facebook rss twitter

Review: HEXUS PSU (Power Supply Unit) Roundup - Taoyuan 2005

by Ryszard Sommefeldt on 23 September 2005, 00:00

Tags: Aopen, Enermax (8093.TWO), Zalman (090120.KQ), Akasa, Thermaltake (3540.TWO), OCZ (NASDAQ:OCZ), Tagan, SilverstoneTek, FSP Group (TPE:3015), Hiper, ETASIS

Quick Link: HEXUS.net/qabkp

Add to My Vault: x


I've been sat in my home office at the keyboard for the last hour now, figuring out how to start this article. The sheer scope of the copy I've got to write, encompassing testing and analysis of 34 - yes, thirty four - power supplies, means that I'll struggle to write the main sections, never mind this introduction. So I've been thinking about why the testing came about in the first place. If you want to write a decent intro, you start at the beginning I guess!

It's all because of Q-Tec. In all the years I've been a PC enthusiast, which is over a decade now, scarily, Q-Tec is the brand that crops up time and time again in online tech forums, hardware-led IRC channels and IM conversations with your friendly hardware geek. Your PC is playing up, everything else seems fine and you've tested all of your components in someone else's computer. The only thing left to point the finger at is the PSU. And given Q-Tec's aggressive pricing, at least in the U.K., of supposedly high-output power models that really are nothing but fool's gold, you're likely to find one of their supplies at the heart of many PCs in this country.

Of course Q-Tec isn’t the only PSU vendor that seemingly sells sub-standard PSUs marketed on peak output power. Even vendors of supplies that do meet output power ratings are guilty of it now and again, based on market forces and their competition.

Indeed we empathise with why elements within a respectable company become tempted to market a PSU based on its peak power output. If company ‘x’ advertises it’s offering a 600W PSU for say around 50% less than the price of a true 600W PSU from company ‘y’ (which is genuinely capable of offering a sustained 600W output) then to a misguided end-user (and retailer) the deal looks like a no brainer, and the sales of decent PSU essentially dry up.

The situation also seems compounded by garden variety magazines and websites which publish ‘reviews’, significantly based on the wrong metric and their own ignorance, or worse - the commercial interests of these publications and its advertisers… for instance would you trust any publication which chose not to disclose that its tests were conducted under the guidance of, and within the premises of the manufacturer of the ‘review’ winner?... which begs the question - why would anyone in their right mind, trust anything spewed by such publications?

The bottom line is that HEXUS hasn’t yet seen any PSU review, published elsewhere, which we think our readership should rely on an as guide to handing over your hard-earned.

It's these basic tenets that catalysed the HEXUS autocracy, in discussion with their minions, to set the wheels in motion for this, the mother of all power supply tests.

The basic aim of the testing was to take a whole bunch of the most popular power supplies on the market today and test them to their stated limits to see just what units would stand up to the torture. The HEXUS standpoint is that a power supply should do what it says on the box, as a sustained output power. If you see said box in a shop, you should be confident that the figure for power on the box can be obtained as a load power output. Advertising a peak output power does nobody any good, especially if the sustained output power isn't anywhere on the box or in the literature. What use is advertising to someone 500W or more, if the supply is capable of around half that under sustained load, often with poor rail regulation and lots of generated heat?

The old adage that you get what you pay for really does apply in the power supply world. And it ties in with the larger reasoning that the print press has ruined the margins of the PC building industry by forcing ever lower price points over time. The £299 PC? Nowadays that's something to have some confidence in, but in days gone by when the print titles forced massive price gouging, OEMs needed to cram in cheap, sub-standard components to hit those price points, or suffer bugger all margins in the aim of keeping your quality, in order to satisfy the figure on a magazine cover.

PSU vendors have played up to that with models that mislead, peak power telling the consumer little about what's really inside. 550W PSU for under £20? You're joking, really. So we put the feelers out to find a suitable lab to properly test a bunch of units to see what they could really do. What we needed was a lab with the equipment to really push PSUs and monitor what was going on, but a lab we could use unsupervised and free of interference, especially if it turned out that the lab and the equipment we needed was the testing lab of a PSU vendor.

What better place to fully test PC power supplies than where the vendors test them? Of the guys and gals we got in touch with, only one was entirely happy for us to sit in their labs, unsupervised, for as long as we needed it. FSP Group, a company you might also know as Fortron Source, are the largest PC PSU supplier on the planet. With their own range of units on retail sale and a simply massive industrial presence, they also have a huge OEM business, making supplies for a bunch of other component vendors that you may have heard of.

They've got a large lab area in their Taoyuan City offices, near Taipei in Taiwan, that was perfect for our needs. Large and employee free, bar the guys that perform acoustic testing round the corner, they were happy to set it up with the equipment that we needed and give us walk-in access to the building so that we could spend a week at our own leisure testing supplies we'd bought and brought to Taiwan ourselves, without FSP influence. Perfect.

So with Computex looming large, we went out and bought a whole bunch of power supplies and had them shipped out to Taoyuan City. PD and I then flew out to Taiwan a week before Computex started to test as many of the supplies as we could in the 5 days we had in the lab, to see what would shake out. The results were surprising, especially given our testing methods. Read on to find out what supplies we bought and shipped, how we used and abused them and how they all stood up to the intense scrutiny.

PD breaks stuff in Taoyaun
The sexiest man in IT, Paul 'Captain Stitch Me Up' Dutton, plots up ways to undo my hard work in the labs