Active Topics

 


Reply
Thread Tools
Posts: 256 | Thanked: 939 times | Joined on Jun 2014 @ Finland
#21
Sounds odd. Usually the BNF stops working with BME replacement as/if battery is not calibrated, but starts working when you modprobe -r the battery modules. I think it's intended to work that way. And with bme replaced, you should be able to recalibrate the battery with bigger capacity. I have now switched to Fxtec Pro1, but when I was using n900 I used 1600mAh batteries and they were calibrated ok.

I still think that your device does not actually faint, but shuts down instead. Where the difference is that shutting down is intended and fainting is not. Fainting happens because the voltage have dropped too low for the device to operate, which makes the device to faint immediately, without for example unmounting partitions etc. And this causes very likely some damage. On the other hand, shutting down is intended, and is caused by the os recognizing too low (but not low enough to faint) voltage and shutting down before anything bad happens. This shutting down have set to happen before the voltage drops to dangerously low, and also before battery gets calibrated. Modprobe -r should prevent those safe systems working, but somehow you still have them running. I still have no skills to help you with this, but a pity that you gave up Did you try to use the bq2700.sh script instead of bnf? Don't know why it would make a diffence though. Yoy can find it here, along with some useful info http://talk.maemo.org/showpost.php?p...&postcount=846
Those instructions to calibrate battery are not succesful, however. You probably know this already, but here's the working procedure: http://talk.maemo.org/showpost.php?p...93&postcount=2
 

The Following User Says Thank You to Koiruus For This Useful Post:
Posts: 8 | Thanked: 14 times | Joined on Sep 2019
#22
Thanks Koiruus,

I tried again, and finally success!!! Reinstalled Pali's BME removed the modules and let the battery slowly drain. I actually left it overnight by itself and I think it worked because with modules loaded BNF works and "calibration needed" is 0! The trick seemed to be a very very very slow drain. I actually let it drain overnight without using it at all, no sound, no display. It seemed that was just the low enough drain to get that 15s needed. Of course I'm not sure since I was not watching it!



having advanced power installed was confusing things, but now that I removed that things look to be working well! thanks for the help everyone.
 

The Following 2 Users Say Thank You to bdbogart For This Useful Post:
Posts: 256 | Thanked: 939 times | Joined on Jun 2014 @ Finland
#23
Originally Posted by bdbogart View Post
Thanks Koiruus,

I tried again, and finally success!!! Reinstalled Pali's BME removed the modules and let the battery slowly drain. I actually left it overnight by itself and I think it worked because with modules loaded BNF works and "calibration needed" is 0! The trick seemed to be a very very very slow drain. I actually let it drain overnight without using it at all, no sound, no display. It seemed that was just the low enough drain to get that 15s needed. Of course I'm not sure since I was not watching it!

http://www.ekran.org/tmp/Screenshot-20191223-105738.png

having advanced power installed was confusing things, but now that I removed that things look to be working well! thanks for the help everyone.
Great! But, is that screenshot after calibration? Because it says Battery capacity is 1140mAh. Shouldn't it be more?
 
Posts: 8 | Thanked: 14 times | Joined on Sep 2019
#24
hmmm. The battery says 2450mah on it... I tried calibrating again but it says the same thing, not sure how that # is calculated... I see the following:

Code:
~ $ cat /sys/class/power_supply/bq27200-0/charge_full_design 
2056320
~ $ cat /sys/class/power_supply/bq27200-0/charge_full        
1138830
~ $
So the "design" charge looks correct. I used this battery without re-calibrating for quite some time (months and months) so perhaps I've damaged it's capacity by never letting it totally decharge after quite a few charge cycles.

Last edited by bdbogart; 2019-12-30 at 17:55.
 
Community Council | Posts: 683 | Thanked: 1,227 times | Joined on Sep 2010 @ Mbabane
#25
@bdbogart - the charge_full_design doesn't really mean anything on N900 (unless i'm too mistaken).

You will still need other discharge cycles to get your capacity properly calibrated, since the chip is believed to not allow capacity changes greater than 6% at a time.

Unless you're unlucky to have a fake battery.
__________________
RX-51
 

The Following 2 Users Say Thank You to sicelo For This Useful Post:
Posts: 8 | Thanked: 14 times | Joined on Sep 2019
#26
Originally Posted by sicelo View Post
You will still need other discharge cycles to get your capacity properly calibrated, since the chip is believed to not allow capacity changes greater than 6% at a time.

Unless you're unlucky to have a fake battery.
hmmm. I did one more cycle and now the full charge value is 996mah! A difference of 13%? (1140-996)/1140 = ~0.126

The difference between 996 and 2450 is 6% increments is a quite a few cycles, especially if (as just happened) some cycles have the potential to decrease the apparent capacity...

Hopefully not a fake battery, though there does not seem to be any in stock any longer: https://www.amazon.ca/gp/product/B07B46XKZC
 
Reply

Tags
battery, bme replacement, cssu thumb2


 
Forum Jump


All times are GMT. The time now is 23:33.