Search

This article presents a multi-input piezoelectric energy harvester (PEH) with the multi-input series synchronized switch harvesting on inductor (MIS-SSHI) rectifier and noninterruptive maximum power point tracking (MPPT) technique. The proposed MPPT solution tracks the vibration states of multiple piezoelectric transducers (PZTs) using the fractional cumulative peak voltages method, ensuring the maximum power output for the MIS-SSHI rectifier. In addition, the proposed method captures the peak working voltages of multiple PZTs for MPPT without interrupting the energy harvesting process for open-circuit voltage detection. To validate the effectiveness of this multi-input PEH, an experimental platform is built for performance evaluation. The experiment results confirm the feasibility of the proposed MPPT solution for multi-input rectifiers, achieving a peak MPPT efficiency of up to 98%.

A capacitorless digital low-dropout (D-LDO) voltage regulator with built in adaptive sampling clock is proposed for wireless sensor’s power management applications. The D-LDO regulator adopts a coarse-fine control scheme to accelerate the load transient response. In steady state, the regulator operates at fine-tuning mode with a small switch strength and a low clock frequency. Once the output voltage exceeds the boundaries, a coarse-tuning mode is triggered to quickly find out the desired control bits with a large switch strength and a high clock frequency. The sampling clock is generated with the inherent delay of the internal circuit, thus eliminating the dependence on the external clock. In addition, an active analog-assisted loop is introduced to improve the voltage droop. Fabricated in 180 nm CMOS process, the D-LDO regulator regulates the output voltage range from 0.65 V to 1.15 V. At a VOUT of 0.75 V and VIN of 0.85 V, the proposed D-LDO regulator shows a transient response time of 195 ns with an undershoot of 133 mV at a step load alternating between 5 and 30 mA. It attains a peak current efficiency 99.97% and figure of merit 0.016 ps.

Boolean satisfiability (SAT)-based exact synthesis has made significant progress in recent years, particularly in logic rewriting for the identification of potential subnetwork replacements. However, existing rewriting algorithms suffer from two major drawbacks: 1) inflexibility due to precomputed potential replacement candidates and 2) high-computational complexity of off-the-shelf conjunction normal form (CNF)-based SAT solvers. In this article, we propose a novel semi-tensor product (STP)-based exact synthesis approach for logic rewriting. The STP-based exact synthesis encodes Boolean functions into logic matrices and uses circuit-based all solutions SAT (AllSAT) solver to obtain all optimal replacement candidates with a single pass. Additionally, we improve the subnetwork selection strategy to allow flexible rewriting by selecting the most cost-effective implementation of all optimal candidates. Experimental results, compared with the state-of-the-art logic synthesis tool ABC, show that proposed STP-based exact synthesis reduces 41% runtime on average and solves all instances within a time limit. Moreover, after mapping into 6-LUT FPGA technology and standard cells, we obtain average improvements in the area of 6% and 18%, respectively.

A novel rectifier-less synchronous electric charge extraction and inversion (ReL-SECEI) interface circuit for Piezoelectric energy harvesters is proposed in this article. At the peak voltages of the Piezoelectric Transducer (PZT), a portion of electric charges is inverted on the clamped capacitance Cp, so that the PZT is automatically regulated at a higher output voltage state to improve extracted energy from the PZT. In addition, the self-powered implementation of the ReL-SECEI circuit is presented. It reduces the number of components by optimizing the positive and negative synchronous switches. The rectifier-less and simplified designs reduce the internal power loss and, hence, improve the energy harvesting efficiency. The ReL-SECEI circuit is verified by the experiment, the harvested power is load-dependent and the peak output power is obtained with the optimum load, which reaches 4.2 times that of the standard energy harvesting circuit.