Effect of DC bias on capacitiors vs their DC ratings

[beng]

I have always been told that when substituting capacitors, the general rule was to choose caps with the same or next higher voltage rating and the same or +20% or so capacitance.
The voltage rule was because the effective capacitance varies significantly vs the DC bias voltage between two caps with different voltage ratings.
However I was surprised to learn recently that the effective capacitance of electrolytic capacitors was not as badly affected.(compared to ceramics for eg)

So I think a fundamentals episode on this topic, comparing various capacitor types/voltage ratings and capacitance values would be very informative.

Rgds Ben

[rs20]

Dave has already done a video on DC bias effects on capacitance in certain types of ceramic capacitor. There's not much need for a further video, since other types of ceramic, plus all electrolytic, polypropylene, mica, tantalums etc etc, do not exhibit the DC bias effect to any typically notable degree. The high-value ceramic capacitors are the odd ones out here.



Edit: This is my 1000th post! Hooray! Only 24 more till a real milestone!

[Circlotron]

I did some measurements on some 10nF 50V multi layer ceramics some time ago, taking them up to several times their rated voltage, and the capacitance reduces so much with voltage that the stored energy varies almost linearly with applied voltage rather than the square of the voltage.

[T3sl4co1l]

Yes, outside of the initial high-C region, energy storage is essentially linear (or, C(V) goes as 1 / V).

Which goes to show, if you need bulk energy storage, and it has to be ceramic, you're better off with C0G, which has constant C and therefore quadratic E(V).  This isn't unreasonable for high voltages (the energy density is better than electrolytics), but is rather impractical in cost!

Tim