Pressure Transducer - Natural Frequency vs. Frequency Response

The upper -3 db response will be slightly higher than the natural frequency:

http://www.cs.wright.edu/~jslater/SDTCOutreachWebsite/intro_freq_resp_functions.pdf

Oh, and btw, you don't want any of your source signals' spectra anywhere near the natural frequency of the sensor for linear data acq.

It will depend somewhat on the internal damping of the transducer. However, a rule of thumb is that the transducer should be linear to about 10% of its resonant frequency. Some do better than that but should be tested to determine how much better.

Hi,

a pressure transducer has a flexible membrane that is integral to or attached to the rigid housing.

This membrane has a stiffness c (N/m) and a mass m (Kg). So it is a spring - mass - damper system.

Damping from non-ideal mounting or construction of membrane to housing attachment and material and air-movements.

Any spring-mass system has a natural frequency: ω²=c/m.Frequency f is 2.pi.ω. pi=3.14...

So you will have an amplitude response and a phase response of this transducer as in a system of second order. (assumed that the membrane-movement to electrical signal conversion does not introduce additional changes).

Depending on accuracy requirements the amplitude response is often said to be constant up to 10 ... 40% of the natural frequency but this is an approximation.

The phase is very much dependent on the damping.

http://www.tf.uni-kiel.de/matwis/amat/elmat_en/kap_3/basics/b3_3_1.html

http://commons.wikimedia.org/wiki/Category:Harmonic_oscillation

More in these links and many others.

Back to your question: you need the natural frequency and the damping to calculate frequency and phase response.

From DC to 10% of natural frequency there is little change in amplitude response but may be severe phase change if damping is high.

From 10% to 40% of natural frequency there is - in low damped oscillators - considerable amplitude response. ....

Look to the plots or make your own plot then you will see the situation in much better clarity.

RHABE

Yes, the phase response is very much affected by the damping. We had an application once where we needed to second source a pressure transducer used in a hydraulic servo loop. The natural frequency/scale factor/impedance of the replacement transducer was exactly the same as the one it was replacing but for some reason systems using the replacement would break into oscillation every now and then. ( The control was for yaw of a 30 ton turret on an M1A1 tank so this was a big deal. ) After much head scratching we arranged a test of both frequency and phase response of the two transducers. Very expensive test. What we found was that because the replacement transducer used a tiny filter at the inlet port while the original did not there was a huge difference in the phase response even though the frequency response was the same. The additional phase lag was enough to make the system marginally unstable. We removed the filter and never saw the problem again.

How did you test for the frequency response of the pressure transducer?