Interesting PSU Stress Test, Many PSU's Put to the Test.

Printable View

11-07-2006, 05:13 PM
Kunaak

2 Attachment(s)

Interesting PSU Stress Test, Many PSU's Put to the Test.

now, normally I dont find PSU reviews too interesting, cause they tend to be alittle too vague and general, and leave out details I am actually interested in, like efficiency, and load readings.

but these guys seemed to go alittle above the call of duty for this test.

using some serious hardware, they put a number of PSU's from cheap to high end to the test, and its quite interesting at what they find.

some notable suprises...

TAGAN, and Seasonic, Enermaxs do great, at both efficiency, and stress testing...

while Antec (a name I previously thought was high quality) did pretty damn bad in just about all areas.

and the Fortron 700 didnt pass either...

but one in particular surprises me the most...
the OCZ GameX 700.

its efficiency rating is not even close, and it fails the Stress test.

only thing is, I wish they had more PSU's to test.
it would have been interesting to see how the HX620 from corsair, the 650 from mushkin, and the 610-750 silences, and the 1000k's from PC Power and Cooling would have done.

http://www.activehome.co.uk/personal...g-power-supply

http://www.xtremesystems.org/forums/...1&d=1162948297

http://www.xtremesystems.org/forums/...1&d=1162948297
11-07-2006, 05:18 PM
vowkilmer

Geez where is Johnny when u need him. :)
11-07-2006, 05:24 PM
perkam

In the general hw section :p

Perkam
11-07-2006, 05:25 PM
nn_step

I'm pretty sure he has some useful data to such ends
11-07-2006, 05:27 PM
Cooper

So did they tested each rail individualy or put them at max load at once ? :confused:
And imo 3.3V and 5V rails failing at max load usually doesn`t mean much to end-users cos it`s almost impossible to fully load them in regular rig
11-07-2006, 05:30 PM
irev210

Quote:

Originally Posted by vowkilmer

Geez where is Johnny when u need him. :)

no kidding!

Check out www.jonnyguru.com

best PS reviewer out there!
11-07-2006, 05:36 PM
filmbot

Makes me proud to have an Enermax in my rig.
11-07-2006, 05:48 PM
WeStSiDePLaYa

they screwed up the antec true control testing. i dont think they plugged the front panel in.

if you dont plug the front panel of the true control in, it automatically sets all of its rails to the lowest possible values.

it has the same internals as the independently regulated antec tested here, except the true control lets you adjust the values above and below spec, and from johnnys testing, it held up very well.

so id have to assume they did not properly test the antec.

http://www.jonnyguru.com/review_details.php?id=33

not to mention no other reviews of the true control have encountered such problems, nor has their been reports of this.

so kinda makes me wonder about their testing if they cant even properly set the psu's up... hmmmmmm
11-07-2006, 05:48 PM
LordofDoom

I think my PSU is a rebranded and souped up Enermax :D
11-07-2006, 05:49 PM
WeStSiDePLaYa

Quote:

Originally Posted by LordofDoom

I think my PSU is a rebranded and souped up Enermax :D

too bad enermax doesnt build psu's:rolleyes:
11-07-2006, 07:02 PM
Dr_at_Home

46% efficiency measured on that RaptoxX? I've never even heard of them, but damn:p:
I would of liked to see the Mushkin in the tests though. :)
11-07-2006, 07:06 PM
Kunaak

haha, just be glad you dont have one of them Mercury PSU's....
that thing blew up, haha....
11-07-2006, 07:08 PM
syne_24

Quote:

Originally Posted by Dr_at_Home

46% efficiency measured on that RaptoxX? I've never even heard of them, but damn:p:
I would of liked to see the Mushkin in the tests though. :)

i never even heard of that brand either but damn that thing draws nearly double for what it outputs. That's like the worst efficiency ever..
11-07-2006, 08:00 PM
breakfromyou

i have an antec neopower 480w...according to Antec's customer support, i need to RMA this POS. i dont think i can...

