← NAOAT News

Optical control explores simpler circuits around power switches

An MIT research interview examines vertical gallium nitride devices and the remaining challenge of faster turn-off.

Conceptual illustration of advanced technology research
AI-generated illustration. Conceptual image; not a photograph of the reported event.

In a September 30 MIT Research Laboratory of Electronics interview, researcher Sharon Jung Han Hsia describes vertical gallium nitride power devices controlled by light. The approach aims to simplify isolated control of switches operating at different voltages.

Hsia reports improved performance across successive device generations, including blocking voltage above 600 volts in follow-up work. Faster turn-off remains a major objective. Cost, reliability and circuit-level benefits still need evaluation. The interview identifies solid-state circuit breakers as a possible nearer-term application and multilevel converters as a longer-term direction; it does not announce a commercial product.

NAOAT analysis

The attraction for a circuit designer is a potential change in the supporting hardware around each switch. A device that requires a different control path could alter board layout, insulation planning and the placement of sensitive electronics. Its value should therefore be compared at the converter or breaker level, where those changes affect the complete design.

One useful comparison would hold the electrical task constant and account for the entire control chain in both alternatives. The optical version would need its light source, delivery path and associated power included. The conventional version would need its gate drivers and isolation components included. That accounting would reveal whether complexity disappears or moves into a different subsystem.

The choice of initial application also matters. A research team could compare the switching requirements of a protective interruption task with those of continuous conversion, then identify which requirement the current device can credibly meet. This would make the first demonstration an application decision rather than a contest for the largest headline number. Such a comparison is our proposed evaluation framework, not evidence of qualification for either use. The most persuasive next result would connect device behavior to an identified circuit requirement and show a measurable benefit after all supporting components are counted. That is how an unusual control mechanism could become a practical option for power electronics.