Sweet, finally a quiet power supply from seasonic with modular cables. It looks good and performs very well. Now if only they made a 430W version. I don't like to pay for stuff I don't use.

It will be nice to recommend a different power supply than seasonic. I was getting tired of recommending them to the new guys. Now I can say Corsair and even though they still get a seasonic, I don't have to keep saying it.

So when will Antec rebrand their cases so I don't have to keep recommending the P150, Solo, P180, NSK2400, and NSK3300?

In the UK, these PSU's are not cheap, and are overkill for a large number of SPCR systems. Yes, if you are a gamer w/ an X1900XT or SLI then they are appropriate but for low-power systems Seasonic S12 is still unbeatable from a value/quality perspective. Is there a way to tell for certain if these PSU's are made by Seasonic?

Looks like some nice PSUs..

Typo patrol trikes again, the last bit in the first paragraph of "temperature & cooling" says These are cooling running PSUs.

Aside from the fact that they look and sound identical, you can confirm that it is Seasonic made by the three letter code listed on the main transformer. VRL means Seasonic.

Greetings,

I'll echo what others have said: very nice, but very expensive, and too powerful for almost all systems. Sell a 300-430watt unit(s) for under $100US, and then that would be a unit I would buy.

Thanks for this nice review, I was looking at these models since they were out. Now I will buy the 620W with confidence for my next setup!

I believe that Seasonic made them on purpose 2 db louder than their S12E+ ones, so they don't share the same market.
They seem to have no extra vents that let heat return into the case, one thing that I don't like in my S12E+650.

I doubt that.

The voltage to the fan at startup was about the same (3.8V), and while we measured slightly lower SPL with some of the S12s, you have to remember that the ambient of our lab also changes a bit. When it is a bit noisier, our overall measurements tend to go up. (And the reverse is also true.) A 1-2 dBA difference is not significant, really.

The fan for the S12-550 Energy+ was a 0.33A model, compared to the 0.37A for the 650e+ and the Corsairs. BUT we did not hear or measure any significant difference between the 550 and 650 at default. This is about the only area where the 550e+ might measure better, but it's really not worth sweating about. It's not significant.

I had my doubts about Corsair. I was pleasently suprised. It's a modular seasonic, that doesn't exhaust the heat back into the case. The 5 years warranty is just a topping on the cake.

Looks like nice PSUs, if very familiar.

Any reason only the HX620W has ~20w data and not the HX520W? I would like to see ~20w data from more PSUs myself.

wow, sounds like a great PSU. definitely seems like a good alternative for the ol' seasonic S12, and being a corsair it should be easier to find in more countries that dont have much in the way of seasonic.

Guess again

IMHO a 20W DC output data point would not be useful; so few (if any) non-laptop systems can achieve such an incredibly low power draw, and those that do are probably using a DC-DC board + an AC-DC brick; conventional switching power supplies are very inefficient at such low loads.

This Corsair is priced very "competitively" (at least here), our biggest e-tailer has the 520W version for 139€ compared to 159€ for the Seasonic M12 500W.

The link to the power supply testing platform on page for points to v3, but says v4. Looks like a couple of nice power supplies, but too bad they're rated so high.

Then why provide it for one PSU and not the other?

The 620 was tested some days after the 520; the 20W test point was added on a lark. We may keep it in just for the heck of it. It's such a simple test because the settings are just one switch per line. It's only relevant for very low power PSUs, tho.

Would a 520W always draw more power and produce more heat than a 430W, even at the same load? If not, the HX520 looks real nice: not crazy expensive, modular cables, quiet, and fairly future-proof (for 'mainstream' systems).

Not always.

It depends on the individual PSUs' efficiency curves and the power demand of your system. If your system mostly draws <100W, then a similarly high efficiency but lower max power PSU will generally give you lower real power usage. Such as, for example, the Seasonic SS-300SFD, which is >80% even as low as 40W load. But not the Seasonic S12-330, which has an efficiency curve much like the Corsair, not reaching 80% efficiency till close to 150W.

For what this means in real world terms (watts, not percentages), let's assume 70W as idle power draw -- fairly typical for a modern system with a decent but not major gaming card.

