09-10-2016, 12:32 AM
(This post was last modified: 09-10-2016, 03:36 AM by pfeerick.
Edit Reason: looking at wrong datasheet!
)
The other thing that may be a factor is that most 'automatic' charge controllers are not dumb... they don't just charge at their set current... but ramp it up based on how the battery voltage responds... if the voltage rises too quickly... it drops the current as it senses it's charging the battery too hard.
Since you're using the pine64 batteries with thermistors and I'm not, in the absence of any other data, I'd be thinking it's the lack of thermistor that is altering things... although from how the datasheet reads, connecting the temp sensor pin to GND should disable the temperature sense part of the charge algorithm.
Anyway, my battery is charging again after a full cycle down to flat with a shutdown script triggering at 5% remaning capacity, and on the recharge it is back up to 3.7v, 6% capacity charging at 415ma (so says the pine) with total load of 580ma on the 5v side.
Edit: I had to go back and edit my previous post, as I was reading the datasheet for the PMIC on the CHIP, not the pine64... and unfortunately the chips PMIC has some more sane defaults... like the charge led being controlled by the charger by default, whereas the 803 on the pine64 defaults to register (so software) control :-(
Since you're using the pine64 batteries with thermistors and I'm not, in the absence of any other data, I'd be thinking it's the lack of thermistor that is altering things... although from how the datasheet reads, connecting the temp sensor pin to GND should disable the temperature sense part of the charge algorithm.
Anyway, my battery is charging again after a full cycle down to flat with a shutdown script triggering at 5% remaning capacity, and on the recharge it is back up to 3.7v, 6% capacity charging at 415ma (so says the pine) with total load of 580ma on the 5v side.
Edit: I had to go back and edit my previous post, as I was reading the datasheet for the PMIC on the CHIP, not the pine64... and unfortunately the chips PMIC has some more sane defaults... like the charge led being controlled by the charger by default, whereas the 803 on the pine64 defaults to register (so software) control :-(
(09-09-2016, 05:51 AM)MarkHaysHarris777 Wrote:(09-08-2016, 11:04 PM)pfeerick Wrote: Also, I don't think the charge current is anywhere as high as you think it is... the AXP209 AXP803 is capable of charge currents of up to 2.8A, but I haven't seen the pine64 report anything above 600ma charge current, so it certainly isn't configured that way on mine, and I haven't touched the AXP registers. The AXP209 AXP803 defaults to 1200mA charge current, ... but then again, the register defaults also indicate the charge LED should work out of the box, so some register settings certainly appear to have been changed to less than optimal settings!
This seems to be another inconsistency between these boards. But , if my 8000 mah batt is being fully charged in 3.5 hours (that is consistent between the two machines) then the current has to be somewhere around 2.3 amps.
I think the charge times are definitely affected by what else is plugged in, whether the hdmi is running, and other factors... but, also I think that the machines are not consistent in this regard either; I'm hearing all sorts of conflicting statements (and I only have three boards to test on).
