In a previous study, displacement self-sensing using driving current has been proposed for a synchronous-type macroscale electrostatic motor under alternating current (ac) operation; however, it cannot be used in practice due to its low noise immunity. This work improves the position self-sensing method for electrostatic motors under amplitude-modulated ac operation, in which the driving signal is amplitude-modulated using a high-frequency carrier signal. The modulation transfers the signal frequency to a higher region, which results in a much larger driving current and high noise immunity. In the modified self-sensing method discussed in this work, the high-frequency signal is demodulated, and then the slider displacement information is distilled through Clarke and Park transformations. The systematic error of the method is discussed to show that it depends on the performance of the cutoff frequency of the low-pass filter (LPF) in demodulation. Theoretical analysis shows that the accuracy is possibly improved by decreasing the cutoff frequency of the LPF. Simulation results confirmed that the noise immunity of this modified self-sensing method is much higher than that of the previous method, which means it can be used in practice. Verification experiments have been done under various conditions by using prototype electrostatic motors, which validated that the modified self-sensing method realizes a high accuracy/resolution of around $20\,\mu$ m in practical situations.

This work proposes a self-sensing method for a synchronous electrostatic film motor. The method utilizes the relation between the driving voltages and currents of the motor. Through several signal conversions on the measured motor currents, the system distills the information of the slider position to realize displacement sensing. The resulting system can track the slider position both for synchronous and non-synchronous operations. The system assumes stepwise motions of the slider and fully exploits the characteristic current features of the motions. Two different signal conversion methods have been tested using numerically simulated motor currents and real motor currents. The two methods differ in terms of accuracy and noise immunity. Considering that the currents of the electrostatic motor are relatively small, the method with high noise immunity and relatively low accuracy was found useful for practical applications. The position resolution achieved by the method was in the order of the electrode pitch, which is 200 μm in the prototype motor used in this work.

Several kinds of exclusive displacement sensor systems have been proposed for electrostatic film motors, to fully profit from their thin and flexible structures. The existing sensor systems share the same underlying principle, which utilizes a multi-phase high frequency sensing signal to realize the robust and precise displacement sensing. However, high frequency oscillators and amplifiers are required in them, which can increase the complexity of the motor system. This paper proposes a new method to estimate the slider position for the electrostatic film motors based on the characteristics of the motor itself. The proposed method has lower sensing resolution than previous sensor systems, but it can be realized in a much simpler way, contributing to the simple system structure. The displacement will be estimated directly from the driving voltages and currents, by which the sensing system can be simplified. This paper focuses on a synchronous-type motor and analyzes its current characteristics during synchronous operation in theory, based on the capacitance network model of the typical electrostatic film motor. From the theoretical results, it has been known that pulse signals in current features, which can be obtained by removing base waves of the source currents, correspond to the stepwise motion. Because pulse signals change periodically, a new pulse train can be obtained by adding the absolute values of these current features. The slider position can be calculated with a resolution of the electrode pitch, by counting the number of pulse signals in pulse train after binarization. A verification experiment has been done, which has shown that the slider position can be estimated by the proposed method with a resolution of the electrode pitch.

Electrostatic motors are promising forms of actuation for future robotic devices. The study of their different implementations should accelerate their adoption. Current models for resonant electrostatic induction motors were found not to be able to properly describe their behavior, namely, with regard to changes with position. This paper reports a new analytical model for these motors, aiming to address this issue. The model is based on identification of all capacitance harmonics, through a simplified method. Using these, equations for different motor parameters, notably, thrust force, were obtained and compared to previous literature. The new equations model position dependent properties, such as force ripple. The outputs of this model were validated through experimentation with a prototype, with the results confirming the new model better describes motor behavior. An analysis into how to decrease this ripple was also discussed and tested. We concluded that the use of a higher number of harmonics resulted in a much more accurate model, capable of adequately characterizing motor outputs with changes in position.