摘要
Metal oxide (MOX) gas sensors typically require precise temperature regulation for selective and stable gas detection. However, conventional temperature control methods often rely on open-loop scanning or multiple sensor channels, increasing system complexity and power consumption. This article proposes a closed-loop adaptive temperature control circuit for a single MOX gas sensor that automatically searches for the optimal catalytic oxidation temperature based on its real-time response characteristics. Implemented in a 0.18-μm CMOS process, this design integrates temperature regulation, comparison, and readout functions into a single loop, reducing circuit complexity and area. The overall die area is 1.2 × 1.2 mm, operating at 1.8 V, and it supports a wide sensing resistance range from 9 k Ω to 90 M Ω. Experimental results demonstrate that this system effectively identifies the optimal temperature for varying gas concentrations, achieving high selectivity and sensitivity. It shows great potential for portable gas sensing applications such as environmental monitoring, industrial safety, and noninvasive medical diagnostics. Future work will focus on reducing power consumption and incorporating machine learning techniques to enhance the ability to distinguish complex gas mixtures.