Issue 573: Battery Hotswap

Reported by Nikolaus Schaller, Mar 23, 2014

Maybe this is a solution?

Comment 1 by Joerg Reisenweber, Mar 24, 2014

Nice chip. The supposed config would be:
USB Vbus -> $<charger> (TWL4030/bq??? charger) out -> 
main battery -> LTC3355 V(in) 

LTC3355 V(out) -> TWL4030 (ex-Vbat) MAIN BAT V???.IN [[doc: 
SWCS032E; p.33 Figure 4.1]]

The TPS65950/TWL4030 (if I understand correctly) is shutting down 
hard to "off" state at 2.7V(bat) the latest, and has the 
following quiescent current "Ground current I(q)" on 
V???.IN, in "Off at 30°C" [[doc: SWCS032E; p.35 Table 4-3 
3 uA : Table 4-3. VDD1 dc-dc Regulator Characteristics
1 uA : Table 4-4. VDD2 dc-dc Regulator Characteristics
1 uA : Table 4-5. VIO dc-dc Regulator Characteristics
1 uA : Table 4-6. VDAC LDO Regulator Characteristics (("Off 
mode at 55°C"))
1 uA : Table 4-7. VPLL1 LDO Regulator Characteristics
1 uA : Table 4-8. VPLL2 LDO Regulator Characteristics
1 uA : Table 4-9. VMMC1 LDO Regulator Characteristics
1 uA : Table 4-10. VMMC2 LDO Regulator Characteristics
1 uA : Table 4-11. VSIM LDO Regulator Characteristics
1 uA : Table 4-12. VAUX1 LDO Regulator Characteristics
1 uA : Table 4-13. VAUX2 LDO Regulator Characteristics
1 uA : Table 4-14. VAUX3 LDO Regulator Characteristics
1 uA : Table 4-15. VAUX4 LDO Regulator Characteristics
(1uA : Table 4-21. VRRTC LDO Regulator Characteristics; this will 
stay on anyway, to power RTC from main bat)
10uA :   "Sleep mode, I(OUT) = 0";  11uA for "I(OUT) 
= 1 mA"

"Table 4-22. Power Consumption" states:

Mode: "Wait-on"
Description: "The phone is apparently off for the user, a main 
battery is present and well-charged. The RTC registers (registers in 
the backup domain) are maintained. Wake-up capabilities (like the 
PWRON button) are available."
Typical Consumption: 64uA * 3.8V = 243.2 μW

Yet unclear to me if all those quiescent currents would drop even 
further when main battery goes <2.7V. For sure the VRRTC power 
source gets switched from main bat to backup bat, thus reducing load 
to main bat and thus reducing the risk to quickly drain it to an 
irrecoverable state.

Now for the LTC3355 we'd need to keep this chip active (when using 
above sketched config) to power the RTC and other stuff in backup- 
and wakeup-domain. The datasheet states [[doc: p.3]]

I(Q)   V(IN) Quiescent Current   
CONDITIONS:" Charger Off, Not Switching, VOUT = 3.3V, No Load 
in Regulation, Supercapacitor Charged"   
min 60, nom 120,  max 215 μA

we need to consider carefully if those 0.22mA constant inevitable 
load are accepatble even for drained batteries that must not get 
drained further, from - say - 3.0V down to 2.5V which would destroy 
the cell. The remaining capacity in such LiIon cell at 3.0V for 
discharge down to 2.5 is only a few tens of mAh, making for a few 
50s of hours a drained battery may sit in device.
we may consider if we can find a solution that reliably stops 
further battery drain when battery reaches 3.0V. Pulling "PFI 
(Pin 1): Input to the Power-Fail Comparator" to GND may result 
in chip significantly reducing I(Q)

 Integrated Power Management/Audio Codec
 Silicon Revision 1.2
 Data Manual

Comment 2 by william james, Sep 26, 2019

You should add data manual because in this way you know what you are 
doing with code. On the other hand, pick API from for quick 
working if you are looking for some of them.

Comment 3 by Nikolaus Schaller, Dec 14, 2020

Status: WontFix

Created: 10 years 2 months ago by Nikolaus Schaller

Updated: 3 years 6 months ago

Status: WontFix

Followed by: 2 persons