Parallel capacitor affecting voltage divider circuit

Question

Ran into a bit of a strange issue with a voltage divider. I take the voltage of a single cell li-ion and divide it down to a 0 - 3V range so it can be read by an ESP32, I then multiply this voltage by a fixed constant to know what the battery voltage is.

This same circuit has been built on 50+ PCB's but I'm only seeing the issue on 2 specific boards. There is a parallel capacitor for some filtering, but on these 2 specific boards, with this capacitor fitted the voltage reported by the ESP32 is incorrect. Not by a huge amount, in the region of ~200mV.

If I remove the capacitor the voltage is reported correctly.

What I'm struggling to understand is how a parallel capacitor which should in theory be open circuit at DC, is affecting the voltage divider circuit.

Is anyone able to shed any light on why this is happening?

EDIT: Capacitor used is an 0402 X5R 10uF 6.3V MLCC - Kemet C0402C106M9PACTU (Datasheet) (Spec sheet)

Answer 1

I can think of two possible reasons-

1. the capacitors could be damaged and leaky.

2. the ESP32 ADC input (like most unbuffered SAR converters, though this is not a particularly impressive one) draws a certain amount of current every time you make a measurement. There's a small capacitor that needs to be charged through a resistor (14pF and 1kΩ if memory serves). You have a relatively high impedance divider (16kΩ+), so depending on how you are using the ADC (measuring rate etc.) you can have effects like this.

Answer 2

Seems like quite a big capacitor for this use case. Maybe the internal resistance of the capacitor influences the voltage divider. Is the voltage too low or too high?