Algorithm for compensation of external influences upon metal oxide gas sensors using the datasheet characteristics

Gas sensors, like any other type of sensors, are affected by external influencing factors among which the most aggressive are the ambient temperature and humidity. If the influence is small, their effect on the global accuracy of the sensor is reduced, and the error caused by these factors is included in the admissible error provided in the datasheet. However, if the influence ais significant, their effect can no longer be neglected and compensation of these errors is necessary based on the known influence characteristics found in the datasheet of the sensor. Unfortunately, these characteristics are not linear and the compensation must be accomplished according to an analytical relationship, if it can be known, or based on look-up tables implemented in the memory of the measuring device. Things get complicated when there are several influence factors. The paper describes a method for compensating the influence of ambient temperature and humidity on an MQ7 metal oxide gas (MOG) sensor, mainly dedicated to measuring carbon monoxide (CO), by mathematically modelling the surfaces of the characteristics given in the sensor’s datasheet and their implementation on a microcontroller platform. Experimental data show that, for a temperature variation between 10 and 50 Celsius degrees (°C) and a relative humidity (RH) variation between 30 and 90%, a reduction of the total amount of error is obtained by compensating the influence quantities resulting in an accuracy improvement of more than 60%.