Hi! > Some comments on the end-of-charge behavior. The rt5033 chip offers three > options. In the Android driver, a forth option was implemented. Hmm. I'm working on that on motorola-cpcap driver, and I guess this is going to be common problem for many drivers. > - By default, the rt5033 chip charges indefinitely. The current goes down but > there is always a charge voltage to the battery, which might not be too good > for the battery lifetime. > - There is the possibility to enable a fast charge timer. The timer can be > set to 4, 6, 8... 16 hours. After that time has elapsed, charging stops > and the battery gets discharged. This option with a timer of 4 hours was > chosen by Beomho Seo in the patchset of March 2015. However, that option > is confusing to the user. It doesn't initiate a re-charge cycle. So when > keeping plugged in the device over night, I find it discharging on the > next morning. > - The third option of the rt5033 chip is enabling charging termination. This > also enables a re-charge cycle. When the charging current sinks below the > end-of-charge current, the chip stops charging. The sysfs state changes to > "not charging". When the voltage gets 0.1 V below the end-of-charge constant > voltage, re-charging starts. Then again, when charging current sinks below > the end-of-charge current, the chip stops charging. And so on, going up and > down in re-charge cycles. In case the power consumption is high (e.g. tuning > on the display of the mobile device), the current goes into an equilibrium. > The downside of this charging termination option: When reaching the end-of- > charge current, the capacity might not have reached 100 % yet. The capacity > to reach probably depends on power consumption and battery wear. On my mobile > device, capacity reaches 98 %, drops to 96 % until re-charging kicks in, > climbs to 98 %, drops to 96 %, and so on. Not reaching 100 % is a bit > confusing to the user, too. Is the system powered from the battery in the not-charging case? Anyway, we should teach userspace that "full battery" does not neccessary mean 100%, as keeping battery at 4.3V wears it down quickly. Best regards, Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
