Dataset from switching loss measurement test bench by opposition method

📝 Description A detailed description of the dataset: Purpose: To share preliminary data measured from an open source, collaborative test bench Content: Measurements (data points, oscilloscope pictures, measured DC values with multimeters) and configuration file, included post processed data to extract Turn OFF switching losses Sources: Measured at Laplace Lab in Sept. 2026, with IRFP 240 200 V 20 A Si MOSFET, and implemented in two half bridges from Infineon EVAL-1ED3122Mx12H - see settings. Target Audience: researchers, developers, general public, people involved in switching loss analysis and design of power converters More information can be found in: https://github.com/owntech-foundation/test_bench_code (repository, the current branch is "release candidate") and https://forum.owntech.org/t/power-tuesdays-automated-switching-loss-measurement-bench/ The dataset has been generated experimentally by an automated test bench, build at Laplace, in collaboration with LAAS-CNRS and Owntech. We thank all persons involved in this project since 2024: professors, scientists, engineers, students at Université Toulouse. 🔧 Experimental Settings & information Settings: (main key parameters) Parameter Description Sa Number of DC measurement samples per Step NPLC Number of Power Line Cycles delay_com Duration of each phase step InitialPhaseShiftPWM Initial phase shift during DC voltage ramp up Imin min inductor peak current during steps Imax Max inductor peak current during steps For this version, two automated test sequences were produced, for two values of dead time, and each, including 2 DC bus Voltages, 4 peak inductor current, 3 switching frequencies. Parameter Description Sa 30 NPLC 0.5 delay_com 0.6 [s] InitialPhaseShiftPWM 3 [°] Imin 0.5 [A] Imax 3.5 [A] Vmin 40 [V] Vmax 60 [V] frequencyPWM [40,50,60] [kHz] DeadTimePWM [300,600] [ns] InductorL 40 [µH] Vgs +20 / -5 [V] Rgon, Rgof 10 , 4.7 [Ohm] Limits: At the moment, the calculation of Eoff based on total measured power is still under fine tuning and calibration. The measured power is however accurate and fairly reliable. 🗃️ Dataset Structure The whole dataset is compressed in a single ZIP file. The config files used to generate the dataset is included in a separate json. Folders Folder Name Description [dt300ns] [Automated test sequence for fixed 300 ns dead time] [dt600ns] [Automated test sequence for fixed 600 ns dead time] sub-folders Folder Name Description [excel] [Extracted values from the DC measurement of DC bus voltage, and DC current ] [Oscillo] [Additional collected pictures from a second oscilloscope throughout the related automated step sequence] [PhaseShift] [Raw data obtained for several sweeps of switching frequencies, DC bus Voltages, max inductor current] [sauvegarde] [Post processed data to extract Eoff from raw data] The zip file also contains oscillosope pictures acquired at a lower sampling frequency (fixed DC bus voltage, stepped max inductor current) Oscilloscope waveforms Oscilloscope at switching frequency: Channel Description Ch1 Vgs at low side, Half Bridge 1. 350 MHz Passive probe 10:1 Ch2 Inductor current. 100 MHz Hall effect probe, 0.1 V/A Ch3 Inductor differential Voltage. 150 MHz diff probe, 50:1 Ch4 Vds at low side, Half Bridge 2. 150 MHz diff probe, 50:1 Digital D[0,2,5,7] Logic gate signals from µController. D2 Low side HB1, D0 High Side HB1, D5 Low side HB2, D7 High side HB2 Oscilloscope at automated test sequence time window: Channel Description Ch1 Inductor current. 30 MHz Rogowski probe, 0.1 V/A Ch2 Multimeter trig Out signal (DC current DMM) Ch3 DC bus differential Voltage. 150 MHz diff probe, 50:1 Ch4 Multimeter trig Out signal (DC voltageDMM) 📚 Usage How to Access Download all files here. Example Use Cases Compare with datasheet values Compare with simulation values Compare with other measurement methods Citation If applicable, provide a citation format for the dataset: [1]N. Rouger et L. F. Lavado Villa, « Dataset from switching loss measurement test bench by opposition method ». Zenodo, oct. 08, 2025. doi: 10.5281/zenodo.17296029. 🤝 Contributing Please share feedback or contribute to other data sets. Contribute to "open power tuesdays" at Laplace in collaboration with LAAS-CNRS and Owntech. 🚀 Next Steps / Next Developments Priority Task/Development Description Status Expected Completion Medium ["Add new data sets"] [Generate other data sets at higher load currents] [Planned] [Expected date, e.g., Q1 2027] High ["Improve documentation"] [Better description on the Github repo.] [In Progress] [Expected date, e.g., Q4 2026] Low ["Includes Eon"] [Add measurement of Eon] [Planned] [Expected date, e.g., Ongoing] How to Get Involved Report Issues: Contact person Contribute: Use the data set, improve the measurement test bench Discuss: https://forum.owntech.org/t/power-tuesdays-automated-switching-loss-measurement-bench/