OK......I'll reask the question. Since this power supply is simply an FSP-Epsilon unit, and the FSP Epsilon has a well known issue with out-of-spec ripple under loads, how did you manage to only get 18mV ripple in the +12V rail at full 600W loads?

That sort of ripple, what you referred to as "modest", is in fact....if the measurement is accurate...outstanding ripple control of the sort found in PCPower & Cooling Turbo-Cool power supplies, which are probably the absolute best at controlling ripple on the market.

Instead, you have a power supply whose maker has a well documented history of being unable to control +12V ripple under serious loads. You can see out of spec ripple in their own supplies and other rebranded Epsilons, such as the OCZ GameXstream. I've seen ripple hitting over 150mV at full load.....so how did you get only 18mV on this one?

Just curious.....is your testing equipment in good order or your settings set properly or are you using an incorrect capacitance in your tester, or are you actually not recording/testing for ripple at full load as you indicate? It'd be almost impossible for this particular power supply, esp. at this price point and from this particular manufacturer, to achieve the low, almost non-existent, ripple value you represented in your testing.

I don't know if SPCR's testing is above reproach, if your comments hold water, or some combination of the two. Let me summarize the testing SPCR has done by date of review:

You say 18.x is good but only one power supply in that table did worse (the Tagan EasyCon). Do you really think almost last place is that out of line compared to what you expected?

OK, lets see if sorting by load gives us any insight:

Hmm, higher loads tend to have higher ripple. Though the Ultra X-Pro stands out as a very good performer on the top end.

Finally sorted by 12v ripple:


I'm not sure that this sort gives any new insight. We see the Seasonic relationship easier between the Corsair and Antec but most of us knew that already...

Those numbers are too small. Take a look at jonny's ripple measurements, they seems more realistic.

OK, jonnyguru on the HX520 "overall ripple was not more than .05V (50mV.)" which seems kind of vague, he is rounding because it isn't an exact measurement? So the SPCR number is what one third the jonnyguru number? 14.1 vs 50?

I don't see a Nexus power supply review on jonnyguru to compare to. I did however find the thread where jonny said he didn't know why the numbers are different and gave a couple of guesses. Can you point me to a review on his site that is supposed to be the same power supply as the Nexus NX-8060?

Nice review, as always. SPCR never fails to fully review a product, and give you the kind of info you really need to build a silent system.

One small request though. Would it be possible to test PSUs with a real system, in idle load and S3 sleep states?

The reason I ask is that it would be helpful to see how a PSU performs and sounds in an actual case (say a popular one like the NSK-6000) compared to other PSUs, with other popular silent components like a Sythe Ninja and recommended quiet HDD. Also, the S3 state power draw and efficiency stats would be handy. Recent issues with whining power circuits in S3 state could be eliminated too.

I'd just like to add another vote to the "should have reviewed a different PSU in the series" movement.

Any of the 8040, 8050, 8060 are sufficient to power any sane system, yet alone a quiet one. The 8040 is the cheapest, the 8050 is (according to Nexus) the quietest. Given these facts, no matter what your priorities (noise or cost) you wouldn't chose the 8060.

I'd have done a quick test - is the 8050 noticably quieter than the 8040 at 200W? - and then chosen which of those two to review based on this.

while we are at nexus-tec

anyone planning on reviewing there newest case

http://www.nexustek.nl/morpho_premium_a ... c_case.htm

(some odd thing: why use a push pull at the bottom of the case,
why not just two push and at the very top a pull by the psu?)

Getting a lot of info from SPCR I built a new quiet system. The PSU was a toss-up between the NX-8050 and the HX520. I could only find one review that pitted the two against each other and I had to have Google translate it. The summary was that both are very good but that the Nexus was quieter (I'll post the link when I find them again).

Ultimately the Nexus was "somewhat" quieter than the Corsair unit. I also liked the fact that Nexus claims that of the three 80X0 PSUs, the 8050 is the quietest and doesn't start ramping up the fan until about 300w. My system idles at around 215w and has peaked at 308w from the wall.

The end result is that I am extremely pleased with the sound level of the 8050 in my Antec P182 case. If anyone is wondering, I do have a Noctua 120mm 800 RPM fan blowing in the lower HD/PSU chamber which I'm sure helps.

My only concern is that the power ratings are not as solid as I would like. Currently, at idle I am getting 12.95 on the 12v line, 5.54v on the 5v, and a flat 3.30 on the 3.3v.

If I had to do it again, I'd still call it a toss-up. I certainly like the Nexus sound level (or lack-thereof) but would prefer the Seasonic build of the Corsair as that seems to be rock solid. Definitely more head-to-head comparisons are needed here.

