bit-tech.net

What is the best 400-599W PSU?

Comments 1 to 25 of 37

Reply
Zinfandel 13th September 2011, 14:30 Quote
FINALLY BIT TECH.

FINALLY.

My faith is restored.

Although... Lack of conclusion confusion :|

And given the price range of £30ish to £110 ish I can't help feel there are a lot of PSUs missed out that I'd like to have seen included. But very nice to see a round up of any kind from BT :D.
supermonkey 13th September 2011, 14:37 Quote
Nice review, guys. My only criticism is that the graphs on Page 11 (12V rail test) might benefit from either additional labels on the Y-axis or separate colors to differentiate the 12V rails. As it is, according to the legend every rail is identified by the same blue color.

I'm pretty certain that my assumption is correct, and that the graph is read from top to bottom (i.e. 12V1 is at the top, then 12V1, 12V2, etc.). However, in the interest of clarity it would have been nice to see the rail designations along the Y-axis instead of the power supply model, which is already identified in the graph header.
lehtv 13th September 2011, 14:59 Quote
A disappointing review, for two reasons:

1) No fan noise level comparison charts. A really big part of a power supply's quality is how much noise it makes at different load levels (e.g. 50% and 100%).

2) No overload tests. The point of this is to see exactly how much can be pulled from the PSU over its rated wattage, and how the PSU reacts to being overloaded - i.e. are its protections fully functional. It's a pretty big no-no if a PSU sparks up when overloaded, as opposed to simply shutting down without endangering the components.
Zurechial 13th September 2011, 15:15 Quote
That's much more like it! Good roundups are immensely useful and this price-range is a smart one to test.
Introductory text to explain concepts and testing methodologies, summary on each product tested; graphs, performance results and individual product scores - That's what I expect from bit-tech and please let it continue. :)

Some of the scores are a bit odd (75% for a PSU that one should leave on the shelf because of only just meeting the ATX spec?!) and the lack of a conclusion is odd too; but then again the results are clear and speak for themselves so I don't really have a problem with that.

The results from the Seasonic go to show why we so often see Seasonic PSUs recommended by people in the know.
I've been using a Corsair CX430 in one of my builds here and I confirm that it's a nice, quiet PSU. It doesn't get any heavy loads in my usage, but it's nice to see the touted reliability & quality of Corsair's PSUs confirmed in testing.
Zinfandel 13th September 2011, 15:32 Quote
I made my own conclusion.

Corsair wins! On a price to performance basis anyway.

http://i.imgur.com/i23tH.png
bigkingfun 13th September 2011, 16:20 Quote
Lizard 13th September 2011, 16:28 Quote
Quote:
Originally Posted by bigkingfun
Hintyhint

We'd love to do sound pressure tests, but most of the meters aren't sensitive (sub 20dBA) to measure PC components and the ones that are cost and arm and a leg.

Plus, as the Chroma load tester (and its PSU) make a ton of noise, all the meter would be listening to is them, not the PSU.
mucgoo 13th September 2011, 16:39 Quote
A conclusion page please?
hyperion 13th September 2011, 18:00 Quote
Very nice review. Would have liked a chart comparing the amps on the 12v rail for convenience but otherwise a really helpful review.
Lizard 13th September 2011, 18:03 Quote
Quote:
Originally Posted by lehtv
1) No fan noise level comparison charts. A really big part of a power supply's quality is how much noise it makes at different load levels (e.g. 50% and 100%).

Please see my answer to this question, also asked by bigkingfun. The reviews do however tell you of our observations about how noisy each PSU is at different load levels.
Quote:
Originally Posted by lehtv
2) No overload tests. The point of this is to see exactly how much can be pulled from the PSU over its rated wattage, and how the PSU reacts to being overloaded - i.e. are its protections fully functional. It's a pretty big no-no if a PSU sparks up when overloaded, as opposed to simply shutting down without endangering the components.

