I'm getting some crazy values in the logging, both in draw and in charging.
On battery power I can see draws up to 4,2A
And when charging there are loggins up to 1,2A if I understand this correct.
Both values I think is very high.
I have the original 2000mAh battery and I also experience, after a few cycles that the capacity have risted, from the first cycle, from 1895mAh to now 2121mAh. The last one would be great, if I could trust it
Have others experienced odd readings from this widget?
Edit: KI8 with Siyah kernel 2.0, no changes in OC,UC,UV and charger mAh.
hey,
do you have the "mA corrections" enabled in the options menu?
have you calibrated the phone?
Yes the battery is calibrated. Full charge, plug out, batttery out, wait min. 90 sec. etc.....
Full cycle, up,down, and up (for the widget)
No mA corrections.
nilste said:
No mA corrections.
Click to expand...
Click to collapse
you should maybe enable it? this is my last idea which could help.
My question is what tells the OS what the battery voltage and percentage is? Is this hardware on the mainboard or is it in the kernel? To understand why I ask then read this thread.
http://forum.xda-developers.com/showthread.php?t=1353417
If its in the kernel then could it be possible to have a custom kernel built that will report the correct voltage for us tbolt users having this problem?
So I reckon I'm stuck with half battery life?
Have you tried a different battery?
The charger IC is controlled through a kernel driver, once the battery hits a certain voltage it will stop charging.
Typically when these batteries start to die they will drop voltage a lot faster than normal (making it look like it didn't charge correctly).
I'd try using a different battery to see to see if the results are the same. If they are the same this IC is probably faulty.
This will show battery voltage in mV:
Code:
cat /sys/class/power_supply/battery/batt_vol
drebbe said:
Have you tried a different battery?
The charger IC is controlled through a kernel driver, once the battery hits a certain voltage it will stop charging.
Typically when these batteries start to die they will drop voltage a lot faster than normal (making it look like it didn't charge correctly).
I'd try using a different battery to see to see if the results are the same. If they are the same this IC is probably faulty.
This will show battery voltage in mV:
Code:
cat /sys/class/power_supply/battery/batt_vol
Click to expand...
Click to collapse
The battery isn't the problem. The phone actually charges the battery all the way it just doesn't read the correct voltage. It reads about .3 volts lower than what the battery actually is which makes the phone shut off .3 volts too early. Full charge is 4.2 volts and that's what the battery reads with a meter after an overnight charge but in the phone it only reads 3.9xx volts. It should shut off around 3.5 volts and it does but the voltage will actually be around 3.8xx when checked with a meter. See what I'm saying? It reads .3 volts lower in the phone.
What is IC and is possible to have a custom driver to offset for the wrong voltage being reported? This is a common problem with alot of tbolts. I tried PMing a couple kernel devs and got no response.
I have developed a kernel patch that allows controlling the charge voltage of the battery. Lowering the charge voltage will typically increase battery cycle life by 2x for 0.1V and potentially much more at high temperatures. The default charge voltage is 4.3V, so with 4.2V you can expect 2x the battery life; with 4.1V 4x the battery life (the battery won't be fully charged).
The patch adds a new sysfs control: '/sys/devices/i2c-0/0-006a/float_voltage'. The setting is in milli-Volt; 3800mV - 4300mV is allowed. The charger supports 20mV increments.
For setting the charge voltage permanently, you can add something like:
echo 4100 > /sys/devices/i2c-0/0-006a/float_voltage
to an init script.
For my N7:
4.0V -> 73% charge
4.1V -> 83% charge
4.2V -> 93% charge
4.3V -> 100% charge
(The patch is against the stock kernel.)
The patch is included in the ElementalX kernel:
http://forum.xda-developers.com/showthread.php?t=2389022
and the Glitch kernel:
http://forum.xda-developers.com/showthread.php?t=2449919
anyone care to try first? is it any different than undervolting with custom kernel?
adichandra said:
anyone care to try first? is it any different than undervolting with custom kernel?
Click to expand...
Click to collapse
This is very different. It isn't supposed to increase battery runtime, it's supposed to decrease it.
It allows allows to not completely charge the battery. In return, the battery won't deteriorate as quickly. Making up numbers here: if the battery would normally survive 300 charge cycles (charged to 4.3V); if you only charge to 4.2V (or about 93% capacity), the battery would last for 600 charge cycles before it's dead.
ah i thought this tweak was about to give double battery life in one single charge. might be usef for others but not me though since i never use a gadget more than 18 months. thanks
adichandra said:
ah i thought this tweak was about to give double battery life in one single charge. might be usef for others but not me though since i never use a gadget more than 18 months. thanks
Click to expand...
Click to collapse
No it extends your battery life as in the longevity. The battery life that we usually think of is how long the battery will last on a charge like you were thinking. That's not what this mod is for. Such a mod that you were hoping for doesn't exist.
tiny4579 said:
No it extends your battery life as in the longevity. The battery life that we usually think of is how long the battery will last on a charge like you were thinking. That's not what this mod is for. Such a mod that you were hoping for doesn't exist.
