A transducer modelling technique for the identification of the transfer function and driving-point impedance

Abstract

A new modelling technique for ultrasonic transducers is proposed in this paper. It consists of a systematic approach that considers the transducer as a two-port black box characterized only by its electrical driving-point impedance and electromechanical transfer function. This transducer modelling technique is therefore suitable for front-end hardware designers because, unlike most widespread models like Mason, KLM and so on, it does not require knowledge of the internal manufacturing characteristics. Driving-point impedance and axial acoustic far-field measurements are experimentally derived from broad-band excitation and stored in a computer memory. With these data sets the transducer model is carried out by identifying two polynomial ratio functions in the s variable domain, which approximates the two experimental data sets. A mean-square-error minimization algorithm has also been developed in order to optimize the identified functions. This algorithm gives good agreement of the identified function with the experimental data. This procedure is applied to different transducers and leads to an effective model since all the actual acousto-electric characteristics are taken into account. The advantages of this technique are exploited when used in conjunction with analog circuit simulation software for the design of front-end electronics. The results presented are relative to the application of this procedure to a transducer operating in the frequency range 3–8 MHz and with characteristic impedance from 10 to 150 ω.