Found the review:

The review from the begining.

Or just skip to the conclusion:

if this is a software reading it's unlikely to be accurate. check with multimeter first.

Oh really? I didn't know that. I used Everest Home Edition 2.20. I don't have a multi-meter. So maybe my PSU is "ok" after all?

Is it better to look at the 8050 than the 8040 then? Does the Nexus come with 6/8pin PCIe?

Thanks,

Mike

As is often the case, I repeat myself:


The noise/load data Nexus provides is clearly NOT at normal system temperature, but at some very low (<25C) room temperature. The SPCR test shows how utterly irrelevant this is when the working ambient temperature is higher, more realistic.



What this indicates is that whatever the differences in starting fan speed, in a real system under a normal conditions, all three of these Nexus PSUs would ramp up to a fast enough speed that noise/speed differences would be erased.

Sorry Mike, but Confuciius was quoting the Nexus is "quieter" than the Corsair HX so that goes completely against your review. I was trying to get to the bottom of it thats all.

Hmmm

He was quoting a review, not speaking of his own experience. I see nothing in that review that suggests any kind of simulated real conditions noise testing. Looks like they just turned them both on w/o load on the testbench to listen.

His Nexus 8050 is in a P180, which is the very best setting for it. It should stay quieter in that case than in any other because its working environment will stay cooler.

Yes, just to clarify, I have only used a Nexus NX-8050, I have not had my hands on a Corsair HX520 (although that was my "second" choice). That one review, translated from German by Google, was the only head to head review I could find.

Also, as Mike pointed out, my PSU is in a P182 with a mid-case fan which is the optimal cooling environment for any PSU. So the Nexus' fan is probably running at the bare minimum if it is temperature controlled. Hence it appears to perform nice and quiet for me!

Ok thanks, I would be putting the PSU in my P180 so I guess (and hope) I would get similar results.

Cheers,

Mike

Ey Mike, as usual, could you please tell me if the fan of this unit is plugged or welded to the pcb?

Thanks!

Soldered as described in the review. Pls read the article!

As can be read in the article:

The fan is one of Adda's, which has become popular among quiet PSU makers. It's from the same series used in Seasonic and Corsair PSUs, for example. The fan is soldered directly to terminals on the main PCB.

You are right Mike, sorry about that.
It seems that after 8 hours reading one doesn't know what he reads

Nexus makes one of the most respected (on SPCR anyway) and well performing fans there is .. yet they use some ADDA fan in their PSU? Makes me wonder if the Nexus fan is actually worse in terms of dB(A) to air-flow rate compared to the ADDA.

Same here... Really strange when you think about it... you'd think they'd try to improve their fans based on the ADDAs. Or maybe the ADDAs are just so so cheap compared to theirs that it's not worth the extra money.

No, that's not it. The Addas used in PSUs are invariably ball-bearing variants which survive the heat better than sleeve bearings, apparently. All the Nexus and Yate Loons we like are sleeve bearings. Besides, the Nexus fan just doesn't have enough speed/flow capability to keep the PSU cool at high power, and that's needed for it to pass safety testing.

Good point MikeC.

Actually this is very misleading : according to their website (http://www.nexustechnologyusa.com), the 8060's fan is a "Real Silent 12cm fan" (http://www.nexustechnologyusa.com/c/ntu ... 060_1.html)
And "Real Silent" is just the name of their fans (http://www.nexustechnologyusa.com/c/ntu ... of_20.html). Note the capital letters in Real Silent.
So I was kind of expecting one of their fans to be in there.


Therefore I'm guessing that ABBA is Nexus' fan manufacturer... right ? Or that they are just using the name of a good fan to advertise their PSUs.
Or maybe it's just a typo and I'm a bit paranoid

The retail box Nexus "real silent fan" is a modified Yate Loon w/ sleeve bearings. It may also be cherry picked from the assembly line, but that's not necessarily true.

Quite the opposite, I'd say.
The decibel scale isn't linear, but logarithmic. The zero-level is therefore at negative infinity, not at 0 dB.

Just for fun I've recalculated the data to a linear scale, with the (filtered) sound intensity reaching from 56 (new design at min power) up to 3,631 pikowatt per square meter and the x-axis at 0. (0dB=1pW/m2, 30dB=1,000pW/m2)
At max load the old design is rated as 3,631 whereas the new design is only rated at 977, quite a visible difference.



Cheers
Olle

Did you measure with a powermeter or by software? If by software then those readings might be off somewhat...

Hey if we can hook 'em in with high power PSU reviews, then show them the light, and send them off running pico's with suspended laptop drives, its all good!