Help! I'm being repressed!

On second thought lets NOT go to Camelot, tis a silly place.

The grail was a good addition. But after getting into the meat of the article, there was too much good stuff that it got forgotten.

The article says that the fan kicked in after about 30 minutes of 300W. In addition, there was a high pitched whistle noise, both in standby mode and during normal operation.

On an open bench top -- not in an extreme hotbox, where it would probably kick in at 100~150W. Note the user report I cited earlier -- once the fan starts, it never stops.

LMAO!!!!!!



BTW, I was rather hoping to see a pico psu in this roundup. Like a DIY fanless solution. A few people on the forums seem to get very good results combined with a Dell DA-2. You could wire up a PCIe 6 pin, a 4 pin for the CPU, and then 12v into the pico to power the rest of the PC. Providing you stick to 65W TDP CPUs and graphics cards with 1 aux power connector, you will not overload the pico or a DA-2 and can make a very decent spec PC with that power budget. Should be extremely efficient too. An AMD 5770 would be a good candidate since it has been measured to use at most 75W under load. Maybe it gets 95%+ of the power from the 6 pin aux power input making it a perfect choice for this setup.

On the contrary, a picopsu has no place in this torture test. Yes, it is fanless but it is only half of a psu. The other half, which creates most of the heat, is your own choice, not an intrinsic part of the picopsu -- any number of ac/dc adapters can be used -- and it is not subject to the heat of the PC's, being outside the case. Hence, the torture test is simply not relevant for the picopsu.

Well, then both companies (FSP and Etasis) are sourcing the chassis and the heatsink in the same place? Do the inside components look different?

Neither was opened up -- but I think the two companies make their all own stuff, they are both manufacturers (not just marketing brands like Antec or Corsair). But you could be right, the casing looks very similar. Will see if Sliverstone can clarify -- maybe they made a supplier change to FSP.

I'm still on the hunt for obscure fanless PSUs that haven't been mentioned yet!

Deviating from the ATX standard there has been a Shuttle fanless FlexATX PSU.

FSP also make 120W and 150W fanless FlexATX PSUs:
http://www.fsp-group.com.tw/english/1_p ... &proid=559
http://www.fsp-group.com.tw/english/1_p ... &proid=558
Both are listed as being 85% efficient so should be modern designs and I can certainly see these as being workable within a lower powered Shuttle with enough airflow through the case.

There's a frustrating number of older 'semi-fanless' PSUs which call themselves fanless. Nowadays we're seeing more high end PSUs which don't start their fans until needed and that's all that some of these do. At what level these start at, how loud the fans are and whether they operate stabily is still debatable.

"80mm spare fan for safety" - Is it spare!?! If it's for safety surely it isn't spare.

Thank you for this article.

I think you've set the bar pretty high for a holy grail / silent pc with 200W sustained usage; I think a modest 60W sustained power is a good target for a holy grail PC (acheivable with modern CPU/video card).

I will look forward to more "torture test"/holy grail articles, perhaps for motherboards running in ~40-45C torture temperature (or higher).

Compare the Silverstone/Silentmaxx with Fortron, look at the design and the hole pattern of the heatsink.
You can't call that a coincidence.

Aurthur...


AURTHUR!!!!11

I stand corrected, Neil, you are right. Silverstone confirmed it is a FSP 400. I was fooled by the different coloring. I was also informed of immediately pending price change to $130-140 and that the warranty is 3 years. Changes to be made in article soon. [/u]

Great article! Thank you very much!

As always - a great review. Even tho I'll probably never have use for any of these PSUs it's always good to know

Someone mentioned the NorthQ as another fanless ATX PSU.

Well, there's one more. Here's the gist.

A good few months ago the Polish PC geek community first caught glimpses of a 600W fanless PSU made by a Polish company - AZO Digital Silent Power 600W. Some time later the PSU hit the shelves and was recently "tested" by a PC enthusiast website. The PSU is certainly not a copy of any other PSU, fanned or fanless; features a patented DC/DC and SSR converter (don't ask...)

Here's the test linky - click

Major features: 600W, 0 dB, ATX 12C 2.2, 92% efficiency (as measured/claimed by the manufacturer - no certificates, mind you), 4 12V rails, 2 x PCI-E 6pin, APFC...

To cut the long story short, the PSU proved to be a moderate failure. Biggest quibbles, confirmed by several posts by actual users, are: parts of the test PC blown by the PSU (along with the oscilloscope), chirping & ticking noises under typical (100-200W AC) load, average efficiency, high price. Two PSUs were tested, both switched off under moderate load (see test PC) and required heavy additional cooling to operate for longer than 2 hrs. The PSU scored a little over 80% efficiency, loosing significantly to Enermax Pro87+ 500W Gold. The PSU was not tested with a PSU testing contraption to actually reach full load (instead of, say, 300W). If this was the case, we would probably see lotsa fire.


All hail Seasonic FL!

EDIT
Come to think of various stuff, I'd like to know what tests have they run in the factory to check the PSU. And what one has to do to actually hurt* a Seasonic FL
* not just shut it down, I mean - huRRRRt

Great article, I had the SilentMaxx350, well, actually two over 3 years. The first one died (silently and was replaced by a second from the company that eventually started producing black smoke when one of the ceramic capacitors burned through its 105C pvc sleeve (surprising, it still worked while smoking).

