SDG #186 A Real Example of Microphonic Ceramic Capacitors!

Microphonic behaviour in multilayer ceramic capacitors is an effect that is mentioned in datasheets and application notes but rarely demonstrated on a real board. This video does exactly that, using audio recording and probing to trace an audible whine back to specific MLCCs rather than to the usual suspects such as inductors or transformers.

The underlying mechanism is piezoelectricity. Class 2 dielectrics such as X5R and X7R physically deform when a voltage is applied across them, so any AC ripple or switching waveform on a rail mechanically flexes the part. Because the capacitor is soldered to the PCB, that movement couples into the board, which acts as a sounding surface and amplifies the vibration into something audible. Two separate components are identified as the noise sources, and the fix is tested rather than assumed, which is the useful part: it shows whether changing the component or its mounting actually eliminates the noise or merely changes its character.

There is also a worthwhile detour into what one of the unidentified components on the board is actually doing, which helps explain why the noisy capacitor sees the waveform it does in the first place.

For anyone designing or repairing switching supplies, the practical conclusions are clear. Suspect ceramics when a board sings, prefer dielectrics with lower capacitance change or larger case sizes where audible noise matters, keep high ripple capacitors away from flexible board areas, and remember that a mechanically quiet layout is a design choice, not luck. It is a good reminder that a capacitor is not purely electrical.