Modeling and Design of High-Power Medium-Frequency Planar Transformers

Abstract:
Wide bandgap (WBG) semiconductors are developing very rapidly in the recent years, such as Silicon Carbide (SiC) and Gallium Nitride (GaN) semiconductors. There will be higher wider bandgap, higher critical breaking down field, higher thermal conductivity, and higher saturation electron velocity in WBG semiconductors compared to conventional semiconductors, like Silicon (Si) semiconductors. All these physical characteristics will result in reduced switching losses, higher switching frequency, higher withstand voltage, and higher operation temperature. Therefore, WBG semiconductors are gradually applied in the whole electric power system, like wind turbine, solid state transformer (SST), electric vehicle (EV) charger, etc. HEART project in Aalborg university will develop and commercialize a next generation fast EV charger based on SiC semiconductors that will have class leading efficiency and power density.
With the increased frequency, design of medium-frequency transformer (MFT) becomes more critical for whole system. Although smaller volume and higher power density can be achieved, severe eddy currents will occur in the winding conductors, which can be solved by litz wire, cooper foil, printed circuit board (PCB) as windings. PCB-based planar transformers, in particularly, are widely used in low-power applications and show better performance compared to all other structures. Low profile, high surface to volume ratio, accurate mass fabrication, and low cost in PCB-cased planar transformer draw lots of attention to apply it in high-power MFT. But high-power PCB-based MFT will face several challenges in practical applications.
The first one is the large parasitic capacitance due to large width of conductors, which will cause severe electromagnetic interference (EMI) issues. The second one is about thermal performance. Despite the high surface to volume ratio, compact structure still can lead to worse thermal performance, which will influence the materials of PCB and core on performance and even degeneration. The last one is muti-objective optimization. There are too many variables and constraints, case study or brute-force algorithm should be correctly chosen in in different cases.
Based on these challenges, in this Ph.D. project, a general analytical model of parasitic capacitance in PCB-based MFT considering electric field fringing effect based on six-capacitance equivalent circuit will be proposed. Then, this analytical model will be extended based on simplified ten-capacitance equivalent circuit to identify the influence from grounded core and housing on parasitic capacitance. For thermal performance, a more accurate thermal model assuming nonuniform temperature of conductors in the same layer of PCB will be proposed. At last, multi-objective optimization will be conducted based on case studies and the optimal design of PCB-based MFT will be obtain for fast EV charger in HEART project. All models and optimal design of PCB-based MFT will be verified by small signal tests and high-power tests based on dual active bridge (DAB) converter. The main research outcome of this Ph.D. project is expected to contribute efforts to overcoming some engineering challenges in high-power PCB-based MFT, which can give some basic references to the future design of MFT.

Funding: Project 222807: HEART (Super High Efficiency Fast EV charger with energy storage and Grid support functionality: a heart of the future green transport)