比较器
整流器(神经网络)
电气工程
CMOS芯片
电压
能量转换效率
二极管
功率(物理)
电子工程
计算机科学
材料科学
工程类
物理
循环神经网络
机器学习
随机神经网络
量子力学
人工神经网络
作者
Amin Rashidi,Kjeld Laursen,Seyedsina Hosseini,Hai-Au Huynh,Farshad Moradi
标识
DOI:10.1109/tbcas.2019.2949154
摘要
This paper presents a novel micro-scale ultrasonically powered optogenetic microstimulator with the vision of treating Parkinson's Disease. This system features a power-efficient active rectifier benefiting from a novel powering approach for its comparators. The main basis of the idea is to lower the Rail-to-Rail supply voltage of the comparators, thereby lowering their propagation delays. This technique improves the power conversion efficiency of the active rectifiers in two ways. First by decreasing the propagation delay of the comparators, and second by reusing the consumed power by the active diodes. The proposed system including the active rectifier, a novel double-pass regulator, a current reference, and a burst detection circuit is designed, simulated and fabricated in TSMC 0.18 μm CMOS technology with a total silicon area of 0.09 mm 2 . Based on the experimental results, the proposed active rectifier exhibits a voltage conversion ratio of 92.22-93.57% for input voltages of around 3 V, and a power conversion efficiency of up to 94.51% for a load of 2.84 mW and over the frequency range of 1-7 MHz. A proof-of-concept system including the fabricated chip, a 560 × 560 × 490 μm 3 -sized lead zirconate titanate (PZT-4) piezoelectric receiver, and a custom-designed 280 × 180 × 100 μm 3 blue μ LED is designed and measured in a Water tank. For an acoustic intensity of 28.8 mW/mm 2 , the available electrical power at the crystal terminals, the output DC power, and the output light intensity were measured equal to 2.454 mW, 2.276 mW, and 9.29 mW/mm 2 , respectively. The quiescent current of the chip in absence of power bursts is measured equal to 8.4 μA.
科研通智能强力驱动
Strongly Powered by AbleSci AI