Cooler Master GX 650W Power Supply Review

Cooler Master is breaking new maketing ground this time around with a "Gamer Xtreme" PSU at 650 watts. Marketing aside, it is packing a 52 amp single rail and a 5 year warranty, so maybe all is not lost on kitschy sales speak. Gag me with a spoon.

continued...

Ripple Testing

Since voltage output is not the only concern when it comes to quality DC output we next examined the ripple and regulation characteristics of the Cooler Master GX 650W. We examine these points since unnecessary ripple can cause premature failure of sensitive components in a number of different PC subsystems.

The DC output quality was logged via our digital oscilloscope and the EasyScope II software package. Each divider horizontally represents 2ms while each divider vertically represents 0.05v or 50mv. The ATX specification states that a unit should remain at or below 120mV of ripple and noise on the 12v rail while under 50mV on the 3.3v/5v rails.

Control Test Graphing

Article Image

This image is the blank background control test on an unused connector from our SM-8800 during the Cooler Master GX 650W testing. This lets us determine what the background noise looks like during testing. If at any time a trace deviates from this reading that is the noise/ripple being logged by the oscilloscope for that rail. As you can see the trace is flat and shows as a blue line obscuring the axis. If during a test the axis becomes visible but a waveform is hard to discern it is most likely due to the amplitude of the trace being small in relation to our voltage divider.

120v and 100v Input

Article Image Article Image

Test #1 is equal to approximately 25% of the rated capacity of the GX 650W at 45c. This makes Test #1 equal to 173W by loading the 12v rail to 12a, the 5v rail to 1a, the 3.3v rail to 1a, the +5vsb to 2a, and the -12v to 0.5a. In Test #1 we see the GX 650W starting off with active traces as the peak amplitudes are ~25mV for the 12v rail, ~20mV for the 5v rail, and ~30mV for the 3.3v rail.

Article Image Article Image

Test #2 is equal to approximately 50% of the rated capacity of the GX 650W at 45c. This makes Test #2 equal to 324W by loading the 12v rail to 24a, the 5v rail to 3a, the 3.3v rail to 2a, the +5vsb to 2a, and the -12v to 0.5a. Test #2 sees the 5v rail stay even with Test #1 values while the 12v rail moves up to ~25mV and the 3.3v rail moves up to ~35mV.

Article Image Article Image

Test #3 is equal to approximately 75% of the rated capacity of GX 650W at 45c. This makes Test #3 equal to 481W by loading the 12v rail to 36a, the 5v rail to 5a, the 3.3v rail to 4a, the +5vsb to 2a, and the -12v to 0.5a. Test #3 sees more increases in trace amplitude with the 12v rail hitting ~35mV and the 5v rail hits a peak trace amplitude of ~30mV. The 3.3v rail, however, peaks the highest at the ATX12v specification limit of ~50mV.

Article Image

Test #4 is equal to approximately 100% of the rated capacity of the GX 650W at 45c. This makes Test #4 equal to 633W by loading the 12v rail to 50a, the 5v rail to 5a, the 3.3v rail to 3a, the +5vsb to 2a, and the -12v to 0.5a. In the final regular test, at 120v, we see the 12v rail's peak trace amplitude grow to ~50mV while the 5v rail holds steady at ~30mV. The 3.3v rail however has gone completely out of specification now as it hits ~65mV.

DC Output Quality Summary

Like many other things today the DC Output Quality of the Cooler Master GX 650W is disappointing, and ugly. Test #1 starts off the show with active traces that peak at ~25mV for the 12v rail, ~20mV for the 5v rail, and ~30mV for the 3.3v rail. This only grows throughout testing until we hit the ATX12v specification limit for the 3.3v rail in Test #3 at 50mV. This is with only a 4a load as well. When we switch to Test #4 we see the 3.3v rail's trace amplitude grow once more as it hits ~65mV at 120v. The 100v result was actually larger than this as well, however the unit died before the data was able to be logged. So, had the GX 650W not been a total failure at load testing, it would have failed the DC Output Quality portion of today's review as these results just plain suck.