddity, Charge to 100%, Unplug to 97%

Search This thread
By:
Advanced…
L

lawilson2

New member
Sep 15, 2010
3
0
[Q] Wifi Tether: Subject to $15 T-Mobile Charge?

Curious about this. Currently I'm connected to my laptop for the past 15 min just fine; no alert came up that said I had to get the $15 fee that T-Mobile charges for the "privilege" of tethering on my current plan. Does that mean that an unlocked phone isn't subject to T-Mobile's tampering?

If so, is this the only value for rooting right now? I don't see any other need for root with this phone except if it fools T-Mobile.

Thanks in advance for your answers.
 
A

allen099

Senior Member
Feb 22, 2010
297
14
New York
I know there are numerous threads about battery life, but this one isn't about the life itself really.

I can charge my phone to 100%, but when I unplug it, the battery indicator shows 97% right away. I plug it back in, rinse and repeat, and get the same issue. I even tried charging with the phone off.

Ha anyone had luck with using a full 100%?
 
slowz3r

slowz3r

Senior Member
Jun 10, 2010
2,360
104
Malmstrom AFB, MT
allen099 said:
I know there are numerous threads about battery life, but this one isn't about the life itself really.

I can charge my phone to 100%, but when I unplug it, the battery indicator shows 97% right away. I plug it back in, rinse and repeat, and get the same issue. I even tried charging with the phone off.

Ha anyone had luck with using a full 100%?
Click to expand...
Click to collapse

This is normal, search some threads
 
slowz3r

slowz3r

Senior Member
Jun 10, 2010
2,360
104
Malmstrom AFB, MT
Id keep it, there was an article posted in one of the threads stating why phones do this...the N1 did a similar thing where after it charged going from 100-96 percent went really quickly


i would equate it when you buy a 16 gig SD card and it shows as 15.6 gigs
 
B

brianbrain

Senior Member
Sep 1, 2010
234
56
Chicago, IL
allen099 said:
I know there are numerous threads about battery life, but this one isn't about the life itself really.

I can charge my phone to 100%, but when I unplug it, the battery indicator shows 97% right away. I plug it back in, rinse and repeat, and get the same issue. I even tried charging with the phone off.

Ha anyone had luck with using a full 100%?
Click to expand...
Click to collapse

Mine did this until I calibrated the battery. Charged it to 100%, then drained it all the way. Charged it back up to 100%, and now not only does the battery last longer, but the percentage seems to go down in a more linear fashion.
 
A

allen099

Senior Member
Feb 22, 2010
297
14
New York
brianbrain said:
Mine did this until I calibrated the battery. Charged it to 100%, then drained it all the way. Charged it back up to 100%, and now not only does the battery last longer, but the percentage seems to go down in a more linear fashion.
Click to expand...
Click to collapse

Same here, I've done this two nights in a row, but no change.

Thank you for the links, guys. Pretty insightful actually. I've got it charging right now at 99%. As soon as it hits 100%, I'll take it off and see if it drops instantly. I'm guessing it shouldn't as it JUST hit 100%. But we shall see..

Thanks again!
 
A

allen099

Senior Member
Feb 22, 2010
297
14
New York
Well...I mean, it's not something anybody is "worried" about, but it does suck because I use my phone heavily as I'm in the IT field ... so every percentage counts. After I just unplugged, it went to 99% which my N1 did as well. I'm fine with 99%, but 97% seemed a bit low.
 
C

cosimoss

Senior Member
Dec 25, 2010
510
111
London
Full charge oddity

When the phone has been switched off and charging the phone shows the fully charged icon on screen when charging appears to be done. Battery Indicator Pro displays only a 95% charge when the phone is switched on. The same happens when the phone is on and being usb charged. When 100% charge is shown and the usb cable is removed the charge is then shown as 95%. This does not happen on my other devices running 2.2. They will show 100% or 99%. A device design or 2.3 bug?
 
A

A_Flying_Fox

Senior Member
Aug 10, 2010
1,610
170
The epic 4g has this problem. I think its something with samsung. You have to unplug it, plug it back in, let it charge to full again and rinse repeat until its really at 100%

Then reset your battery stats in cwm

Sent from my SPH-D700 using XDA App
 
unremarked

unremarked

Senior Member
Jun 18, 2009
934
203
Out There
As noted in the stickied FAQ and further discussed in the link provided by nxt (Beat me to it!), this is by design folks.

unremarked said:
Q: I unplugged my phone, and my battery dropped from 100% to 95% immediately, or it won't/takes along time to charge past 99%, what gives?
This is by design. Your phone will slow down and eventually stop pulling charge at or slightly greater than ~95% regardless of what your battery indicator says. This is to extend the overall life of the battery, as constant 100% to 0% charge/discharge cycles will cause it to fail prematurely. If you're interested in really pushing it to 100%, you can use a technique that is called "bump charging" and is better detailed here: http://www.androidpolice.com/2010/12...battery-drain/
Click to expand...
Click to collapse
<- Link has more or less the same info as nxt's.
 
deyna

deyna

Senior Member
Dec 14, 2010
94
9
Hernando, MS
Weird battery reading

My Nexus S was at about 16% when I put it on the charger then I did a quick reboot and it said it was at 52%. I unplugged it, powered off, pulled the battery, then rebooted and it read 28%. Even that is hard to believe after less then 5 minutes of charging. Has this happened to anyone else?