Click to expand...
Click to collapse
Oh how they can wish though. Though still good work regardless, and very useful.
Anyone know the actual estimate for the amount of charge cycles the battery can handle before giving up the ghost?
I tend to agree with adichandra.. In today's consumer market, most gadjets are usually obsolete after 18 months..
Sent from my Nexus 7 using XDA Premium 4 mobile app
So charge cycles are extended by never allowing the battery to charge to 100%? Sounds similar to a feature on my thinkpad laptop that let's you set the charge and discharge thresholds.
Thank you for posting this!
OJ in Compton said:
So charge cycles are extended by never allowing the battery to charge to 100%? Sounds similar to a feature on my thinkpad laptop that let's you set the charge and discharge thresholds.
Click to expand...
Click to collapse
Yes. Same idea. Some Samsung laptops have this as well.
How can we apply this patch? Sorry for such a noob question...
marcus6999 said:
How can we apply this patch? Sorry for such a noob question...
Click to expand...
Click to collapse
If you are not building your own kernel, there will hopefully be a few kernels in this section that will include this.
If you are building your own kernel, you can take the posted 'smb345-charger.c' and replace 'drivers/power/smb345-charger.c' in your kernel tree. It's based on the stock kernel.
tni.andro said:
This is very different. It isn't supposed to increase battery runtime, it's supposed to decrease it.
It allows allows to not completely charge the battery. In return, the battery won't deteriorate as quickly. Making up numbers here: if the battery would normally survive 300 charge cycles (charged to 4.3V); if you only charge to 4.2V (or about 93% capacity), the battery would last for 600 charge cycles before it's dead.
Click to expand...
Click to collapse
Most kernels on all my phones have ability to set precisely the % you want battery charged up too. Haven't seen it for flo/razor
Sent from Pimped N7 (2013) XDA PREMIUM
Hmm that's a good suggestion. Truly to increase the battery charge cycle life by almost 2x you just have to charge it until 90%.
So basically you don't charge it to max voltage and you don't let the charger begin the high voltage stabilization.
So with normally (good Lithium batteries these days) 500 charge cycles you go for 1000 and so on.
I was searching for something like that, because it's a tablet that can last through many years.
I think it's a great feature and all kernels should implement and have it at stock. And let the user decide and choose voltage with an easy script if he wants it.
Battery life year extender
Thank you
tni.andro said:
If you are not building your own kernel, there will hopefully be a few kernels in this section that will include this.
If you are building your own kernel, you can take the posted 'smb345-charger.c' and replace 'drivers/power/smb345-charger.c' in your kernel tree. It's based on the stock kernel.
Click to expand...
Click to collapse
I'm going to add this in the next update to ElementalX
CTCaer said:
Hmm that's a good suggestion. Truly to increase the battery charge cycle life by almost 2x you just have to charge it until 90%.
So basically you don't charge it to max voltage and you don't let the charger begin the high voltage stabilization.
So with normally (good Lithium batteries these days) 500 charge cycles you go for 1000 and so on.
Click to expand...
Click to collapse
There is no good reason to use battery percentage to terminate the charge, instead of using a lower charger float voltage. Likely the only reason a percentage cutoff was used in other cases is that the charger chip didn't support setting the voltage.
The voltage setting is also much more predictable - the battery gauge can be quite unreliable in terms of charge estimate and easily jump 10% when it recalibrates itself.
Can anyone post the source of info for 2x battery cycle if only charge up to 4.2v? I read that not charging lithiums to full charge can help extended life but never seen anyone give estimates like 2x at 4.2v and 4x for 4.1v. Estimates seem a bit high to me. I agree with others here that this tablet can last a few years if the battery holds up so I was planning on replacing the battery after a year or two but if lowering the battery voltage really works this well then I will try it. There is a pretty big downside of having almost 10% less battery life all the time.
neotekz said:
Can anyone post the source of info for 2x battery cycle if only charge up to 4.2v?
Click to expand...
Click to collapse
http://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries
http://macomp.ru/wp-content/uploads/2012/04/1607542.pdf
http://www.eetasia.com/STATIC/PDF/200806/EEOL_2008JUN16_POW_TA_01.pdf
tni.andro said:
There is no good reason to use battery percentage to terminate the charge, instead of using a lower charger float voltage.
Click to expand...
Click to collapse
Of course there is. The last 10% to 20% are causing the most aging of a lithium ion battery.
If you really want to prolonge the live of your battery charge it only up to 80%.
There is no real reason to limit the voltage as the stock hardware and drivers handles this already really good by adopting the charging voltage dynamically.
There's no reason to change charger's voltage. The only thing there's meaning in changing in the charger side is output current.
For example with the size of Nexus battery you can use a 1A one or 800mA. It will charge slow but you can maintain some more charge cycles.
Still the most important is to not let the battery reach it's maximum voltage (100%) and let the charger do it's voltage stabilization.
As also said from others before that's what takes the battery charge life away.
The best for good juice and year protection is 90%.
