Power Supply Comparison

[Farkash]

Hi,

I have just read the power supply comparison, and first, I would like to commend THG for a job well done.

Regarding this article, THG has recommended Fortron's FSP power supply. However, I did not find any data in the test results section regarding this power supply. Furthermore, all of Fortron's power supplies' model numbers begin with FSP, and there is no specific model number. Third, I noticed that the Verax' power supply's model number suspiciously resembles a model number from Fortron's catalog, and I was wondering perhaps the Verax P/S is actually a Fortron aswell.

Thanks,

 

[LtBlue14]

also, the enermax sure didn't get much attention, and they're recommended so much on these boards. i guess from this review one can surmise that they did not perform spectacularly (except that they came with good documentation...) but only average. hmm maybe antec is a better choice to go with after all

 

[selander]

I am surprised the article did not mention PC Power & Cooling. They have made high quality power supplies since 1985. It was founded because the owner wanted a quieter power supply.

Check out the inside story on power supplies, which shows a cutout of their supply and a description of the high quality components that go into it.

http://www.pcpowercooling.com/products/power_supplies/insidestory/index.htm

http://www.pcpowercooling.com

 

[upec]

I cannot find any place that is selling Verax PSU in US. Anybody know where I can get one.

 

[LtBlue14]

yeah i couldn't find verax or TSP anywhere, and again Fortron was not listed in any of the tests...i would actually say this review was NOT very well done

 

[AntecRep]

At a guess PC Power and Cooling didn't send them a PSU (maybe the request got mishandled etc.) and they are a bit expensive too.

Also I think some of the authors are in Europe so some models available there may not be available here and vice versa.

Just somethings to keep in mind.

AntecRep

 

[lhgpoobaa]

Yes. It was a good review. Andmuch more emphasis needs to be put on PSU's.

Time and time again i see people with big PC problems, both intel and amd, that can be traced back to the PSU.

I dont know why but PSU's seem to be a phsycological blind spot. When they get their UBER computer made it has all the top notch components, then wreck it with a UBER-shite PSU of crappyness :frown:

Something else which i dont think the review touched on enough was <b>PSU build reliability.</b>
Sure one PSU worked fine, but would 3 in a row behave the same? Ive seen a glowing review of the codegen 350W psu elsewhere on the web, yet my friend had one month of sheer misery with his codegen before <font color=blue>puff!</font color=blue> it died in a pretty cloud of blue smoke.

Enermax on the other hand do what they were designed to do, day in, day out. I have one, as do 3 friends, and none of us have ever had a single problem with our Enermax Units.

Build reliability is just as important as being able to deliver the goods.

<b>And if you gaze for long into Toms Hardware Forums, The Forum gazes also into you! </b>

 

[upec]

I think their should also test more PSU that is 400 Watt and above. The three recommended Power supplies - Fortron FSP, Verax and Herolchi. 2 are 300 Watt and 1 is 350 Watt. I think most people come to this website have a power supply that is 400 Watt or above.

 

[upec]

I agree. I also think build reliability is very important. I brought an IBM 75GXP series hard drive because it was the fastest drive at the time according to many reviews. The drive died 6 month after I brought it and I found out this particular hard drive have very high fail rate.

 

[lhgpoobaa]

True, most high end PC's today wouldn't be seen dead with anything less than a quality 350W Unit.

However as the tests performed as very challenging and occasionally distructive the costs associated with buying alot of high end PSU's is probably prohibitive. Toms Hardware isn't made of money unfortunately.
And i doubt you could rent PSU's for such tests.

<b>And if you gaze for long into Toms Hardware Forums, The Forum gazes also into you! </b>

 

[lhgpoobaa]

hehe
You should have seen the PSU i started with.
I was one of the many that got duped with a store built system with a cheap ass PSU.

It was a 300W Omni unit, came with the omni case. With my original system it was running so close to its max just with the 1200C processor.
It pumped out twice the heat of my enermax, and could barely hold 4.8v on the 5v rail.

The enermax also
A. weighs in at 1400g+, compared to 550g for the omni
B. has 11 molex cables, 1 floppy and the p4 atx. The omni had 4 molex 1 floppy and thats all.
C. one of the voltage rails could supple DOUBLE the current.