Frankly, given how many PSUs still don't do what they claim do on the label, we're more concerned about how PSUs behave at their stated level at the moment, let alone what they may do when overloaded.
edzieba 13th September 2011, 18:06 Quote
Unless you're running multiple grunty GPUs and an overclocked CPU along with an impressive bank of HDDs, you'd be hard pressed to draw more than a couple of hundred watts peak.
The 'required PSU wattage' ratings listed by Intel, Nvidia and AMD are hugely overestimated, probably to counter the fact that common el-cheapo power supplies can only supply their rated wattage for a fraction of a second before catching fire, so this give them a large buffer to prevent angry consumers who bought a £10 '500W' supply from complaining. With a half-decent brand-name supply that can support it's rated capacity indefinitely (and can peak to quite a lot more for a few seconds at a time) you can easily get away with half the 'required' supply power, maybe even less if you don't intent to be running Prime95 for 24 hours a day.
e.g. under load, my relatively modest rig (e8400, gtx460) draws a little over 200W running full-pelt. Going by manufacturers 'minimum system power' ratings, I should need at least 450W!
lehtv 13th September 2011, 19:25 Quote
Quote:
Frankly, given how many PSUs still don't do what they claim do on the label, we're more concerned about how PSUs behave at their stated level at the moment, let alone what they may do when overloaded.

Given that all of the power supplies in this comparison well perfectly able to output their rated wattage, I hardly see this as a valid point. Fair enough regarding the dBa tests though.
trig 13th September 2011, 19:37 Quote
the "conclusion page" is at the end of each product review page...don't be lazy, read the review...might learn something...

and to say this review is disappointing because it lacks noise level and overload testing is a joke...
1) the noise a fan on the unit makes has very little to do with the quality of the unit. it might be a highly desired feature, for the 1% of pc builders who need absolute silence in their builds...otherwise, any case with 2+ fans in it (gpu fan, cpu fan, case fan) is going to cover up the psu fan noise...
2) this isn't guru, it's a roundup of the best 400-600 watt units that people might buy. not "the best roundup of 400-600 watt units that people who need 700 watts should buy...

i get that it may be missing things you would like to see, but to call it disappointing...well, how about you show 'em how to do it...

nice job bt
GuilleAcoustic 13th September 2011, 19:54 Quote
This test confort me to go with the Seasonic X series. Nice review ;)

ps: Could be nice to have a noise comparison at idle and load, but that is not the most important.
Rai 13th September 2011, 20:57 Quote
On the last graph of page 12, the Antec High Current Gamer HCG-550 Plus shows a + 11.94 v on the -12v rail. I am assuming this is a typo.

The Seasonic X series is impressive. If my machine didn't require more than 600w, I would be tempted to buy it.
Aracos 13th September 2011, 21:25 Quote
This is why I don't subscribe to CPC anymore, because the articles end up on bit-tech :-\
sotu1 13th September 2011, 22:54 Quote
Interesting piece - would be great to know what your test rig was set up to.

Cheers :)
Lizard 13th September 2011, 23:46 Quote
Quote:
Originally Posted by sotu1
Interesting piece - would be great to know what your test rig was set up to.

It's all in the article:
Quote:
Originally Posted by ARTICLE, PAGE 1
To test each PSU, we programmed the load testers to drain the amount of power that each manufacturer claims its PSU can deliver. The voltage of each rail was measured at 10, 50 and 100 per cent load to determine if it was within the ATX spec. We then left each PSU running at 100 per cent load for 15 minutes to see if it could produce stable voltages over an extended period.
sotu1 14th September 2011, 00:14 Quote
Quote:
Originally Posted by Lizard
Quote:
Originally Posted by sotu1
Interesting piece - would be great to know what your test rig was set up to.

It's all in the article:
Quote:
Originally Posted by ARTICLE, PAGE 1
To test each PSU, we programmed the load testers to drain the amount of power that each manufacturer claims its PSU can deliver. The voltage of each rail was measured at 10, 50 and 100 per cent load to determine if it was within the ATX spec. We then left each PSU running at 100 per cent load for 15 minutes to see if it could produce stable voltages over an extended period.

eh....forgive the stupidity, that don't mean much to me :D
Lizard 14th September 2011, 00:46 Quote
Quote:
Originally Posted by sotu1
oh....forgive the stupidity, that don't mean much to me :D

Fair enough, I'll rephrase for you.