Antec has already been added to the "Brands to never buy again" list. any psu that gets hot enough to burn you has to be a joke.
11-07-2006, 08:01 PM
Dr_at_Home

Quote:

Originally Posted by Kunaak

haha, just be glad you dont have one of them Mercury PSU's....
that thing blew up, haha....

Yeah. I looked on google for more info on the thing, and its probably the cheapest power supply ever. :smoke:
11-07-2006, 10:55 PM
Eastcoasthandle

silencer 750 review
11-08-2006, 02:12 AM
Helmore

Quote:

Originally Posted by WeStSiDePLaYa

too bad enermax doesnt build psu's:rolleyes:

Uhm, that's not true, Enermax does have thier own fabs and thus designs and builds thier own PSUs.

@LordofDoom: Your Corsair HX620 is build by Seasonic, not by Enermax.
11-08-2006, 02:15 AM
XSAlliN

Wish they had "Fortron BlueStorm II" on that test.
11-08-2006, 02:25 AM
Eastcoasthandle

Quote:

Originally Posted by Helmore

Uhm, that's not true, Enermax does have thier own fabs and thus designs and builds thier own PSUs.

you gotta check the UL number here. Yes, there is some level of politics with how a company reports it but they are suppose to be honest by law.
See if you can find it here
http://www.phoronix.net/image.php?id...erty_label_lrg
11-08-2006, 02:27 AM
dinos22

doesn't surprise me that Enermax is the way to go :toast:
11-08-2006, 02:37 AM
damnnit

dam ocz 700w, it failed the atx test.
11-08-2006, 03:13 AM
gundamit

After two RMAs on Antec True(ly No)Control 550 a long time ago, I'm not at all surprised to see it fail miserably. OCZ is surprising, but considering its price after rebate it might be worth it if its just a random QC issue. Critical 12v lines are strong.

Maybe yet another lesson on paying for quaility on the PSU.
11-08-2006, 03:27 AM
kiwi

Not sure how they tested but my FSP Epsilon 700 (same as OCZ 700) 5V line never dropped below 4.90V powering conroe rig with 5 hard drives.
11-08-2006, 09:50 AM
jonnyGURU

Quote:

Originally Posted by vowkilmer

Geez where is Johnny when u need him. :)

Flying off the handle at half ass PSU reviews:

http://www.jonnyguru.com/forums/show...73&postcount=6

Yes I'm a pompous arrogant ass, but I have my name plastered across the top of my website. Would you expect anything less?!

That and I'm sick as fsck and really just want to slam some Jagermeister and go back to sleep.

And Enermax does make their own PSU's. They even have relabeled Enermaxes in some parts of Asia.
11-08-2006, 12:59 PM
Revv23

Quote:

Originally Posted by Cooper

So did they tested each rail individualy or put them at max load at once ? :confused:
And imo 3.3V and 5V rails failing at max load usually doesn`t mean much to end-users cos it`s almost impossible to fully load them in regular rig

I agree, though it would be nice to know that it can meet specs.
11-09-2006, 12:37 AM
XSAlliN

Personally I don't care about those details, for a good & clean review you need one page to add the facts and 3 pages whit pictures to show those facts - I don't need a brief history on how the PSU works, but the thing is that review wasn't made for "PSU Stress" - the title shows it: "Understanding power supply units" and if they "don't Understand" how the PSU works why do they try to teach some one? :)
11-09-2006, 01:35 AM
Turtle 1

Well If your going to test a bunch of PS. and you omitt AC power&cooling there has to be a good reason for it not being in the test. To expensive. no that can't be it. none available no that can't be it. I wonder why no AC P&C PS.
11-09-2006, 04:02 AM
Mats

Quote:

Originally Posted by irev210

Check out www.jonnyguru.com

best PS reviewer out there!

