On the issues of linearizing a sensor characteristic over a wider response range

In todaypsilas automation systems transducers are making core elements in the instruments and the circuits used for measurement, control and industrial applications. The task of a transducer is to reproduce a physical quantity as an electrical signal which with the help of conditioning circuits, is transformed into a form that suits a corresponding ADC requirement before a digital equivalent output of the required physical quantity is produced. In the most ideal cases a digital quantity is a true replica of the physical quantity when the transducer has got a linear response. However, in most of the cases the transducers characteristics are nonlinear, and hence at very points along the whole range of the transducer characteristics, the corresponding digital output is an exact replica of the concerned physical parameter. This work is about how a physical read more accurately in the case of nonlinear sensor characteristics, and then a microcontroller is programmed with the same technique while reading from an input over the entire range. The data of the microcontroller reading shows very closely matched with the actual sensors response. Further, the reading error is considerably reduced to within 10% of the actual physical which shows the utility of the technique in very sensitive applications.