A Method to Determine the Electret Charge Potential of MEMS Vibrational Energy Harvester Using Pure-White Noise
A Method to Determine the Electret Charge Potential of MEMS Vibrational Energy Harvester Using Pure-White Noise
复制标题
利用纯白噪声确定 MEMS 振动能量采集器驻极体电荷电位的方法
DOI：
10.1109/tsm.2020.2983442
复制
发表时间：
2020
影响因子：
2.7
通讯作者：
H. Toshiyoshi
中科院分区：
文献类型：
--
作者：
H. Mitsuya;H. Ashizawa;N. Shimomura;H. Homma;G. Hashiguchi;H. Toshiyoshi
一种高速测量方法已被开发出来，用于通过电气方式确定 MEMS（微机电系统）振动能量收集器中形成的永久电荷或驻极体的电位电压。传统的导纳方法需要很长时间才能观察时域输出功率的波形，当内置驻极体电势被外部电压补偿时，输出功率的波形会减小。本工作提出了一种新方法，通过对白噪声电压激励的振动能量收集器进行FFT（快速傅里叶变换）分析，独特且快速地确定驻极体电势；当驻极体电势被施加的外部偏置电压补偿时，频域中的谐振峰值减小。传统方法中使用白噪声来测量线性特性器件的频率响应。然而，由于输入信号的随机性，测量结果中固有地引入了随机噪声，这促使我们重复测量并进行时间平均以平滑噪声。因此，我们使用了一种新开发的白噪声，即所谓的“纯”白噪声，它具有完全平坦的功率谱，因为它是由恒定幅度和随机相位的波形合成的。我们计算这种纯白噪声的傅里叶逆变换，无需使用数学平均即可立即获得功率谱，因此测量吞吐量提高了十倍。
A high-speed measurement method has been developed to electrically determine the potential voltage of the permanent charge or electret formed in a MEMS (microelectromechanical systems) vibrational energy harvester. While the conventional admittance method requires long time to observe the waveforms of the output power in time domain that diminishes when the built-in electret potential is compensated by an external voltage, this work proposes a new method to uniquely and promptly determine the electret potential by using the FFT (fast Fourier transform) analysis on the vibrational energy harvester excited by the white-noise voltage; the resonant peak in the frequency domain diminishes when the electret potential is compensated by the applied external bias voltage. White noise has been used in conventional method to measure the frequency response of devices of linear characteristics. However, due to the randomness of the input signal, random noise is inherently introduced to the measurement result, which urges us to repeat the measurement and to perform time-averaging to smooth out the noise. Therefore we used a newly developed white noise so-called “pure” white noise, which has a totally flat power spectrum, as it is synthesized from waveforms of a constant magnitude and random phases. We compute the inverse Fourier transform of such pure-white noises to instantly obtain the power spectrum without using mathematical averaging, and thus the throughput of measurement is enhanced ten-times faster.