Each PSU was tested by drawing the amount of power that the manufacturer claims it can produce. i.e. if the manufacturer says 450W, then we drew 450W from it.
Device Unknown 14th September 2011, 01:58 Quote
meh, if you want real, unbiased professional reviews of hardware, goto tomshardware. I just come here to see project logs.
fluxtatic 14th September 2011, 06:17 Quote
Quote:
Originally Posted by Device Unknown
meh, if you want real, unbiased professional reviews of hardware, goto tomshardware. I just come here to see project logs.

You're kidding, right? Tomshardware has sucked for years. If you want to geek out on PSU reviews, get thee to HardOCP or Anand - they'll get down to brand and series of caps, soldering quality, etc, along with all bit has here. For the rest, check many sites - eventually you'll figure out who to trust for what.

That said, appreciate this guide, BT - just wished you'd had one up a year ago when my 400W fried and I had to replace it (with a 450W BFG - decent, but if I hadn't been in a hurry, I would have gotten an Antec Earthwatts.)

Would love a Seasonic, but they're a little on the dear side for me yet - maybe on the Bulldozer build (btw, where's my FX, AMD?)
mav2000 14th September 2011, 07:56 Quote
Great review and good testing...I wish there was a proper conclusion to it though.
Adnoctum 14th September 2011, 10:26 Quote
Quote:
Originally Posted by bigkingfun
Hintyhint

Although I agree in principle to your hint, I disagree for two reasons:
  • In most cases PSU fans are amongst the quietest in a system, especially true when the fan is a 120+mm one, as all these are. My PSU is a PC P&C Silencer with an 80mm fan. It is inaudible in an otherwise VERY quiet gaming system. My other low-power-optimised gaming system (for when I want to play older games) has slow, low powered fans, and has a Corsair HX with a 120mm fan. It is also inaudible over the quiet 120mm CPU/system fans.
  • There is more to a fan's quality than volume. Often the motors or blades give off annoying buzzes or drones or hums that a meter just won't measure.
Which is why I am satisfied with the subjective evaluation given in each PSU entry. If the fan or component noise is note worthy, it is mentioned and given some analysis (what kind of noise and under what circumstances).

I'm pleased that BT has done this round-up, and hope that it is semi-regular as PSU go through product refresh cycles. Can we look forward to other segments? 600-999W and 1000+W perhaps?

I'm not so bothered with the lack of final summary, as each PSU is given a separate write up with a subjective look at how it fits within the round-up. The problem with a final summary is that even within the limited span of capacities the person thinking of buying a $40 Thermaltake isn't the same person who is thinking of buying a $126 Seasonic. The individual summaries allows the reviewer to deal specifically with the PSU at hand as well as put the PSU into the wider context. I'm happy with the results.

I think that these types of reviews aren't being done as much any more, with a focus on the sexy CPU, GPU and SSD reviews to the detriment of the enthusiast who like to keep informed about ALL the components in their system. I may be in the minority, but I spend at least a day researching like mad before I impulse buy. Often I will mull a purchase for months before committing. Some people like turtles, I like fan reviews and the like.

There have been a distinct lack of reviews of PSU and mechanical HDDs over the last couple of years (all over, not just at BT), and I have really felt this lack of hard numbers with HDDs. HDD speeds can vary quite a bit depending on capacity, number of platters and even which part of the platters are used. A review of a 2TB model often isn't very useful when you are in the market for the 750GB version.
At the moment I'm in the market for a 2TB 7200rpm HDD, and finding a review that tests current model Western Digital, Seagate, Samsung and Hitachi drives was quite difficult. Perhaps this is something BT could do instead of that 6th GTX560 review with different cooling or 10MHz speed bump ("XXX eXtreem Super OCers Maxx WHUT UP! Edition"). Even here among enthusiasts, most people buy HDDs over SSDs.

I recently had to buy a new DVD-RW for my media PC, my first DVD purchase in over two years. Have DVD drives become so commonplace and commoditised that there is no distinction between manufacturers and implementations? It used to be that you looked for one model because the *chipset* was better than the other.

I feel old and cranky and I keep wanting to throw out the odd "In my day...".
Vo0Ds 14th September 2011, 12:02 Quote
Quote:
Originally Posted by Aracos
This is why I don't subscribe to CPC anymore, because the articles end up on bit-tech :-\

This is the reason I read Bit-Tech, because the CPC articles end up here.

...not really, but I thought it was an amusing angle.
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums