Is it only me, or does that fan speed controller dial (seen here) look a lot like the one on a Scythe Samurai (seen here)?

I think its just you...

Actually, it looks like a lot of people's fan controller...pretty much right out of the generic component stock bin. You can buy those pot's at Radio Shack for about a dollar. Not that that's a bad thing...if's it's not broke....

Well, the NH-U9 kicks the butt of the admittedly smaller Scythe Katana, with a 3,3,2 and 6C smaller heat rise at 12,9,7 and 5V respectively. Shame it's not available for Socket A or it'd be one of the best coolers around for that platform. Guess the weight wouldn't be such a great idea though...

I wonder how well the NH-U9 would fit into the various HTPC cases available as it looks ideal for that application?

Well written and informative review as always.

I was wonder if you think it's possible to bend the lowest fins with a pair of pliers so the mosfets get more air? On a 939 board the fan blows to the back exhaust right?

Typo in the table at the price row: "NH-U12: ~€40 (~US$45)" should be "NH-U9: ~€40 (~US$45)" i think.

not a bad little cooler, just a pitty they dont seem very available

It's just a shame that orientation is limited on the S478 & K8 platforms, unlike the Scythe Ninja. What's impressive is how good the smaller of the 2 varieties performs. I wonder if this might give Scythe the idea of making a smaller version of the Ninja that would fit in a great many more peoples' cases/motherboards.

The concern that there was a lack of cooling for the VRM was mentioned again. But even in a recent review of the AC Freezer 7, which was compared to a XP-120 & Ninja, there wasn't a noticeable difference between the coolers with regards to the VRM. So I think that there is some conflict between SPCR's logical argument versus their own real-world findings. I guess we have to wait patiently for the upcoming article on the subject?

It's not that I’d want to put your results in question, but I’m a little bit surprised that the NH-U12 falls behind the Skythe and the Thermalright so clearly. There’s no doubt that these are excellent coolers, but in most of the other reviews that compare these coolers, the NH-U12 is about on par with the Skythe and beating the Thermalright (e.g.: http://www.teccentral.de/artikel/artike ... -Test.html). In some reviews, the NH-U12 even outperforms the Ninja: http://www.matbe.com/articles/lire/267/ ... page16.php

Well, different platforms yield different results.

The included CoolLink fans look very much like the Akasa Amber series fans, built by Y.S. Tech.

I think what it might come down to is different methodologies.

The reason it does so well in the Matbe article is because they're using dual fans, whereas the Ninja only uses a single fan (at least it's the same fan...)

I don't know enough German to determine what fan was used in the TecCentral article, but it seems that this testing was done in a case, and with their "Quiet" fan at full speed (correct me if I'm wrong). I would speculate that it's their in-system test that changes things.

Also, the way that SPCR tests heatsinks — undervolting fans to find the threshold of acceptable cooling — tends to exaggerate the differences IMO.

They already do, although performance is worse. On the bright side Katana is orientation independant on all sockets except Socket A.

But is very relevant to the way SPCR readers will use the heatsinks, so fully appropriate for the particular user base. It's horses for courses isn't it. We test them in the conditions that we will likely use them in, they do the same.

I don't think different methodologies explain it all. If you look at the Matbe graphs, the NH-U12 is 2,5°C ahead of the Skythe Ninja with both coolers using the very same _one_ Papst fan. The Noctua gains another 0,5° using a second Papst fan.

You're right that TecCentral tested the coolers within a case. They actually didn't use any case fans at all, so this may actually come close to some real world silent rigs. Note that they used an Abit KN8 though, so the Noctua had a big disadvantage compared to the Skythe because the Skythe could blow the hot air out towards the back of the case whereas the NH-U12 had to blow up towards the PSU. You're also right that they used their silent fan (Noiseblocker SX2, 1,52V - which isn't too silent by the way) at full speed. Now I completely agree that undervolting the fan may (depending on how good the coolers tend to perform with low airflow) "exaggerate the differences" as you put it. But I wouldn't agree that this has to be in favour of the Skythe. If you look at the Matbe test for example, at 5V the NH-U12's lead over the Skythe increases to 4,1°C.

wouldnt it simply be because those other review sites and SPCR used different fans? nexus is a very low CFM fan, and as such shows heatsink performance under minimal airflow conditions
the other sites use higher airflow fans, and as you know, some heatsinks perform alot better with high CFM fans and they perform terrible with low CFM fans. (due to design of the heatsink)
i hope i got across my idea??

