PSU Tier List [OLD]

[vanished]

20 minutes ago, STRMfrmXMN said:

Software readouts of voltage suck. Unless it's Corsair Link or a Legacy/EFI tied software then it's gonna be inaccurate and, if memory serves me, that's AIDA64 which is inaccurate. 

This is HWMonitor, but thanks  

[STRMfrmXMN]

1 minute ago, Ryan_Vickers said:

This is HWMonitor, but thanks  

Oh yeah, HWMonitor, super inaccurate, don't trust it for voltages.

[quan289]

8 hours ago, Ryan_Vickers said:

This is HWMonitor, but thanks  

Just downloaded HWMonitor to check my voltage reading, and it says my RM650x's 12V rail is at 0.048V... lol At least AIDA64 have given me a more reasonable reading of 12.096V.

 

If the software isn't compatible with your motherboard and not getting the correct readout, the voltage reading is simply a random number generator as shown above.  Even if it getting it from the correct readout, the IC that these software is getting the information from is not reading it the point where the PSU connects to the MB in which the overall MB design will affect the readings, and that's not considering the accuracy / calibration of the sensors themselves. So if one software is inaccurate, then any other software would be inaccurate as well (including the BIOS and the manufacturer's software like Asus Suite), as they should be getting the readout from the same IC.

 

If you have dealt with Asus MBs, you probably have set off the "Asus Anti-Surge Protection" which is design to monitor the readout of the voltages of the rails. The protection is so finicky, that it's notorious for giving false reports and shutting your system down especially during sudden changes in loads which happens a lot in gaming computers.

 

If you want to have an accurate reading, it is best to check it using a digital multimeter or get a PSU that allows you to monitor the voltages right at the source to minimize any variables that may skew the readings (Corsair Link, Enermax ZDPMS, Thermaltake DigiApp, etc).

[vanished]

4 hours ago, quan289 said:

If you have dealt with Asus MBs, you probably have set off the "Asus Anti-Surge Protection" which is design to monitor the readout of the voltages of the rails. The protection is so finicky, that it's notorious for giving false reports and shutting your system down especially during sudden changes in loads which happens a lot in gaming computers.

I did have that go off once on two different computers during a momentary powerloss during a thunderstorm.  That was before I had my UPS

[ApolloFury]

How come the Newton R3 is ahead of the Cooler Mater V-Series when the V-Series are 100% Japanese and a rebrand of Seasonic while the Newton uses a mixture of Taiwanese/Japanese?

[STRMfrmXMN]

1 hour ago, ApolloFury said:

How come the Newton R3 is ahead of the Cooler Mater V-Series when the V-Series are 100% Japanese and a rebrand of Seasonic while the Newton uses a mixture of Taiwanese/Japanese?

Newtons are an ATNG design, not Seasonic.

[Bananasplit_00]

awesome  i always have a tab with the whitelist open for suggesting first timers PSUs and to know if the PSUs i have are not super shit  

[ApolloFury]

21 hours ago, STRMfrmXMN said:

Newtons are an ATNG design, not Seasonic.

I didn't say Newtons are Seasonic, I said the V-Series are Seasonic which are 100% Japanese while ATNG is a combination of Taiwanese and Japanese.

[Mr.Meerkat]

1 hour ago, ApolloFury said:

I didn't say Newtons are Seasonic, I said the V-Series are Seasonic which are 100% Japanese while ATNG is a combination of Taiwanese and Japanese.

http://www.kitguru.net/components/power-supplies/zardon/fractal-design-newton-r3-1000w-psu-review/

 

I don't see a problem with the Newtons but in a way, I guess you're correct in stating why the Newtons are T1 when it's probably closer to being a T2 unit  (teapo caps aren't that Chinesum you know ). 

[STRMfrmXMN]

2 hours ago, ApolloFury said:

I didn't say Newtons are Seasonic, I said the V-Series are Seasonic which are 100% Japanese while ATNG is a combination of Taiwanese and Japanese.

I mean, caps are not the be-all and end-all for quality tiering, 

 

9 minutes ago, Mr.Meerkat said:

http://www.kitguru.net/components/power-supplies/zardon/fractal-design-newton-r3-1000w-psu-review/

 

I don't see a problem with the Newtons but in a way, I guess you're correct in stating why the Newtons are T1 when it's probably closer to being a T2 unit  (teapo caps aren't that Chinesum you know ). 

but should I move the Newton?

[Mr.Meerkat]

6 minutes ago, STRMfrmXMN said:

but should I move the Newton?