Enermax EG701AX -- ~72% eff -- 97W
Corsair 520 -- ~75% eff -- 93W
S12-330 -- ~75% eff -- 93W
FSP Zen -- ~81% eff -- 86W
SS-300SFD -- ~84% eff -- 83W
PicoPSU -- ~85% -- 82W

The Enermax is not high eff by today's standards, hitting a peak of only ~78.5%. The last two have the best low power efficiency of all PSUs tested.

The worst and best eff. (at 70W) are 72% vs 85%. It looks really significant, yet the actual power difference is just 15W. Swapping a 60W incandescent with a 15~17W high efficiency bulb will save you a lot more energy with much less expense and trouble.

That assumes your PC will be at idle most of the time. Mine is. I know from monitoring my system AC over the years (now that I no longer do any distributed computing).

My take on all this is....

1) If you can match the PSU to your system needs, and get a PSU that's high efficiency in the range where your system draws most of the time, go for it.
2) Still, make sure you have at least 25% headroom beyond the maximum power draw of the system
3) Don't go crazy trying to get the highest efficiency -- consider the real WATTS, not just the percentage.

One idea I have is to request a special SPCR variant of the SS300SFD 80+from Seasonic:

1) same quiet 80mm fan used in Antec NeoHE
2) longer sheathed cables, perhaps a couple more than is supplied, and one 6-pin PCIe power plug & a 20/24-pin ATX plug
3) ATX mounting adapter included

Alternatively, the SS300SFD 80+ circuitry in a standard S12 casing...

please Seasonic, pretty please... these would sell like hotcakes among the SPCR faithful (depending on price, obviously).

20W DC isn’t that unachievable. With a system using the following setup I managed less:
Core Solo 1.83 GHz, 1GB DDR2-533, Gigabyte S479.

120GB 2.5â€

That truly is stunningly low SC; was the Core Solo undervolted at all? It does make you wonder why the idle power consumption of the C2D is so high (26W) relative to the previous Core generation. It goes without saying that at 16W system power consumption true passive operation is a piece of cake.

Single core VS Dual core?

This too would be my dream PSU. The fan in my SS-300SFD was really clicky. Plus it's a tight fit in the SFF case. If you stuck the board in a standard ATX case, you could fit larger more widely spaced fins. I thought about trying it myself as I have plenty of old PSUs that I could steal the case and throw out the crappy parts. But getting the parts out of the SS-300SFD would not be so easy. Looks like some were assembled after it was already mounted in it's case. I haven't felt that adventurous yet.


Looks like one of the 240W ones is some custom PSU that will only fit a specific case. The other looks like BTX. The 300W one they have listed is ATX.

Yes, it was set to 0.95V with RMClock.

The extra core certainly doesn’t seem to help. I compared a Core Solo versus a Core Duo and even when only using one core the Core Duo is penalised. Even a Celeron M at stock voltage consumes less power than an under-volted Core Duo when only one core is under load.
I think this will be addressed in future generations when they introduce separate frequencies per core and possibly even individual VCores. AMD are introducing the first feature next year, I don’t what Intel’s plans are in this area.

Greetings

I have been considering these supplies for a while, but I have one question prior to purchase.

I am looking to build a 775 system, initially with an Allendale or Conroe, but later to swap out the cpu for a Kentsfield.

Do you think the 520 would have enough grunt for the Kentsfield variant of this build, bearing in mind we are talking 4 cores at 100% load 24/7 running Folding@home?

If the supply markings colour schemes were the other way round I'd get the 620 without a thought (I have an aversion to the colour red, and yellow is my favourite colour )

At the rigs fullest config I would envisage:

Board and Kentsfield.
2 gig o ram.
2 x 120mm case fans (possibly a third case fan)
2 x sata hard discs.
DVD writer and possibly a second rom drive.
Upper mid gpu...eg: Radeon 1900xt.

My main concern is 24/7 full load stability for the Kentsfield.

No problem, you could run two quad core Xeons with one of those.

If you are folding, might as well do it with your ATI gpu... It should be quite a lot faster than even a Kentsfield. I think simultaneous CPU and GPU folding was not possible, but i might be wrong.

Hi

I already have a 1900xt folding in my AMD X2 rig.......not particularly interested in more gpu folding at this time.

The 1900 listed in my original post was just an example of the top spec I'd be looking at.