Four killer PSUs go head to head

I’ve tried to pen the introduction for this latest power supply round-up at least half a dozen times now. Each time I sit down to write, I draw a blank. The stack of PSUs to my left provides little inspiration, and although our beastly load generator remains one of the most impressive pieces of hardware in my lab, the sight of its chunky switches, imposing banks of resistors, and massive fans has failed to get my creative juices flowing. I’ve even tried imbibing various intoxicants, to no avail. Such is the tortured life of a hardware reviewer.

The problem, I think, is the subject matter at hand. PSUs just aren’t that exciting, especially when compared to CPUs, graphics processors, chipsets, motherboards, hard drives, notebooks, netbooks, sound cards, enclosures, coolers, and just about everything else we cover here at TR. Don’t hit the back button on your browser just yet, though, because what PSUs lack in excitement they more than make up in importance.

A power supply may look like little more than a glorified AC-to-DC converter, but the flow of electrons it generates is the lifeblood of a modern PC. If that stream of current is dirtied, either by too much AC content or sagging DC voltages, system stability and even overclocking potential can suffer. Substandard PSUs can fail in epic fashion, too, sometimes taking other components with them in a puff of very expensive smoke. I guess that qualifies as excitement, just not the sort you’d actually want.

As in many markets, the new hotness comes to PSUs at the high end before trickling down to lower-wattage units. Today, we’ve gathered four new enthusiast-oriented models from Corsair, Enermax, Seasonic, and XFX to see how they fare against our beastly load generator. With wattages between 750 and 850W, these PSUs easily have enough power to fuel a high-end system. Read on to find out which one does the best job.

Rounding ’em up

We’ll of course look at each of these PSUs in greater detail, but first, let’s set the stage with a wider view of how they compare. Many important attributes make up a good PSU, and we’ve crafted a simple comparison chart that summarizes some of the basics for the models we’ve assembled.


Wattage

Cooling

Modular?

80 Plus?

Warranty

Price
Corsair HX750W
750W
750W 140 mm bottom Yes Silver 7 years
Enermax
Revolution85+ 850W
850W 135 mm bottom Yes Silver 5 years* $219.99
Seasonic X
Series 750W
750W 120 mm bottom Yes Gold 5 years $179.99
XFX Black
Edition 850W
850W 135 mm bottom Yes Silver 5 years

As you can see, we’re looking at 750 and 850W units. Corsair and Seasonic will square off at 750W, while Enermax and XFX will do battle 100W up the line. Don’t put too much stock into these total output ratings, though. Each PSU divides its power differently, and where the watts flow is arguably more important than the sheer number available.

But what about the fact that we’re testing 750W units against others with a 100W advantage? Worry not, because our testing methodology is designed to take into account such differences. In addition to putting each PSU inside a real system, we’ll be probing its performance at 25, 50, 75, and 100% of its rated capacity.

A power supply’s efficiency is one of its most important attributes and something we’ll test ourselves in a moment. However, the 80 Plus program also provides efficiency certifications for the industry. Three of these four units are 80 Plus Silver certified, which means they achieved 85-88% efficiency in the tests conducted by the program. Only the Seasonic X Series has an 80 Plus Gold rating, denoting an efficiency of 87-90% in the program’s standardized tests. You can view the 80 Plus testing protocol here (PDF).

The energy lost due to less-than-perfect efficiencies generates heat that must then be expelled from the PSU. All the models we’ve gathered employ bottom-mounted fans, but the sizes of those fans range from 120 to 140 mm. Interestingly, what should be the most efficient PSU also has the smallest fan. In a moment, we’ll see how these slightly different approaches to PSU cooling affect system temperatures and noise levels.

We can easily measure degrees and decibels, but quantifying a PSU’s longevity is considerably more difficult. A good power supply can last for years, persisting through multiple upgrade cycles in a primary system before ending its days tucked away in a closet file server. We can’t test for long-term durability and still produce reviews in a timely manner, but we can get a sense of how long each company will stand behind its product by looking at warranty coverage. Seasonic and XFX both cover their PSUs for five years, which is about standard for high-end enthusiast models. Corsair kicks in an additional two years of coverage for the HX750W, while Enermax skimps with a three-year warranty.

The Revolution85+’s shorter warranty would be easier to forgive if it were one of the more affordable models, but it’s actually the most expensive by $30. The Corsair, Seasonic, and XFX units are all priced within $30 of each other, with the HX750W ringing in as the most affordable of the bunch.

*Update 01/15/10 — Enermax has decided to extend the warranty coverage on a number of its PSUs, including the Revolution85+, to five years. What’s more, it seems that this very round-up may have inspired the change. Regardless of the motivation, we applaud Enermax for improving the Revolution85+’s warranty coverage.

Counting connectors

All of the units we’re looking at have modular cabling, but each offers a different assortment of plugs and connectors. We’ve untangled the mess and summarized the connector counts for each PSU below.


Main power

Aux 12V

PCIe

4-pin peripheral

SATA

4-pin floppy
Corsair HX750W
750W
20/24-pin 4/8-pin 4 x 6/8-pin 6 12 2*
Enermax
Revolution85+ 850W
20/24-pin 4-pin, 8-pin 4 x 6/8-pin 6 12 1
Seasonic X
Series 750W
20/24-pin 4-pin, 8-pin 4 x 6/8-pin 8 8 2*
XFX Black
Edition 850W
20/24-pin 4-pin, 8-pin 4 x 6/8-pin 8 11 2*

First, the easy stuff. All of the units have hybrid 20/24-pin primary power connectors, and they can all provide auxiliary 12V power with a four- or eight-pin plug. The HX750W’s auxiliary 12V line uses a hybrid 4/8-pin connector, while the others have separate plugs for each.

On the PCI Express front, four hybrid 6/8-pin connectors extend from each PSU. Official CrossFire and SLI certifications aside, you should be able to power a couple of high-end graphics cards from either camp with any of these PSUs.

Things get a little more interesting when we turn our attention to SATA and peripheral connectors. All but the Seasonic are biased towards SATA connectivity, with the Corsair and Enermax units offering twice the number of SATA connectors that they do Molex plugs. The Corsair and Enermax PSUs can make use of all the Molex and SATA connectors listed above. However, the Seasonic and XFX models don’t have enough modular plugs for all of their included cables. You’ll have to sacrifice two Molex or two SATA plugs with each PSU.

Floppy drives are rare these days, but PSU manufacturers still support their mini four-pin connectors. The Revolution85+ has a built-in floppy connector, while the others come with Molex adapters. Corsair’s adapter will cost you a single Molex plug for each floppy. The adapters included with the Seasonic and XFX PSUs are less wasteful, attaching two floppy connectors to a single Molex plug.

Cable reach

The number of connectors provided is only one part of a power supply’s cabling equation. How far those connectors reach is also important, especially if you’re running a larger tower enclosure or one of those exotic upside-down cases that puts the PSU below the motherboard. Even if you have a relatively small, traditional case, longer cables can provide more routing flexibility.

