power supply advice needed

[jonnyGURU]

Originally posted by: Bobthelost
So would it be fair to say that the verdict is that the Corsair is a very nice rebrand, but it's overkill for your system. A 500W Seasonic S12 (instead of a M12 620W variant) would be sufficent. You're overspending.

You know what's good? A Corsair 520W. I think Corsair screwed up and priced it too cheap. That's why they gave 620W to all of the reviewers: get the word out on the 620W. Unfortunately, the rep at Corsair was foolish enough to give me his MSN account info, so I kept PM'ing him for a 520W and now I'm going to do a review on it. Cliffs: It's the same thing but with slightly lower 12V rails.

 

[Bobthelost]

Sounds good, now if they do me a 430W version i might consider it for my next build

 

[acegazda]

Originally posted by: jonnyGURU

Originally posted by: DLeRium
Neg. Forget the Corsair. Get Seasonic.

Have you looked at the Corsair at all or even read any of the posts after the first three?

The Corsair IS a Seasonic, only it's got better support because it has a 5 year warranty and you can actually ring someone at Corsair and get an RMA if necessary. Ever try getting an RMA from Seasonic? It's not impossible, but if you're an end user it's not easy either!

The only Seasonic I would suggest over the Corsair is the PCP&C Silencer 750W. I can't see suggesting anything Seasonic branded as they don't offer anything more powerful and when the M12 comes out, which will then offer a 650W version, the only thing they have to offer is an extra 60MM fan. Wait a minute... Do I WANT that?

there was someone on the forums who had a s12 blow up on him and ruin all his parts. Sesasonic RMA'd the psu and paid for everything else too. I don't know if this is a good thing because of their service or a bad thing becasue it blew up... but that was only one person...

 

[Pabster]

PC Power & Cooling.

The power supply is the most critical part (and oft overlooked) part of a rig. Since you're dropping a wad on a new Conroe build, an extra $50 to $100 won't be missed.

Edit: Go with the 850 if budget allows.

 

[jonnyGURU]

Originally posted by: acegazda

Originally posted by: jonnyGURU

Originally posted by: DLeRium
Neg. Forget the Corsair. Get Seasonic.

Have you looked at the Corsair at all or even read any of the posts after the first three?

The Corsair IS a Seasonic, only it's got better support because it has a 5 year warranty and you can actually ring someone at Corsair and get an RMA if necessary. Ever try getting an RMA from Seasonic? It's not impossible, but if you're an end user it's not easy either!

The only Seasonic I would suggest over the Corsair is the PCP&C Silencer 750W. I can't see suggesting anything Seasonic branded as they don't offer anything more powerful and when the M12 comes out, which will then offer a 650W version, the only thing they have to offer is an extra 60MM fan. Wait a minute... Do I WANT that?

there was someone on the forums who had a s12 blow up on him and ruin all his parts. Sesasonic RMA'd the psu and paid for everything else too. I don't know if this is a good thing because of their service or a bad thing becasue it blew up... but that was only one person...

Any power supply can blow up and potentially take out components. It even happens with PCP&C units sometimes. Simply put: Electronics aren't infallible. It is good that Seasonic stepped up. But I wouldn't expect any less from Corsair. I would say to give Corsair a lot of credit for the same reason people give OCZ a lot of credit. They add value to a product by giving it a good warranty, they have support forums and reaching someone at Corsair is a toll free number away. That's not always the case with some brands.

 

[viramor]

Considering the huge price difference I may indeed go with the Corsair 520watt. It's currently listed at $116 on zipzoomfly but they haven't got any in stock yet.

Should the 520w be plenty for my system?

Here is my newegg wish list. Currently I have the seasonic in there. I won't be buying for a about a month so I hope the conroe prices come down and nvidia releases their new nforce mobos.

http://secure.newegg.com/NewVersion/wis...cWishDetail.asp?WishListNumber=3709904

 

[orangat]

Originally posted by: jonnyGURU
.......
Well... come on..... Tom's Hardware??? Let's be realistic.

But I agree. It's come to my attention that some (not all) of the better PSU testers out there have an "agenda" or "industry alliance" that consistantly taints what on the surface appear to be good reviews.

On the other hand, most tests are "slap it in and turn it on."
........

Tom's did good job on its psu review and the stress testing in the review was way better than any big site I've come across including your own. Some articles and sections on THG may be iffy but not every reviewer there is tainted.

 

[jonnyGURU]

Originally posted by: orangat

Originally posted by: jonnyGURU
.......
Well... come on..... Tom's Hardware??? Let's be realistic.

But I agree. It's come to my attention that some (not all) of the better PSU testers out there have an "agenda" or "industry alliance" that consistantly taints what on the surface appear to be good reviews.

On the other hand, most tests are "slap it in and turn it on."
........

Tom's did good job on its psu review and the stress testing in the review was way better than any big site I've come across including your own. Some articles and sections on THG may be iffy but not every reviewer there is tainted.

Ok, Tom.

Look, I'll be the first one to admit that my reviews aren't perfect. I get WAY more credit than I deserve. I can't measure sound and I don't even measure the length of cables. Frankly, I don't care. My reviews are rather rushed because I'm married with kid and have a life outside of working on computers. I have fun with it. That's all that matters to me. I have fun. Am I doing it for money? You think any of those advertisers pay? They're there to make the site look professional. They're commission based banners. Buy an Ionic Breeze and I might make a buck.

But before you hold anyone else in high regard:

Be skeptical when someone tells you they "put a 300W load on the power supply" without telling you what the actual amperage load is on each rail.

Be skeptical when someone measures ripple from the PWM. If you see ripple numbers that are less than half of what the manufacturer actually specs, the testing methodology needs to come into question.

A few things I've noticed, but won't name names...

Assumptions made about construction of power supplies, placement of components, grade of components, etc. I've seen a lot of ignorant remarks made on other sites. Obviously, I can't address them one by one because then I'm seen as "jealous" or worse, a shill or viral marketing.

Tests that can not be duplicated on actual load test equipment. I don't expect people to run out and spend a few grand on a load tester, but if a claim is made that this much wattage was put on these rails and this much wattage is put on this rail and this is the voltage I got and I can't duplicate it, I call fud. A bank of resistors is totally adequate for load testing. I love the rig TH built. But if you can't give me hard number that can be emulated by another party, one has to wonder what the motivation is behind these numbers.

 

[viramor]

Whats your take on modular vs fixed cable power supplies Jonny. I know your about to release your article about it in your blog but I'm wondering what you recommend. I see Seasonic is going to be releasing their own modular power supplies soon as well.

Oh and thanks for all the input, its been very helpfull.

 

[jonnyGURU]

I have nothing against modulars. In fact, I'm using what could be considered a modular times two.

The SevenTeam ST500EAZ external power supply I have has modular connectors coming out of the back of it and then modular connectors again where it goes into the back of the case.

http://www.jonnyguru.com/PSU/ST-500EAZ/P1010002.jpg

Hopefully the site I contributed the article to will have it up by this weekend. My friend Supersix said that it was overly technical and monotonous, but I think that even if you read the first and last pages and skim the middle you can get the jist of it.

 

[orangat]

Originally posted by: jonnyGURU

Originally posted by: orangat

Originally posted by: jonnyGURU
.......
Well... come on..... Tom's Hardware??? Let's be realistic.

But I agree. It's come to my attention that some (not all) of the better PSU testers out there have an "agenda" or "industry alliance" that consistantly taints what on the surface appear to be good reviews.

On the other hand, most tests are "slap it in and turn it on."
........

Tom's did good job on its psu review and the stress testing in the review was way better than any big site I've come across including your own. Some articles and sections on THG may be iffy but not every reviewer there is tainted.

Ok, Tom.

Look, I'll be the first one to admit that my reviews aren't perfect. I get WAY more credit than I deserve. I can't measure sound and I don't even measure the length of cables. Frankly, I don't care. My reviews are rather rushed because I'm married with kid and have a life outside of working on computers. I have fun with it. That's all that matters to me. I have fun. Am I doing it for money? You think any of those advertisers pay? They're there to make the site look professional. They're commission based banners. Buy an Ionic Breeze and I might make a buck.

But before you hold anyone else in high regard:

Be skeptical when someone tells you they "put a 300W load on the power supply" without telling you what the actual amperage load is on each rail.

Be skeptical when someone measures ripple from the PWM. If you see ripple numbers that are less than half of what the manufacturer actually specs, the testing methodology needs to come into question.

A few things I've noticed, but won't name names...

Assumptions made about construction of power supplies, placement of components, grade of components, etc. I've seen a lot of ignorant remarks made on other sites. Obviously, I can't address them one by one because then I'm seen as "jealous" or worse, a shill or viral marketing.

Tests that can not be duplicated on actual load test equipment. I don't expect people to run out and spend a few grand on a load tester, but if a claim is made that this much wattage was put on these rails and this much wattage is put on this rail and this is the voltage I got and I can't duplicate it, I call fud. A bank of resistors is totally adequate for load testing. I love the rig TH built. But if you can't give me hard number that can be emulated by another party, one has to wonder what the motivation is behind these numbers.

I don't work for Tom or Xbitlabs so I'm not sure who you are referring to.

If you have something more substantial instead of simply saying "Well... come on..... Tom's Hardware??? Let's be realistic." I'd like to hear it. Have you even posted any comments on theor website and interacted with the reviewers from Tom's or Xbitlabs regarding their testing?

