Page 5

Load testing the PSU

How do you test a PSU? In theory it's not that difficult, but in real-life situations it's very difficult with such a highly rated PSU. But the fact is that we do not have the gear in house to be able to stress this PSU to it's full capacity, we'll get darn close though.

We can connect everything we want, add more devices and overclock, but even then we'd peak at ~550 Watts. But hey, let's try that out and see what happens.

So here's what we did. We took an eVGA Force 680i SLI mainboard and equipped it with Conroe core 2 quad Q6600 Processor, 2x GeForce 8800 GTX cards setup in SLI. After the installation we loaded the latest drivers, and enabled a seriously funky gaming experience and power hungry system. Pretty spiffy setup, don't you think?

Now to stress a little more we overclocked the 4 cores on the CPU towards 3.0 GHz, had the DDR2 memory running at 1147 MHz and the two GeForce 8800 GTX cards toward 625 MHz on the core and 2100 MHz for its memory.

A small issue though since we used two GTXes. This PSU only comes with two dedicated PCI-Express connectors; which is a little silly. So we took two 6-pin to molex converter cables (looks just ugly) and connected the second card. I really fail to see why only two 6-pin PCI-Express connectors are used. 

We now enabled a seriously funky gaming experience at a monitor resolution of 1920x1200 (a good balance between hefty GPU usage whilst using the CPU at maximum as well). Now for the test we used X3: the reunion as it's proven to be a little more demanding than 3DMark. It's a total bitch on graphics cards; it makes them sweat! We set the resolution at 1920x1200 and enabled 8xAA and 16xAF to make sure the graphics cards were working hard! The balance of these settings also ensures me that that the CPU is utilized as much as can be, although not one title is really optimized for Quad core .. we did push the bar.

Now at this point I realized that we're stressing the PSU slightly as it's didn't get even slightly warm. So we added fan's, some extra lighting coming from the chassis, and enabled water-cooling on the CPU draining it's power from the PSU.

Two things we monitored ... We monitored peak PC power consumption with a wattage meter. And we tapped the voltage lines with a multimeter to see if we could detect any fluctuations.

Okay you get the idea already, stable as a rock and all that without any harsh noise from the fans. Pretty "vigorous" stuff. Let's place our findings in a chart. We pushed a maximum of 571 Watts with this new system.

Here we took some of the voltages in both IDLE and LOAD (fully utilized) modes. We noted down the lowest and highest value we saw and that is the fluctuation. If a PSU is unstable we'd see some fluctuation, differences and discrepancies which can result in system instability.

For gamers with SLI/Crossfire setups, the most power draw will be at the 12 volts rails. Above we can monitor the voltage distribution lines we monitored.

We have measured all voltage lines with a multi-meter directly at the device or mainboard. These are not BIOS measurements. We measured two "states". The system pretty much doing nothing (idle) and the system stressed out 100% (load).

Now, what you need to understand is that on each of the voltage rails, a 5% tolerance would be the accepted as normal according to the ATX spec, meaning the power-supplies actual voltages should stay somewhere in-between the values.

When we look at total PC system power consumption we need to realize we are pulling roughly 570 Watts on this system. Excessive yes, but the PSU didn't seem to push really hard. The PSU remains cool, the voltage rails stable.

This little toddler is stable for sure, and remember this was tested on a very expensive 2500 USD NVIDIA High-end SLI and overclocked system.

Sound levels - we have to include a quick stroll regarding noise levels these days. I've had cheap PSU's in this office that likely could have functioned as active rotating helicopter blades when they got a little hot. This PSU doesn't need much wording in this regard though, it simply is quite silent, yet audible. The heat on the inside of the PC is being sucked outside the PC, creating airflow with the help of a 140mm fan for quiet cooling due to its ability to move a large volume of air at low RPM.

dBA levels are hard to measure, we did however, and the results: Reasonably silent at ~38 dBA which pretty much can be classified as inaudible / normal noise levels.