With the help of our trusty measuring tape, we’ve determined the maximum cable reach for each PSU. The measurements shown for SATA and four-pin peripheral connectors refer to the reach of the last connector on the longest cable.

Only a few inches separate the longest motherboard power cables from the shortest ones. Corsair provides the longest reach here, outstretching the Revolution85+’s cables by half an inch. The XFX unit’s primary power cable is a little on the short side, which could create problems in upside-down cases in particular.

The PCIe power cables provided by each PSU are closer in length than the motherboard power leads. Again, Enermax and Corsair lead the field.

Despite having the shortest primary motherboard power cable, the XFX Black Edition’s Molex and SATA plugs offer several more inches of leeway than the competition. Enermax pulls up particularly short when we look at four-pin Molex connectors, but at least it offers plenty of reach for SATA devices.

Rated capacities

We’ve already discussed the total output wattage rating of each of the PSUs in the spotlight today, but that’s only one component of the output rating story. Modern PSUs divide power across three main lines at 3.3, 5, and 12 volts, with each of those lines carrying a maximum current rating. As if that weren’t complicated enough, many PSUs also spread 12V power over multiple individual lines.


Maximum output current (Amps)


DC Output

+3.3V


+5V

+12V
Corsair HX750W
750W
25 25 62
Enermax
Revolution85+ 850W
25 25 30, 30, 30, 30,
30, 30
Seasonic X
Series 750W
25 25 62
XFX Black
Edition 850W
24 30 70

Of these four models, only the Enermax splits its 12V line between multiple rails—a whopping six of them, to be exact. The rest of the units consolidate their 12V power on single rails.

Don’t get married to these current ratings, though; they’re only the tip of the iceberg. To get a handle on each PSU’s true capacity, we have to determine the maximum output wattage of each voltage line. The math is easy enough thanks to Ohm’s Law, which allows us to calculate wattage given voltage and amperage. However, power supply units with multiple rails are typically limited by how much power can be spread across those multiple lines—a total that’s usually less than the sum of each rail’s output capacity. Most PSUs also place limitations on how much power can be shared between the 3.3V and 5V lines. And then there’s the maximum output wattage across the 3.3, 5, and 12V lines, which often falls short of the PSU’s total output rating in order to reserve capacity for lesser-used voltage lines like the -12V and 5V standby rails.

To make sense of it all, we’ve put together a table showing the maximum output power for each PSU’s 3.3, 5, and 12V rails. Where applicable, we’ve also indicated the maximum combined 3.3 and 5V power, the maximum combined 12V power, and how much wattage the PSU can spread across all three main rails.


Maximum output power (W)


DC Output

+3.3V


+5V

+12V
Corsair HX750W
750W
82.5 125 744
150
750
Enermax
Revolution85+ 850W
82.5 125 360, 360, 360,
360, 360, 360
160 840
850
Seasonic X
Series 750W
82.5 125 744
125
750
XFX Black
Edition 850W
79.2 150 840
150
850

Interestingly, the Corsair and Seasonic PSUs have identical output ratings for their individual 3.3, 5, and 12V lines. However, Corsair is more optimistic about how much power can be shared between the 3.3 and 5V rails. The maximum sustained output across all three rails is identical at 750W, though.

The Revolution85+ has a little more combined 3.3 and 5V capacity than the 750W units. About 100W of additional 12V output power is also spread over its half-dozen rails. That total 12V capacity neatly matches the XFX Black Edition, whose single 12V line can likewise handle up to 840W. The Black Edition actually has a less powerful 3.3V rail than even the 750W units, though. It also has less combined 3.3 and 5V power than the Enermax PSU.

As one might expect from a collection of high-end PSUs with loads of PCI Express graphics card connectors, there’s a clear bias toward 12V power. Each model is capable of reaching very close to its total output wattage on just the 12V line alone. CPUs and graphics cards demand more from the 12V line than any other component, a fact clearly not lost on these vendors.

Loading up The Beast

A cornerstone of our power supply testing methodology is a custom-built load generator created by forum regular just brew it! that we like to call The Beast. If you’re unfamiliar with its frightening array of resistors, I strongly suggest checking out this page from our first round-up to use The Beast, which details the rig’s internals and our testing methodology.

We recalibrated each and every one of the variable-load resistors inside The Beast prior to testing for this round-up. None of the resistors needed more than a minor adjustment, but since we did tweak things some, the results in this article aren’t directly comparable to those from our previous PSU round-ups.

Behold The Beast!

We use The Beast to push each PSU to 25, 50, 75, and 100% of its output capacity while measuring DC voltage, AC ripple content, and gathering data that can be used to calculate overall efficiency. This creates a kind of level playing field on which each PSU is pushed to its individual—and advertised—limits. We should note, however, that our testing methods are not identical to those used by the 80 Plus program or by vendors who publish their own efficiency ratings. The Beast is a different breed of load generator than those used by the folks at 80 Plus; it only taps the 3.3, 5, and 12V rails, leaving -12V and standby 5V lines unused. To ease confusion, we’ll be referring to efficiency ratings gleaned from our own testing as Beast efficiencies. Those figures shouldn’t be compared to efficiency ratings posted by PSU makers.

The Beast is also limited to applying loads in 2A increments, so we borrow a page from The Price is Right and use amperage loads that come as close as possible to our targets without going over. The chart below shows the amperage loads applied to each PSU.


Total loads (Amps)


25%

50%

75%

100%

3.3V

5V

12V

3.3V

5V

12V

3.3V

5V

12V

3.3V

5V

12V

Corsair HX750W 750W
2 2 12 6 6 26 10 10 38 14 14 52

Enermax Revolution85+
850W
4 4 14 8 8 28 12 12 44 16 16 58

Seasonic X Series 750W
2 2 12 6 6 26 8 8 40 12 12 52
XFX Black
Edition 850W
2 4 14 6 8 30 10 12 44 12 16 60

When testing with The Beast, each power supply was hooked up using its primary and auxiliary 12V connectors, and when available, two PCIe power connectors and six 4-pin peripheral connectors. We used a Pico ADC-212 digital oscilloscope to probe the 3.3 and 5V wires on the primary power connector. The 12V lines were probed at the primary power connector and also at one of the PCIe power connectors. In the graphs on the following pages, 12V power from the primary connector will be marked 12V1, while power from the PCIe connector will be marked 12V2.

Rather than calculating efficiency based on static 3.3, 5, and 12V, er, voltages, our calculations take into account the actual DC voltage delivered on each line during testing. This should compensate for any voltage fluctuations that some PSUs exhibit under load.

Our testing methods

Testing was conducted in two parts. First, the PSUs were run in the system detailed below for a series of power draw, temperature, and noise level tests. They were then hooked up to The Beast to test power delivery and overall efficiency.

All tests were run three times, and their results were averaged.

