Access Official Technical PDF

Share Summary:LinkedInEmail

SiC inverter design can help address the competing demands of efficiency, power density and thermal control in electric motor drives. A three-phase prototype developed for an IST Formula Student vehicle illustrates the engineering process: compare wide-bandgap semiconductor options, characterize device losses under inverter-like conditions, then integrate switching, sensing and cooling into a complete system.

The work is relevant beyond racing. Industrial automation engineers and system integrators face similar design questions when specifying or developing variable-speed drives: how will a power stage behave across changing loads, what losses should be expected, and how can heat and electrical transients be managed within a compact assembly?

Why evaluate wide-bandgap devices?

A motor inverter converts a DC supply into controlled three-phase waveforms. By adjusting the output, the drive can regulate motor torque and speed. The power semiconductor technology affects conduction losses, switching losses, operating frequency and thermal requirements.

Silicon IGBTs have been widely used in high-voltage converters because they combine useful voltage-blocking and current-handling characteristics. Their turn-off behavior, however, includes a tail current that contributes to switching losses. Silicon Carbide (SiC) MOSFETs and Gallium Nitride (GaN) HEMTs offer alternative switching characteristics, but their suitability depends on the system voltage, topology and operating conditions.

In the prototype evaluation, SiC MOSFETs and Schottky diodes were selected for the inverter. GaN HEMTs were also tested and compared. The study noted a practical voltage constraint for the commercially available GaN devices considered: their voltage rating made them unsuitable for direct use in a two-level inverter on a 600 V DC bus without changing the topology. A split-bus, open-winding arrangement was considered as an alternative, but would add system and control complexity.

SiC inverter design starts with realistic loss estimates

Semiconductor losses are commonly grouped into conduction and switching losses. Conduction loss reflects the electrical drop across a device while it carries current. For a MOSFET, the on-state resistance is central to this calculation; for a diode, forward voltage and resistance contribute. Switching loss is the energy dissipated as a device turns on and off.

Datasheet values and manufacturer SPICE models are useful inputs, but they do not automatically describe every real operating point. In a motor drive, current and voltage change throughout operation, and switching energy depends on the conditions at each transition. A single loss figure can therefore be an incomplete basis for estimating heat generation across a dynamic duty cycle.

To build a more application-relevant picture, the devices were tested in inverter configurations. Double-pulse testing was used to examine switching behavior, while conduction behavior and temperature response were also characterized. Experimental results were compared with simulations based on manufacturer-provided SPICE models. The measured data was then used to develop dynamic loss models that account for changing operating conditions.

This approach has a practical lesson for drive development: use datasheets and simulation to guide component selection, but validate important loss assumptions with measurements that reflect the intended circuit and operating range.

Integrating the power stage

Choosing the semiconductor is only one part of building an inverter. The prototype also required gate-drive circuits, a DC-link capacitor, voltage and current measurement, temperature sensing, a printed circuit board and a cooling plate.

Gate-drive design must support controlled switching while helping protect devices from fault conditions. The PCB layout also matters: parasitic effects in the power circuit can influence switching behavior and voltage overshoot. The project therefore treated reduced parasitic effects as a design objective, alongside the physical arrangement of power components and sensing circuits.

Instrumentation provides the information needed to evaluate and operate the converter. Measuring DC and phase-related electrical quantities, along with temperature, allows engineers to assess performance and observe thermal behavior during testing. These functions must be designed into the assembly rather than treated as an afterthought.

Thermal management is part of electrical design

Losses become heat, so thermal design is inseparable from semiconductor selection. A compact inverter must provide a path for heat to move from the device through its package and mounting interface to the cooling structure. The prototype included a purpose-designed cooling plate, and thermal analysis was used as part of its development.

Thermal models can help estimate both steady-state temperatures and transient response. Steady-state analysis addresses the temperature reached under sustained dissipation, while transient analysis examines how temperatures evolve as load changes. For a drive with a variable duty cycle, this distinction matters: the power stage may experience short high-load intervals as well as longer periods of lower or moderate demand.

For operations teams, the broader implication is that thermal limits should be evaluated against the actual operating profile, not only a nominal or average load. Loss estimates, cooling design and temperature measurement need to work together.

Testing the inverter under load

After assembly, the prototype was tested with an inductive load and a three-phase motor. These tests verified operation beyond component-level characterization and provided a basis for examining the inverter’s output and efficiency. The reported testing indicated favorable efficiency and power density, though the available summary does not provide numerical performance figures for general comparison.

Testing with different loads serves distinct purposes. An inductive load can support controlled evaluation of switching and current behavior, while a motor introduces the operating context of a real drive. Together, they help expose issues that isolated device measurements cannot reveal, including interactions among modulation, current, switching and thermal behavior.

Practical implications for drive developers

The prototype points to several useful practices for industrial inverter projects:

  • Match device technology to the bus and topology. Voltage rating and system architecture determine which semiconductor options are practical.
  • Estimate losses across operating conditions. Dynamic loads require more than a single datasheet operating point.
  • Combine simulation with measurement. Manufacturer models provide a starting point; circuit-level tests help validate their relevance.
  • Design thermal paths early. Device losses, packaging, mounting and cooling must be considered as a connected system.
  • Include sensing and protection in the architecture. Instrumentation supports validation and makes operating behavior observable.

SiC inverter design is therefore not simply a component substitution exercise. The practical value of SiC MOSFETs depends on how well the devices are modeled, driven, laid out, cooled and tested in the intended application. For engineers developing compact motor drives, a disciplined path from device characterization to system-level testing provides a stronger basis for design decisions—and helps ensure that efficiency and power density targets are considered alongside thermal and electrical constraints.

Access Official Technical PDF

About The Author