Piezoelectric Effect of Ceramic Capacitors

I am just wondering if there is anybody here that has had this issue with their circuits. Also how likely is it that this effect will add distortion/noise to the output of an audio amplifier even if it is not used in the audio path but only for decoupling purposes of MOSFETS, op amps, and also used as power supply capacitors.

Kemet claims that their **broken link removed** of ceramic capacitors significantly reduces this piezoelectric effect. Does anobody have experience with this?

For decoupling purposes they should be OK as the voltage they introduce will be very tiny compared to the voltage they have across them and it would also be likely that other capacitors are in parallel.

I had a problem once in a digital office telephone product that would occasionally dial wrong numbers or drop the line while transferring a call to a different extension. For a long time it was assumed to be a software bug but nobody could trace it. After detailed analysis it turned out to be a ceramic capacitor in the clock generator of the phones microprocessor circuit. When keys were pressed, the flexing of the PCB under the keypad caused the clock to jump frequency and the data stream to the line was disrupted. Changing the capacitor to one with less microphony fixed the problem completely. I eventually tracked it down using a stable signal generator and mixer circuit. When mixing the clock and the signal generator at the same frequency, the phase difference output seen on an oscilloscope was just as though a microphone was attached, even shouting at the PCB showed the audio waveform!

Brian.

I have had many piezo and memory effect and dC/dV effects in ceramics in my career.

For this reason ceramic should never be used as a reference cap for touchscreens, sample&hold capacitors and are very microphonic in low signal conditions with vibration.

I prefer plastic film in these applications.

Audio circuits have an excellent Supply Voltage Rejection Ratio where they ignore power supply noise. A ceramic capacitor is frequently used ads a high frequency supply decoupling capacitor.

The main consideration of decoupling caps is the self resonant frequency SRF, and the ESR, which is the best case minimum impedance at SRF. For this reason several different values are used in wide band amps such as class D,E . When used in parallel it widens the SRF bands to stretch the bandwidth of the minimum impedance.

Then the source / load impedance ratio of the decoupling caps becomes the % noise level imposed on the supply. At these speeds, the CMMR is diminished greatly, but normally above audio range.

Adding a small choke or ferrite bead improves the noise rejection to a low ESR distributed decoupling caps by LPF response.