400-450W PSU Roundup
by Christoph Katzer on November 6, 2007 4:00 AM EST- Posted in
- Cases/Cooling/PSUs
Corsair DC Outputs
The DC output has minor flaws on the 3.3V and 5V rails, as we have seen with other similar models. The 12V rail stays within 3% of 12.00V. We have frequently heard complaints that our output results are lower than other reports on the net, particularly on the 3.3V and 5V rails. Please compare the load on the rails, however, as we are testing in accordance with the ATX12V v2.2 guidelines which requires higher loads. We think the loads on these rails might be a little outdated and hope that Intel will adjust these levels with an updated specification, but until then we will continue to push these rails to their limit even though it might be different in actual computer systems.
Efficiency
The efficiency reaches its maximum 85% at 220W, with the 230VAC input achieving 1.5% higher than the 120VAC input. With a 90VAC input, the unit only reaches 81% efficiency. While this may not be the most efficient power supply we've ever tested, it still manages to stay above 80% efficiency at most loads and is definitely a cut above the numerous budget power supplies that are on the market.
Ripple & Noise Results
| Corsair 450VX 450W | ||||
| PSU Load | 3.3V | 5V | 12V1 | Wattage All Rails |
| 10% | 1.31A | 1.31A | 2.76A | 45W |
| 20% | 2.62A | 2.62A | 5.52A | 90W |
| 50% | 6.54A | 6.54A | 13.79A | 224W |
| 80% | 10.47A | 10.47A | 22.06A | 354W |
| 100% | 13.09A | 13.09A | 28.58A | 440W |
| 110% | 14.40A | 14.40A | 30.33A | 480W |
The DC output has minor flaws on the 3.3V and 5V rails, as we have seen with other similar models. The 12V rail stays within 3% of 12.00V. We have frequently heard complaints that our output results are lower than other reports on the net, particularly on the 3.3V and 5V rails. Please compare the load on the rails, however, as we are testing in accordance with the ATX12V v2.2 guidelines which requires higher loads. We think the loads on these rails might be a little outdated and hope that Intel will adjust these levels with an updated specification, but until then we will continue to push these rails to their limit even though it might be different in actual computer systems.
Efficiency
The efficiency reaches its maximum 85% at 220W, with the 230VAC input achieving 1.5% higher than the 120VAC input. With a 90VAC input, the unit only reaches 81% efficiency. While this may not be the most efficient power supply we've ever tested, it still manages to stay above 80% efficiency at most loads and is definitely a cut above the numerous budget power supplies that are on the market.
Ripple & Noise Results
Facebook
Twitter
RSS
37 Comments
View All Comments
zeroidea - Wednesday, November 7, 2007 - link
The Antec PS featured in this article is currently on sale at staples.com for $30.http://www.staples.com/webapp/wcs/stores/servlet/S...">Link
SilthDraeth - Wednesday, November 7, 2007 - link
I know it falls outside of the 450Watt max, but it is still below the 500 watt barrier. I just bought the psu for a midrange system I built my mom, I know it works well, but I don't have the ability to test everything.One can wish right?
Noya - Tuesday, November 6, 2007 - link
If you look around (buy.com), the Corsair 450vx can be had from $51-61 pretty regularly, and at that price it's untouchable. I must say I've had one for about two months and haven't had any problems with it.smthmlk - Tuesday, November 6, 2007 - link
Can we have a full list of caps in each unit? Noting the primaries is nice, but what about the others? Thanks.Talcite - Tuesday, November 6, 2007 - link
Thanks for finally putting up O-scope readings, it's quite nice to see them. One thing I'm concerned about though is the lack of explication or analysis.For example, is the entire o-scope range 200mV in the 12v readings or is it one division? Also, there's a number of strange spikes in all the o-scope readings. I'm pretty familiar with the 450VX o-scope readings (mostly from other sites) and I haven't noticed any spikes of that nature in their readings. It probably isn't, but is the equipment faulty?
Thanks for putting the readings up anyways though, they're a nice addition.
Super Nade - Tuesday, November 6, 2007 - link
Those strange spikes are probe noise. If the probe is coupled incorrectly, you will see this artifact. This cold be due to any number of factors like EMI for instance. Following the ATX guideline on using 0.1 uF output coupling caps will minimize this to a great extent.MrOblivious - Tuesday, November 6, 2007 - link
You have to look at the sampling time on the readings to compare between sites. Different sampling times will make the traces appear a bit different.phaxmohdem - Tuesday, November 6, 2007 - link
I just recently built a rig for a buddy using a Thermaltake PurePower (or ToughPower.. can't remember) 480W Unit and it works like a champ powering a Core2 Quad Q6600 CPU, 2GB RAM, 2HDD's, two Optical Drives, 8600GT, TV Tuner and a few fans. (Wish I knew about the Corsair PSU when I spec'd that system out).But yeah, Unfortunately most people equate Watts to overall quality.... not unlike the MHz war of days gone by. Power supplies are one of the hardest components to convince people to spend extra on for some reason. FFS, electricity is kind of the basis of the whole dam computer!
Eventually it ends up as "Oh well. Go ahead and get that 600W $30 power supply and let me know how that goes for ya. I could use a good laugh, and don't say I didn't warn you."
strikeback03 - Tuesday, November 6, 2007 - link
Maybe Anandtech should go ahead and list the weight of each PSU. That has long been used as an estimate of quality.jonnyGURU - Tuesday, November 6, 2007 - link
Not any more. Topologies have changed to the point where you can have very light quality units and very heavy crap. Weight is NOT a factor.