INCREASING THE ACCURACY OF MEASURING AMPLITUDE IN CONTACTLESS CONTROL DEVICES OF THICKNESS POLYMER FILMS

Abstract

The massive use of polymer films of different types and thickness determines the relevance of measuring their parameters, primarily the thickness, especially in the production process.Methods for measuring the polymer films thickness were analyzed. The most promising ones are acoustic and optical methods, in which one of the main parameters of the output signal of the primary transmitter is an amplitude value. During measuring the thickness of thin and ultrathin polymer films the variation range of signal amplitude decreases, which reduces amplitude measurement accuracy by increasing the impact of noise.The purpose of this work is to analyze and develop methods to improve the accuracy of measuring signals amplitude for control devices of polymer films thickness.The methods to improve the accuracy of measuring amplitude were considered. According to the results of the analysis it is proposed to use the structural method based on the amplitude and averaging coefficients. Their values are known for standard form signal, and can be defined for constant signal shape, typical in measuring the polymer films thickness. In this case, the value of the amplitude is determined by measuring average rectified (ARV) and/or root mean square (RMS) values and related coefficients. It significantly reduces the impact of noise.Structural diagram for the proposed method was developed and model researches on its basis were made. They indicate that when signal/noise ratio is in the range of 100 to 5, amplitude measurement accuracy increases more than 30 times. When signal/noise ratio is less than 5, accuracy decreases, but still 5 times higher than in traditional methods of measuring amplitude.Methods comparison for determining the amplitude by RMS and ARV values shows that when the signal/noise ratio is more than 10 the difference is insignificant. When signal/noise ratio is less than 5, better results can be obtained with ARV values and averaging coefficient. Particularly when the signal/noise ratio is 1, the error is less by almost 7 times.Thus, researches have confirmed the high efficiency of the proposed method under the impact of noise and interference, compared with traditional methods.