Optimized Metal–Oxide Semiconductor Sensor Readout Techniques for Quantitative Gas Sensing

This work presents a framework to optimize the readout interface for Metal-Oxide Semiconductor (MOS) gas sensors considering the power-law model that describes the sensor response, both for reducing and oxidizing analytes. Considering the two excitation schemes, i.e. the Constant-Current Excitation (CCE) and the Constant-Voltage Excitation (CVE), the proposed analysis aims to optimize the sensor signal acquisition, thus minimizing the system complexity and the number of bits of the Analog-to-Digital Converter (ADC) required to achieve the desired gas concentration resolution across the gas concentration range of interest. The theoretical analysis is supported by an experimental validation, wherein the resistance responses of two commercial n-type sensors were emulated using digital potentiometers. The results show that it is possible to achieve the same gas resolution over a given gas range using a less performance-demanding ADC. Under the considered experimental conditions, the CVE scheme requires an ADC resolution 7 bits lower than the CCE scheme for the reducing gas scenario. Conversely, in the oxidizing scenario, the CCE scheme requires 5 bits less resolution than the CVE approach. Furthermore, the required conversion chain gain is reduced in the latter case, thereby reducing the front-end requirements.