A Reconfigurable Multisensor Based on Printed Circuit Board Technology for Measuring Moisture Content and Temperature in Stored Grain

Abstract

We present the design and fabrication of a reconfigurable smart-sensor for measuring moisture content (MC) and temperature in grains stored in silos, using a capacitance fringing field interdigitated sensor and using a resistance temperature detector (RTD), both implemented on a single conventional printed circuit board. The sensor was tested in laboratory, using corn kernels with percent MC in the range of MC <inline-formula> <tex-math notation="LaTeX">$= 8\%$ </tex-math></inline-formula> to MC <inline-formula> <tex-math notation="LaTeX">$= 32 \%$ </tex-math></inline-formula> and, using a simple relaxation oscillator and a frequency-to-voltage converter, we measured a capacitance variation of <inline-formula> <tex-math notation="LaTeX">$\Delta C = 4.99$ </tex-math></inline-formula> pf, with a sensitivity <inline-formula> <tex-math notation="LaTeX">$S = 0.21 \, \rm {pF}/\%$ </tex-math></inline-formula>MC. The RTD sensor, which uses the copper resistance of the PCB tracks, was characterized in the 10 <inline-formula> <tex-math notation="LaTeX">$^\circ \text{C}$ </tex-math></inline-formula>&#x2013;<inline-formula> <tex-math notation="LaTeX">$50~^\circ \text{C}$ </tex-math></inline-formula> temperature range and the measured results presented a very linear behavior (<inline-formula> <tex-math notation="LaTeX">$R^{2} = 0.99888$ </tex-math></inline-formula>) when compared to the measured temperature values using a commercial sensor (LM135). A modified bridge circuit was developed, where a linear behavior of the differential output was obtained. The developed configuration allowed for the independent adjustment of both the offset and the gain in the output voltage <inline-formula> <tex-math notation="LaTeX">$V_{\text {out}}$ </tex-math></inline-formula> of the bridge. The bridge signal processing circuit was calibrated using an endpoint method, and the measured points between the calibration points presented a maximum endpoint nonlinearity error of <inline-formula> <tex-math notation="LaTeX">$|E_{\text {nl}}| =\,\,0.64\,\,^\circ \text{C}$ </tex-math></inline-formula>.