I'd say SPCR is better, and they give ou an idea of how the fan + fan controller works.
Check out their Mushkin review, not really a good PSU. Very inefficient.
11-09-2006, 04:21 AM
WeStSiDePLaYa

Quote:

Originally Posted by Mats

I'd say SPCR is better, and they give ou an idea of how the fan + fan controller works.
Check out their Mushkin review, not really a good PSU. Very inefficient.

they dont provide much for performance numbers, they dont even use an o-scope to show ripple!

personally, i dont care how a psu sounds, or how its fan controller works.

i want to see how well the rails are regulated, and that they have proper filtering so there is minimum ripple.
11-09-2006, 05:07 AM
Mats

Quote:

Originally Posted by WeStSiDePLaYa

they dont provide much for performance numbers,

Exactly what performance numbers are you talking about?

Quote:

Originally Posted by WeStSiDePLaYa

they dont even use an o-scope to show ripple!

Really? Well I don't see it in johnnyguru.com's review either.
It's true that SPCR haven't included it in every test earlier, but now they do. They also checks out what happens if the VAC falls down to 80 - 90 V, if the PSU still can deliver a proper output. I've never seen that anywhere else.

Quote:

Originally Posted by WeStSiDePLaYa

personally, i dont care how a psu sounds, or how its fan controller works.

I'm surprised that you're not interested if the fan can keep the PSU cool and work properly under heavy loads. You saw what happened to that Mercury PSU (not saying that it must be because of the fan). The best way to test it is to feed the PSU with the heat from the components it's powering. Not all heat though, since some of it usually goes out through a case fan. But if you never put a PSU in a case it's less of a problem. I'm not sure if Activehome put the PSU's in cases, but if they didn't then it wouldn't surprise me if some more PSU would fail during the tests. The intake heat makes all the difference.

Quote:

Originally Posted by WeStSiDePLaYa

i want to see how well the rails are regulated, and that they have proper filtering so there is minimum ripple.

SPCR always gives you the values (V & A) for each rail at different loads, up to maximum load.
11-09-2006, 05:45 AM
perkam

Quote:

Originally Posted by Mats

Exactly what performance numbers are you talking about?

Really? Well I don't see it in johnnyguru.com's review either.
It's true that SPCR haven't included it in every test earlier, but now they do. They also checks out what happens if the VAC falls down to 80 - 90 V, if the PSU still can deliver a proper output. I've never seen that anywhere else.

I'm surprised that you're not interested if the fan can keep the PSU cool and work properly under heavy loads. You saw what happened to that Mercury PSU (not saying that it must be because of the fan). The best way to test it is to feed the PSU with the heat from the components it's powering. Not all heat though, since some of it usually goes out through a case fan. But if you never put a PSU in a case it's less of a problem. I'm not sure if Activehome put the PSU's in cases, but if they didn't then it wouldn't surprise me if some more PSU would fail during the tests. The intake heat makes all the difference.

SPCR always gives you the values (V & A) for each rail at different loads, up to maximum load.

Not that many people that can open a psu up and tell you everything you need to know rail and component wise...

Perkam
11-09-2006, 05:57 AM
Mats

Quote:

Originally Posted by perkam

Not that many people that can open a psu up and tell you everything you need to know rail and component wise...

Agreed, and I'm not one of them...;)
11-09-2006, 06:07 AM
jonnyGURU

This is really stupid. Comparing "who has the best PSU review?!?" Give me a break. You should allow yourself the opportunity to read ALL of the good reviews and take something home with you from each one.

Quote:

Originally Posted by Mats

I'm surprised that you're not interested if the fan can keep the PSU cool and work properly under heavy loads. You saw what happened to that Mercury PSU (not saying that it must be because of the fan). The best way to test it is to feed the PSU with the heat from the components it's powering. Not all heat though, since some of it usually goes out through a case fan. But if you never put a PSU in a case it's less of a problem. I'm not sure if Activehome put the PSU's in cases, but if they didn't then it wouldn't surprise me if some more PSU would fail during the tests. The intake heat makes all the difference.

And the reviews at jonnyGURU.com have both intake and exhaust temperatures at each load. If the fan is quiet, GREAT! But if it's not moving any air, stuff's going to blow up. You've never read a review at jonnyGURU.com, have you?

Quote:

Originally Posted by Mats

Really? Well I don't see it in johnnyguru.com's review either.

