Cooler Master Hyper 212 Plus is awesome

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

ClockHound

Golden Member
Nov 27, 2007
1,108
214
106
After searching a few places, noticed that EVGA now calls all their high end video card coolers ACX as well. So...if it's tough to find the CPU cooler ACX, one could just buy one of those $500+ 980's, rip off the cooler, throw away the useless video card and you're good to go.

Or...could search a little more. Is this it B. Duck?
http://www.newegg.com/Product/Produc...82E16835288004
 

MongGrel

Lifer
Dec 3, 2013
38,751
3,068
121
Still have one of these TRUE Coppers on my HTPC with a couple Noctua fans odd as that sounds, might change things soon.

Sucker weighs about 5 lbs.

img0475on0.jpg


Be more a collectors item soon I imagine.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
After searching a few places, noticed that EVGA now calls all their high end video card coolers ACX as well. So...if it's tough to find the CPU cooler ACX, one could just buy one of those $500+ 980's, rip off the cooler, throw away the useless video card and you're good to go.

Or...could search a little more. Is this it B. Duck?
http://www.newegg.com/Product/Produc...82E16835288004

I'm pretty sure that's it.

There are only 3-egg, 4-egg and 5-egg reviews. Just to make sure, I scanned through them again to see if those were the ones I'd read together with the Hardware Secrets test review. That has to be "the one."

There was one remark about the fan, disparaging "red-bling" but especially explaining the fan's noise profile. This is true. I can identify my AP-30, but I notice the hum of another fan I can't hear in the other system. A 120 or 140mm Akasa Viper is in order -- possibly a 140R, but I'll read more of the specs on the 120's as I look for the Akasa 140mm square fan. The square fan is worth about 7 more CFM than the R fan. I can't really hear them at top-end. I could re-test to refresh my primate memory.

Another option could be a Noctua "industrial" iPPC 120mm PWM fan with top-end 3,000 RPM, or a bigger 140mm iPPC model.

After that, you group them, hook them up and start playing around with fan profiles until you know . . . and hear . . . what you're doing.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
Still have one of these TRUE Coppers on my HTPC with a couple Noctua fans odd as that sounds, might change things soon.

Sucker weighs about 5 lbs.

img0475on0.jpg


Be more a collectors item soon I imagine.

I was also wrong about the ACX cooler. It's just over a kilogram. It is more than possible that this includes the fan-weight. But it's not too much for a budget-level motherboard with thinner PCB. It is a fairly rigid assembly. You can't much make a Jew's-Harp by running your fingers across the fins, but that's exactly what happens when you flex a D14 or 212+ fin.

Your standard 120x38mm fan with a beefier motor is going to weigh about 250 gm. A good quality 140x25mm weighs less than 200gm, except possibly for Noctua, but then I think I used one of those to get my 7 oz measurement.

If you can't feel easy about putting a Delta Tri-Blade 120x38mm 150CFM 3-pin on your ACX, I can think of a few elegant ways to make it a pusher without hanging it on the cooler. In the process of doing that, it would be rendered pretty quiet. Now I'm thinking I should give it a try.

Easier to get an Akasa 140mm and adapter to connect to the "isolators" that hold the red-LED ACX fan to the cooler. The ACX fan is your standard clear-plastic 120x25mm, and the isolators are just custom rubber fan mounts for the cooler. You very soon learn hout to install it and remove it, and RAM doesn't pose much of a problem until you start using high-profile Tridents or the Corsair equivalent -- what? -- "Dominators?" "Vengeance?"

Anything like a Ripjaws model can be installed without even removing the fan for all four slots. You just have to practice one or two times before you "get it."
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
I've still never had any of those grey industrial Noctua fans, but they sure look nice.

http://www.relaxedtech.com/news/2014/noctua-introduces-two-new-fan-product-lines-and-accessory-kits/

Had been eyeballing em awhile now, but haven't been buying any fans lately.

You could control those things pretty well with something like Fan Xpert.

I've even been sloppy about following up all my "hunches," but Noctua has a reputation -- even a name like "Noctua" -- that may indicate how much they'd taken pains to make the fan more quiet. The consumer fans seem to have heavier motors, sculpted fins for better airflow, but only run at around 0.1A more or less. These could be 0.3A -- even more. But they're promoted as being lean on power consumption. And 3,000 RPM gives wiggle room to work with. I was trying to remember the CFM throughput with the 3,000 RPM model, but it was more than acceptable -- I guess you'd say "promising."

If I buy any, I'll buy two. Probably just as good to find the spare Akasa I have in the parts locker. I bought two of those -- as well. But the only reason you'd install it: to get rid of the noise from the ACX LED fan and possibly raise throughput by significant CFM. Except for the obstructive reduction in fan-area posed by the 140-to-120 plastic adapter. The ACX proprietary fan mounts would just fit through the holes of an adapter plate, and snap into place like any rubber fan mount does.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
Here's a thought, and I'd bet money on it.

