High-Frequency Power Conversion & FPGA Control

Context: University of Minnesota · Prof. Jungwon Choi Research Lab
Period: 2022–2023 · Collaborative university research
Focus: Power Electronics & FPGA
Tools / methods: Intel Quartus Prime · FPGA · MATLAB/Simulink · LTspice

Engineering question

High-frequency power conversion benefits from reduced magnetic component size, but switching loss, timing precision, electromagnetic effects, and gate-drive coordination become increasingly consequential. A programmable digital controller can provide reproducible timing signals for experimental converter studies.

My work

  • Contributed to research on power-conversion circuits, resonant operation, and wireless-power applications.
  • Worked with simulation and prototype-oriented analysis of high-frequency converter behavior.
  • Participated in FPGA-related digital-control work using Intel’s Quartus design software; the relevant control concepts included PWM timing, synchronization, and timing constraints.

Technical approach

The research connected power-stage behavior with digital-control timing. In an illustrative FPGA-driven converter architecture, a clocked counter and phase comparator can generate switching events; a dead-time stage separates complementary gate commands; and synchronized outputs can be related to the converter’s electrical response.

For a clocked digital timer, a representative timing resolution is

$$\Delta t = \frac{1}{f_{\mathrm{clk}}},$$

and an integer-valued dead-time setting corresponds to

$$t_{\mathrm{dead}} = N_{\mathrm{dead}}\Delta t.$$

These equations describe general digital timing, not a disclosed lab design. Intel Quartus Prime provides an FPGA development workflow spanning design entry, synthesis, device fitting, and timing analysis.

Engineering considerations

  • Switching operation: dead time influences body-diode or third-quadrant conduction, switching transitions, and potential cross-conduction.
  • Resolution: clock frequency bounds simple counter-based PWM timing granularity.
  • Control-to-power interface: gate-driver delays, isolation, and propagation mismatch complicate ideal synchronous assumptions.
  • Magnetics: resonance, frequency-dependent losses, and coupling affect the valid operating envelope.
  • Verification: RTL simulation, timing-analysis reports, and measured gate waveforms establish different forms of evidence.

Evidence and maturity

Research participation involving resonant power conversion and FPGA design tools. The timing architecture above is an explanatory reference, not a claim of verified timing closure, a particular Intel FPGA device, a novel PWM IP core, or measured MHz-range converter efficiency.

Public scope

No university-owned RTL, hardware design, unpublished measurements, source project, or project-specific timing reports are reproduced.

Tool-specific features are cited as background; their availability is not evidence that every feature was used during the research.

Shivam Vashishtha
Shivam Vashishtha
Engineer

Electrical Engineering and Computer Science student working across power, AI, energy systems, and emerging technologies.