A QCM SENSOR ARRAY-BASED ELECTRONIC TONGUE WITH THE OPTIMIZED OSCILLATOR CIRCUIT USING FPGA

Abstract

In this project, an electronic tongue (ET) based on a quartz-crystal microbalance (QCM) sensor array was designed and the methods for sensor data analysis were described. In this regard, as chemical sensors, QCM crystals coated with three different phthalocyanines were employed to discriminate toxic chemicals in liquid media. For sensors’ preparation, an automatic coating instrument was developed based on the airbrush technique. The oscillator circuits used in the coating instrument and measurement setup were optimized to minimize the noise and mutual interference of the other oscillator circuits in the sensor array. The designed measurement system, which was responsible for collecting and storing data, communicating with a computer, consisted of a field-programmable gate-array-based control unit. The measurement system collected training data from the sensor array, and then computer application using MATLAB used an artificial neural network to find the composition of toxic chemicals in the sample water. The measurement system was tested for three different toxic chemicals, and the average success rate in predicting them was 4.08%. The desired chemicals can be detectable by preparing the necessary sensor array with the proposed ET.

Existing System

In the existing system ETs like voltammetric, potentiometric, optical, or enzymatic sensors are used.

Proposed System

In this project a simple and cost-effective solution to detect chemicals in liquid media was developed. The developed measurement system consisted of an optimized oscillator circuit with automatic gain control (AGC) in order to interface with the QCM sensor array. Then, by using the designed oscillator circuits, both the coating instrument and the field programmable gate-array (FPGA)-based measurement setup were developed. The designed FPGA-based measurement instrument is able to be reconfigured easily based on the measurement system requirements or the specific characteristics of the QCM sensors. Discrimination of the chemicals in water was achieved with an artificial neural network (ANN) after the training algorithm was applied. Finally, an ET system that was capable of monitoring the chemical levels in water was achieved. Even if, in this paper, three different chemicals were distinguished by using three sensors, it is possible to detect the desired chemicals by making necessary sensor coatings with this system.

Architecture

Block Diagram of Measurement Setup