Ceramic capacitors and microphonics: a real-world example.

[jdbakker]

Hi,

Download http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1003,C1040,C1019,P2168,D1258 .
Go to page 10.
Have a look at figure 4 (the scope plot). Anyone looking for a cheap microphone capsule ?

Not a new story, I know. Still, I find it an eye-opener to see such a clear example. >1mV peaks, that's pretty bad. It would have been nice had the data sheet included the exact make/model of the cap, but I suppose that wasn't the goal of the example.

Another datapoint: one of my latest projects for work involved a 25W pulse laser. 10ns 30Amp pulses, with rise/fall times of 2ns, and a 1kHz repetition rate. Everything worked fine, but the circuit would emit a LOUD whine. After dusting the board with fine non-conductive powder the vibration was shown to come from the bank of 0603 X7R ceramic decoupling caps.

Funny thing, some manufacturers of high-end audio ADCs still specifically recommend decoupling their high-impedance Vref nodes with... X7R capacitors.

[Note that I do not claim to know enough about the innards of said ADCs to predict the effect of piezo-induced noise on the conversion process; it may well get cancelled out as a common-mode effect. Still, I'm not taking any chances]

JDB.

 

[bcarso]

And X7R is relatively benign. Try Y5P, Z5U etc. for some serious piezo action.

 

[jdbakker]

[quote author="bcarso"]And X7R is relatively benign. Try Y5P, Z5U etc. for some serious piezo action.[/quote]

Indeed. In all fairness, I have the (unsubstantiated) feeling that the plot in the Linear Devices datasheet I linked to is for a Y5V cap.

Hm, none of the boards I have on hand have large Y5V caps, although I know I have used them as PSU input caps in some designs. I'll do some measurements if/when I find one (+ some time).

JDB.