So I moved onto pico-psus; finally settling on the 150W. Was recently thinking about wiring up two of these together to get some more power (trouble is it would have two bricks) but now the Seasonic X-400 gives me another choice.

But neither my desktop or htpc case has a vent which would allow the X-400 to dissipate up heat. One case has the psu at the top rear but does not have the top vent. The other case has its psu sideways at the bottom, my guess is that the X-400 may not work in either case. Can SPCR test the x-400 for different installation scenarios?

Alternatively, can anyone recommend cases where the X-400 would work?

Another great review. Thanx Mike.

Like the X400 review, this one is yet another indication of the level of quality and careful attention to detail from Seasonic. They are the only ones that were not willing to release a fan-less model that fails under full load. Until now, when they know they can do it.

Thank you for proving it.

For what its worth I don't think we will be seeing many fanless models from other companies following this as a trend. It might just be beyond most their capabilities.

uh... probably not. Just too time consuming when it's only for one or two products (x-400/460). Just remember that in a top/back conventional PSU position case w/o venting above it, the open vents on the back panel of the x-400 is where most of the heat will go. Or if there's strong enough pressure from a back mounted case exhaust fan, then outside air will pull through the PSU into the case (and out that fan), cooling the PSU in the process. Airflow & hot air follows the path of least resistance. The top panel of the case will also absorb some of the PSU heat over time, which will also dissipate heat to the outside of the case. There's just not that much heat to dissipate -- 10% of the load, typically.


Most decent mid-tower cases with systems that draw <300W* peak. Chances are, this is 98% of non-dual-vga gaming rigs that use single-GPU cards. Most bottom-placement PSU tower cases. Most positive pressure airflow cases. Most HTPC cases with a bit of room above the PSU and some venting around it.

*Aside: A good reason to get an AC power meter, btw. Just check your PSU efficiency curve on SPCR's test review (or another good PSU test site if we haven't reviewed it), plug the PC into the AC meter, and monitor the power while playing your favorite games or doing your most hardware-intensive tasks. Then it's easy to calculate the max power demand & the heat the PSU has to dissipate.

A few changes were made in the article to reflect the above. Here's the most interesting issue:

Could a Seasonic X-400 run a overclocked Core i7 and a single GTX 480 in a Silverstone FT02?

Mike, thanks for the suggestions. Keep up the good work.

Cheers

A core I7 can have anywhere from 18 to 130W TDP (depending on which i7 you mean), so the answer I guess would be "depends", given that the GTX 480 has a TDP of 250W.

Assuming you're going fairly hardcore (i.e. i7 980X extreme), I would think that under full load, the answer would be a definate no.

Given that:

Anandtech ever tested (http://www.anandtech.com/show/2977/nvid ... e-wait-/19), and runs at nearly 100degrees C under load to boot, I'd suggest using a Radeon 5xxx series instead, although I imagine the system will still be pretty loud.

MikeC said

"Most decent mid-tower cases with systems that draw <300W* peak. Chances are, this is 98% of non-dual-vga gaming rigs that use single-GPU cards."

Didn't he also say the X-400 could go up to 700W peak with some nice decent airflow going through it? (Silverstone FT02 with AP181 penetrator fans should do nicely.)

Hardware secrets pull 602.2 W from this power supply in overload test

"Conclusions

The new Seasonic X-400 Fanless is an impressive power supply. During our tests it achieved efficiency always above 87%, peaking 91.5%, voltages very close to their nominal values (3% regulation instead of the traditional 5% regulation), and extremely low noise and ripple levels. On top of that, due to its highly overspec’ed components, we could easily pull 600 W from it with efficiency still above 88%. This is remarkable. Seasonic could easily sell this power supply as a 600 W product, if they added more connectors, of course."

Yeah this is what I gathered from MikeC's review. It can go a lot higher than 400W.

Also the Seasonic Rep posted this on HardOCP,

"The X-460FL will hit the market shortly in USA & EU, maybe 2 or 3 more weeks and it is based on the same topology & technology as the 400FL but with upgraded components to meet the 460w rated output."

Sounds like win

Naturally, I was assuming you'd want the system running silently since it's a fanless psu you're paying the $ for. I'd probably opt for an x-650/750 if going for a system with that kind of power draw. No way the cpu or gpu is going to be fanless in such a system, so then why bother with a fanless psu?

What about running a i7-980X at stock speeds with say a Thermalright HR-02 running passively and just letting the AP181 penetrator fans do the cooling in the bottom of the FT02?

I read a review somewhere saying it worked quite well.

I want to try and make everything passive in my system except for the fans in the FT02.

Thats why I was interested in the X400/460

Do you think its doable Mike?

Just try it -- only way you'll know for sure.

but a X-460 would be able to power a system like mine correct? Even if I wanted to be dumb and put a fanless PSU in a non-silent rig.

Overclocked i7 would be what? 200w? Personally, I would go for an i5 750. At 4GHz would be 150W usually.

So for an i7 overclocked and a GTX 480, that would pull around 37 Amps. And the PSU is spec'd for 33A? For some reason the table says Seasonic SS-400NF... http://www.silentpcreview.com/Seasonic_ ... anless_PSU But was pretty easy to work out tho.