Now I know you've never read a review at jonnyGURU.com. Because you pick the ONE REVIEW (the newest one and therefore featured on the front page) that doesn't have the ripple and noise results because it was done by Matt and not myself.

Take a look at any of the other reviews and you'll see screenshots of EasyScopeII getting it's data from a USB Instruments Stingray Oscope.

http://www.jonnyguru.com/reviews.php

And for the record, the review does make mention of ripple, but there are no screen shots of it because Matt uses a standard desktop Oscope and not a USB unit like I do.

From http://www.jonnyguru.com/review_deta...=27&page_num=3

Quote:

Originally Posted by Madmat

At this point we see that the loss was a minimal 1.8W overall but the ripple had basically doubled from 25mV on the 12V to 50mV and the 5V rail was hitting 30mV. The 3.3V rail hung in there with spikes hitting 20mV but the majority was under 10mV.

Look.... I don't care if your favorite PSU reviewer is Tom's Hardware. But before you start saying "well I like so and so because he does this and that and the other guy doesn't" at least take a look at the other sites reviews before you start making statements about what you THINK they do and do not do.

And WeSTSiDE, as Mats points out, SPCR has begun measuring ripple and actually uses the same device I use. I'm not sure how and where they measure the ripple though. Perhaps someone can point that out to me. I know of a couple other sites that measure ripple, but they do it at the PWM and not at the load, so the results are often exponentially inaccurate.

For example: I'd like to check that Zalman out for myself and confirm the ripple because that PSU is FSP Epsilon based and if the ripple is really that low, the FSP must have "fixed" whatever problem they had that caused the excessive ripple on their previous units. Either that or the measurement is wrong.

I like SPCR's reviews. They do proper load tests and they measure noise. I wish I could measure noise, but the load tester is too loud to be able to actually measure anything accurate.

But really, to argue "who's better?" Come on guys. :rolleyes:
11-09-2006, 06:28 AM
Lestat

but wouldn't the argument of who's better, in reviewing, be a valid argument.?

if they dont properly test hardware, and if they are well known to not properly test hardware, they could and would be tagged as bad or incompetent reviews.
where as someone such as yourself or the rare few others who go thru all the trouble of reviewing the aspects of hardware that truly matter, would be given a tag of good and proper reviewers.

if all your going to do is stick a volt meter on a psu to show droop, alot of good that does, sure it shows the raw data of a solid rail or not but a psu is much more than its rails.

i think there is a valid argument in who reviews hardware the best.

and i lost all respect and trust for Antec after their True Power series. total crap, but good at the time. i had one that constantly made my system reboot under load. and after switching to a better brand.. all reboots went away,, needless to say i thru away the antec and didnt look back.

i personally have had my OCZ 600w powerstream since they came out and have never wanted another psu. it has never caused 1 single issue and its adjustable rails are great. not to mention the fact that it its droop is almost nil.

i recently purchased a Sunbeam Tech 550w SLI power supply that i must say, from what little testing i have done is fantastic. the rails are overvolted out of the box and the 5 and 3.3v rails dont move an inch while the 12v rail only bounces 0.01~0.02v at most, with an average of about 0.01v.
now does it make it a great psu,, no as i dont have time to bring it into work and measure the unit with the scope here at work. but it suites my needs quite well.
i have seen some random rebooting with the system its hooked to but i can not say at this time its the psu. further testing is needed.
11-09-2006, 06:30 AM
Mats

Quote:

Originally Posted by jonnyGURU

This is really stupid. Comparing "who has the best PSU review?!?" Give me a break. You should allow yourself the opportunity to read ALL of the good reviews and take something home with you from each one.

Yeah, it's stupid. I thought that I could tip people about SPCR, didn't want it to turn out like this.

Quote:

Originally Posted by jonnyGURU

You pick the ONE REVIEW that doesn't have the ripple and noise results because it was done by Matt and not myself.

Actually, that's just because I picked the last review so it was just by coincidence.

