Ice detection with coplanar capacitive sensor

Unwanted icing on the cooling appliances is very important in view of cooling performance and energy efficiency. Accumulated ice mass on the freezer decreases the cooling transfer by prevent the air flow contact with cold side of the evaporator. Covering by ice of the coldest zone as called evaporator of the appliance is effect to energy efficiency negatively. A kind of heater is activated for prevent the ice accumulation on the evaporator and it melts the ice by heats the evaporator. In this algorithm as called Defrost, activation of the heater depends on the time, temperature and appliance usage parameters. Activation time and activation duration of the defrost algorithm changes by appliance usage period by user. These durations are calculated by estimation of the algorithm and not be able to detect icing correctly. Therefore, in some cases, even if evaporator zone is not covered by ice, defrost algorithm activates the heater just like evaporator zone is covered by ice and effects the energy efficiency negatively. In this study, dealed to designing of a sensor which detect the icing. A lot of different coplanar capacitive sensor designs have been designed in simulation tool ANSYS Maxwell Software and ice detection state has been simulated. Fringe effect fields have been used for detecting the icing on the sensor surface. For decreasing the effect of the parasite from environmental factors on sensor and detecting the realistic icing, shield has been applied to rear surface of the sensor. The determined optimum sensor geometry according to ANSYS Maxwell Software results has been realized and measurements have been done with signal conditioning circuit. Protection layered coaxial wire harness has been designed for decrease the parasite effect of the environmental factors to cable which between sensor and signal conditioning circuit. In signal conditioning circuit, LM555 timer has been used for measuring the sensor charging time and instant capacitance has been determined by time constant instantaneously. An algorithm was created for eliminating the unwanted situations and preventing the misdetection of the icing. Measurement uncertainty analysis has been done according to measurement modeling and considering the uncertainty sourced from components (resistance, timer, etc.) which used in signal conditioning circuit. Besides, each measurement result in signal conditioning circuit has been determined by average of the 30 consecutive measurements.