Processor

Athlon 64 X2 5000+ 2.6GHz
System bus HyperTransport
16-bit/1GHz
Motherboard

Asus M2N32-SLI Deluxe Wireless Edition
Bios revision 0906
North bridge nForce 590 SLI SPP
South bridge nForce 590 SLI MCP
Chipset drivers ForceWare 9.35
Memory size 1GB (2 DIMMs)
Memory type

Corsair CM2X512A-5400UL

DDR2 SDRAM at
742MHz
CAS latency
(CL)
5

RAS to CAS delay
(tRCD)
5
RAS precharge
(tRP)
5
Cycle time
(tRAS)
12
Audio codec Integrated nForce
590 SLI/AD1988B with 5.10.1.4530 drivers
Graphics
GeForce 8800 GTS 640MB
with ForceWare 162.18 drivers

Hard drives
2 x

Western Digital Caviar RE2 400GB
SATA

OS


Windows XP Professional

OS updates
Service Pack 2

We used the following versions of our test applications:

The test systems’ Windows desktop was set at 1280×1024 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.

All the tests and methods we employed are publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

Corsair’s HX750W 750W
A surprising bargain

Manufacturer Corsair
Model HX750W 750W
Price (Street)
Availability Now

Corsair’s HX620W took home a TR Editor’s Choice award in a massive 11-way PSU round-up some two years ago. Seeing Corsair come out on top in a round-up is hardly surprising; the company has been building quality enthusiast-grade products for years. However, the HX620W was the company’s very first power supply, making its excellent overall performance and smart feature set all the more impressive.

There are now 14 different models in Corsair’s PSU lineup. The latest to arrive in our labs is the HX750W, which, as you’ve astutely observed, has a 750W output rating. Among our group of four PSUs, the HX750W is the cheapest of the bunch by a good $20.

Despite being the most affordable unit in the pack, the HX750W doesn’t cut any corners. Heck, its seven-year warranty is the longest of the lot by two years, and Nvidia’s even blessed the unit with an official SLI endorsement. The HX750W sports the largest fan of the bunch, too—a variable-speed 140-mm fan occupies the unit’s bottom panel accompanied by plenty of venting at the rear.

PSUs rarely offer much in the way of visual flair, but Corsair’s use of matte black finish throughout gives the HX750W a stealthy look. I like the understated aesthetic, although you’re only going to see it if your case happens to have a window—and a tall one, at that.

More important than the HX750W’s choice of exterior colors is its modular cabling. The primary motherboard connector and auxiliary 12V line are the only cables you can’t remove from the unit. They’re also the only cables that you really can’t build a modern system without, so it doesn’t make much sense to have them modular, anyway.

All of the PSUs we’re looking at today have some degree of modular cabling. However, only Corsair does something original with the cables themselves. The blacked-out leads match the PSU’s Vader motif, and Corsair says their thin profiles are designed to maximize airflow within a system.

The HX750W delivered solid DC voltages across all four load levels we generated with The Beast. The 12V rail trends a little high and the 3.3V line a little low, but only by the slimmest of margins. We’re talking about deviations of less than one tenth of a volt here.

Regardless of the load level, the Corsair PSU’s AC ripple content remains flat at around 45 millivolts, which is well within tolerances.

Under our Beast-generated loads, the HX750W never dips below 89% efficiency. Impressive, especially when you consider that the PSU’s efficiency doesn’t change by more than a percentage point across our 50, 75, and 100% loads.

Enermax’s Revolution85+ 850W
Talkin ’bout a revolution

Manufacturer Enermax
Model Revolution85+ 850W
Price (Street) $219.99
Availability Now

With a $220-280 street price, Enermax’s Revolution85+ is the most expensive PSU in this round-up. We’re not averse to pricey hardware here at TR, just so long as you’re getting what you paid for. The question for the Revolution85+ is a simple one, then: is it worth the extra scratch?

At first glance, one might be inclined to say no. Sure, the Enermax brings 100W more than the Corsair and Seasonic units. However, the XFX Black Edition also packs 850W, yet it’s at least $30 cheaper. You’re not paying a premium for additional warranty coverage, either. The Revolution85+’s three-year warranty looks positively stingy next to the five- and seven-year coverage offered by the competition.

Update 01/15/10 — Enermax has decided to extend the warranty coverage on a number of its PSUs, including the Revolution85+, to five years.

Dive a little deeper into the Revolution85+’s design, and you can see why Enermax might be inclined to charge extra. Most high-end PSUs these days consolidate their 12V power on a single rail, but Enermax spreads it across half a dozen lines, a design the company claims offers the best safety and compatibility. The PSU’s dedicated 3.3 and 5V DC-to-DC conversion circuitry is said to offer better efficiency than generating 3.3 and 5V power with a transformer, as well.

The Revolution85+ also supports zero-load states on all of its rails. This capability is apparently necessary for future CPUs that will employ ultra-low-power C6 sleep states and upcoming graphics cards that will offer “hybrid” hibernation modes that cut power substantially.

Even the Revolution85+’s variable-speed fan has received special attention. A small lip on the fan’s frame is designed to reduce turbulence, which should lower noise levels. Another nice touch: the fan will keep running for 30-60 seconds after a system is powered down to ensure that hot air doesn’t stagnate inside an enclosure once the case fans stop spinning.

The Revolution85+ has a textured, almost speckled finish that’s unique, if nothing else. I like the indentations on the side panels, although again, you’ll never notice unless you happen to have a case window. You will notice the status light at the rear, however. It glows in three different colors to indicate normal operation, standby mode, and failure.

From here, we can also see a healthy dose of ventilation on one of the PSU’s internal panels. This is the only model in the bunch that offers internal and external ventilation holes.

Most of the Revolution85+’s cables are modular, but the primary motherboard, auxiliary 12V, and two PCIe power connectors are not. If you’re building a system and don’t need at least two power cables running down to PCI Express graphics cards, you probably shouldn’t be using an 850W PSU.

Although Enermax sheathes all of the Revolution85+’s cables, it leaves a rather large gap between the sheathing and the first plug on each line. There’s no sheathing between multiple connectors on a line, either.

DC voltages are nice and stable on the Revolution85+, and very close to the target value for each line.

The Revolution85+ keeps its DC lines relatively free of AC ripple content. We’re looking at only about 50 millivolts of AC content here.

Interestingly, the Enermax unit’s Beast efficiency peaks under a 75% load, just topping 89%. The Revolution85+ does no worse than 86%, reaching that mark under our lightest load.

Seasonic’s X Series 750W
One of its very own

Manufacturer Seasonic
Model X Series 750W
Price (Street) $179.99
Availability Now

Seasonic is one of the largest PSU manufacturers around. In fact, Seasonic builds PSUs for a good number of other firms that then sell those units under their own names. The X Series is Seasonic’s very own product, though, and its most advanced model to date. This is the first power supply we’ve tested with 80 Plus Gold certification, too—the highest standard the program sets for ATX PSUs.

