Akasa PaxPower 460W Power Supply

How We Test Power Supplies:

In order to thoroughly test a power supply we have built a special rig that is capable of dynamic load levels on each rail. A short article covering our power supply testing methodology in detail can be found here. Please read over the guide if you haven't already, you might just learn a thing or two. Simply plugging a power supply into a regular PC for "testing" is not accurate by any means, nor can any useful numbers be extracted by testing that way.

Also worth reading is our power supply primer article titled, Powering Your PC: Truths and Misconceptions - A Guide To Power Supplies, which covers how power supplies work in great detail.

If you have read the power supply testing article, you know that we publish four (4) groups of numbers for each power supply to give a simplified overall representation that can be compared to other power supplies. Many other load ranges are tested (but not published unless there is a serious problem) to ensure voltage levels remain within ATX12V specifications. I like to refer to it as "working the switchboard" since it involves flipping lots of switches tied to resistors on each rail to achieve various loads. Unfortunately we can't always dial in exact maximum numbers because of the fixed levels of resistance, but load levels we test are often far in excess of any load a regular PC will create. Voltages are typically measured at the SATA & +12V connectors, however other connectors may be used depending on availability.

AC Ripple Test Results:

Using a Tektronix 2215 Oscilloscope, we check for AC ripple that is present in the DC power. We only checked the +3.3V,5V,+12V1, and +12V2 rails (not the -12V or 5Vsb). Potential frequencies could be in the 60Hz range (which is what the AC line power is in the US), or in the higher KHz range closer to the PSU's switching frequency. The ATX & EPS guideline tolerance is 50mV p-p on the +3.3V & 5V rails, and 120mV p-p on the +12V1 & +12V2 rails.

I've decided against displaying a table due to emails I've received from people. Just because one power supply has maybe 10mV more ripple on a rail than another does not necessarily mean it's not as good. As long as ripple is within spec then one should not worry. It's when you see a lot of noise (far exceeding specs) on the rails that something is wrong, and in those cases it's probably just a faulty PSU that needs to be RMAed.

Everything checked out okay with the PaxPower 460W power supply. Ripple was well within acceptable ranges.