Study On Dielectric Measurement Method And System Based On Fringing Electric Field Sensors

Fringing electric field (FEF) sensors have different characteristics from thetraditional capacitive sensor, which the electrodes are in the same plane or curvedsurface. FEF sensors are widely used in non-destructive measurement of productcharacteristics in industrial process control, including porosity, moisture, density,and viscosity, with the advantages of single-side access, adjustable signal strengthand tomography capability. Dielectric constant is acquired by the capacitance andresistance between the electrodes of FEF sensors.The thesis focus on three-wavelength FEF sensor, a measurement system basedon digital signal processor (DSP) is proposed. The system included DSP, drivingcircuit, conditioning circuit, the simultaneous sampling circuit, man-machineinterface circuit. The driving circuit generate sinusoidal drive signal with differentfrequency.The conditioning circuit amplify and filtering the output signals fromthe different channels. The sampling circuit acquire the output signals and deliverthe data to DSP.According to the short circuit current measurement mode, the formula of theFEF sensors’ capacitance and resistance are deduced. The frequency, amplitude andphase of the signals are calculated by using spectrum analysis. In order to guaranteethe measurement accuracy, the sampling frequency is selected in the differentsignal division and the spectrum correction is used. The permittivity is calculatedby the multivariate nonlinear regression method.The system software is designed, including the main monitor program, theinitialization module, the driving control module, the sample module, the calculatemodule and the man-machine interface module. It achieved providing drivingsignal, measurement multi-frequency signals, processing data, display and settingthe parameters.The experiments for thickness measurement, PCB plate flaw and power cableinsulation testing are done. The results show that the measurement system canidentify the thickness of the samples and detect sample damage effectively.