Ive seen a range of PSU's, and ive realised the easiest way to judge the quality of a PSU is by physical WEIGHT. Heavier allways = better.

<b>And if you gaze for long into Toms Hardware Forums, The Forum gazes also into you! </b>

 

[Black_Cat]

Hopefully THG will update this article. I've dealt with decent power supplies from Nspire and Enlight. Now Vantec makes an aluminum power supply. What about Sparkle? Are the guts of these power supplies actually Forton or Verax? Perhaps is the US they are sold under different names.

To start press any key. Where's the "any" key? --Homer Simpson.

 

[balzi]

I think it's possible that branding changes between countries.
I myself would go for Antec and Enermax because I've heard from people that actually use them.

IMO the article was a debacle. They somehow avoided the Enermax supply.. maybe it was cost.. and then they give the title "adequate" to some supplies that didn't even meet spec. let alone run reliably <b>at</b> spec.
On top of that.. there were countless bar graphs with no explanation of what was going on..

I looked at the "Power Co-efficient" graph and noticed that Antec scored 89%, while Enermax got 68%.. I thought this was a relationship between actual power out and measured maximum output [the article says this is what the co-efficient means].<b>BUT</b> the numbers are confused later on when Enermax is listed as 389W measured/350W claimed & Antec got 471W measured vs. 380W claimed.. so both supplies exceeded expactations.. btu its never mentioned. To my mind Enermax should get 389/350 = 111% and Antec would get 471/380 = 123%. The winners of the article would get Fortron->129%, Verax->130%, and Herolchi->117%. bear in mind that all these supplies are rated differently - two 300W, two 350W, one 380W

I tried to get the 89% and 68% to mean something like maybe it's related to the max. specs on the BEST PSU being tested. nothing made sense. Those numbers (and possibly the entire article/debacle) are just fluff, no substance.. the only thing I believe is that some PSUs really did smoke, and that they probably did intend to write a decent article on PSU performance.

Balzi

P.S I'm probably being harsh to my non-English speakign friends in Germany (where the article was written) but this topic is of interest to me and I was bitterly disappointed to find that the facts didn't get represented in way that allowed a value judgement on the PSUs on offer.
Please THG, write this article again.. and outline why you recommend anything at any point and even why some weren't listed. And please stop being lazy with no explanation after a pretty picture.


"Meet me on Platform 10 in 3 minutes... or Platform 3 in 10 minutes, whichever you prefer!!"

 

[merlin42]

What is the 'Power Coefficient'? There is a graph on page 7 of the review with this heading, but NO explanation! It sort of seems that 'higher is better' is this graph but not even that is documented. I really apreciate tom's trying to objectively review power supplies since they are a poorly understood component, and can cause many headaches when they go bad. But the review was not as transparent as it should have been (IMHO). In particular the formula for 'overall score' should have been documented so that users could adjust and recalculate it for their own situations.

 

[camieabz]

Rather pleased with the comparison. I bought an Antec case with a 430W True Power supply 2 days ago.

<b><font color=blue>~ <A HREF="http://www.btvillarin.com/phpBB/viewtopic.php?t=324" target="_new">My System Specs</A> ~<font color=blue></b>

 

[whenrik]

I have purchased 2 different PC Power & Cooling PSU. One Silencer 300W and one Silencer 400W. I have not had any problems with them, and I think they have a 5 yr warranty. The site I purchased from appears to have many of the specs referenced here, such as 3.3V + 5V combined outputs. I believe that THG has mentioned them many times. www.directon.com

I will almost definitely purchase another PC Power & Cooling(they have a 600W version) for the next PC I build.

 

[jamarno]

<b>Ive seen a range of PSU's, and ive realised the easiest way to judge the quality of a PSU is by physical WEIGHT. Heavier allways = better.</b>

I've found that very light = bad, but heavier does not always = better, an example being Leadman/Powmax. Their supplies have fairly large inverter transformers and probably the largest heatsinks, but they're also known for low output power, as Tom showed with a 500W Leadman.

 

[ejsmith2]

Another Kudos! for the tiger team of lab rats in Germany. I know setting up that testbed took some time, and some cash for those resistors.

An excellent, excellent article; long overdue, but done so outstanding, that it makes the timing a moot point.