If you slap on a Hyper 212+/EVO and duct it to the rear exhaust port with something like a Noctua iPPC PWM 120mm, I would expect to get load temperatures of maybe 75C (on my SB-K's) on the stronger stress-tests for something between 110W and 140W of thermal dissipation.

That would be the average of maximums, and I only take this from my SB-K experience of sig and the other extravagance of old-tech I'm working on right now.

For the latter -- 2700K -- cheap, 4+1 phase-power-design ASUS board -- "totally stable" @ 4600 Mhz and 1.33V and 50% LLC -- gives about 62C under Prime95 Blend. I think the sFFT is higher, but not that much.

Feel free to extrapolate . . . I'm just wondering how much I beat the last 68.5C average on the D14 system. Now I'll have to test the sFFT just to find out, but I bet it would still show around 63.5C. Uhhh? Maybe . . 64C.

Now I must have the complimentary gift prize of a Swiftech X240 . . 240X, $25-worth of foam-board and glue, and about $20-worth of Spire pads. And two $30 Noctua iPPC's!! Yeah. Get me Vanna White fo' one night!



Different processor models -- some with putty or "polymer" TIM -- will show higher temperatures, but in the comparison, I can get the lowest air-temperatures unless all you air-cooler guys turn and bow down to my new Duck religion.

. . . . Geez!! I was right!! 63.5C !!!! Room-ambient is only 1C lower than it was when I took the 68.5C reading on the other system. Currently 78F.

But also -- possible to make a 212 EVO about 2C better than a STOCK D14 installation. That doesn't come from this experiment; but the experiment proves the extrapolation.
 
Last edited:

ClockHound

Golden Member
Nov 27, 2007
1,108
214
106
Still have one of these TRUE Coppers on my HTPC with a couple Noctua fans odd as that sounds, might change things soon.

Sucker weighs about 5 lbs.

img0475on0.jpg


Be more a collectors item soon I imagine.

That is the most insane HTPC cooler ever! Congrats. But, you wouldn't dare put that in a rubber duct. But, a copper duct...hmmm...
 

ClockHound

Golden Member
Nov 27, 2007
1,108
214
106
. . . . Geez!! I was right!! 63.5C !!!! Room-ambient is only 1C lower than it was when I took the 68.5C reading on the other system. Currently 78F.

But also -- possible to make a 212 EVO about 2C better than a STOCK D14 installation. That doesn't come from this experiment; but the experiment proves the extrapolation.

So are we to deduct that a ducted cooler is better for CPU temps? What about the VRMs? Make some baby ducts?
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
So are we to deduct that a ducted cooler is better for CPU temps? What about the VRMs? Make some baby ducts?

Ducting the tower heatpipe -- if it doesn't come bundled with its own duct(s) like the Macho Zero -- is good for about 5C at something around 110W of thermal power -- maybe more. Depends on the cooler, but that's what I get with the D14.

In these tests I'd posted here, I wanted to test the ACX cooler's Hardware Secrets review conclusion that ACX was 6C cooler than D14. Both coolers are ducted the same way -- essentially identical. There is actually more intake fin area for the D14 to pull in extra air. But I still get about 5C improvement in temperatures with the ducted ACX and the ducted D14.

so we're talking about two types of improvements: 5C approximately for ducting the rear side of a heatpipe-tower to a strong exhaust fan, and 5C +/- give or take for the difference between the ACX and D14, whether they're both ducted or "unducted."

Now you bring up a point I'd been thinking about and actually had some epiphanies about it today. How to cool the VRMs? And while we're at it, how about the RAM?

To provide a little inspiration, I've just been through the happy acquisition of a perfect Sabertooth Z77 motherboard, and the painful loss of the board because I had two successive accidents with it -- one which crush a socket pin, and the other which broke off the nipple on the pin after I'd straightened it and tried to make it "more perfect."

If you've looked at the various Sabertooth boards, you should've concluded that they put the "TUF Armor" plastic plate on the board to provide enhanced cooling. It's not there to "protect" anything. There are two 40mm fans which push air into the crevice between the board and the duct-plate, forcing air out the hole around the Northbridge heatsink and the bottom of the plate.

Now I'm waiting to hear from a company that "acquires" IT assets and sells them in turn to consumers. They do repairs that include replacement of soldered chips on PCB boards. So I asked them if they can repair my socket pin, because ASUS just wants me to pay for a new motherboard. Surely I digress, but shall continue.

If the Sabertooth is now cyber-junk, It may be possible to simply modify the duct-plate to fit another ATX motherboard. But if you can't do that -- if I can't do it -- it should be the easiest thing in the world to cut an ATX-sized rectangle of black foam art-board and literally build the same duct to approximately same dimensions as the Sabertooth.

I have this idea of forcing airflow from a Sunon 40mm Mag-Lev fan at the top of the mobo and next to the I/O plate, which then travels through a duct that passes over the top of the heatpipe-cooler base, continues toward the front of the board over the top of the RAM, with a box fitted at right angles to the duct-plate directing the airflow into the intake stream of the heatpipe cooler.