The best for year extension and low juice is 60%.
And of course, when you really need 100% juice (trip, flight, beach, whatever) you just charge it full and don't care for some cycles.
Good batteries have -+500 Full charge cycles. So almost one and a half year if you go from 100->0->100 everyday, before the capacity (mAh) of battery drops.
tni.andro said:
There is no good reason to use battery percentage to terminate the charge, instead of using a lower charger float voltage.
Click to expand...
Click to collapse
TDO said:
The last 10% to 20% are causing the most aging of a lithium ion battery.
Click to expand...
Click to collapse
That's correct.
If you really want to prolonge the live of your battery charge it only up to 80%.
Click to expand...
Click to collapse
That's exactly what lowering the charger float voltage can do. The original post has numbers for different voltages. E.g. with a 4.1V charger float voltage, the charger will stop charging at 83% charge.
There is no real reason to limit the voltage as the stock hardware and drivers handles this already really good by adopting the charging voltage dynamically.
Click to expand...
Click to collapse
The stock driver sets the charger float voltage to 4.3V. Given that there is current limiting (1200mA input current limit), the voltage will be lower initially, but the voltage regulation is 4.3V once the current is low enough.
Like I mentioned, using the battery percentage as cutoff isn't great, since the battery gauge can be pretty inaccurate (can be off by >10%). Setting the float voltage lower results in the same charge cuttoff every time, even if the battery gauge makes an inaccurate estimate.
I am using the latest LineagOS. I thought Le Max 2 has QC3.0 support which should let it run at 12v charge voltage, but it does not.
The best it does is 9V and 1.6A (measured using usb meter) totaling to 14.4W. I tried the phone at room temperature, also cooled it down in fridge and tried again. Results are same. The battery was 1% - 3% at start.
Is this normal? I calculated and 14.4W if batteryt was charged to 4.35V makes roughly 3300mA which sounds rather low. I expected phone to go around 4100mA at least while charging.
More importantly, it never switches to 12V charging voltage. Is that normal?
my phone is going up to 4100mA while charging. On some roms you have to enable quick charge in your phones settings.
@LivingLing did you test with a program like Ampere or using a USB meter device? Because your ROM may be giving you incorrect values and it can't be trusted.
There is no quick charge option in LineageOS but I can see that phone is using quick charge because charger increase voltage to 9V. I am trying to understand why it wouldn't go to 12V
yurtesen said:
@LivingLing did you test with a program like Ampere or using a USB meter device? Because your ROM may be giving you incorrect values and it can't be trusted.
There is no quick charge option in LineageOS but I can see that phone is using quick charge because charger increase voltage to 9V. I am trying to understand why it wouldn't go to 12V
Click to expand...
Click to collapse
No i didnt used anything like that, just mentioned the values given by the Rom. But my phone charges Form 10% to 100% in like 15 minutes. So thats actually pretty fast in my opinion.
That sounds impossible, make a screenshot from the charging graph please. That would be something about 10A charging current...
I get about 4,5A when I activate QC on the EUI
@LivingLing Your phone must be broken if you can charge 0 to 100% in 15 minutes. You can't really charge a lithium battery very fast after about 80% and thats why quick charge implementations never talk about 0 to 100% charge time.
With lithium battery, most implementations set the input voltage to about 4.35V and battery charges as fast as it can pull power. This slows down when battery is near full.
I recommend you to get a USB meter, they cost few dollars anyway and quite cool devices.
Hi guys does anyone know how we can limit voltage when we charge ur phone?
What i use is rooted phone with ACC + AccA ront end but some bug you always have to check because most time AccA doesnt apply limit voltage as i ask.
Any better solution?
Why you need to limit the voltage? If you want to help battery to stay in good health for longer you should limit the charge current instead. Easiest way to do that is to use older charger with lower wattage. I usually charge my pixel 6 from my laptop which has type c ports with usb power delivery support and are limited to 12 watts so charge current never exceeds 2.1 ampers. I have also tested my old power bank and cable and there power is limited to even lower 7.5 watts or 1.5 ampers of current.
On battery university they said, the voltage from the charge is the problem not the amp speed, for example when your phone is at 3.8v(30%) and you charge with a quick charge 3amp at 4.05v(not 4.25v) you quickly recharge until 60%(4v) and you slowly and you reach 4.05v(+-70%) and you can let your phone charging longer you want the voltage never go over 4.05v and at 4.05v the said atfter 2000 recharge cycle you still have 90% of brand new battery capacity.
Your phone said full charge at 4.25v but with lithium ion it overcharging and that really dommage you battery life.
So in real is not the % of capacity because when you charge your phone to 80% diring the charge process the voltage use to charge is 4.35v and that is bad for the battery
BU-808: How to Prolong Lithium-based Batteries
BU meta description needed...
batteryuniversity.com
"Every 0.10V drop below 4.20V/cell doubles the cycle but holds less capacity. Raising the voltage above 4.20V/cell would shorten the life. The readings reflect regular Li-ion charging to 4.20V/cell."