Zalman have changed the UL number to their own, although I don't think it makes any difference, same PSU anyway:

Quote:

Originally Posted by SPCR

We know that the previous Zalman PSUs were made by Fortron-Source Power, but surprisingly, the UL file number on the UL/CSA certification sticker is Zalman's own. The primary transformer in the middle of the PSU is marked with the letters SPI, which we know to be the acronym for Sparkle Power Inc, a sister company to FSP. So Zalman remains with the same supplier that they've used since their very first PSU some five years ago.
11-09-2006, 06:43 AM
jonnyGURU

Quote:

Originally Posted by Lestat

but wouldn't the argument of who's better, in reviewing, be a valid argument.?

if they dont properly test hardware, and if they are well known to not properly test hardware, they could and would be tagged as bad or incompetent reviews.

where as someone such as yourself or the rare few others who go thru all the trouble of reviewing the aspects of hardware that truly matter, would be given a tag of good and proper reviewers.

if all your going to do is stick a volt meter on a psu to show droop, alot of good that does, sure it shows the raw data of a solid rail or not but a psu is much more than its rails.

i think there is a valid argument in who reviews hardware the best.

Amazing how you managed to take my point and completely throw it in the :banana::banana::banana::banana:ter.

I'm not even talking about guys that can't even do a half ass review. But I'll take a review where they throw the PSU in a killer rig and measure voltages any day of the week.

If all you do is go to one source for reviews and continually regurgitate the data from that site, good site or bad site, then you're nothing but a bot. People need to get off their ass and find information from multiple sources and put that information together into something that's important to them.

Go to my site to see the load regulation, SPCR for noise, HardwareSecrets to see the unit completely dissected and LegitReviews if you want to see how it does in an X1950XTX Crossfire rig.
11-09-2006, 06:47 AM
jonnyGURU

Quote:

Originally Posted by Mats

Yeah, it's stupid. I thought that I could tip people about SPCR, didn't want it to turn out like this.

No tipping off necessary. ;) I think it's pretty safe to say that people that want a good PSU review know to go to SPCR. That's why I link to SPCR every time they put a new review up. I'd like to think Mike and I have a mutual respect for each other.. but I don't know. Mike hasn't talked to me. :)

Quote:

Originally Posted by Mats

Zalman have changed the UL number to their own, although I don't think it makes any difference, same PSU anyway:

Yeah... I'm dosed on cold medicine and feeling like crap, so I didn't notice this, but Oklahoma Wolf pointed this out at my site: The unit is FSP based, but the caps are Teapo and are of a larger value. So it seems that the ripple issue may have been addressed.
11-09-2006, 06:51 AM
Mats

jonnyGURU: Yeah it's true that I never read your reviews, and that's because I didn't know they existed (even though SPCR calls you a "worthy competitor"). I saw your site for the first time today, but I'm telling you it won't be the last...
11-09-2006, 07:59 AM
jonnyGURU

Quote:

Originally Posted by Mats

jonnyGURU: Yeah it's true that I never read your reviews, and that's because I didn't know they existed (even though SPCR calls you a "worthy competitor").

Cool.

Although Mike spelled my name wrong. He also says I'm the latest addition even though I've been reviewing PSU's with a load tester since 3/17/2005. :fact: :cool:
11-09-2006, 08:09 AM
Lestat

Quote:

Originally Posted by jonnyGURU

Amazing how you managed to take my point and completely throw it in the :banana::banana::banana::banana:ter.

I'm not even talking about guys that can't even do a half ass review. But I'll take a review where they throw the PSU in a killer rig and measure voltages any day of the week.

If all you do is go to one source for reviews and continually regurgitate the data from that site, good site or bad site, then you're nothing but a bot. People need to get off their ass and find information from multiple sources and put that information together into something that's important to them.

Go to my site to see the load regulation, SPCR for noise, HardwareSecrets to see the unit completely dissected and LegitReviews if you want to see how it does in an X1950XTX Crossfire rig.

