Volume 125, Issue 2, February 2009
Index of content:
- TRANSDUCTION 
125(2009); http://dx.doi.org/10.1121/1.3056560View Description Hide Description
Analysis of electromechanical transducers employing axially symmetric vibrations of piezoelectricceramic thin-walled tubes of arbitrary aspect ratio is presented based on application of the energy method [B. S. Aronov, J. Acoust. Soc. Am.117, 210–220 (2005)]. The suggestion made by Giebe and Blechshmidt [Ann. Physik, Ser. 518, 417–485 (1933)] regarding representation of a tube vibration as a coupled vibration of two partial systems is used to choose the assumed modes of the piezoceramic tube vibration, and the energy of the flexural deformations accompanying the extensional vibration of a tube is also taken into consideration. The Lagrange type equations describing the transducer vibrations are derived, and the equivalent electromechanical circuit is introduced that conveniently represents the solution. The resonance frequencies, effective coupling coefficients, and velocity distributions for differently poled piezoceramic tubes as functions of their height-to-diameter aspect ratio are calculated. The validity of the equivalent circuit is extended to the case that a piezoceramic tube is mechanically and acoustically loaded on the ends and acoustically loaded on its outer surface. The results of calculations are in good agreement with the results of experimental investigations.
125(2009); http://dx.doi.org/10.1121/1.3058639View Description Hide Description
This work aims at the development of an expert diagnostic system for moving-coil loudspeakers. Special emphasis is placed on the defects resulting from loudspeaker nonlinearities. As a loudspeaker operates in the large signal domain, nonlinear distortions may arise and impair sound quality. Analysis of nonlinear responses can shed light on potential design faults of a loudspeaker. By exploiting this fact, this expert diagnostic system enables classification of design faults using a defect database alongside an intelligent fault inference module. Six types of defects are investigated in this paper. A large signal model based on electromechanical analogous circuits is employed for generating the defect database, through which a neural-fuzzy network is utilized for inferring the defect types. Numerical simulations and experimental investigations were undertaken for validating the loudspeaker diagnostic system.