Evaluation of power losses in traction converters equipped with different types of MOSFET SiC switches
Reference number | |
Coordinator | Bombardier Transportation Sweden AB - Bombardier Transportation AB Västerås |
Funding from Vinnova | SEK 727 000 |
Project duration | June 2014 - May 2015 |
Status | Completed |
Important results from the project
The purpose of the project was to estimate the possible efficiency improvement using SiC based power semiconductors instead of Si in traction converters for trains. Two SiC device structures, planar and trench, have been evaluated using loss data derived from FEM simulations on 1.7 and 3.3 kV devices. For a typical traction load cycle the converter efficiency increased from 97,1% to 98,8%/99,0% using 1.7 kV devices and from 99,1% to 99,5%/99,6% using 3.3 kV devices relative the Si case. In both cases the trench SiC MOSFET gave better result.
Expected long term effects
The project has led to quantified estimations of conduction and switching losses for planar and trench SiC MOSFET device structures for 1.7 and 3.3 kV devices. The project has also evaluated existing MOSFET compact circuit simulation models and their ability to properly describe SiC based MOSFETs. Comparison with measurement results show about 10% deviation in turn-off losses and 30% in turn-on losses. Results show that the MOSFET model for SiC needs improvement, e.g. the description of transition from linear to saturated region of the on-state characteristic.
Approach and implementation
The work has been carried out by Bomabardier and Acreo. Acreo has made FEM simulations of 1.7 and 3.3 kV SiC MOSFET device structures, planar and trench. Bombardier has evaluated compact simulation models for SiC MOSFETs, compared against measurements on 300A/1.7kV SiC MOSFETs and calculated converter losses and efficiency for traction converters using a typical load cycle. Competence wise Bombardier and Acreo matched each other very well in this project which led to high quality results in an efficient manner. Several ideas for future work were also identified.