Print this article

E- Power Xscale 1000W Power Supply


Date
: 12/17/08 – 05:19:14 PM

Author
:

Category
: Power Supplies

Page 1 : Introduction

Not too long ago, E-Power showed up at our doors with the Thunder 650W power supply, and while we were impressed with what E-Power had to offer, especially at such a low price point, but sometimes 650W just does not cut it.

Returning for another round is E-Power with their Xscale series. The Xscale series is rather… scaleable, coming in variants from 470W up to a jaw dropping 1200W. Today, we are going to look at the Xscale 1000W power supply.

Under the larger umbrella of Topower, E-Power, Topower's North American arm, was founded in 1990 as a company with a sole focus: PC power supplies. Once a small provider of consumer grade power supplies, Topower has now expanded its product line to include server, workstation and industrial grade power supplies. With so much amounted experience, the Xscale 1000W should be very interesting!

Page 2 : Features and Specifications

Here is what E-Power has to say about the Xscale 1000W power supply:

Twelve SATA connectors? That's probably a record. Not that I am complaining – in fact the more the merrier!

For a 1000W unit, 900W on the 12V rail alone is pretty impressive! These are the kinds of numbers we like to see!

Page 3 : Package and Content

The Xscale 1000W comes in a fairly typical cardboard box.

Pretty much each and every surface of the box is plastered with text, charts of graphics.

The back of the box provides a brief summary and a run down of key features.

Inside, we have the power supply, manual, AC cord, Velcro ties and screws.

Starting with the exterior, the Xscale 1000W is truly a beautiful unit. I am a sucker for brushed finishes, and the Xscale is finished in dark titanium with some light brushing.

The cables are also fully sleeved in either black sleeving or plastic shielding with black, blue and white Molex connectors.

The Xscale 1000W is a dual 80mm power supply, with one fan in the front and one in the back. Even with a 'Silent Engine' badge on the fan, I am skeptical. I have ever only seen a handful of truly silent 80mm fan cooled power supplies.

Towards the end of the unit, we have the rear exhaust, AC connector, power switch, and a particular looking red switch. This switch allows the power supply to run with either split 12V rails or a single combined rail.

E-Power may have gone a bit overboard with the connectors, but then again, its always better to have more than less! For the motherboard, we have a standard 20+4 pin connector, a 4 pin and two 8 pin connectors.

For video cards, we have two 6+2 pin PCI express connectors, and four 6 pin PCI express connectors.

Equally in excess, which I mean in the most endearing way, 16 SATA connectors!

Surprisingly, we still have room for four standard peripheral connectors, along with two floppy disk connectors.

And finally, our fan connectors. That's right, leads dedicated to fans, and only fans. As strange as it may be to be excited over various connector configurations, I trust my fellow wire management obsessed to understand! These leads topped with 3 pin connectors and wired with much thinner cables allows one to easily route and hide wires. And there are three of these leads!

Finally, my favorite part. Let's cut that sticker and void the warranty!

No wonder it is so heavy!

Generally, when something looks this packed, it usually means good things.

Two 12V transformers, cylindrical heatsinks, clean wiring job, and beefy components!

On the primary side, we have two TK capacitors, and on the secondary side a mix of Teapo and CapXon.

Page 4 : Testing Setup

Quite unlike any other PC component, properly evaluating a power supply involves much, much more than running a battery of tests or sitting down in front of a shooter for a few hours armed with a pen and a pad of paper. Although, we have the utmost appreciation for a few headshots coupled with explosions, power supply reviews call for much more. The general mantra for testing power supplies tends to fall somewhere in the ranges of loading the power supply up to the top, and letting it suffer. We here at Overclockers Online are inclined to agree. Accordingly, our testing methods have evolved to allow for more sophisticated and accurate testing.

