As in our last power supply test, Inadequate and Deceptive Product Labeling: Comparison of 21 Power Supplies , we used a special power supply testing device that simulates the power consumption and load of a typical PC system.
THG was able to test the efficiency of high-end power supplies with the help of a special measuring instrument. Here, defined resistive loads represent the power consumption of typical PC components. Reducing the resistance applied to the +3.3-, +5- and +12-Volt lines raises the Amperage (and thus power) requirements at a constant output Voltage.
We presented the ratio of load to the individual lines based on the percentage of power consumed by the components in a high-end PC (see table).
Our lab engineers then measured whether the power supplies kept to their specified operating Voltages, based on an output rating of 400 Watts for power supplies in the 400- to 449-Watt class. In the 450-Watt-and-above league, the output rating was 450 Watts.
Each power supply's maximum output was also determined disregarding manufacturer specifications. The maximum load is an indication of the performance and quality of the components in the power supply. Efficiency is also included in the test results. This is calculated from the relationship between the input and output power.
Another important criterion for evaluation is noise. The volumes measured account for 60% of the ergonomics evaluation. The Mediator 2238 measuring device from Brüel & Kjaer measures the noise generated by the fans at a distance of 50 centimeters.
Since all the devices tested are fitted with temperature-controlled fans, the noise level of each power supply depends largely upon the load placed on the power supplies. For this reason, the noise level is measured shortly after switching on, at a low load, and then measured again under full load once it has had time to warm up.