PSU tier list 2.0

[synphul]

My point was just that from a consumer standpoint, the list results from sources like that are confusing. How many consumers went out and bought a honda passport because of a brand name and despised isuzu's rodeo for being an isuzu when they're the same vehicle with different badging rolling off the same line. I won't even get into corsair, whether they set their specs or not - one wouldn't find it's way into my systems even by accident. Maybe corsair has low return numbers because when they fail, the user finds it's not worth the postage stamp and simply buys a different unit.

Same could be said for any of these, these are only verified return/warranty repairs if I understood correctly. So it's still only a small part of the overall picture regardless. It's like trying to rate a restaurant based on verified complaint forms submitted to management. Just because a restaurant chain has a low percentage of filed complaints from their patrons doesn't mean they don't have a high percentage who just couldn't be bothered and never returned without that data being recorded anywhere.

 

[Karadjgne]

I guess my train of thought on all this was due to some posts I was reading on another site. It was about which batches of 3770k cpus were actually shown to have had good OC and which failed in the lottery. What I got from all that was a sheer sense of amazement at the amount of fake RMA's where ppl RMA'd a cpu simply because they received either a lousy OC or a bad batch number.

And then I read some posts on returned psu's and it struck me as really odd the Seasonic had such a high return rate. Why? And I guess I just combined the 2 forums and figured that for some reason, the units were probably not defective, or not all anyways, but just more expensive, when there is a much cheaper unit of the approximate size needed, built by a widely known pc component builder, so of course it must be good. Fake RMA. Doesn't seem like a stretch of the imagination.

But I do agree the data is 'iffy' at best. For all I know there were only 100 Seasonic units sold, so 3 returns isnt bad, but over 1000 CoolerMaster units sold, so 1% is really bad. I dunno. Too many unknown variables.

 

[FALC0N]

I'm not silencing anybody for making criticisms. As I have stated, this list is community managed. But if you're gonna make such questionable views on what units should go where, based on style and price, instead of reading word by word reviews by reputable sources, then i'm sorry, I cannot change it.

Questionable? Based on what? That I disagree with the peanut gallery? You keep claiming that your ratings are based on reviews. Yet I gave 3 examples of PSU's in tier 3 that clearly review at tier 2 levels and 1 example of a tier 5 PSU that reviewed at tier 4 level, and you just ignored it and referenced the peanut gallery. Either your rankings are based reviews or they are not. You say one thing, then do another.

The RM series of units aren't great. Their first lineup had some pretty poor capacitor choices (aside form the japanese ones), on both sides of CWT and Chicony. Ltec's used in the Chicony units and CapXon capacitors used in the CWT models, on very expensive units. The second lineup had overheating problems, thus Corsair was recalling loads of them.

I already address this issue earlier. Lets review.

HardOCP managed to obtain a thermal shutdown at 40% load by running it at 45c ambient, which is 5c over its temp rating. Same at 100%. This is NOT a fail by any reasonable measure. Any PSU that shuts down safely outside its performance envelope has done exactly what it is supposed to do. The test did reveal a small issue with the timing of fan engagement that would only affect systems operating at 40% load, in the Gobi desert without fans, thus the recall. It wasn't a big deal BEFORE the recall, let alone now with issue corrected.

The Caps are a legit issue, but 99% of PC PSU's have Chinese Caps and run just fine. I'm not a big fan of it, but it isn't enough to drop it a full tier. And that is also the position of most pro reviewers. The RM reviewed well.

The third refresh hasn't had issues, but it's safer to leave them in Tier 3, just in case something happens.

This is NOT a test based position. This is personal bias. Exactly what I have been saying is wrong with your list.

 

[FALC0N]

Oklahoma Wolf was providing his opinion on the cleanliness of the circuit boards from Chicony and Delta. If you're going to provide a reviewer's opinion as fact then you should provide some evidence to back up those opinions.

Reviews don't provide the whole story. They are only a very miniscule sampling of the product. There's no long term testing to determine reliability. The loads used during testing are not real component loads so any potential problems like coil whine are not revealed.