Stock 2.3.1 unrooted

Sent from my Nexus S using XDA App
 
You must log in or register to reply here.

Similar threads

jlevy73
[INFO] [i9020] CyanogenMod 7 Nightly
Replies
9K
Views
1M
A
avi.avis
oldblue910
[I9020T][I9023][I9020A][M200][D720] Nexus S 4.1.1/JRO03E Jelly Bean Update Thread
Replies
3K
Views
979K
punjabijack
punjabijack
oldblue910
[REF] ALL Nexus S OTA Updates for ALL Variants, Plus Full ROM Packages
Replies
519
Views
389K
EthanFirst
EthanFirst
Z
[INFO] Nexus S Official ICS 4.0.3 Released Dec 16!!! (download links)
Replies
2K
Views
567K
B
blazingwithfire
K
[GUIDE] Step by Step ROOT Nexus S via ClockworkMod Recovery image (i9023 and i9020A)
Replies
615
Views
387K
A
agilua

Top Liked Posts

24 Hours All time
  • There are no posts matching your filters.
  • 1
    R
    RogerPodacter
    deyna said:
    My Nexus S was at about 16% when I put it on the charger then I did a quick reboot and it said it was at 52%. I unplugged it, powered off, pulled the battery, then rebooted and it read 28%. Even that is hard to believe after less then 5 minutes of charging. Has this happened to anyone else?

    Stock 2.3.1 unrooted

    Sent from my Nexus S using XDA App
    Click to expand...
    Click to collapse

    well i modified the nexus one battery driver and got it accepted into cyanogen ROM source. So i just did some digging for the nexus s driver, and found some interesting differences which likely are the reason why you see this jump in capacity.

    as a quick high level summary. the nexus one battery contains an all-in-one fuel gauge chip solution which calculates its own percent readings, and just sends that to the OS to display. however looking at the nexus s driver, it seems they actually have a voltage table. see below:
    PHP: 
    /*
17  * Battery Table
18  */
19 #define BATT_CAL                2447    /* 3.60V */
20 
21 #define BATT_MAXIMUM            406     /* 4.176V */
22 #define BATT_FULL               353     /* 4.10V  */
23 #define BATT_SAFE_RECHARGE      353     /* 4.10V */
24 #define BATT_ALMOST_FULL        188     /* 3.8641V */
25 #define BATT_HIGH               112     /* 3.7554V */
26 #define BATT_MED                66      /* 3.6907V */
27 #define BATT_LOW                43      /* 3.6566V */
28 #define BATT_CRITICAL           8       /* 3.6037V */
29 #define BATT_MINIMUM            (-28)   /* 3.554V */
30 #define BATT_OFF                (-128)  /* 3.4029V */

    Interesting. When your near empty, around 3.5 volts, and you plug in, the voltage jumps considerably. but this should be accounted for. HOWEVER, i'd bet that because you immediately reboot your device, the driver must reload from scratch, and all previous info is wiped. so now it sees it's voltage increase in absolute terms, and thinks it's charge is that much higher. just a hunch anyway.

    This might also give hint as to why it drops from 100% to 97% so quickly. Because battery full is likely being used as the full point, but while the charger is still plugged in, there is no voltage drop yet until the charger is removed, immediately dropping the voltage a tad. just look at the voltage table values.

    Also interesting which i've not seen before is this in the driver:
    PHP: 
    #define OFFSET_VIBRATOR_ON              (0x1 << 0)
55 #define OFFSET_CAMERA_ON                (0x1 << 1)
56 #define OFFSET_MP3_PLAY                 (0x1 << 2)
57 #define OFFSET_VIDEO_PLAY               (0x1 << 3)
58 #define OFFSET_VOICE_CALL_2G            (0x1 << 4)
59 #define OFFSET_VOICE_CALL_3G            (0x1 << 5)
60 #define OFFSET_DATA_CALL                (0x1 << 6)
61 #define OFFSET_LCD_ON                   (0x1 << 7)
62 #define OFFSET_TA_ATTACHED              (0x1 << 8)
63 #define OFFSET_CAM_FLASH                (0x1 << 9)
64 #define OFFSET_BOOTING                  (0x1 << 10)
    Above you see these different "tasks" classified into these rough groupings, and each one is bit shifted. but below is the same info taken from the driver as well. it seems to me like samsung has some pre-defined weighted system for each of these tasks, and they are using it in the driver. rather interesting, i've not seen this before. i wonder how accurate it is.
    PHP: 
    #define COMPENSATE_VIBRATOR             19
71 #define COMPENSATE_CAMERA               25
72 #define COMPENSATE_MP3                  17
73 #define COMPENSATE_VIDEO                28
74 #define COMPENSATE_VOICE_CALL_2G        13
75 #define COMPENSATE_VOICE_CALL_3G        14
76 #define COMPENSATE_DATA_CALL            25
77 #define COMPENSATE_LCD                  0
78 #define COMPENSATE_TA                   0
79 #define COMPENSATE_CAM_FALSH            0
80 #define COMPENSATE_BOOTING              52

    anyway i dont know how this battery compares to say the nexus one, which does all of the battery work inside the chip itself. but the nexus s driver also seems like we could do some experimenting with these values and altering them to see if its possible to increase usable capacity, fix some of these "bugs", etc. kinda wish i had a nexus s to compile the kernel and test stuff out.
    View
    1
    unremarked
    unremarked
    mishkasf said:
    Hi,

