Closing the Loop: Contained Liquid-Coolers from Corsair and NZXT Compared
by Dustin Sklavos on December 26, 2012 12:01 AM ESTEase of Installation
As I mentioned in the introduction, despite having six new closed-loop liquid coolers to test, there are only two actual procedures as far as installing the waterblocks; the rest comes from the variations in mounting the fans to the radiators and in turn to the case itself.
NZXT and Corsair both provide detailed, fairly easy to follow instructions for assembly, and they both make the same suggestion that's great for them but potentially impractical for end users: they suggest mounting the fan as an intake in the back of your case, bringing cool outside air directly into the radiator. This is a wonderful idea in theory, but in practice something I've rarely seen implemented. With rare exception, cases are designed to bring cool air in through the bottom and front and exhaust it out of the top and back (where the radiator will go). Assuming you don't have a dedicated video card in your system, this is a great idea, but the instant you start putting components into your case you run the risk of severely mucking up the intended airflow design of the case. For what it's worth, I don't think I've seen any boutique systems in for review that have actually oriented their closed-loop coolers (or even any of their radiators) in this fashion.
As for the installation order, it's going to depend on how roomy your case is: if you're working in cramped quarters, you may want to install the waterblock first and then mount the radiator. If you have room, doing it the other way couldn't hurt. I typically mount the fan to the radiator before installing the radiator itself whenever possible.
.jpg)
Seen above is the Intel mounting system for the Asetek coolers. They employ a backplate that fits smartly around the socket's backplate, and from there installation is handled in one of two ways: the NZXT way, and the right way. NZXT includes a retention ring that plugs into the bottom of the waterblock to keep the piece on the left in place, but this isn't actually how the Asetek waterblocks are designed to be installed. The piece on the left screws into the mounting backplate, but you keep it loose. From there, you insert the waterblock between the notches, then twist it so the block is held in place by the notches. Then you tighten the screws, and it's held securely and evenly into place. Do not use an electric screwdriver; the plastic holding the mounting posts inside the backplate isn't the most durable, and it's very easy to strip it.
.jpg)
You can see how the block mounts into place in this installation of the NZXT Kraken X60.
.jpg)
Seen above two parts of the mounting system for the CoolIT/Corsair blocks. The piece on the left is a backplate that mounts behind the motherboard, but you have to slide the posts into position and unfortunately you can get an uneven installation due to the backplate potentially pressing against the socket's backplate (and the screws therein). From there, you install retention screws from above the motherboard into the posts. The bracket on the right then goes over the waterblock, and four screw caps then twist onto the retention screws.
.jpg)
You can get an idea of how it comes together from the image above.
Neither one of these mounting systems are perfect, but I can tell you personally that I do prefer Asetek's solution. The CoolIT one is a bit more prone to an uneven fit, which resulted in my actually doing some retests while testing these systems. Asetek's mount is ultimately simpler, easier to work with, and more likely to evenly press the waterblock against the heatspreader. At least as long as you follow the instructions Corsair/Asetek provide, and not the ones NZXT provides, which include a superfluous retention ring that has the open round piece attach to the waterblock instead of the backplate.
Facebook
Twitter
RSS
68 Comments
View All Comments
schulmaster - Friday, December 28, 2012 - link
The most aberrant result of this review, at least in my opinion, is how the H80 and H100 compare. Presumably similar 120mm Fans, but differing by a significant factor in cooling area, the H100 should not be performing at H80 levels under load, let alone below.Here: http://www.youtube.com/watch?v=u8HsjgtMzEg
Linus runs a test series, with the variable of 120mm Fan quality removed, using top-of-the -line cooling on all contenders; and the h100 unsurprisingly performs 15+ degrees better than the H80 under OC load. I don't know wherein the testing parameters, on either side, emerged the resulting discrepancy of this magnitude. However, I would diversify the bonds of my research before discounting the H100 as a pragmatic enthusiast cooling solution. Perhaps Linus' using of a real- world-esque environment, ie only average cable management in a closed chassis contributed?
Foeketijn - Friday, December 28, 2012 - link
For all the people who may be too enthusiastic about high end air coolers.These things are harder to compare than you might think, in all the reviews I've seen it's all about the temp delta /Noise ratio. Funny thing is, a watercooler throws it's heat right out your case while a lot of warmth of your air cooler is circulated back to cool your heat-generator with preheated air. That's probably why almost every watercooler review has some comment about "why you should go for an similarly priced air cooler" and that being followed up by an actual owner of an watercooler (probably "closed loop" since the few DIY watercooler owners really understand the amount variables in hand) countering with his or her trial and having much better real life results.