The X Series carries a bit of a premium thanks to its Gold certification; the 750W model costs $20 more than Corsair’s HX750W. The Seasonic unit is also considerably harder to find online. Our price search engine doesn’t list it, but Newegg does have units stock, with free shipping, to boot.

Seasonic PSUs have never been particularly flashy, and neither is the X Series. The combination of matte black with gold trim reminds me a little of old-school stereo equipment for some reason. The look is a classy one, although there’s less of it, simply because the X Series is the shortest PSU of our pack by about an inch. None of the PSUs in this round-up are too large to squeeze into most mid-tower cases, though.

Perhaps because the X Series has a smaller casing, Seasonic uses a Sanyo Denki San Ace fan that measures just 120 mm across. This is the smallest fan in the bunch. Still, given Seasonic’s pedigree, it’s hard to question the decision. Seasonic’s focus seems to be on lower noise levels, which is why the fan doesn’t spin up at all until the PSU is loaded up to at least 20% of its capacity. From there, the fan spins at a constant “silent” speed until the PSU load reaches 50%, at which point the fan really starts to ramp up.

Internally, the X Series uses a patented DC connector module with an integrated voltage regulator that the company claims offers “near perfect” DC-to-DC conversion for 3.3 and 5V power. According to Seasonic, this design allows the PSU to run lower currents to supply its 3.3 and 5V lines, reducing loss and improving overall efficiency.

All of the PSUs in this round-up may be modular, but the X Series is the only one that’s completely so. Each and every one of its cables, including even the primary motherboard connector, can be completely detached.

While this arrangement may offer questionable utility for end users who are probably always going to need a primary motherboard and auxiliary 12V line connected, the lack of attached cables does make the X Series a little easier to squeeze into tighter enclosures. The cables are nicely sheathed, too—not just down to the first plug on each line, but in between the subsequent ones, as well.

The X Series 750W offers consistent DC voltage delivery that’s well within the company’s +/- 3% tolerances.

AC ripple is low, as well, averaging about 50 millivolts across all rails and load levels.

Like the HX750W, the Seasonic X Series is most efficient under the lightest load generated by The Beast. There’s little difference in efficiency between loads at 50, 75, and 100% of total capacity.

XFX’s Black Edition 850W
Another one jumps onto the bandwagon

Manufacturer XFX
Model Black Edition 850W
Price (Street)
Availability Now

For the most part, we’ve seen a great deal of consolidation in the PC market in the last few years. The PSU market, however, has exploded with new entrants. The latest player to throw its hat into the ring is longtime graphics card maker XFX, which is starting at the high end with the Black Edition 850W. Go big or go home, I guess.

Speaking of going big, XFX has gone with quite a bold design for its very first power supply product. The slate-gray enclosure has lots of angular detailing, especially around the fan guard, which seems to be inspired by a spider’s web. Behind the guard lies a 135-mm cooling fan that’s been dipped in an apparently radioactive shade of lime green. There’s a real sense of distinctive industrial design here, and while it’s not the sort of look I want in my living room, it should fit right in within the confines of a high-end gaming rig.

The Black Edition has the perfect color scheme for SLI certification, which it boasts alongside a CrossFire-ready badge. Energy Star 4.0 certification also solidifies this PSU’s, ahem, green credentials.

With the aid of DC-to-DC voltage regulator modules for the 3.3 and 5V lines, XFX says that the single-rail Black Edition offers up to 90% efficiency. The PSU’s advertised voltage tolerance is +/- 3% for all rails, which is pretty standard for high-end units.

XFX has wisely gone the modular route with the Black Edition, and it leaves the primary motherboard, auxiliary 12V, and two PCIe power connectors permanently tied to the PSU. Even the modular connection panel has some subtle cosmetic detailing, complete with another shot of that neon green.

The modular cables are a little more run of the mill. They’re sheathed nicely in black, but look like they could’ve been built for any other PSU. I suppose you can’t have everything, especially when this 850W unit costs just $10 more than the Seasonic 750W. And XFX delivers where it counts, serving up the five-year warranty one expects from a premium power supply.

Even a budget PSU should deliver impeccable DC voltages, and the Black Edition has no problem on that front.

The PSU’s AC ripple content is also minimal, hovering around 50 millivolts like all the other units we’ve looked at today.

The Black Edition does even better on the efficiency front, hitting nearly 98% when subjected to the lowest load generated by The Beast. Once more, we see the Beast efficiency level off across 50, 75, and 100% loads.

Efficiency

We don’t want to make too many direct comparisons between test results gathered from The Beast, because each PSU was run under different loads corresponding to its individual specifications. However, efficiency is important enough to make an exception. Here we’ve graphed the efficiencies of each PSU at 25, 50, 75, and 100% capacity to see how they stack up against each other.

Yeah, so these PSUs are all pretty efficient when connected to our load generator. The Seasonic and Corsair units fare the best overall and offer nearly identical efficiencies across all four load levels. The Enermax and XFX units aren’t far behind, either.

Only the Revolution85+ fails to reach its highest efficiency with our lightest 25% load. Interestingly, it’s also the only one that uses multiple 12V rails rather than a single beefy one, which could be responsible for what we’re seeing here.

System temperatures

Our next batch of tests deployed the PSUs in a real-world system with a pair of GeForce 8800 GTS cards in SLI and a couple of hard drives. For these tests, the systems first sat at idle for 10 minutes. They were then subjected to a 10-minute load consisting of Prime95, Oblivion, and an IOMeter file server test pattern. Finally, the systems were allowed to idle for a 10-minute cool-down period.

System temperatures were measured using Everest Ultimate Edition.

At idle, the Seasonic PSU seems to be doing the best job of keeping the chassis cool, which is somewhat amusing considering that it has the smallest fan of the bunch. The HX750W isn’t far behind, but the Black Edition doesn’t appear to be moving nearly the same amount of air as its competitors.

The Black Edition registers the highest system temperatures under load, too. The others are a little more closely matched, with the Enermax offering the lowest CPU temperatures and the Corsair keeping our motherboard the coolest.

After ten minutes of cool-down, the field actually looks pretty even. The Revolution85+ cools the CPU faster than any of the other PSUs, and even the Black Edition drags itself into first place in motherboard temperatures. However, the XFX PSU still leaves our system’s hard drives a few degrees warmer.

Noise levels

Noise levels were measured at idle and under load using a TES-52 digital sound level meter placed 1″ from the rear of the system and out of the direct path of airflow.

Only a couple of decibels separate the field at idle, with Seasonic and Enermax tying for the lead. The HX750W is the loudest of the lot here, although not by enough of a margin that my ears noticed the difference.

The spread shrinks down to just over a decibel under load, and although the HX750W redeems itself according to our sound level meter, it’s very difficult to actually hear differences in noise levels this small. None of the fans have high-pitched or otherwise annoying noise characteristics, either.

Power consumption

Power consumption was measured at the wall socket for the entire system, sans monitor and speakers.

