BQ24123: BQ24123 temperature limiting equation ?

What is the resistor divider values you have? Can you please show schematic?

What is the voltage on TS pin during the test?

How much did the temperature window shift compare to +5 to 45? 

I had a second look. The min temperature window is 45 degree. The window can not be narrowed.

I used values 22050 and 4160 ohms on NTC.
RT2 is 8355616 ohm, RT1 is 7929 ohm.
Closest values i can get with resistors are 8357949 and 7905 ohms. Exel calculation say that when NTC is 4160ohm output on TS pin should be 1.085732V, when ntc is 22050ohm output should be 2.317088V. So pretty close.
Next i place the resistors to the board. I use potentiometer to simulate the NTC.
I measure TS pin voltage. When charging is ongoing i have to drop the TS pin voltage to 0.9 volts before the charging ends. Why?
When i try to start charging i need to rise the voltage again over 1.08volts.
Why such behavior? Isnt it logical that when voltage drops under 1.08v charging ends?

Additional information is that the NTC resistor value in previous situation is 3120 ohm which eqials to nearly 60degrees in Celcius whis is 10 degrees higher than calculated. This means that even the limit is set to 50 battery keeps charging even close to 60. The limiting voltage range given on datasheet is not correct? Battery charging start point is between 1.0836-2.31525 but once started the charging can continue even outside this range. Is that correct?

When using potentiometer, the step size is not easy to control. 0.9V is close to the spec. Please see the electrical characterization table.

There is a 40ms max deglitch time on ts fault. This means when the threshold is reached, the charging will suspended after 40ms. Also, are you measuring using the multimeter? Can you show scope capture?

There is a hysteresis on falling edge. Thus you see the different voltage on suspending and resuming charging. The reason it is higher when resuming the charging is that we want to make sure the temperature has dropped to below the hot threshold.

Regarding to your last question. If you plot out the thermistor resistance curve vs temperature, you can see that it is not linear. The bq device is voltage based this means internally the IC is comparing the voltage on TS vs the internal reference voltage. To clarify my previous post, the temperature delta corresponding to 0-45 degree C is not the same as 5 to 50C. Thus if you would like to shift the temperature window, the temperature delta will not be exactly the same.

I use this thread because we are facing similarity challenges with BQ24120 IC.

Our battery cell specification states that charging temperature should be in 0C-45C. As in figure 9 in datasheet (p.14); the window with V(HTF) is higher than V(TCO). And there comes our problem, the charging initialization temperature range works as should be, but during the charge cycle temperature, V(TCO), can be nearly ~10deg higher than the hot temperature threshold. As Jing Zou repply earlier in this thread " The min temperature window is 45 degree. The window can not be narrowed." <- is it mentioned somewhere in the datasheet? I couldn't find. I assume, this is the same with BQ24120? And this min 45C window, is with the V(LTF)-V(HTH), right?

BQ24130 seems to use V(TCO) threshold in RT2 equation (p.17 in datasheet). Why? (Bq24120 use V(HTF) ) Is there any change, to set V(LTF) - V(TCO) to 0-45C in BQ24120 without using external comparator?

Correct, the temperature window is the 45C for all bq2412x family

The voltage threshold should be corresponding to the resistor. If RTH_hot is used, then VHTF should be used in the equation.