Oklahoma Wolf was talking about "build quality". That encompasses more than just the circuit board. And his opinion IS evidence. He is an experienced, respected, PSU reviewer. Chicony also makes HIPRO, which is a respected OEM brand. Noticed that nobody has offered a compelling rebuttal. So as it stands, my case for Chicony is the strongest that has been presented to date.

True, reviews don't provide the whole story. However, the opinions of end users, many of whom have never touched an RM, don't strike me as anything I should value over the pros.

 

[FALC0N]

The new RM750 and RM850 models made by CWT are now using all Japanese Nippon Chemi-Con electrolytic capacitors. jonnyGURU even posted some pictures of the interior of the CWT made RM850:

http://www.jonnyguru.com/forums/showthread.php?t=10868&page=12

It looks like the new HXi Series with some modifications.

I think this is a very good call. I have said from the beginning that I wouldn't drop it a tier based on this, but I did feel it was a big mistake, given the target market.

 

[FALC0N]

Dottorent sucks. I mean, he has to. Otherwise, why would he go through all the trouble of researching, reading, compiling, scoring and otherwise assigning values to all those power supplies? What is he, nuts? Who would do that? What kind of person pores over details, reviews, specs, more reviews, data sheets, other lists and spends time getting a feel for the general consensus regarding a bunch of stupid power supplies? Anybody that dedicated either has to have a serious case of OCD or is just plain obsessed, right? Right?


C'mon, give the guy a break.
.

I don't disagree with any of this, but I have been very uncomfortable with his responses. Yes, the RM is a contentious issue. But what about the Enermax made LEPA or the Fata1ity, or a few others. I felt his responses were just perplexing at times. I just doesn't feel like he gives different views a fair shake.

I don't know. Maybe I should just let it go. It not a world ending issue.

And for the record Dottorent , I do respect the hard work you put into this, even if I disagree with the outcome.

 

[Ulikedat]

According to HardOCP's review, it didn't. And I read the reviews very carefully and I have some pretty tight tolerances. Just one thing different from the rest can mean I grade it gold.

In the case of the HXi 1000w, it missed Platinum due to 88% Efficiency at 100% load, according to HardOCP's review of the HXi 1000w.

However, this no longer applies to my rules, in which strong evidence is required to justify it's true 80+ efficiency rating, safety of the unit and stability. One review is now no longer enough. It has to be reported in numerous tests from the best of the testers of the internet, not counting reviewers with insufficient equipment.

Maybe I misread the reviews, which is possible. But the HXi is now back as a Platinum unit.

But please, next time don't shout at one mistake. Everyone makes them from time to time.

Thanks alot for this list but i was wondering why the Corsair HXi is in tier 2 and not in tier 1.

 

[LucoTF]

Zebronics should be higher for crossing a zebra with my favourite drink, gin and tonic.

 

[FALC0N]

Zebronics should be higher for crossing a zebra with my favourite drink, gin and tonic.

I am all for this. Sharkoon will need something to drink. Look at him! He's thirsty!

 

[Ulikedat]

Thanks alot for this list but i was wondering why the Corsair HXi is in tier 2 and not in tier 1.

it's not made by seasonic and has the misfortune of being made by CWT, makers of corsair PSU that blow up. So it would be impossible for CWT to make an tier 1 PSU.

Could you direct me to a source?

Cheers,

U

 

[Darkbreeze]

Just ignore that comment, he was in a mood, or something. I'm not really sure why that unit is in Tier 2 since it scored 10's for every category, on every review I have seen, except value, due to it's being overpriced compared to similar units, which really shouldn't be a factor in ranking. I'm sure Dottorent had a reason for placing it there, but aside from it being Corsair (Which I dislike as a matter of course when it comes to their PSUs) I'm not sure what it is.

 

[FALC0N]

Just ignore that comment, he was in a mood, or something. I'm not really sure why that unit is in Tier 2 since it scored 10's for every category, on every review I have seen, except value, due to it's being overpriced compared to similar units, which really shouldn't be a factor in ranking. I'm sure Dottorent had a reason for placing it there, but aside from it being Corsair (Which I dislike as a matter of course when it comes to their PSUs) I'm not sure what it is.