Our 850W units predictably draw more power than the 750W ones, particularly at idle. However, those differences shrink under load, where we’re only looking at a five-watt spread. Corsair and Seasonic trade the lead at 750W, while Enermax seems to have a more consistent edge over XFX at the higher wattage.

Conclusions

Round-ups like this one can be difficult to judge because we looked at four very competent competitors today. Each and every one of them has modular cables and plenty of connectors, including four 6/8-pin PCI Express plugs. Power delivery was excellent across the board, with tight DC voltage tolerances and very little AC ripple. For the most part, all the units were plenty cool and quiet, too.

I wouldn’t have reservations about running any of these units in my own systems. However, I do think that some are better than others, especially when you consider overall value.

The most questionable value in the bunch is Enermax’s Revolution85+. With a $220 street price, you’re looking at about a $30 premium over the XFX Black Edition. Granted, I think Enermax has a little more going on under the hood with support for zero-load states on all rails. But the measly three-year warranty simply isn’t good enough when less expensive competition gets at least five years of coverage. And the competition does a better job of cable sheathing, too. With the Revolution85+ offering largely equivalent performance to its rivals, it’s difficult to justify the price premium for what is otherwise a great PSU.

Update 01/15/10 — Enermax has extended the Revolution85+’s warranty coverage to five years, improving the PSU’s value proposition substantially.

Part of what makes the Revolution85+’s price so hard to swallow is the fact that an XFX Black Edition with the same output rating costs only $190. I love the design direction XFX has taken here, and for a first entry into the market, the Black Edition is a very solid PSU. There is room for improvement, though. The ultra-long Molex and SATA cables are great, but the Black Edition needs some more length on its primary motherboard and auxiliary 12V lines. Based on the higher system temperatures we observed, the fan profile could use some tweaking, too. Otherwise, the Black Edition offers good value for a high-wattage PSU.

Seasonic’s X Series may have a capacity rating 100 watts lower than the XFX and Enermax units, but it should still be capable of powering a high-end system. 80 Plus Gold certification doesn’t come cheap, of course, but the unit’s $180 street price is reasonable considering its low noise levels and power consumption. The completely modular design is also an interesting twist, although I can’t see it delivering any actual benefits in a real-world system unless you’re using the X Series as a secondary PSU to power peripherals or graphics cards exclusively. Still, of all the PSUs in this round-up, the X Series is the one I’d pull off the shelf to build a system with right now just because it’s slightly quieter than the others.

Corsair HX750W
January 2010

With that said, it’s not the PSU I’d go out and buy with my own money. That would be Corsair’s HX750W, which rings in $20 cheaper than the Seasonic yet offers an additional two years of warranty coverage. The HX750W’s performance is comparable to that of the Seasonic unit, but its cables are longer, thinner, and more plentiful overall. Yes, the HX750W’s idle noise levels are a couple of decibels higher at a distance of 1″, but put my ears a couple of feet away, and they don’t notice a difference. While it may not be the most interesting or exciting PSU in the bunch, the HX750W does offer the best mix of features, performance, and value. That earns it our coveted Editor’s Choice award.

Comments closed
    • bcronce
    • 10 years ago

    I think it was an overall good review. One thing that would be nice is if TR got one of those oscilloscopes to see actual switching DC voltage variation. I see AC variation was showed, but PSUs output bursts of DC power ~50,000-100,000 times per second which “averages” out to 3.3,5,12 volts. The voltage distortion during these pulses is also important.

    I know hardwaresecrets is looking into getting a “real” PSU tester, but an entry level one costs about $50k and that’s a really shinny penny.

    whip out the Techreport paypal donations link.. 😛

    • Black Jacque
    • 10 years ago

    I love PSU reviews too.

    I admit it seems peculiar that highest efficiency is at 25% load, and not at 50%.

    However, if I have a complaint, its that “Does the world need yet another HIGH WATTAGE PSU review?

    The trend in power consumption is downward for CPUs, GPUs, drives (including SSDs), and chipsets. TechReport’s own reviews show Intel i3, i5, and i7 Builds using r[

      • MadManOriginal
      • 10 years ago

      I think it’s more likely that these ‘flagship’ units are what the manufacturers send to be reviewed. I don’t think that lower wattage units actually cost a lot less to build in a quality manner (crappy build quality is obviously cheaper) but they sure go for a lot less so of course these are what manufacturers want to promote since they have the higher margins. For example, I hardly think that an 800W PSU costs 2x as much to make as a 500W PSU. Maybe TR can ask for more reasonable size PSUs to review though.

        • Black Jacque
        • 10 years ago

        I disagree.

        I think its a lot harder for a manufacturer to make a high efficiency, low wattage power supply for a low price , then to make a high efficiency, high wattage power supply for a high price.

        Getting the higher performance and reliability at a lower cost makes for a more complicated engineering and manufacturing solution.

        I can’t help but be reminded of Detroit-built American cars. The Big-3 built SUVs, pick-ups and sedans. While the rest of the world built econo-boxes.
        Look what happened.

      • fantastic
      • 10 years ago

      “The trend in power consumption is downward…”
      It was before the Core i7 and Phenom X4 came out.

      SSDs do help, if you can afford them. I was actually dreaming about being able to buy SSDs to replace my hard drives recently. Maybe next year…

    • oldDummy
    • 10 years ago

    Good review, gives a reasonable account of each PSU.

    Well done.

    • Oli
    • 10 years ago

    I’m a little confused (and tired)… someone please correct me…

    P11 – XFX uses the most energy of the pack at idle – draws 232W from the socket. Not miles off a 25% load (DC)

    P9 – XFX is the MOST efficient at 25% load

    • Clint Torres
    • 10 years ago

    Doesn’t Seasonic make Corsair’s PSUs?

      • Clint Torres
      • 10 years ago

      After some research, it looks like Channel Well(CWT) and Seasonic are both OEMs for Corsair.

        • BlackPearl
        • 10 years ago

        Thanks for that information. It would be good if PSU reviews included OEM info in the descriptions.

    • Shining Arcanine
    • 10 years ago

    Your results contradict the figures provided by SPCR, where the maximum energy efficiency they achieved for the Seasonic X-650 was 91.7% at about 38% load:

    §[<http://www.silentpcreview.com/article986-page4.html<]§ While figures like 98% power efficiency would be awesome at 20% loads, where computers usually are for power supplies of these watt ratings, that simply does not happen in real life.

      • mattthemuppet
      • 10 years ago

      I don’t think they care – they’ve blown off pretty much anyone who’s made any comments relating to this before (see Madmanoriginal’s posts above. They’re results aren’t even internally consistent (see Oli’s post #39) so I shouldn’t worry about it too much. I don’t even know why I bother reading TR PSU reviews anymore, the ones on SPCR are far more professional and consistent.

    • JokerCPoC
    • 10 years ago

    I wouldn’t buy another Enermax Revolution 85+ psu as their Molex power connectors wiggle around a lot when one tries to plug them into a device(hdd, etc), r[<*[

      • _Sigma
      • 10 years ago

      l[http://www.enermaxusa.com/catalog/popup_image.php?pID=179&image=2&osCsid=22eee26d5fef84a9076b56a125c98712<]§ clearly has a fan on the bottom.

