So just to add some more characterization measurements.
I’m using a Mayfly Sn 20395, and attempting to characterize what range the readings from the Mayfly ADC Vbat are linear when it is only powered from Vbat. I’ve put a resistor network on the extV to the LiIon batter, to halve the voltage to keep it within the Vcc range for the ADS1115. Initially I used 100K//100K – and verified the output is linear.
I’ve connected an external powersupply to the Vbat, an external Meter (Materman 37XR) and then making measurements
Power supply to Mayfly JP1 Vbat (instead of LiIon)
Adjust power supply voltage to desired V measured on Meter.
My conclusion is that the Mayfly Vbat becomes nonlinear under 4.0V.