Not being paid to be a company shill means that you have more freedom to expose flaws instead of insinuating about testing methodogies, ripple specs or component quality.

 

[Pens1566]

Originally posted by: DLeRium
Neg. Forget the Corsair. Get Seasonic. Look, if you want top notch best PSUs, we're talking ZIppy or PCP&C. If you can't afford those and want to drop down 1 notch, there's Seasonic.

Enermax, Corsair stand right below the Seasonic at a different notch. Maybe OCZ's high end PSUs can compete, but are dfeinitely a notch lo wer.

Then comes Fortron, Antec and OCZ.

Like jonnyGURU said, did you read the rest of the responses? It is a seasonic!!!

NM, you listed Antec ahead of OCZ. :roll:

 

[jonnyGURU]

Originally posted by: orangat

I don't work for Tom or Xbitlabs so I'm not sure who you are referring to.

I know you don't. That's why the wink.

Originally posted by: orangat
If you have something more substantial instead of simply saying "Well... come on..... Tom's Hardware??? Let's be realistic." I'd like to hear it. Have you even posted any comments on theor website and interacted with the reviewers from Tom's or Xbitlabs regarding their testing?

Not being paid to be a company shill means that you have more freedom to expose flaws instead of insinuating about testing methodogies, ripple specs or component quality.

I already said I'm not going to point fingers or details or whatever. You can follow whoever you want to follow, read whatever you want to read. Makes no difference to me. All I'm saying is that you should really know and understand a product before you judge it, or know and understand how something works if you're going to compare what one person does versus another.

Personally, I think the best reviews are the ones where someone has a killer build and they simply slap a power supply into it. There you can get realistic noise assessments, operate in actual operating temperatures, etc. Unfortunately, a good deal of sites that do this are using an Athlon XP 2500+ with a Radeon card (exaggerating, but still.) But take a site that does do this properly and then add to it the ability to measure the actua load per rail, properly measure ripple at the load, measure total power used, calculate efficiency, etc. then we're talking about a really good review site. Too bad it hasn't happened.

The "other guys" are using seemingly technical means to produce questionable results. That's all I'm saying and that's all I'm going to say. Reader beware. Now quit trolling me.

EDIT: Oh, and I have interacted with certain reviews. And unfortunately, they tend to not like to listen to criticism. Constructive or not. When I get responses like, "Well I've tested hundreds of power supplys" as an answer to a question, then I know I'm not getting anywhere.

People, myself included, need to learn humility. I don't know everything. But I learn with everything I do. Some people are so set in there ways, they refuse to learn OR are so misguided by their motives that they don't feel the need to learn.

 

[rise]

Originally posted by: orangat
I don't work for Tom or Xbitlabs so I'm not sure who you are referring to.

If you have something more substantial instead of simply saying "Well... come on..... Tom's Hardware??? Let's be realistic." I'd like to hear it. Have you even posted any comments on theor website and interacted with the reviewers from Tom's or Xbitlabs regarding their testing?

Not being paid to be a company shill means that you have more freedom to expose flaws instead of insinuating about testing methodogies, ripple specs or component quality.

thats funny.

hey, tom's failed the antec tpII 550 until antec reps arrived to show them how to run the test

so you can consider tom's idiots for not knowing how to test it or, well, you fill in the blanks.

 

[jonnyGURU]

Originally posted by: rise

Originally posted by: orangat
I don't work for Tom or Xbitlabs so I'm not sure who you are referring to.

If you have something more substantial instead of simply saying "Well... come on..... Tom's Hardware??? Let's be realistic." I'd like to hear it. Have you even posted any comments on theor website and interacted with the reviewers from Tom's or Xbitlabs regarding their testing?

Not being paid to be a company shill means that you have more freedom to expose flaws instead of insinuating about testing methodogies, ripple specs or component quality.

thats funny.

hey, tom's failed the antec tpII 550 until antec reps arrived to show them how to run the test

so you can consider tom's idiots for not knowing how to test it or, well, you fill in the blanks.

Please link. I missed that.

I've heard of similar things happening. You can have a bank of resistors and know that these resistors will create "X" load, but if you don't know how to read a power supply label, you can potentially crap out a power supply by delivering a load not meant for it.

 

[rise]

Originally posted by: jonnyGURU
Please link. I missed that.

I've heard of similar things happening. You can have a bank of resistors and know that these resistors will create "X" load, but if you don't know how to read a power supply label, you can potentially crap out a power supply by delivering a load not meant for it.

i'll find it. they even had some happy photos of the antec reps shaking hands with the toms crew.

 

[jnmfox]

jonnyGURU

Would you recommend the Corsair 520w over the Fortron Epsilon FX600-GLN?

Thanks

 

[GalvanizedYankee]

It was in the big round-up of PSUs at THG. Nearly a year ago.

Antec sent a tech from Taiwan and the USA to Germany. Then the unit passed


...Galvanized

 

[rise]

here ya go

Important Update - Antec technicians from Taiwan and the USA visited the Munich THG laboratory for just this purpose. The reason for the 3-hour visit was the divergences we discovered in the ripple test, which were not within spec - although the power supplies did not fail. The technicians from Antec performed new ripple tests on their power supplies together with the THG laboratory engineers. It turned out that the tolerances and high-frequency vibrations with their PSU can lead to varying results in the ripple tests. However, we didn't observe this phenomenon in any of the other candidates.

Because we determined that the ripple tests can remain entirely within spec in the present models and that the other values were flawless, the Antec devices still earned positive test grades.

 

[rise]

ok, they were'nt shaking hands, but its certainly a happy picture :laugh:

and hey, i have a tpII 550. as Galv will tell you, i'm no antec hater. but facts are facts.

 

[Bobthelost]

The technicians from Antec performed new ripple tests on their power supplies together with the THG laboratory engineers. It turned out that the tolerances and high-frequency vibrations with their PSU can lead to varying results in the ripple tests. However, we didn't observe this phenomenon in any of the other candidates.

Ok i may have dropped out of my Electronic Engineering degree for CS, but i know enough to know that doesn't make an iota of sense. Tolerances and high frequency vibrations? Double speak for "it's still shite but after they flew a bloke out we can't say that"?

 

[rise]

they didn't publish the pic with the giant check like happy gilmore collected in the back of his car

 

[jonnyGURU]

Originally posted by: Bobthelost

The technicians from Antec performed new ripple tests on their power supplies together with the THG laboratory engineers. It turned out that the tolerances and high-frequency vibrations with their PSU can lead to varying results in the ripple tests. However, we didn't observe this phenomenon in any of the other candidates.

Ok i may have dropped out of my Electronic Engineering degree for CS, but i know enough to know that doesn't make an iota of sense. Tolerances and high frequency vibrations? Double speak for "it's still shite but after they flew a bloke out we can't say that"?

I agree. Ripple is what it is. Whatever ripple your going to read with a scope is going to be the ripple your components are going to experience.

I didn't experience too much ripple on the TPII 550W I tested, so I don't know why they got the reading they got.

And in that review, they actually did measure ripple properly, at the load and not at the PWM. When I was at the Ultra office in Taiwan, I saw that exact same set up on their test bench: the row of Chroma ATE's. And they even had the PCB with the master ON/OFF switch that all of the PSU connectors plugged into. The BNC outputs of the ATE feed data to an oscope based on readings taken at the load. So the reading would have been very accurate regardless of "tolerances and high frequency vibrations."

I wonder why the Tom's staff didn't stick to their guns. How did Antec convince them that $5000 worth of equipment is wrong. :disgust:

 

[orangat]

Originally posted by: jonnyGURU

Originally posted by: Bobthelost

The technicians from Antec performed new ripple tests on their power supplies together with the THG laboratory engineers. It turned out that the tolerances and high-frequency vibrations with their PSU can lead to varying results in the ripple tests. However, we didn't observe this phenomenon in any of the other candidates.

Ok i may have dropped out of my Electronic Engineering degree for CS, but i know enough to know that doesn't make an iota of sense. Tolerances and high frequency vibrations? Double speak for "it's still shite but after they flew a bloke out we can't say that"?

I agree. Ripple is what it is. Whatever ripple your going to read with a scope is going to be the ripple your components are going to experience.

I didn't experience too much ripple on the TPII 550W I tested, so I don't know why they got the reading they got.

And in that review, they actually did measure ripple properly, at the load and not at the PWM. When I was at the Ultra office in Taiwan, I saw that exact same set up on their test bench: the row of Chroma ATE's. And they even had the PCB with the master ON/OFF switch that all of the PSU connectors plugged into. The BNC outputs of the ATE feed data to an oscope based on readings taken at the load. So the reading would have been very accurate regardless of "tolerances and high frequency vibrations."

I wonder why the Tom's staff didn't stick to their guns. How did Antec convince them that $5000 worth of equipment is wrong. :disgust:

So what are you implying when you say -
""Well... come on..... Tom's Hardware??? Let's be realistic." ? before knowing about this issue?

Maybe you should get more info. instead of simply expressing disgust.

 

[GalvanizedYankee]

orangat, you have been a troll in PSU threads for the last six months now.

You seem to contribute little and only enter threads to attack. Are you angry at your
father and taking it out here. That's called leaking...Contain yourself sir!


...Galvanized