Home page » Peripherals

Search:

Refine search for Peripherals

To properly test the abilities of a power supply one needs to draw constant and measurable loads on it. This requires some sophisticated and expensive equipment, which we don't have in our offices. So to facilitate our testing we arranged to use the test equipment at FSP's UK office.

The equipment is standard stuff, made by Chroma System Solutions Inc. We used a 61500 AC power source to ensure we had an ultra clean source of power then two 6310 series DC loads to provide constant loads for the power supply. These are basically large variable resistors that just use up the energy being supplied to them by dispersing it as heat. You simply plug the power supply in then dial in the number of amps you'd like to be drawn from each voltage line and turn everything on. If it doesn't go bang, you're off to a good start. Assuming everything seems to be working alright, we then record the voltage being supplied by the power supply to make sure it's in the required bounds.

Now there are various ways you can load the power supply. In the past we've tested at a nominal 40, 60, 80, and 100 per cent of the PSUs total rated power output. This gives a very thorough impression of how it performs across a wide variety of potential system configurations. However, it requires at least four sets of tests, so is very time consuming. Also, because the test equipment is so noisy it completely drowns out any noise emanating from the PSU at low load levels, making it impossible to judge how good or bad the PSU noise level is - making it somewhat moot to test at all these various low levels.

So, this time around we've limited our testing to just 50 and 100 percent load. This way we can ensure the PSU can work at its stated maximum and also double check its efficiency holds when used at a lower load level. As well as this, we've plugged the PSUs into our office test bed, which consumes about 160W of power at idle and 370W under load, to see just how quiet these PSUs can go when a system is idling and how noisy it gets in typical load conditions.

In order to test this I constructed a rudimentary sound proof box in which to mount the PSU. Despite it's rather thrown together appearance this contraption successfully dropped the sound level by between 10 and 15 decibels depending on the type and direction of external noise. Ambient noise in the box in our office at night (just me and the servers) was around 30 decibels. We measured sound levels from a distance of about 30 centimetres from the PSU, ensuring the PSUs fans weren't blowing directly onto the sound meter. We then recorded numbers for the system idling and when under load.


Office Test Bed

* Asus P6T Motherboard
* Intel Core i7 965 Extreme Edition
* AMD ATI HD 4870 X2
* 3 x 1GB Qimonda IMSH1GU03A1F1C-10F PC3-8500 DDR3 RAM
* 150GB Western Digital Raptor
* Pioneer BDC-S02BK Blu-ray reader / DVD/CD writer

One final note before we get onto the group test itself; there are a surprisingly vast number of power supply manufacturers out there and each one has multiple lines of power supplies. So to cover them all would require a round up of about 40 power supplies at one go. This is obviously not practical when we're borrowing another company's equipment so instead we've taken a selection of 10 power supplies (as many as we could test in one day). So, it would be remiss of us not to at least mention a few other companies that we've tested previously and found to also make high-quality PSUs, but that we didn't have space to include today. So, if you're buying then also keep an eye out for Enermax, Thermaltake, Tagan, Antec, Akasa, Corsair, Hiper, and Zalman.

Add your comment

© 2003-2009 TrustedReviews & IPC Media
About Us | Contact Us | Legal | Privacy | RSS