Hey, the question does have merit. NP0=C0G dielectrics are defined by having capacitance basically independent of temperature (temperature coefficient α within ±0.00003 per degree Kelvin or Celsius), and no aging effects, but the voltage-capacitance relationship is somewhat separate.
As I discovered different types of dielectrics and their properties, I encountered the
Temp and voltage variation of ceramic caps article at EDN (and a variant at
analog.com).
That shows that even for X7R and X5R dielectrics, the dependency varies depending on the voltage rating, but also on the capacitor physical size. The information was taken from public information provided by two different reputable manufacturers (Murata and TDK), so it is credible.
I did not find any curves, but
all capacitor manufacturers do claim that C0G (aka NP0) capacitors have basically no voltage dependency in their capacitance, within the maximum voltage the capacitor is rated for.
The underlying question, then, is whether this is true, or if it is a common misconception stemming from their neglible temperature coefficient.
As I am only a hobbyist in electronics, and a scientist, to me an experiment with a few different C0G caps would nicely resolve the question: measure the capacitance of the caps at different bias voltages. (Can one DC-bias only one leg of a Wien bridge?)
Any suggestions for such a circuit, measuring the capacitance of a DC-biased capacitor?