The only thing I would like to have seen them touch briefly on is "tweaking" the psu. Out of all my psu's that I've used, I've not had one burn out on me; a couple of them needed the tweaks so very, very badly it was just crazy that they got QA'd out of the plant. And I know I've put the stress on a couple of them, strictly judging by that column of current ratings for the different components. I've got a couple of 300w "Powman" psu's running a couple too; both have "PC Power and Cooling" (or whatever it is that the board seyz back by the heatsinks) boards.

I have a couple of Leadman 500w psu's, and the 12v rail always runs low when it comes in the mail (11.7), and the 3.3rail runs hot enough to make me sweat sometimes (3.5-3.55). By edging that up to about 12.3, and down to 3.4, I have to say I've been very happy with them both. I've not done anything about the default fans, though. That's just a wire snip away from an $8 fix. Which would bring my cost up to $45 for that psu, including shipping.

Good enough for me, my computers, and my pocketbook...

[Jedi mind trick] You LOVE Palladium. [/mind trick]

 

[HammerBot]

Yes, a fairly nice job. However, I was looking forward to read about the power conversion efficiency (since this indicates how much power is lost in the PSU and hence the need for more cooling) and the Power Factor.
The Power Coefficient is a stupid term. I never heard it before. THG probably invented it. Suffice to say is that PSU x only supplies y percent of the rating.

 

[balzi]

What?!!?!?

I can't believe anyone thought the article was well written.
The results were there, but the conclusions were random peices of dribble... and the recommendations were worse. Apparently coming out near the top of the graphs means you win but what do those graphs say... who knows??

I suppose the noise ones made sense. but beyond that.. well, I'm just amazed that someone thought the article was 'excellent, excellent article; long overdue, but doen so outstanding'.
Are you off your nut!?!? hmm. sorry....

well, enuff complaining..

"Meet me on Platform 10 in 3 minutes... or Platform 3 in 10 minutes, whichever you prefer!!"

 

[balzi]

Suffice to say is that PSU x only supplies y percent of the rating.

Hammerbot,
This rule doesn't hold up.
Read the figures in my other post.
The figures for the top two PSUs do not match this "percent of rating delivered" formula.

I am still wondering what those figures meant.. maybe they fudged them to put their 'favourite' supplies in top spot.


"Meet me on Platform 10 in 3 minutes... or Platform 3 in 10 minutes, whichever you prefer!!"

 

[7up1n3]

The article was a good start at informing the general user of the need for a high-quality power supply. However, it did seem to gloss over several items. I also am unclear as how the review came up with the "winners".

One omission that surprised me was the non-mention of the fact that several mfgs, including Antec with their TruPower line, separate the 3.3v and 5v lines. This prevents one line from stealing power from the other when under load.

Seems worth mentioning...

 

[goblin]

I agree, I wish they would have included a PC Power and cooling unit. They are used by many people with dual cpus (I'm one of them) Funny how they thought enermax was a decent supply. Enermax maybe ok for single cpu systems. But I read tons of dual users that could not get the 450+ enermaxes to hold constant voltages under load. The PC Power and cooling supplies are rock stable under load and also include line conditioning.

It would have been nice for Tom to include some graphs that showed just how much the voltage dropped across the amp load range.

Oh well overall a decent review that should open some newbie eyes as to why their system maybe hanging at times.

 

[goblin]

I agree, I wish they would have included a PC Power and cooling unit. They are used by many people with dual cpus (I'm one of them) Funny how they thought enermax was a decent supply. Enermax maybe ok for single cpu systems. But I read tons of dual users that could not get the 450+ enermaxes to hold constant voltages under load. The PC Power and cooling supplies are rock stable under load and also include line conditioning.

It would have been nice for Tom to include some graphs that showed just how much the voltage dropped across the amp load range.

Oh well overall a decent review that should open some newbie eyes as to why their system maybe hanging at times.

 

[balzi]

oops Double post dude.

I actually thought I saw a PC Power PSU included in the mix.

And to 7up1n3 - I agree, at least it opens the eyes of those that think the PSU is just a spot where you can save money by skimping on quality without any effect.

let's see if Tom comes back and gives a good overview to the experienced who want some good figures and good conclusions.

"Meet me on Platform 10 in 3 minutes... or Platform 3 in 10 minutes, whichever you prefer!!"