How We Tested

Maxpoint is the distributor for several power supply brands including Tagan, Seasonic, Etasis and previously Enermax and CoolerMaster. Tagan is Maxpoint’s own brand, designed in-house and then out-sourced for manufacturing. In Hamburg, Germany it has a room stacked with incredibly expensive power supply testing equipment and not to mention the skills and expertise of its engineers. Maxpoint was kind enough to allow me to use this equipment to test our power supplies – as done so in our earlier group tests.

I decided to put my emphasis more on graphics cards and the 12V rails than anything else, as CrossFire and SLI has proven to be quite tricky to get running in labs with certain power supplies.

I decided to keep the load on the 3.3V, 5V, -12V and 5VSB (Stand By) consistent at all times, independent of load. Unlike the 12V rails, which would increase proportionally to the load on the system (such as playing a game), these rails tend to be very similar at idle as they are at load. I loaded these up at 4.5A, 4.5A, 0.2A and 1A respectively, totalling 44.75W of power.

Testing the 12V rails involved several tests and up to four separate loading machines, all running off a continuous regulated 230V. The main test result was obtained by first calculating the maximum wattage available to the 12V rails once the other rails were loaded up. In well designed PSUs, this meant I could fully load the 12V rails, while in others I was forced to deduct 44.75W from the overall amount.

Once the amount was calculated, this was then divided up proportionally amongst the rails to match the distribution, so that each individual rail was loaded up by the same factor. This was done at 40,60,80 and 100 per cent.

The next test, I worked out which rail(s) were powering the PCI-E power (indicated in bold on the specification tables) and loaded these to 100 per cent to simulate a high powered graphics cards. The other 12V rails had a 7A load placed on them to keep things balanced and realistic. Finally, I tested the overcurrent protection on the same rail, to test the safety aspect of things, but also as a form of overclocking. When at 100 per cent, each test was left for 20 minutes to ensure stability and any voltage drop was noted.

Using the same machines, we also tested power efficiency. This is calculated by comparing the total power taken from the mains by the power supply to the total power outputted on the rails. This is represented as a percentage. The higher the percentage, the less waste there is, the cheaper the power supply will be to run and generally the cooler the unit will be.

Finally, each of the fourteen reviews in this group test is followed by a second page containing the specifications of the power supply as seen on the label, and the testing configuration used at each percentage of load. At the end are all the results for each rail at each level of load. I hope you have as much fun reading it as I did writing it!

comments powered by Disqus