Methods to Measure Power Electronics’ Efficiency

Validating the system efficiency of a power-electronics circuit is essential in evaluating the overall system performance, design optimization, and sizing of cooling systems. Figure 1 shows the conventional method of performing efficiency measurement. The power electronics system operates at the rated output-power level, and, by measuring the input power and output power, you can calculate the systems efficiency using the equation η=(POUT/PIN)×100%, where POUT is output power and PIN is input power. In other words, the measured input power is equal to the output power plus the power loss of the system.

However, measuring the efficiency of a high-power system that delivers power to loads such as motors, generators, or industrial-computer equipment requires a source that delivers the rated power. The infrastructure therefore should comprise a suitably rated source and an equivalent load that can support the rating of the power-electronics system you are evaluating. These requirements can drive up the facility’s infrastructure cost; for one-time design-validation measurements, this cost is difficult to justify.

This Design Idea describes alternative methods of measuring the efficiency of a high-power power-electronics system that simplifies the test-infrastructure requirement by eliminating the test load and using a source that must support only the loss of the power-electronics system. Figure 2 shows the proposed method, which eliminates the test load by shorting the output/load terminals. The system’s control algorithm maintains the required input- and output-current amplitude and frequency by developing circulating reactive power. IGBTs (insulated-gate bipolar transistors) and magnetic components dominate the system’s losses, which are functions of the amplitude and frequency of the input and output currents. The loss is also less sensitive to the power-factor and PWM (pulse-width-modulation) index.