Do you have any recommendations on measuring the voltage of a lithium battery (e.g 4.2V) using nRF52?
Obviously you can't measure directly as Vbatt > Vdd.
To use a simple voltage divider you would need a maximum resistance of about 10k to achieve 1% accuracy (because of 1M Rinput). This would draw 420uA which is probably unacceptable if your processor is asleep and using only 1.2uA.
You could have a low-resistance voltage divider that is switched on and off, but you have to ensure that the FET isn't leaking when Vgs=(Vdd-Vbatt). This might be the cheapest solution, but it still draws 420uA when you're measuring.
You could have a high-resistance voltage divider with a voltage follower op-amp, but your voltage divider will now draw power all the time (maybe <1uA if you're lucky) and the op-amp will also draw power (probably >1uA). If you can switch off the op-amp, that may be your lowest power solution.
I'm wondering if maybe there is some clever alternative solution that someone has already invented that is cheaper, lower power, lower part-count, etc. than the ideas I've listed above.
