The batteries used to power wireless sensor nodes have become a major roadblock for the wide deployment. Harvesting energy from mechanical vibrations using piezoelectric cantilevers provides possible means to recharge the batteries or eliminate them. Raw power harvested from ambient sources should be conditioned and regulated to a desired voltage level before its application to electronic devices. The efficiency and self-powered operation of a power conditioning and management circuit is a key design issue. In this research, we investigate the characteristics of piezoelectric cantilevers and requirements of power conditioning and management circuits. A two-stage conditioning circuit with a rectifier and a DC-DC converter is proposed to match the source impedance dynamically. Several low-power design methods are proposed to reduce power consumption of the circuit including: (i) use of a discontinuous conduction mode (DCM) flyback converter, (ii) constant on-time modulation, and (iii) control of the clock frequency of a microcontroller unit (MCU). The DCM flyback converter behaves as a lossless resistor to match the source impedance for maximum power point tracking (MPPT). The constant on-time modulation lowers the clock frequency of the MCU by more than an order of magnitude, which reduces dynamic power dissipation of the MCU. MPPT is executed by the MCU at intermittent time interval to save power. Experimental results indicate that the proposed system harvests up to 8.4 mW of power under 0.5-g base acceleration using four parallel piezoelectric cantilevers and achieves 72 percent power efficiency. Sources of power losses in the system are analyzed. The diode and the controller (specifically the MCU) are the two major sources for the power loss. In order to further improve the power efficiency, the power conditioning circuit is implemented in a monolithic IC using 0.18-μ­m CMOS process. Synchronous rectifiers instead of diodes are used to reduce the conduction loss. A mixed-signal control circuit is adopted to replace the MCU to realize the MPPT function. Simulation and experimental results verify the DCM operation of the power stage and function of the MPPT circuit. The power consumption of the mixed-signal control circuit is reduced to 16 percent of that of the MCU. / Ph. D.

Microcantilever sensors and microelectrode arrays (MEAs) are currently used in microphysiological body-on-a-chip systems for the assessment of mechanical and electrical function of human cardiac muscle tissues. However, existing microcantilever devices used in these systems do not acquire direct electrical signals and often utilize imaging or optical detection methods to measure the contractile force of human cardiac tissues and the use of MEAs most often focuses only on unstressed cardiac tissues. New biomedical microelectromechanical systems (bioMEMS) chip-based systems have been developed for more advanced cardiac therapeutic evaluation. This work describes the design, fabrication, and characterization of a piezoelectric microcantilever device, for both sensing and actuation, and a piezoresistive microcantilever strain sensor along with their custom amplification electronics for the ability of high-throughput, real-time, continuous force measurements for in vitro cardiac systems. These devices were developed in a format that would enable integration into a multi-organ body-on-a-chip system. In addition, a new method of using MEA technology for the electrical stimulation of cardiac cells was developed to create a stress test-on-a-chip for functional measurements of cardiac muscle, specifically during exercise-related stress for the detection of arrhythmogenicity at both resting and elevated heart rates. The incorporation of a microcantilever strain sensor and an MEA into a single microphysiological system, with the ability to create a stressed cardiac environment, will offer increased functional testing with the evaluation of contractile force in cardiac muscle and electrical activity in cardiac tissue. This has the potential to allow for a more complete and clinically relevant model including basic physiological investigation, pharmaceutical compound development, cardiotoxicity and efficacy studies, and predictive toxicology.