Doctoral Defense

Ultra-low Power Adaptive RF Energy Harvesting and Power Management in Integrated Sensory Systems

Puyang Zheng

May 3 , 2024
3:00 PM
Light Engineering, Room 250
Advisor: Milutin Stanacevic

Radio Frequency (RF) powered backscattering sensors lower the energy cost of the communication in sensory networks and therefor present a promising technology for ubiquitous deployment of Internet of Things (IoTs) devices. However, ultra-low and intermittent power available from ambient wireless source(s) presents a challenge in the design of the harvester and power management system that provides supply voltage(s) for sensing and computational tasks in RF sensors.

We propose a RF energy harvesting system implemented in the standard 180nm CMOS technology that operates at 915 MHz ISM band. A system is designed to achieve the maximum power conversion efficiency (PCE) within the input power range of from -30
dBm to 0 dBm. A tunable L matching network is modeled and optimized at the antenna interface. We propose a dual-channel Dickson rectifier that uses diode connected native NMOS transistors and delivers 1uW of power to the resistive load at input RF power of -27 dBm. A power management unit (PMU), optimized for the operation modes of RF sensors, enables system to achieve an overall efficiency of 25% at -13 dBm input power.

We also present analysis of a conventional fully integrated NMOS low-dropout (LDO) voltage regulator with ultra-low quiescent current tailored to RF sensors. Voltage regulators are a key building block in the design of PMU. Subsequently, an advanced NMOS LDO architecture amenable to integration in RF energy harvesting sensors is proposed to address the limitations of the conventional architecture. The transient enhanced inverter (TEI) is introduced to improve the load transient response. The start-up assistant RC (SARC) improves the settling time for the step input voltage to LDO. The proposed LDO achieves state-of-the-art performance for ultra-low incident RF power.