Quote:

This is really stupid. Comparing "who has the best PSU review?!?" Give me a break. You should allow yourself the opportunity to read ALL of the good reviews and take something home with you from each one.

you are correct in that we should take home something from each review, as most of us do, i thin kwe all go to multiple sites,, its just 1 site in particular that may stick in our minds, or one we know better, or one that has a review thats easier to navigate or understand. but i dont know one person who sticks with 1 site only all the time.
unless its here :)

and what 2 or 3 words did you take away from my statement that lead you to believe i took your words and threw them down the sh|tter?
talk about not taking something away from it?

who does better reviews is a valid argument and you yourself just backed up that argument by saying "go here for this,, there for that, and my site for this"
you just backed up my statement 100%.

those sites do good reviews so your telling us or recommending that we go there... ok.. how does that differ from "who does the better reviews"
its not localized to one site, but its the exact same thing.

:stick:
11-09-2006, 08:29 AM
Metroid

Quote:

Originally Posted by Mats

jonnyGURU: Yeah it's true that I never read your reviews, and that's because I didn't know they existed (even though SPCR calls you a "worthy competitor"). I saw your site for the first time today, but I'm telling you it won't be the last...

Hey how its going mats, long time no see around.

I hope the 7900 is working well.:slobber:

I'm glad you are back.:up:
11-09-2006, 09:10 AM
Mats

Quote:

Originally Posted by jonnyGURU

He also says I'm the latest addition even though I've been reviewing PSU's with a load tester since 3/17/2005. :fact: :cool:

Two reasons:
- It's hard for me to find any review more than half a year old. If you have more reviews somewhere else I don't know.

- I think you're the latest one compared to the other sites mentioned, so I don't see anything wrong with that.
11-09-2006, 09:20 AM
Mats

Quote:

Originally Posted by Metroid

Hey how its going mats, long time no see around.

I hope the 7900 is working well.:slobber:

I'm glad you are back.:up:

Good to see you too!!! But have we ever talked before?:confused: Sorry, maybe my bad memory...
I don't even have a 7900...
11-09-2006, 09:54 AM
jonnyGURU

Quote:

Originally Posted by Mats

Two reasons:
- It's hard for me to find any review more than half a year old. If you have more reviews somewhere else I don't know.

Right. But you've already admitted that you're "new to me." ;) So I'm sure all you did was just look at my site and scrolled down to the oldest review. :cool: The site just got a redo this month and any PSU reviews that did not use the current testing methodology did not get linked from my reviews index (although they're easily found via Google so if someone wants to know about a particular PSU, they can find it.)

Before I did them on my own site, I did them for SLCentral. But Dave sold that site and I haven't been back there in a while (thought the reviews are still there.)

Quote:

Originally Posted by Lestat

you are correct in that we should take home something from each review, as most of us do, i thin kwe all go to multiple sites,, its just 1 site in particular that may stick in our minds, or one we know better, or one that has a review thats easier to navigate or understand. but i dont know one person who sticks with 1 site only all the time.
unless its here :)

and what 2 or 3 words did you take away from my statement that lead you to believe i took your words and threw them down the sh|tter?
talk about not taking something away from it?

Sorry about that. I'm sick and I think the pounding in my head is drowning out reason.

All I was saying is that people should take into account all of the (decent) reviews. It doesn't matter which of those are better. There are reviews that are absolutely stupid and should never be put on the web in the first place. But I wasn't even addressing those. Those should just be ignored.

To try to validate which review does a better job when they all have an effective testing methodology and have a different objective (like SPCR testing for noise) is kind of silly.
11-09-2006, 01:17 PM
Helmore

Quote:

Originally Posted by jonnyGURU

Go to my site to see the load regulation, SPCR for noise, HardwareSecrets to see the unit completely dissected and LegitReviews if you want to see how it does in an X1950XTX Crossfire rig.

That's just pretty much what I do, but it's still pretty hard to find this 'perfect' PSU. Although I guess this can't be blamed on the guys that do the reviews, it's more of a manufacturer's problem.
11-09-2006, 01:52 PM
jonnyGURU

Quote:

Originally Posted by Helmore

That's just pretty much what I do, but it's still pretty hard to find this 'perfect' PSU. Although I guess this can't be blamed on the guys that do the reviews, it's more of a manufacturer's problem.

It's more of a user problem. Everyone wants something different.