Your call, I feel it's probably fine as a T1 unit  (just at the very bottom of T1) 

[STRMfrmXMN]

Just now, Mr.Meerkat said:

Your call, I feel it's probably fine as a T1 unit  (just at the very bottom of T1) 

I feel like it's fine to keep it there.

[Mr.Meerkat]

1 minute ago, STRMfrmXMN said:

I feel like it's fine to keep it there.

First world problems with tier-ed lists...personal opinions and judgement become a factor  (but yeah, it's probably fine)

[STRMfrmXMN]

Just now, Mr.Meerkat said:

First world problems with tier-ed lists...personal opinions and judgement become a factor  (but yeah, it's probably fine)

I do it based on performance and quality and although I don't know how subjective it is I'd say it's a pretty damn good unit. There are different versions of the unit that might hit tier 2, might all be tier 1, haven't looked into each individual unit.

[Space Reptile]

On 19.9.2016 at 5:41 AM, STRMfrmXMN said:

Oh yeah, HWMonitor, super inaccurate, don't trust it for voltages.

i mean i had -14 volts on my +5 in that program , im p shure its innacurate  

[CAPCOM UK]

Silverstone Strider Plus 1000 watt and above models need to be Tier 2 these are amazing PSU builds.

[Mr.Meerkat]

4 minutes ago, CAPCOM UK said:

Silverstone Strider Plus 1000 watt and above models need to be Tier 2 these are amazing PSU builds.

Hmmmm...I think I'll disagree  

[CAPCOM UK]

Just now, Mr.Meerkat said:

Hmmmm...I think I'll disagree  

https://www.techpowerup.com/forums/threads/psu-longevity.226061/page-2

 

A PSU than can last over 10 years.

[CAPCOM UK]

Since OP is classifying all units in Tier 4 as uncpable and only suited to a GTX 950, he has placed 1000watts + units in the same Tier that are above and beyond a lot of those in Tier 2.

 

Silverstone's SST-ST1000-P being one

 

"I believe we can all agree that a complex device is as strong as the weakest component that comprises it.

In this case, Enhance Electronics and SIlverstone opted for a dual ball-bearing fan, the most resilient kind, which easily serves for 10+ years 24/7 at full blast with the hub kept at 30°C. This figure is a common denominator between various manufacturers, and I can confirm it's true for the likes of Delta, Sanyo Denki, AVC, EBM Papst, Sunon and other such manufacturers whose products are regularly found in workstations, servers and industrial equipment. I've had a lot of hands-on experience with those over the years, and can personally attest that the figure is very realistic. So the fan should check out; YLTC may not be the best ever, but they're good enough for 10 years with "normal" use (i.e. not full load 24/7).

Semiconductor parts are also fine as long as they stay cool. The bridge is a Vishay part - among the best ones, MOSFETs on the primary are Infineon (hard to get better ones), and the secondary rectifier diodes are ST Micro - not the best ever, but will easily last 15+ years if kept under 90°C. Secondary MOSFETs for 12V are International Rectifier parts, also very good and very durable. So the power delivery silicon won't be the first to go. Regular and zener diodes used throughout are of least concern, as they're not power-carrying parts. Instead they're used for signaling and so will typically long outlast your interest in the PSU.

Enhance isn't one to skimp on the passives and magnetics, so the coils, transformer, X/Y caps and the resistors should last longer than the paint job on the PSU case 

I've seen the integrated 5VSB PWM+FET solutions go up in flames, and typically that happens when the input filtering is lacking (on this Strider it very much isn't) and there are numerous spikes in the power grid and/or due to poor placement and thus overheating. Mostly this happens on cheap lower-midrange designs, where the component density is relatively high, but component quality is kept low for pricing reasons. Sometimes one of the smaller critical timing/decoupling capacitors will go bad (dry out) which will push the 5VSB IC into high duty cycle making it blow itself up. The platform used in this PSU is rather far from a cheap skimped-out one, so this shouldn't be too much of a concern. Besides, if the 5VSB blows, the PSU won't work anyway and thus won't be in a position to kill your components...buuuut it may take some motherboard ICs with it depending on too many factors. Not a lot you can do about this one, but you shouldn't worry either. It's the unlucky lottery; almost as if you walked the streets worrying about a meteor hitting you in the face.

Controller ICs very rarely go bad and usually as a result of failed 5VSB which powers them, so these shouldn't be a concern, normally.