    I'm wondering if anyone else noticed this -- as soon as I unplug my Nexus S after it has fully charged, the battery indicator [using Battery Indicator Pro] goes from 100% to 96% (sometimes 97% or 98%), and when I plug it in, the lock screen says "Charging - 96%". It takes only a few minutes to charge fully, but as soon as I unplug the charger it immediately loses 2-4% of the battery.

    Is that a Gingerbread bug or is that a Nexus S hardware/battery issue? Am I the only one experiencing this?
    Click to expand...
    Click to collapse

    SEARCH.

    It is your friend.

    Answered in sticky FAQ in the Nexus S Q&A section

    unremarked said:
    Q: I unplugged my phone, and my battery dropped from 100% to 95% immediately, or it won't/takes along time to charge past 99%, what gives?
    This is by design. Your phone will slow down and eventually stop pulling charge at or slightly greater than ~95% regardless of what your battery indicator says. This is to extend the overall life of the battery, as constant 100% to 0% charge/discharge cycles will cause it to fail prematurely. If you're interested in really pushing it to 100%, you can use a technique that is called "bump charging" and is better detailed here: http://bit.ly/f6xiZ0
    Click to expand...
    Click to collapse
    View
    1
    R
    RogerPodacter
    i know the FAQ attempts to explain this behavior, but i dont buy that explanation really. other phones like my nexus one do not drop 3% immediately upon unplugging the charger.

    after looking at the nexus S battery driver vs the nexus one battery driver, the two battery chips are different. the n1 uses a variation of "coloumb counting" which counts the current passed thru the sense resistor in the battery. the nexus s seems to be based off of voltage tables to indicate state of charge. as such, voltage tables commonly do this because the voltage drops as soon as you unplug the charger, dipping the voltage and throwing off the percentage reading.

    this table below is pulled from the nexus S source code, and nexus one does not use this method and has no table:
    PHP: 
    /*
  * Battery Table
  */
 #define BATT_CAL                2447    /* 3.60V */
 
 #define BATT_MAXIMUM            406     /* 4.176V */
 #define BATT_FULL               353     /* 4.10V  */
 #define BATT_SAFE_RECHARGE      353     /* 4.10V */
 #define BATT_ALMOST_FULL        188     /* 3.8641V */
 #define BATT_HIGH               112     /* 3.7554V */
 #define BATT_MED                66      /* 3.6907V */
 #define BATT_LOW                43      /* 3.6566V */
 #define BATT_CRITICAL           8       /* 3.6037V */
 #define BATT_MINIMUM            (-28)   /* 3.554V */
 #define BATT_OFF                (-128)  /* 3.4029V */

    at full charge my phone levels off at 4.17 volts for a few hours. soon as i unplug it it drops to 4.11 volts, then fluctuates and decreases from there. if my phone followed the voltage table above, it's easy to see why a few percentage points would drop immediately.
    View
    1
    R
    RogerPodacter
    so i'll expand on why i think this is bull crap. my "1000 dollar bet" was an obvious set up, i've seen all the talk about bump charging and all. i was able to "hack" the nexus one battery by editing its driver, and got it included into cyanogen. since then, we wrote an app to go along with this.

    in the n1's battery are about 80 memory address registers to control all of the batt parameters. one of them, register 0x65, is the minimum charge current register. it was set to <80mA to terminate the charging indicating battery full condition. we experimented setting this to different values such as <40mA and <20mA. this squeezes out extra charge into the battery. it simulates "bump charging".

    the reason bump charging works, is because when the device is powered up, the battery driver obeys these registers, including reg 0x65, so that the charging stops at exactly 80mA. your logs confirm this as we've found out over in the link in my thread. when you lower it, you get more juice into the battery. when you turn the device off and "bump charge" it, the registers are ignored because the device is powered down. so it literally lets current draw go to zero, which is why more juice goes into the battery.

    with our app, we can now accomplish this same thing with the device on, as many of us have done already. but in no way is this "for safety reasons or design". its simply an industry standard that a li-ion battery is considered "full" at C/20 where C is the mAh capacity of the cell in question. you can find this info out in all the technical data sheets linked in my thread. this helps to make the charging time a nice 2.5 to 3 hours, nothing more.

    its impossible to over charge a li-ion battery. the current draw simply gets lower and lower till it is full and accepts no more charge. there is no safety reason to ever stop charging the battery, it doesnt work like that. anyway if you guys want all the details just read the thread;).
    View

New posts