Optimizing MOX sensor array performances with a reconfigurable self-adaptive temperature modulation interface

The temperature modulation effects of metal oxide based gas sensors have been widely investigated for the volatile compounds detection and, to this aim, several interface circuits were designed for optimization of the sensors properties. In this paper instead of improving only the sensors operation, we propose a novel interface that can also maximize the information that can be collected from the sensor, through a feedback-based thermal modulation method and an active digital control. The self-adaptive modulation method exploits the changes of the sensor resistance to drive its operating temperature. This is implemented in a novel digital interface based on a microcontroller that can drive an array of four sensors independently and provides a real-time control over the modulation parameters. In this way, the behavior of the sensor can be changed during the measurements, enabling the drift compensation, the fault identification and more importantly the minimization of the mutual information between similar sensors. The proposed interface has been tested in two different scenarios: a lab calibration experiment, by recognizing different concentrations of known gases and a practical application, by distinguishing complex VOCs to evaluate the stress level of plants. A multivariate data analysis has been conducted, highlighting a clear effectiveness of the proposed approach and better performances with respect to the standard thermal modulation techniques.

Automated summary

This paper proposes a novel interface for metal oxide based gas sensors, utilizing a feedback-based thermal modulation method and an active digital control to maximize information collection from the sensors. By driving sensor operating temperature based on changes in sensor resistance, the proposed interface allows for real-time control over modulation parameters, enabling drift compensation, fault identification, and minimization of mutual information between sensors. Testing in lab calibration and practical scenarios showed better performance compared to standard thermal modulation techniques.