I do wonder if it being Corsair is the reason. I have observed a palpable anti-Corsair bias over the last few years from most posters. I was one for a while, but I just felt that it became too much. Some of it is justified, but it just goes too far. I think people feel betrayed because Corsair PSU's were just out of this world good those first 5 years, and now they are like the rest of the pack.

 

[Darkbreeze]

Actually, I think it's due to the fact that their budget units have a high rate of issues when used in systems with gaming cards requiring supplemental power or are overclocked. And the fact that they are too proud of all their units that ARE worthwhile. (Reads as, overpriced.) Clearly their good units are good, but they're budget units and even some of their higher end products, regardless of reviews or where the internal hardware is made, just seem to have a lot of recurrent issues.

 

[FALC0N]

Actually, I think it's due to the fact that their budget units have a high rate of issues when used in systems with gaming cards requiring supplemental power or are overclocked. And the fact that they are too proud of all their units that ARE worthwhile. (Reads as, overpriced.) Clearly their good units are good, but they're budget units, regardless of reviews or where the internal hardware is made, just seem to have a lot of recurrent issues.

I think we are expressing the same idea in different form. In 2009, you could buy ANY Corsair and it would hold up to just about anything you threw at it, provided it was in the proper power envelope. That included the first CX, the CX400, which was S12II based. Now jump ahead a few years to 2012, and the user accustomed to Corsair being a sure thing, buys a CX500 in an overclocked gaming rig, thinking it's old Corsair reliable, and it blows up on him. Probably not the PSU's fault, as it wasn't really designed with that kind of load in mind. But it has the feel of a trusted friend stabbing you in the back. Corsair established that standard of excellence, and then backed off in a way that felt duplicitous to many.

 

[dottorrent]

HXi Moved to Tier 1. No idea why I placed it in Tier 2.

 

[Karadjgne]

But that's the gimmick now, isn't it, the power envelope. You'd think that a 500w unit with all the standard connectors including 2x 75w 6-pin connectors would be able to perform power-wise the same as any other 500w unit designed with 2x 75w 6-pin connectors and not, in reality, be an over estimated 400w unit.

So it ends up with no minimum standards, no regulations and no accountability. Truly a case of Buyer Beware.

 

[FALC0N]

I think you are thinking of the VX450 on the seasonic S12-II platform, an excellent budget PSU back in the day. The corsair CX430 pails in comparison.

The CX400, not the 430. The VX was S12II based as well, but the cx400 was their lowest end model at the time.

 

[Deleted member 217926]

I think you are thinking of the VX450 on the seasonic S12-II platform, an excellent budget PSU back in the day. The corsair CX430 pails in comparison.

The CX400, not the 430. The VX was S12II based as well, but the cx400 was their lowest end model at the time.

^ This. I put one in a family members build. It died.

 

[Onus]

As this conversation gets as toasty as a questionable PSU, please try to maintain an even strain. Someone with horrible personal experience may not be swayed by statistics, and the reverse is also true. Neither is necessarily wrong.

 

[Darkbreeze]

HXi Moved to Tier 1. No idea why I placed it in Tier 2.

Probably the right move in this case.

 

[dottorrent]

Yeah, had a feeling it would eventually bite back.

 

[Darkbreeze]

Another Photon review, awaiting placement in Tier 3 next to the 750w?

http://www.anandtech.com/show/9084/rosewill-photon-1050w-review

 

[Darkbreeze]

Not sure if this is the same unit as the Dark Power Pro P10, as it's indicated as only the Dark Power Pro 10, no P.

http://www.anandtech.com/show/8952/be-quiet-dark-power-pro-10-650w-850w-psu-review

 

[dottorrent]

I don't count Anandtech reviews. They're decent, but not detailed.

 

[Darkbreeze]

Ok, now I've got to disagree. How is this not in depth testing. Seems like it to me, but maybe I'm wrong:

For the testing of PSUs we are using high precision electronic loads with a maximum power draw of 2700 Watts, a Rigol DS5042M 40 MHz oscilloscope, an Extech 380803 power analyzer, two high precision UNI-T UT-325 digital thermometers, an Extech HD600 SPL meter, a self-designed hotbox and various other bits and parts.