A New Modular Approach for Low-Cost Electronic Noses

Electronic noses are used for important applications as food processing, pollution control and security system, as well as for laboratory works, as new sensors characterization. In the first case the main limit still remains the high cost of the equipment due to the need of powerful systems, as personal computer, for the elaboration process. In the second case a high level of flexibility is required, as possibility to add new and well-tailored hardware (e.g. sensor interface circuits). In this paper we propose a solution for the implementation of low-cost electronic noses suitable for both uses. The system has modular low-noise architecture, thus it can be tailored for the particular application, lowering the overall cost. In this work, two innovative chemical sensor typologies (resonant and resistive) have been taken into account and the related high-performance first conditioning circuits have been designed and experimentally tested. The detection process can be performed directly by the instrument using a neural network approach; in this way, no expensive external elaborators are needed, allowing a further reduction in the system cost