High-Efficiency Three-Phase Current Source Rectifier Using SiC Devices and Delta-Type Topology
A Dissertation Presented for the Doctor of Philosophy
Degree The University of Tennessee, Knoxville
High-Efficiency Three-Phase Current
Source Rectifier Using SiC Devices and
Delta-Type Topology
A Dissertation Presented for the
Doctor of Philosophy
Degree
The University of Tennessee, Knoxville
Ben Guo
December 2014
Abstract
In this dissertation, the benefits of the three-phase current source rectifier (CSR) in high power rectifier, data center power supply and dc fast charger for electric vehicles (EV) will be evaluated, and new techniques will be proposed to increase the power efficiency of CSRs. A new topology, referred as Delta-type Current Source Rectifier (DCSR), is proposed and implemented to reduce the conduction loss by up to 20%. By connecting the three legs in a delta type on ac input side, the dc-link current in DCSR can be shared by two legs at the same time. To increase the switching speed and power density, all-SiC power modules are built and implemented for CSRs. The switching waveforms in the commutation are measured and studied based on double pulse test. Four different modulation schemes are compared for high efficiency CSR considering the switching characteristics of different device combinations. The most advantageous modulation scheme is then identified for each of the device combinations investigated. A compensation method is proposed to reduce the input current distortion caused by overlap time and slow transition in CSRs. The proposed method first minimizes the overlap time and then compensates the charge gain/loss according to the sampled voltage and current. It is verified that the proposed method can reduce the input current distortion especially when the line-to-line voltage is close to zero. The dc-link current will become discontinuous under light load in CSRs, when the traditional control algorithm may not work consistently well. To operate CSR in discontinuous current mode (DCM), the CSR is modeled in DCM and a new control algorithm with feedforward compensation is proposed and verified through experiments.
LINK ORIGINAL WEB
http://trace.tennessee.edu/cgi/viewcontent.cgi?article=4357&context=utk_graddiss