The primary filter caps tend to have an easy job when paired with a good APFC IC and a complete input filter, so with those being NCC SMQ 30x30, 270uF/420V parts rated at 2k hours @85°C, we should see at least 32k hours effective from the temperature drop alone, plus who knows how many more due to low ripple current (the 2k hours rating is given at 1.94 A of ripple, which is high above what's expected).

This leaves just the secondary filtering caps, which do have to put up with up to 50 mV of ripple (which at their internal resistance translates to about 1.3A of ripple current, worst-case, well within the cap's tolerance). Teapo SCs that are used are known to be "a bit" inconsistent (they may happen to last less than half of what's expected in some cases), which is the reason for my previous comment. Otherwise, they're rated for 3k hours at 105°C, netting around 81k hours expected useful lifetime.

The additional caps on the modular board are from Suncon (Sanyo) which are the best, bar none; and some are NCCs AND their duty is super-easy, meaning they are of no concern whatsoever.

It might come across as the primary cap or the 5VSB primary being more of a time-bomb than the secondary caps, but both are proven to be much more resilient in practice.
I suppose that's due to the primary caps having very good airflow over them and the 5VSB usually being competently designed. Plus the secondary caps' proximity to hot coils and heatsink means they really need to be kept free of dust and other airflow-impeding factors."

[79wjd]

25 minutes ago, CAPCOM UK said:

https://www.techpowerup.com/forums/threads/psu-longevity.226061/page-2

 

A PSU than can last over 10 years.

The stock 400w psu that came with my Centurion 532 case, tens years ago, is still running fine. Just because a PSU lasts doesn't mean it's good. 

[CAPCOM UK]

Just now, djdwosk97 said:

The stock 400w psu that came with my Centurion 532 case, tens years ago, is still running fine. Just because a PSU lasts doesn't mean it's good. 

100% modular cables

Efficiency 85%-88% at 20%-100% loading

Class-leading single +12V rail

Strict ±3% voltage regulation and low ripple & noise

Japanese main capacitors

Silent running 135mm fan

PCI-E 8pin and PCI-E 6pin connectors support

Support ATX 12V 2.3 & EPS 12V

Active PFC

24hour continuous power output with 40°C operating temperature

 

 

Now, there are some weird looking scope shots up there. What we're measuring here is the turn on spike present when any SMPS design gets itself going. The 5VSB shot is only average, with a 5V spike. The ATX specification allows for that rail to hit 5.5V, so we're still in spec here.

The story changes rather dramatically for the 12V side of things. Not only is the spike more of a shelf than a spike, it also happens to be the lowest 12V spike I have ever measured on any unit. THREE VOLTS!!! That is so far in spec you could drive twelve big rigs side by side between that reading and the ATX spec for that rail, which is 13.2 volts. Simply extraordinary. Well done, Silverstone and Enhance.

 

 

 

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story5&reid=180

[Mr.Meerkat]

6 minutes ago, CAPCOM UK said:

100% modular cables

Efficiency 85%-88% at 20%-100% loading

Class-leading single +12V rail

Strict ±3% voltage regulation and low ripple & noise

Japanese main capacitors

Silent running 135mm fan

PCI-E 8pin and PCI-E 6pin connectors support

Support ATX 12V 2.3 & EPS 12V

Active PFC

24hour continuous power output with 40°C operating temperature

Is this a silverstone broche?  

 

On a side note, so what? It's a decent PSU but it's not that good either, it's nothing to brag about. 

[CAPCOM UK]

Just now, Mr.Meerkat said:

Is this a silverstone broche?  

 

On a side note, so what? It's a decent PSU but it's not that good either, it's nothing to brag about. 

What makes a good PSU?

 

It seems you and a lot of others here are a bit blinded bu 80 Plus numbers and fancy caps.

[79wjd]

33 minutes ago, CAPCOM UK said:

What makes a good PSU?

 

It seems you and a lot of others here are a bit blinded bu 80 Plus numbers and fancy caps.

No one who knows anything about PSU quality will care about the efficiency rating in terms of determining quality. And Caps/ripple (and whatever else I dont really understand about) are both very important metrics in determining quality. 

[CAPCOM UK]

1 minute ago, djdwosk97 said:

No one who knows anything about PSU quality will care about the efficiency rating. And Caps/ripple are both very important metrics in determining quality. 

Yup which that unit delivers fine on, the minor rails are within spec but the main 12 volt is stupidly stable.

So that makes it Tier 4 and only able to power a GTX 950?

 

Most dumbest Tier list i have ever seen.