        • JokerCPoC
        • 10 years ago

        That psu isn’t even in a case, Do You even own a HAF-932 & a Enermax Revolution psu? I do, Plus I own a Cooler Master HAF-932 case, The fan grill has to be removed, I’m not a liar, I’d like to see You mount one with the grill inside a HAF-932 or a HAF-922 or a HAF-X with that Grill, You will not be able to do It with these power supplies, As unless You own a case like this, You don’t know what You are talking about… And I do…

        §[<http://i110.photobucket.com/albums/n107/JokerCPoC/P5KD%20Pics/HAF932-EnermaxRevolution85-1050a.jpg<]§ No Grill, You know the thing that keeps noob fingers from the fan... §[<http://www.enermaxusa.com/catalog/popup_image.php?pID=179&image=2&osCsid=22eee26d5fef84a9076b56a125c98712<]§ Notice this picture has a Grill(Wire) between Your fingers and the blades???

          • _Sigma
          • 10 years ago

          No need to get upset, I was just trying to help.

          As mentioned below, certainly seems like a problem with the case than the PSU, as that doesn’t look like it would have issues with my bottom mounting PSU case (Lancool k62).

      • Voldenuit
      • 10 years ago

      I’m not sure I follow you here.

      Are you saying that the problem is that the grille on the Enermax PSU protrudes from the profile of the unit? If so, that’s been an Enermax trait for years, and is easily identifiable from pictures. Yes, it makes it problematic to slide in the PSU in some cases, and something every Enermax owner should be (made) aware of.

      If your problem is that there is not enough room beneath the PSU on your case, then that sounds like a serious issue with the case itself, since a bottom-mounted-PSU case should have enough room beneath the PSU for bottom-intake models to breathe.

      -[

    • MadManOriginal
    • 10 years ago

    I’ll be blunt here –

    95%+ efficiency at 25% loads and 97.6% for one of the PSUs is rediculous and completely not believable. It doesn’t jive with any other information from any other source whatsoever including the manufacturers themselves who would love to pimp such numbers. You need to fix the efficiency testing methodology or risk your PSU reviews becoming a joke.

    Maybe you guys need to get a real PSU tester? (sorry JBI)

      • Dissonance
      • 10 years ago

      /[https://techreport.com/articles.x/18295/4<]§ Our recalibration confirmed that The Beast is indeed functioning as intended, pulling the correct number of amps at the right voltages. Under those specific conditions, that's the efficiency we're seeing. If you'd like to donate a "real" PSU tester, I'd be more than happy to take delivery.

        • Sargent Duck
        • 10 years ago

        Thanks for that. You answered my question as well. Reading for the WIN!

        • MadManOriginal
        • 10 years ago

        That’s ok, it’s not like I was expecting you guys to actually change anything or do anything different based on past experience with suggestions that get rejected out of hand without any further investigation or attempts to implement them. I’m used to that by now from TR staff. Making suggestions or pointing out more than just typos almost always seems to fall on deaf ears, especially when it comes to saying results aren’t right or that there are better or different ways to do things. It’s an endearing type of arrogance I suppose.

        I just felt it important to post that those numbers are not what people should expect from their PSU whatsoever. You guys don’t seem to think that’s a big deal :shrug: PSU efficiciency tests are worthless at TR and I will say so every time a PSU test is published that includes such silliness.

        If you’d like to give me a bunch of free hardware maybe I’ll donate a PSU tester 😉 not sure why I should be expected to do that though just because you guys can’t publish efficiency numbers that aren’t rediculous. Other websites seem to be able to invest in proper testing gear. If you feel fine publishing those rediculous numbers because you call them ‘Beast efficiencies’ good for you. Personally I’d feel embarassed publishing them regardless of what they’re called.

          • flip-mode
          • 10 years ago

          Woah. I got love for you but you are waaay off point. TR hooks the PSU up to the test unit and then reports the efficiency according to that test – what is the problem? Are you really going to burn some bridges over that tiny issue? At least pick a worthy issue to torch the author for.

            • MadManOriginal
            • 10 years ago

            The problem is quite simple – someone reading this review might think that a PSU will deliver 95% effciency at 25% load in their own computer and go out and buy one for that reason or as an upgrade from a less efficient PSU when in reality they will not be getting such efficiency. What’s the purpose of efficiency numbers if they only apply to loads delivered by a homemade PSU tester? It’s a bit like relying on synthetic benchmarks except in this case those synthetic benchmarks don’t even match what every other source gets in the same benchmark.

            lol this came up in the last PSU roundup as well and was rejected as ‘Our numbers are right for what they are’ without any consideration of trying to get numbers that make more sense or are closer to every other source out there, not even a ‘Hmm we will look in to this.’ I’m done with arguing about it, I’ll just continue to point out that 95%+ efficiency at 25% load is not what people should expect whatsoever in the real world if TR keeps publishing those silly results.

            I’m not sure what bridges have been burned? But this attitude has been repeated by multiple TR staff both publically and privately (emails) on various topics – essentially “We are right and yours or others’ conflicting or different results are meaningless.” Hey, they still write nicely but it’s not much use even trying to figure out why those differences occur with that kind of attitude. In one case it was simply different software that’s supposed to measure the same simple thing but reporting differently – however I had to figure that out on my own, TR staff wasn’t very interested in doing their own investigation at all. So, rather than trying to figure things out or make suggestions I’ll just point out blatent errors and leave it at that – much simpler.

            • Voldenuit
            • 10 years ago

            25% load on a 750W PSU is 187.5W. I can happily believe that a modern PSU will peak at that rating, whatever its maximum rated capacity. For instance, 3 PSUs from SPCR (reviews chosen at random, and ranging from 500W to 700W) all had peak efficiencies at around 200W.

            §[<http://www.silentpcreview.com/article1012-page5.html<]§ §[<http://www.silentpcreview.com/article936-page5.html<]§ §[<http://www.silentpcreview.com/article879-page3.html<]§ I also expected a hump, perhaps (one of) the problem(s) here is that TR's testing points are too granular, and completely miss the expected poor efficiency behaviour at low wattages. Sub-100W idle configurations are now a lot more common than they used to be, and it may be of interest to readers to know how their PSU will fare under these conditions. Agree with both yourself and dissonance, actually, that the high (95%+) efficiencies are probably not reflective of real-world scenarios in most desktop rigs (especially at 110V). Is it just a matter of intake temperature? But if my earlier hunch is correct, and the 200-300W mark is the peak of the efficiency hump for these PSUs, then it might not be as far off the mark as one might think compared to the 25% load behaviour of a 300W unit.

            • flip-mode
            • 10 years ago

            I’ll partly agree with you that there does _[

            • MadManOriginal
            • 10 years ago

            The harshness was because this has been brought up before at least once if not more but it always gets blown off. I don’t really like being that way but after more than once I feel more frustrated. You’re right though that TR staff is more professional in regards to criticisms than some other sites. Kyle is just kind of an ass anyway though heh.

