Abstract
MEMS electromagnetic energy harvesters face a critical cold-start bottleneck: ultralow voltage combined with high source impedance. Existing startup converters optimize for either low voltage or high impedance, rarely both. This article addresses this challenge by adapting the self-oscillating joule thief topology through closed-form analytical modeling. Unlike prior semianalytical approaches requiring iterative simulation, our framework enables simulation-free design exploration of bipolar junction transistor (BJT)- and N-type metal-oxide-semiconductor (NMOS)-based implementations. For both implementations, the analytical model predicts output voltage within 10% of experimental results over the measured voltage, load, and source-resistance sweeps. The NMOS version demonstrates a 50 mV start-up from a 300 $\boldsymbol{\Omega}$ source impedance. Experimental validation with a 300 $\boldsymbol{\Omega}$ electromagnetic energy harvester shows a tradeoff in practice: zero-threshold NMOS starts at 0.35 g rms acceleration but provides 43% boost; low-threshold NMOS starts at 0.5 g but achieves 93% boost—both outperforming TI BQ25570’s 0.7 g startup requirement. By positioning the joule thief as a specialized startup circuit complementing high-efficiency main converters, this framework addresses a key barrier to self-powered MEMS deployment and high-impedance source energy harvesting.
Funding Information
  • Norges Forskningsråd (299279)

This publication has 26 references indexed in Scilit: