Efficient Energy Harvesting Circuit for Thermoelectric Generators Using Fuzzy-Logic Based Mppt

Abstract

In this work, we present a highly efficient thermoelectric energy harvesting architecture that integrates a fuzzy-logic-based Maximum Power Point Tracking (MPPT) controller with a 8-stage Reconfigurable Dickson charge pump. The MPPT algorithm dynamically adjusts the number of active chargepump stages to maintain operation at the maximum power point across varying input voltages. To minimize area overhead and improve the efficiency, the reconfigurable Dickson charge pump is implemented using a fully inductor-less topology with a total on-chip flying capacitance of 10 pF. The boosted output is further regulated by a capacitor-less low-dropout regulator (Capless LDO), ensuring a stable 1.56 V supply suitable for wearable electronics. Designed in 90 nm CMOS technology, the proposed system operates over an input range of $\mathbf{0. 7 - 1. 0 5 ~ V}$ and achieves a peak efficiency of $\mathbf{8 0} \boldsymbol{\%}$ at $\mathbf{9 0 1. 3 2 2 6} \mu \mathbf{W}$ which is an improvement from the existing architectures. These results demonstrate that the presented architecture is a compact, power-efficient, and robust solution for self-powered wearable platforms.