            • SomeOtherGeek
            • 10 years ago

            Yea, you were a little harsh, but I do understand where you are coming from.

            Maybe a better question is, how can we help to upgrade the system? Is it worth it considering it is a verily rare test (2 times a year?) Would getting a full testbed or upgrading to “The Beast” II would be too time consuming?

            But then, I got the general idea of the efficiency from their review, so it is good enough for me. If I want more detailed info, I would go elsewhere. No biggie. But it is interesting to see that there is not a whole lot of difference, so getting the cheapest one is the way to go.

            • just brew it!
            • 10 years ago

            It is not being “blown off”. As a result of comments after the previous PSU review, TR acquired additional test equipment to allow them to check (and tweak) the calibration of “The Beast”. I’m still having ongoing discussions with Geoff on this issue. We /[

            • MadManOriginal
            • 10 years ago

            Cool. I reread my posts and yeah they are over the top lol. But what you said is far different from what Geoff said 🙂 q[

            • NewfieBullet
            • 10 years ago

            Jbi, check out the link I posted below. I’m sure you’ll find a similar solution will work for The Beast.

            • sluggo
            • 10 years ago

            l[

            • MadManOriginal
            • 10 years ago

            So you’re saying that the -12V and +5Vsb lines, which have maximum loads of 9.6W and 15W on the Corsair HX850 for example, affect the overall efficiency by some 5-10% when at 25% of 840 (the +12V line) is 210W? 15/210 is 7% and 24.5/210 is 11.7%, so maybe, but does the industry standard test fully load the -12V and +5Vsb line even when testing 20% load? Because 25% load on those lines together is only 6.15W which is <3%.

            • just brew it!
            • 10 years ago

            The question is, are those low-power rails using low-power switchers, or do PSU makers take the cheap way out and use linear regulators? (I don’t know the answer to that.)

            Linear regulators have pretty dismal efficiency…

            • Voldenuit
            • 10 years ago

            Well, the Seasonic X series use DC-DC converters, which should in theory be pretty efficient, so I’d imagine that the efficiency curve for the Seasonic would be the least affected by loading or unloading the low power rails.

            Yet TR’s efficiency curve for the X750 does not resemble the curve SPCR got when they tested the X650…

            • sluggo
            • 10 years ago

            The +5sb rail has to be generated before the primary switcher (which is disabled during standby), and is most commonly done with a linear regulator off of a small transformer fed by the ripple on the main bridge, just after the PFC. Cost is the driver, and efficiency typically tops out at 50%, and is often quite a bit lower.

            The -12V rail is basically unregulated, so losses are confined to core, switch, and IR. I’ve haven’t measured -12V losses in over 15 years, but as I recall an incremental Watt on -12V cost you something like 1.4W in the secondary, so call it 70%.

            • sluggo
            • 10 years ago

            The 80+ program tests at 20%, 50%, and 100% load factors. The test method takes into account the fact that the sum of the rated outputs of each individual output is greater than the rated output of the total supply. For example, the Corsair’s +12V output is rated for 60A, or 720W of the supply’s rated 750W. In addition, the 5V ouput is rated for 25A, or 125W, which means the +5 and +12 outputs cannot be tested simultaneously at 100% loading, as their combined outputs (845W) would exceed the supply’s overall 750W output rating.

            The 80+ program applies a derating in these situations such that the outputs whose rating is limited within it’s subgroup (for instance three +12 outputs each with a rating of 20A, but limited to 50A in total), or limited in combination with another output (5V + 3.3V limited to a combined 150 Watts) are derated so that the 100% overall load point represents scaled-back levels from those outputs. At no point in the testing are individual outputs tested at 100% of their rated output.

            The +5SB and -12 outputs are not treated this way and are tested at 20%, 50%, and 100% of their rated outputs.

            • just brew it!
            • 10 years ago

            Interesting, I had not considered the +5VSB and -12V issue. That may be an important factor.

            • Voldenuit
            • 10 years ago

            The fact that the PSU with the highest idle efficiency on the Beast (the XFX) also has the highest power draw at idle in a real system shows quite clearly that the Beast loads are not simulating real world efficiencies.

            I don’t think this (the -12V and +5Vsb issue) is the smoking gun we’ve been looking for, people.

            • sluggo
            • 10 years ago

            At 25% load on the Beast, all of the supplies should be operating near the same point on their efficiency curves, given similar architectures (which may not be the case). At system idle, however, the 750W units are farther up the efficiency curve than the 850W units are, and this may well account for the grouping shown in the system idle tests, where the 750W units are drawing less power then their bigger brothers.

            Your point is well taken, though, as 25% load on the XFX is very close to the power draw at idle, and it would surprise me to find that the efficiency curve in that area is so steep as to account for the difference in the relative standings when comparing 25% load and idle power draw.

            • just brew it!
            • 10 years ago

            Is it possible that the idle power draw test is drawing most of its power from a single rail, thereby skewing the results relative to a test where all of the rails are operating at close to the same percentage of their rated capacity?

            • sluggo
            • 10 years ago

            I’d say that’s definitely possible. At idle you may be drawing a fairly high baseline of 3.3V and 5V current, pushing those supplies up the curve, while the +12 is way down, what with no GPU/CPU loads to speak of.

            This is one of the reasons the 80+ testing program does NOT use a PC as a load during evaluations but requires fixed passive/dynamic loads, something very much like the beast. Test repeatability and delivering comparable results are what’s important to them, and you can’t get repeatability when the load is throttling it’s CPU, doing random defrags/indexing, and spinning fans up and down.

            Characterizing a supply’s performance in every forseeable user profile down to the last Watt is not the goal of the testing program, and I think in this forum we’re getting a little caught up in disagreements over 10 Watts here or there.

            • Voldenuit
            • 10 years ago

            The difference is over 20W, which is in the region of 10% error, and the margins being used to recommend one PSU over the other are even smaller.

            Perhaps in the interests of full disclosure, it might be more appropriate to note that the performance of the tested PSUs fell within the margin of error of the test, rather than continue to split hairs with a meat cleaver to reach unsupportable conclusions.

            You may have a point about scheduled tasks interfering with idle testing, but that’s yet another reason to control for such variables, and I’d be surprised if TR hasn’t already done so.

          • bthylafh
          • 10 years ago

          Feeling grumpy today? That was uncalled for.

            • MadManOriginal
            • 10 years ago

            Nope, perfectly happy 😀 Just disappointed that a high quality site like TR feels it’s ok to publish silly results. The methodology disclaimer excuse doesn’t cut it for me but I guess it’s good enough for TR staff.

          • danny e.
          • 10 years ago

          blah blah blah.

        • NewfieBullet
        • 10 years ago

        I have to agree with MadMan here, the efficiency numbers are so out of line with the manufacturers numbers and other test sites that you have to question their validity.