The dilemma here is: how to test a power supply without a computer system, but emulate a computer system? It is important to not forget that as much as we try to part from throwing a power supply into a computer system and watching how it performs, ultimately, what we are distancing ourselves from is precisely what we are trying to replicate. Simply, the ideal testing platform would be indistinguishable from a computer system, but graced with the accuracies of an adjustable load. However, since we do not have access to state of the art testing equipment, or the funds to purchase such equipment, we have built a custom load tester. While it does not offer adjustments in the thousandths of a decimal place, it does offer enough flexibility to test current day power supplies and beyond.

Using thick high quality 16AWG wires throughout and equally high quality connectors and switches, we have sought to minimize resistance in the lines while maintaining flexibility. All it takes is a flip of a switch to turn a desired resistor on or off.

A proper electronic load is a start, but we are going to be a bit more ambitious. A keystone piece of our testing mythology is the heat box. A system drawing 800W from a power supply produces quite a bit of heat, and to be testing such a power supply in an ambient environment is not only inconsistent with in-system applications but unrealistic and misleading. Our hotbox consists of nothing more than a case graciously donated to us from Silverstone. Recycling heat produced from the load tester using a series of controlled fans and a duct, we can control operational temperatures and push a power supply to its limits, or over.

Using the heat and load, we will run a battery of five tests: three cold and two hot tests. The three cold tests consist of 50% load, 80% load and finally full load. Naturally, we will escalate the strain on the unit by adding heat in the hot tests. The final two tests are 80% load and full load at approximately 50°C or more.

While the purpose and effects of 'burn in' are debated, doing so causes no harm. Prior to any testing, all units are run for a maximum period of one week without load. At best, this will stabilize the unit; at worst it takes a few hours off the unit's total life span.

In our tests, we will do our best to adhere to ATX specifications including cross loading and criteria for testing. The presentations of our results are designed for ease of interpretation and conciseness.

So, without any further delay, let us get started!

Page 5 : Testing

The aim of the cold test is to reflect ideal operational conditions where heat from the computer system is independent from the power supply. Still though, the term cold test would be misleading as things do get quite hot!

These numbers are quite impressive, I even had to make sure my multimeter was working, from Test 1 to Test 3, which is an increase of 500W, the rails barely moved. It is times like these, as a reviewer I am put at a dissonance. While I was excited for the power supply, I was equally as disappointed that the power supply shrugged off 1000W of load. Maybe some heat will set this unit straight!

So again, nothing. I guess we are going to have to settle here. The Xscale is showing some very, very impressive voltage regulation. But, in terms of acoustic level, as each test progressed onwards, the unit became louder and louder. But with a unit of this magnitude, power supply noise will be the last thing on one's mind, especially if one intends to use up those 12 SATA connectors while running two video cards.

Having a look at the Xscale 1000W at full load on a bar graph, we can see how little heat affects the unit (hint: it doesn't).

In terms of size, the Xscale 1000W is longer than the conventional ATX power supply, however, it should have no trouble fitting into most cases, including our small form factor Silverstone case / hotbox, although it was a tight squeeze with the long cables.

Page 6 : Conclusion

So here we are at the end of another power supply review. So far, with E-Power, we have taken a look at a relatively low end unit (at least by price) and a high end server grade power supply, the Xscale. E-Power has shown its ability to deliver performance when it comes to affordability, and when it comes to well, sheer power. But then again, with Topower being the OEM, we weren't exactly surprised.

We have taken a look at a number of Topower built units over the years, and again, they deliver. The E-Power Xscale unit has shown excellent performance with superb line regulation. The build quality is also top notch. While it may not be the quietest unit, for what it offers, noise shouldn't be much of an issue, especially considering the type of computer system the Xscale 1000W is aimed at.

Advantages

• Excellent performance

• Beautiful finish
• Fully sleeved
• A lot of connectors, and excellent configuration

Disadvantages

• Loud

Overclockers Online would like to thank E-Power for providing a review sample of the Xscale 1000W for review and Silverstone Technologies for providing the chassis for use in our load tester.