We request the manufacturer to extract the fan power leads for 230V, 277V/480V, & 380VDC Internal Redundant Data Center power supplies. The external power source should be appropriately labeled with polarity and voltage specifications. Please refer to the attached photo for proper labeling of external fan/cooling power.

In the testing process, most, if not all, power supply connectors are utilized and connected to the load test fixture. Various power supply configurations may have a multitude of connectors. For instance, the 80 PLUS^® desktop test board provides several connectors for connecting loads, including:

Table 2: Available Connectors for Desktop Power Supplies

Open frame units can be tested. The unit must be connectorized, and if it requires specific cooling air, instructions to provide that air must be included with the unit. If an air plenum is required, it must be provided with the unit on submittal.

We test two desktop power supply samples in parallel. Sample 1 is actively loaded at 0%, 2%/10W, 5%, 10%, 20%, 50%, and 100% load levels. Sample 2 is tested in standby mode according to ALPM and ERP Lot 3 & 6 requirements. This setup allows us to capture performance in both active and low-power states within a single test sequence.

When a unit comes within 0.5% of the next higher badge level on any loading parameter, the test of that unit is completed, and a second unit is tested at the failed point. Should the second unit pass the next level, the second unit is thoroughly tested, and that data is used for the 80 PLUS^® report. If the second unit fails to meet the higher level, then the first unit's data is used to create the 80 PLUS^® report.

The test equipment used can have a significant impact on measurements. In most cases, the difference will be due to input power measurement. The accuracy of the input power measurement is dependent on the power factor of the unit under test, as well as the base accuracy of the measuring equipment. If all equipment and setups were identical, it is still possible to have a difference in readings of twice the stated accuracy. For example, if a power analyzer has an accuracy of ± 0.1%, the worst-case difference could be as much as 0.2%.

If a unit fails to meet any 80 PLUS^® criteria for certification, the test for that unit is terminated. The second unit is then tested at the specific condition of failure. If the second unit passes, the second unit is tested thoroughly, and data is used for the report. The test is terminated if the second unit fails, and a report is issued with recorded failure data.

The AC source and DC load banks are manually set and adjusted at the initial start of the 15-minute interval. Adjustments of the AC source or loads are no longer adjusted during the 15-minute interval while the unit is in operation.

Each Unit Under Test (UUT) begins the test sequence at 0% load. For desktop power supplies, the next test point is at 2% load. However, if the desktop PSU is rated below 500 watts, this step is replaced with a fixed 10-watt load. The sequence then continues at 5%, 10%, 20%, 50%, and 100% load.

For server power supplies, the full test sequence includes 0%, 5%, 10%, 20%, 50%, and 100% load levels. At each load point, the unit is operated for a duration of 15 minutes prior to data collection as shown in Table 1.

The ratings shown on the label of the test unit are entered into an Excel worksheet that performs a calculation based on the algorithm explained in the Generalized Test Protocol for Calculating the Energy Efficiency of Internal AC-DC and DC-DC Power Supplies (Version 6.7.2), Paragraph 6.1.1 Proportional allocation method for loading multiple and single-output AC-DC and DC-DC power supplies.