        In order to add a little more constructive criticism here I suggest you read about the problems that SPCR had with this very same issue and their solution §[<http://www.silentpcreview.com/article269-page1.html<]§ I encourage you to read the entire article but in short I believe you'll find the solution is probably the same one they implemented, i.e. measure the actual load being placed on the unit at each voltage line and use that to calculate the efficiency numbers. No need to throw out The Beast!

    • IRQ Conflict
    • 10 years ago

    It’s too bad PSU manufacturers are all going modular.

      • Anvil
      • 10 years ago

      Nonsense, this is a good thing all around except for cost.

      • _Sigma
      • 10 years ago

      What is wrong with modular? I would never purchase a non modular PSU ever again.

    • Voldenuit
    • 10 years ago

    Where are the PSU dimensions? This is especially important for users that have compact enclosures.

    Also, where is the section discussing the internal layout and components of the PSUs? A PSU review without a look at the innards is akin to reviewing a car without ever lifting the hood to see what’s inside.

    Also missing are tests on the speed ramping behaviour (and concommitant noise profile) of the PSUs under varying load conditions. The idle loads of ~200W (at the wall) are high for modern systems with efficient CPUs and GPUs, and the max of 300W would be breached by the type of enthusiasts who purchase these PSUs.

      • tay
      • 10 years ago

      This. Also like to add, testing all voltages while modulating one at a time (3V, 5V, 12V)

    • LordEkim
    • 10 years ago

    Just one ?
    On witch voltage are test made 110V or 220V?
    This is very important to EU readers, because if test is made on 110V then PSU is 2-3% more efficient on 220V, bringing all of the PSU at or near 80plus Platinum certification

      • Sargent Duck
      • 10 years ago

      110. In North America, only the dryer and oven plug into a 220. Everything else is 110.

      Excellent review. Question though. The Corsair had nearly identical efficincies to the Seasonic (gold) but only recieved a silver rating. What am I missing?

      That being said though, 750 is ABSOLUTE overkill for my system. If they could make these exact same power supplies in something more reasonable (like a 500), I’d snap one up very quickly. As I imagine most enthusiasts would as well.

        • anotherengineer
        • 10 years ago

        Well I thought I would step in and make a Correction, in N.A. 120V is the standard, but every household also has 240V. 120V is most commonly found for receptacles for everyday things, and 240V is usually used for Ovens, clothes dryers and other larger things, like electic Sauna heaters, large electric space heaters, single phase welders.

        There are also 220/240V standard size receptacles. I have 1 in teh house and 1 in the garage, and they power electric motors on my saws, and air compressor.

        So I could, if I wanted to plug my PC into this, but then I would have to move it into the shop or garage lol pass on that.

        I am still waiting for a PSU that plugs into 208V 3-phase lol

          • JokerCPoC
          • 10 years ago

          Sure You have 240v, But the 240v here is not like in Europe or elsewhere, Why?

          Our residential 240v is actually two 120v wires in the cable and not true 240v and ours is at 60Hz, everywhere else It’s at 50Hz or maybe 60Hz… No house in the USA is supplied with real 240v, Just pseudo 240v… *[

        • Klopsik206
        • 10 years ago

        They do.
        I am using Corsai HX520, excellent PSU very resonably priced:
        §[<http://www.corsair.com/products/power_supplies.aspx<]§

        • Klopsik206
        • 10 years ago

        They do. I use Corsair HX520
        §[<http://www.corsair.com/products/power_supplies.aspx<]§ Excellent PSU for reasonable price.

    • grantmeaname
    • 10 years ago

    Yay!! I love PSU reviews!!!

    • Meadows
    • 10 years ago

    Talk about an ancient test rig.

      • JoHowdy123
      • 10 years ago

      lol, i was thinking the same thing. Interesting as it is, it really does not need the update like a graphics driving test bed would need, after all these are PSU’s tested, and mainly looking from “at the wall” numbers and metered numbers…

      • Ushio01
      • 10 years ago

      I’m sure Techreport would be happy to accept an upgraded test rig contributed by one of there most avid readers.

      After all it is only used twice a year.

        • Meadows
        • 10 years ago

        I’m sure someone out there has some sort of a useless, outmoded SLI computer that could very well be used to test power supplies.

        • Mystic-G
        • 10 years ago

        My question is… Service Pack 2?

      • bthylafh
      • 10 years ago

      What’s wrong with it?

      Or is this more of your attention-whoring?

      • Krogoth
      • 10 years ago

      Why does it matter?

      The review’s focus was on the PSU and the testing rig was only used for the noise section. Besides, the main reason why the testing rig hasn’t change is to make results comparable from previous reviews.

        • Meadows
        • 10 years ago

        It would make sense to test these power supplies with a computer setup that’s actually likely to receive them in the real world.

        We could see more realistic idle/load numbers, up to date noise statistics about uselessly expensive multi-GPU setups, and most importantly, we’d see whether such an 850 W supply is needed at all, even in such cases.

          • Vasilyfav
          • 10 years ago

          People who purchase “uselessly expensive multi-GPU setups” usually don’t give a toss about reading PSU reviews and just buy “whatever the best stuff is” or a complete computer system from some overpriced retailer like alienware.

          • sluggo
          • 10 years ago

          l[

            • mattthemuppet
            • 10 years ago

            erm, earth to sluggo, I think Meadows was caustically referring to the dual 8800GTS rig, not “the Beast”.

            I imagine a dual 5870, extreme ed i7, 10 HDD gaming storage monster might push one of these PSUs. Can’t imagine much else doing it though.

            • sluggo
            • 10 years ago

            Then he and you are even further off base. There’s nothing magical about newer hardware. New hardware presents the same load dynamics to the PSU that the old stuff does. Finding out how many thousands of dollars worth of gear you have to buy to max out an 850W supply tells you nothing that you can’t learn by connecting it to a sufficiently low impedance and turnig it on and off.

            • OneArmedScissor
            • 10 years ago

            New hardware does not have the same power requirements as old.

            The farkin idle levels on this computer are higher than what a lot of computers see at load.

            • mattthemuppet
            • 10 years ago

            yes yes yes, I get your point. However, what Meadows (and I) were trying to get across is that there’s little point doing “real life” testing if the components selected barely even breach 25% load. That’s the fallacy of all the test sites that test PSUs with rigs that don’t run load at even 50% rated capacity of the PSU (and by extension, the point of having one of those PSUs at all).
            Don’t lets get started on the impedance tester used either as that throws off such screwy figures that I don’t trust the results it produces.

          • grantmeaname
          • 10 years ago

          850 W are never needed. The only time to buy a PSU with that capacity is if its the cheapest one that has all the connectors you need (500W PSUs with 4xPCIe connectors are really rare…)

            • MadManOriginal
            • 10 years ago

            Yeah, but one word: adaptor.

            I’m starting to think that the number of certain plugs is a big way that PSU makers justify their more expensive offerings.

Pin It on Pinterest

Share This