To test power supplies, I like to load the CPU with what I consider to be a heavy real-world load. To do this I conduct a series of four test that are designed to stress the system in ways that duplicate real-world situations. Several instances of these test are ran and the results averaged for a more accurate result.
- Test 1: System at Idle - This test is ran with Windows sitting at idle for 10 minutes with the voltage measurements taken starting at the 10-minute mark.
- Test 2: CPU Load - This test simulates a CPU intensive load such as running a virtual machine. To test this, I load the CPU with concurrent instances of Prime95 on all CPU cores.
- Test 3: GPU Load - This test simulates a graphically intensive scenario such as rendering a 3D scene, or playing a graphically intensive game. To perform this test I run FurMark which places an incredibly intensive load on the GPU.
- Test 4: System Load - This test is designed to stress the entire system to its limits and simulates a load one might see during an intensive workflow session. To perform this test, I run concurrent instances of Prime95 and Furmark to ensure that the system is pulling as much wattage as possible from the PSU.
All measurements are taken with a Digital Multimeter and verified using a Tektronix Analog Oscilloscope. This also allows us to measure the peak to peak ripple on each power rail.
I’m testing the Xigmatek Maverick S 500W in a mini-ITX system based on a BIOSTAR NM70I-1037U motherboard that has an integrated Intel Celeron 1037U Dual-Core CPU running at 1.8GHz. The full specifications can be found below.
- Motherboard: BIOSTAR NM70I-1037U
- CPU: Intel Celeron 1037U Dual-Core 1.8GHz
- GPU: Intel Integrated HD Graphics
- RAM: G.SKILL Ares Series 16GB (2 x 8GB) 240-Pin DDR3
- PSU: Xigmatek Maverick S 500GB
- Case: Xigmatek Nebula C
Current ATX standards require each power-rail in the PSU to conform to a specific voltage rage. The values below represent a 3% variance range, and any result outside of these parameters is considered poor performance.
- 3.3V Rail: 3.135V – 3.465V allowable
- 5V Rail: 4.75V – 5.25V allowable
- 12V Rail: 11.4V – 12.6V allowable
The results of my testing shows that the Maverick S 500W is a true performer in terms of voltage regulation and ripple. I found less than 3mv of variation on any of the power rails during each test. This is quite impressive as I have tested much more expensive PSUs in the past that could not maintain an 8mv ripple limit. Below you will find a chart that list the voltage level averages in each category I recorded in testing.
The voltage regulation on the Xigmatek Maverick S 500W is quite exceptional, and maintained a close consistency across the entire testing range. The 12V rail did have the widest range across the test, but even at its maximum tested voltage of 12.23V the PSU remained well within the 12.6V allowance. During my testing I did not that the fan remained quite silent even when the system was stressed to the max. In fact the single case fan was an order of magnitude louder than the Maverick’s fan. This really impressed me due to the affordable nature of the power supply.