Abstract : The absolute calibration of the microphones for acoustic measurements requires primary standard microphones. These standard microphones are themselves calibrated using sophisticated protocol (reciprocity calibration) according to the current standards. These standards have been improved all-over the past ten years but some questions remain unclear. In the same way, the appropriate characterization of the artificial ears, required for the calibration of the audiometers, has not been developed yet.
This outlines the lacks in the calibration of standard microphones (in terms of precision) and in the settings of the widely used medical devices. The practical, technical and scientific stakes are therefore of great importance and the studies to be carried out require deepened investigations. Therefore the purpose of this work deals with the reciprocity method for both the free-field calibration and the pressure calibration.
The aim of the first part of this work is to adapt and improve the pressure reciprocity method. The adaptation of this method leads to a technique for characterizing the input behavior of small acoustic components such as small tubes, slits, and cavities (used in the artificial ear). Improving the measurement uncertainties on the microphone efficiency in the highest frequency range led to suggest both an improved model for the microphone and a global modeling for the calibration device in order to study the influence of the radial modes in the cavity on the calibration results.
The second part of this work arises from a key comparison, at an international level, dealing with free-field microphone calibration techniques. This key comparison has required a complete revision of the experimental calibration device at LNE, of the acquisition processes, and of the signal filtering methods required by the extremely low acoustic levels. This work led to undertake more advanced works on both analytical and experimental studies on the concept of acoustic center for microphones.
Some of the results obtained here lay the basis for future works which should enable to improve the modeling for reducing the uncertainties and also for foreseeing the implementation of methods dedicated to the metrology of future MEMS sensors.