AI-Based Soft-Switching Inverter Tech Significantly Reduces Power Losses

Check out our PCIM 2022 coverage.

Soft-switching hasn't been previously implemented in dc-ac systems with varying input voltage, temperature, and load conditions. But that's no longer the case: Pre-Switch uses machine learning to constantly adjust the relative timing of elements within the switching system required, forcing a resonance to offset the current and voltage waveforms, minimizing switching losses.

The company recently released its Cleanwave 200-kW inverter reference. Double-pulse test data demonstrates that the Pre-Switch soft-switching platform, which integrates the company's Pre-Drive3 controller board with a Pre-Flex FPGA and RPG gate-driver board, can reduce total system switching losses by 90% or more.

Pre-Switch is enabling customers to create power systems with switching frequencies 4X to 5X faster than legacy hard-switched IGBT systems and 35X faster than hard-switched SiC and GaN systems with half the transistor count. In the case of a SiC-based EV inverter, increasing the F_SW from the ubiquitous 10 kHz up to 100 or 300 kHz creates a clean sine wave without any output filter.