Re: Overvoltage transients on ceramic capacitors?
age 5, figure 4 of your article shows that the instant breakdown voltage for a 50v, 10uF, X7R capacitor is 1000V.
This surely cannot be right?
Figure 4 show the max value 1uF ( 1000nF) at 500V surge withstanding,
the much higher density 10uF part is not shown.
Capacitors of this density are not made with high density dielectric with large gap electrodes they are high K but also very small gaps of many very thin layers.
Like a spark plug will fire at 1V/um in air from exceeding the dielectric strength, but in this case it is not the energy of the surge but the energy of the follow-on power after "crowbar" SCR effect. When it arcs, it melts to a short circuit from the storage capacitance and low ESR. THey are not self-healing like some plastic films.
Also it does not occur on smaller values of ceramics which can tolerate a higher voltage transients, because they have a higher ESR by design.
Although some ceramics are designed for "flexible mitigation" meaning they fuse open after arcing into a failed short circuit so that the Automotive 12Vdc wiring does not melt tracks or blow an inline fuse.
50V caps are suitable for a 12V automotive system.
50V 10uF caps are NOT suitable for a 48V LED system, that has overvoltage transients.
Always derate by minimum 25% unless you can guarantee to have transient sub-microsecond OVP protection have other over voltage protection. ( not TVS on a high current drive)
The only valid way to measure nano or sub-microsecond transients is with a proper calibrated probe using 50 Ohm termination and R divider network or FET buffered probe.
If a reasonably low ESR MLCC Ceramic capacitor fails from overvoltage on a DC bus, it crowbars like an SCR and melts down in a low impedance short, unless specifically designed and rated to mitigate this.