Yes, I think I understand what you're trying to say, but to be honest, I tend to disagree with you on that point. The NH-U12 is at least said to perform great at "minimal airflow conditions" and actually, that's exactly what SPCR says:



Now your totally right that at 12V the Papst fan Matbe used has 67 CFM whereas the Nexus fan used by SPCR has only 28 CFM, but both according to the statement quoted above and to the measurements at Matbe (2,5°C lead of the Noctua at 12V vs. 4,1°C at 5V), this should actually turn out in favour of the Noctua. Or am I getting something wrong here?

Coolink fans made by Noiseblocker's oem manufacturer.

hmm, you could be right jackz

One thing I have wondered about these tower coolers with heatpipes:

Don't heatpipes work best with some degree of vertical orientation? If so, then in your typical tower case, Ninja et al aren't going to be as efficient as they could be, since the pipes will run parallel to the floor.

Has anyone tried bending something like the Ninja so that the pipes are at least diagonally oriented? Seems like this would be agood design for the next rendition...

Very TiredOldMan

Rather than bend the Ninja, it's simpler to just lay the computer on its side so the Ninja is upright. I haven't done this with my Ninja because I'm just interested in the performance with the case upright, and that's fine.

I did try it once with the XP-90. No difference. There is a small difference with the A-C Freezer 64 but I forget which way is better.

Most heatpipes work fine in all configurations except when the evaporator (where the heat evaporates the coolant) is positioned substantially above the condensor (where the coolant becomes a liquid). The heat from the CPU will maintain a steady internal pressure so that the coolant, once condensed into liquid, is forced (usually by capillary action) back to the evaporator, thus keeping the cycle continuous. How effective heatpipes are depends mostly on the effectiveness of heat transfer between the evaporator & the heat source, between the condensor & the air (with aircooled finned heatpipes), and the speed at which the coolant cycles.

Even when the evaporator is above the condensor, if there's enough heat (ie, when the CPU is really under stress and thus running hot), the capillary "return" of the coolant still works, at least to some degree -- it is a closed system in which pressure plays a big part. The coolant will not cycle as quickly, but it will still cycle.

http://www.thermacore.com/hpt_how.htm



http://www.tsheatronics.co.jp/english/technology/

Hmmm, that is one hell of an interesting system.

It almost seems like some variation of it could eventually be used to generate electricity. I don't know, from liquid moving between a sunlit spot and a shady spot. or, er, something like that...

God, I've got to do something about this tiredness! And this age!

TiredOldMan

Otherwise known as solar thermal power:

US Department of Energy Solar Program

Concentrated Solar Power

Well, there is that...

But I was thinking, rather than the "conventional steam generator" that the page talks about, more along the lines of someday capturing the energy of the liquid's motion itself as it travels along the pipe.

Of course, when you harvest the energy, the liquid (or gas) will not move as much, but I wonder: it seems like the process of

1 capturing heat from the sun, 2 transfering it into a liquid 3 using it to heat yet another liquid 4 generating motion in a generator and 5 turning that into electrical energy

might be a little inefficient.

And I just realized that I am in the process of derailing this thread. Sorry. Someday I'll achieve the proper level of restraint to match my lack of knowledge.

What you describe sounds similar to a Stirling engine. Those can indeed be used for generation of electricity from solar.

Just dropped by to say: great review!

Hello,

I checked out the fan when the review came out. I didn't post this as I thought no one was interested:

80mm Coolink/Akasa/Noiseblocker seem to be the one and the same fan and made by YS tech.
120mm Coolink/Akasa/Noiseblocker seem to be the same fan, a wild guess it could be YS tech.

Akasa Amber 80mm and 120mm
http://www.phoronix.com/scan.php?page=a ... m=75&num=1

Akasa Amber / Noiseblocker /YS tech 80mm
Noiseblocker 120mm / Coolink 120mm
http://prohardver.hu/c.php?mod=20&id=665&p=2

Shame! They can't possibly be the same fan. They have different color blades! Coolink is green, Akasa is amber, Noiseblocker is in transition but weren't their blades blue until recently? Why, I may go into business marketing a fan whose blades are puce polka-dotted in mauve.

My 120mm Akasa Amber fan, bought in the past few weeks, sez AK-183-L2B across the top of the back label and "Y.S. Tech KM 121225LB" across the bottom, so you make very good wild guesses.

Felger,

don't you know the magic of relabelling and different coloured blades? Actually if you look at those fan specs and compare them with each other they don't add up. Now that's a surprise...

Good to know for sure that the 120mm is also YS tech, thanks.

<Off-topic>

Common sense would dictate that the less transfers there are the more efficient an energy-gathering technology would be; therefore direct sunlight-to-electricity conversion should be the most efficient process (photovoltaic), but it isn't. Most commercial solar panels have efficiencies in the low teens, whereas thermal solar can reach efficiencies as high as 40% already.

</Off-topic>