Pre-Switch, Inc has created the world’s first AI-based forced-resonant, soft-switching technology allowing ultra-efficient DC/AC, AC/DC inverters running at 100kHz. It has announced new data demonstrating the efficiency of the company’s architecture at low and full loads. Tests performed on its CleanWave2 reference inverter at 100kHz switching speed illustrate that the system has 98.5% efficiency at 5% load and a peak efficiency of 99.57%.
Bruce Renouard, Pre-Switch CEO comments: “Over a year ago, we announced peak efficiency figures of over 99%. Since that time, we have been improving our AI algorithm to improve low-load performance. The results are astonishing. 98.5% efficiency at 5% load is far in excess of what can be achieved by any other approach today, proving that soft-switching and the Pre-Switch architecture will lead the emobility electrification revolution.”
The company stops switching losses, allowing 10X higher switching frequencies, producing huge efficiency gains. In EVs, this leads to enhanced motor efficiency, range and reliability. The architecture meets the challenges that, until now, have stopped the industry from transitioning to soft-switching for EV inverters. The excellent efficiency at low loads and high-frequency, low distortion sign wave to the motor greatly enhances WLTP (EPA’s European milage test system) range performance. The topology is a variation of the Auxiliary Resonant Commutated Pole (ARCP) soft-switching converter design. However, the company employs sophisticated, embedded AI to dynamically solve complex switching system timing calculations to produce accurate soft-switching under changing input voltage, output load, device tolerances, and temperature changes. Adaptations are made on a cycle-by-cycle basis to minimise losses and maximise efficiency.
Adds Renouard: “We believe that our technology is the paradigm shift that EV OEMs have been searching for, enabling essential efficiency differentiation and EV leadership. These results prove how effective our technology is and definitively validate our previous claims.”