A disadvantage is that the radiator blocks the airflow. So without the different fan in the equation there is less air going through your case. How much this shall impact the performance is strongly dependent on how much heat is being blocked (or how much heat is being produced not including the CPU and how much airflow is left)
Bottom line, comparing an watercooler (an air cooler with water as an heat transporting medium ; ) ) with an air cooler is almost impossible without testing several case builds and probably a really vague conclusion.
That being said, this is the right website for such an endeavor! An while I'm at it, I support the comments about missing some feedback about rotor noise and the 30 db floorlevel of the sound-o-meter. Great review tho, when you want to decide which CLWC to buy.
Khenglish - Friday, December 28, 2012 - link
What were the fan speeds with the thicker radiators? The 38mm rads will need either push-pull or have high rpm fans to push the air through.I would have liked to have seen performance with the IHS removed. Someone who decides to go with a water cooling setup is more likely to remove the IHS. It's well known that IVB has CPU die - IHS connection issues, and the variance between IHS quality among IVB's makes it difficult for people to compare their results to yours. Even with a good IHS, removing it still makes a big difference. The thermal resistance drop makes a big difference, especially when pushing the voltage and clocks. I lost ~5C when I removed the IHS from my 580 (1.25V). An IVB's gains should be even larger.
Your results are pretty terrible for liquid cooling setups, and it would be nice to know how much of that was due to the IHS, or if these liquid cooling setups just aren't very good.
riottime - Saturday, December 29, 2012 - link
i had placed 2 orders for the corsair h55 on newegg when this article appeared. :)they just arrived yesterday. i put one on my tuban 1090t amd system (stock speed) and i5 3550 intel system (also stock speed). i got them mainly to reduce the system noise not to overclock my processors with. so far they're doing just that i'm happy to report. :P
the manual/instruction that comes with it is atrocious. i recommend going on youtube and search for 'corsair h55 install' before you start installing it on your system.
TekDemon - Sunday, December 30, 2012 - link
I think you guys might have to check the mounting and test again because the H80i definitely shouldn't be outperforming the H100i assuming the H100i isn't defective. I mean, your own H100 vs H80 review has the H100 clearly winning and they're the same coolers minus the link.I honestly would suggest double checking your mounting and testing with a higher heat load since a i7 @ 4.4Ghz and 1.4V hardly really shows the benefits of an H100. I would point out that in your previous review the CPU was clocked at 4.8Ghz. If the H80i still manages to beat the H100i then either Corsair has upgraded the fan setup on the H80i such that the push-pull is managing to beat out the H100's stock 2 fan configuration or there's just something wrong with your H100i.
I would also point out that the H100i can be upgraded to a 4-fan configuration while the H80i is maxed out in it's stock configuration, I have an internally mounted H100 (non-i) with 4 fans on it and it's probably about as high performing as you can get with an all internal compact setup.
Reservoir_Dog - Sunday, January 20, 2013 - link
Cannot understand the temp results, i mean, whats, 4.5 under load?4.5 deg celcius??
With the H80i and a 2600K @ 4.2 ghz i get 60 deg celcius with the same test
So is it too hot then?
chesbrougha - Thursday, March 7, 2013 - link
I have an X40 and it is not cutting it with a 3930k. I am thinking that it might be because I can't figure out how to change the profile as there is no mac software (install windows then change back to mac??). Anyone have any ideas?I am thinking about changing to the H100i but think I will run into the same issues
soulcipher - Tuesday, September 30, 2014 - link
Apologies for digging up an old thread, however I joined as I wanted to share my experience with the product and the company. I'll be blunt and to the point. I am using NZXT X40.The company lie about case compatibility. Their head of Tech Support has stated in an email to me they do not test these on any cases but their own. This makes their compatibility list totally false as I found out a few months ago. I am using an X40.
I have a Corsair Carbide 500R and their compat guide states rear placement. Their head of tech stated in an email to me he had the case in front of him, and it fits fine. It took me sending him a video and photos to admit he had 'got it wrong' and that it doesnt fit. They have offered no compensation, and advised me to chop my case up to make it fit. I based my purchase of my current case on the statements NZXT made regarding the compatibility. When I first got it, I assumed I was doing something wrong and dumped it elsewhere in the case. However for top placement, it doesn't provide effective enough cooling, it needs rear placement with ideally, a push/pull config, so when I went back and revisited this notion, I contacted NZXT regarding to see what was going on.
Disgusting customer service, and would most definitely never buy NZXT products again. For anyone saying i'm lying, i'll happily disclose the emails. The product works, its quiet and does an ok job at keeping the CPU cool, but they have lied about it being compatible and offer nothing to compensate for making people out of pocket from their false guidance.