Well, I'm coming back ;
I make 2 polls, when the phone was charging, and when it was discharging.
Polls records every 60 sec during 28800 sec (8 hours).
During the 2 polls, the phone was alone w/o any human activity (screen off)
The polls records : UnixTime, CPULoad, RAMUsed(KB), TotalIRQ, QuectelIRQ, PowerCapacity(%), and DPMS State (On/Off)
poll script was run as root, default prio...
* 1st poll "Charging" :
- PowerCapacity :
- CPU Load :
- IRQ : didn't notice any storm, 5K every 60s.
- IRQ Quectel : growth +20 every 60s. ; no storm.
- DPMS : screen stay off during the whole session
* 2cd poll "Discharging" :
- PowerCapacity :
* Conclusions :
The battery/power capa is weird, it seems that some events put it down, but the capacity reported by the system seems to be wrong. We cannot trust what is reported, but it is sure that when the phone is used (calling, sending msg), the power decrease abnormally and the phone goes Off.
Any other ideas ?
Thanks for your help,
Ps : can send .csv data if needed
Val.
Edit : some GNUPlot
I make 2 polls, when the phone was charging, and when it was discharging.
Polls records every 60 sec during 28800 sec (8 hours).
During the 2 polls, the phone was alone w/o any human activity (screen off)
The polls records : UnixTime, CPULoad, RAMUsed(KB), TotalIRQ, QuectelIRQ, PowerCapacity(%), and DPMS State (On/Off)
poll script was run as root, default prio...
* 1st poll "Charging" :
- PowerCapacity :
- starts at 14% and growth smoothly +2% every 60s.
- at ~90%, it growth 1% every 60s.
- at 99%, it growth slowly until 100% during 15mn.
- stay at 100% during 15mn., then decrease to 92% in 3h.
- increase at 100% during next hour
- re-decrease as above ...
- CPU Load :
- average is 0.10% (CPU state fetch and avg during 5sec.)
- max at 3.41%, min at 0.00
- IRQ : didn't notice any storm, 5K every 60s.
- IRQ Quectel : growth +20 every 60s. ; no storm.
- DPMS : screen stay off during the whole session
* 2cd poll "Discharging" :
- PowerCapacity :
- starts at 93%
- for obscure reasons, UnixTime reports +14000s. after 5mn. ; I don't know if it comes from a bad value, or if the time is the real one (so, it means the script run after waiting +14000s.) ; CPU Load is at 18% during the poll, and the screen goes 'On'. Power decrease immediately to 80%. I didn't notice any external event (SMS ...)
- same event appears 5mn. later ; CPU avg at 14%, power capa decrease to 32%
- same event appears 5mn. later ; CPU at 1.60%, power capa decrease to 4%
- power increase smoothly +1% every 3 or 5mn.
- when it reaches 100% (after some hours), it decrease to 92% then growth to 100%, and so on.
- we find +7500 req. every 60s., but sometimes we have +30K ; these don't come from the Quectel Modem.
- I didn't notice anything
* Conclusions :
The battery/power capa is weird, it seems that some events put it down, but the capacity reported by the system seems to be wrong. We cannot trust what is reported, but it is sure that when the phone is used (calling, sending msg), the power decrease abnormally and the phone goes Off.
Any other ideas ?
Thanks for your help,
Ps : can send .csv data if needed
Val.
Edit : some GNUPlot