I have a diagram of the airflow, which I can post as soon as a scan and upload it (and I'll try and remember). I'll append it to this particular post.

I think you could make these assemblies in pieces so they simply slide together and mate up around the heatpipes of the CPU cooler. That is the only obstruction that presents a problem, but such an approach to fashioning art-board might be fairly simple.

In the meantime, I'm fairly confident that MY OWN rig isn't suffering from VRM heatstroke, because the intake fans force enough air through the case, there's enough turbulence inside the case as well, so there isn't a problem in VRM cooling at the moment.

But there are a lot of things you can do to address that. And adding a ducting strategy would reduce the number of fans and simply work better.

With the Sunon (15,000 RPM top-end), it only requires attention to more thermal fan-control. And you could use a slightly larger fan, or modify the airflow across the board so that it has essentially the same effect.

To replicate the Sabertooth Armor, you'd just get a 1:1 schematic of your motherboard, cut rectangles in it for the PCI/-E slots, and make sure the top of the plate is low enough so that it wouldn't interfere with your add-in cards. the rest would be attention to the VRM, CPU HSF base, RAM and other ducting aspects.

But it's like anything else. I can say "Yes, we can! Yes, we can!" If your system will last about three years, it may seem worth it. Then you can sell it to someone: "Comes with custom-built motherboard duct and fan strategy for typical heatpipe tower cooler!!"

JUST AN AFTERTHOUGHT: I had worried about VRM cooling when I was using a TRUE rubber-ductie as a solution for a 212 or TRUE. I cut a rectangle out of one side of the accordion folds, so that the exhaust would draw air off the VRMs while it pulled air from the CPU cooler with a little less force. Since I'm using that rubber-ductie at the moment, I plugged the hole in it with some vinyl packing foam and secured it with dabs of Pit Crew adhesive. You can't see it, but who cares if you could? It's sealed.

I can think there's a better way to do it than that. Yes, I do.
 
Last edited:

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
This OCCT screenie following this text was made during an "OCCT:CPU" test. Everyone here should be familiar with the OCCT options: the OCCT:Linpack option runs a couple degrees hotter than the OCCT:CPU test. For purposes here, it might have been better to post the Linpack results, but it hardly matters.

This single graphic depiction is for CORE #1 exclusively, since there is a graph for each core. Core #1 is the leading core, representing the upper boundary of a range that is +/- 5C around a mean four-core temperatures.

As I also said -- or could prove with another screenie -- all the more severe tests show a mean at 63.5C and room-ambient here of 77F. The ducted ACX trounces the ducted D14 by ~5C. To me, that would mean that a ducted ACX would trounce an H110i installed without similar possible preparations -- by the same number of degrees.

What you could "DO" with a Swiftech 240X, H110i, Nepton 280L, or even has someone is going to show us in this forum with an H60? That's a completely different question. Maybe -- Noctua iPPC 120's, some foam-board and Spire pads. We'll eventually have to find out, and WGUSLER's H60 in a C70 project thread may give "first indications."

Now if you want screenies of a direct test on a CM Hyper 212+ or EVO, I'd have to do all that work again. But reliable reviews would allow us to extrapolate the results. I'd expect a DUCTED 212 EVO to outperform an UNDUCTED D14 by about 4C. The DUCTED ACX cooler most likely outperforms a DUCTED 212 EVO by 6 or 7C. So if I get an average-of-maximums under controlled ambient for the ducted ACX of 64C, the DUCTED Hyper 212+/EVO would come in around 70C for the four-core mean of maximums and same ambient.

For someone else to try it requires only a piece of scrap cardboard and maybe duct-tape; you can assess the results first before spending some nickels and dimes on foam-board and Spire -- whatever you want to use. Or you could buy the TRUE Rubber-Duc[k]tie for whatever chump-change it's currently priced at.

See what you get at room-ambient 77F. I believe with the same thermal wattage and voltage you'd come up with around ~70C. And I predict without ducting -- 75C -- who knows? maybe more.


2014-12-05-19h22-Temperature-Core%20%231.jpg
 

ClockHound

Golden Member
Nov 27, 2007
1,108
214
106
Hey, Mr. Duck, er, Duct.

Will you be posting any pics of your duct work? I know you pushed your thousand word quota over and above, but some images would be nice for those of us conceptually challenged.

Thanks!
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
Hey, Mr. Duck, er, Duct.

Will you be posting any pics of your duct work? I know you pushed your thousand word quota over and above, but some images would be nice for those of us conceptually challenged.

Thanks!

OK -- I'm going to oblige here. Here's my big case mod from 2007 -- the "Chrome Window:

First%20test%20run.jpg


Early, preliminary design of a ducting strategy:

Broadface%20duct%201.jpg


Broadface%20duct%202.jpg


We're now talking about two ducts: one for the heatpipe cooler; the other for the motherboard -- which is simply the bottom of the heatpipe box serving as a duct plate. The ProLiant Server case was fitted with two 120x38mm Sanyo Denki fans -- one to pull air through the heatpipes, and the other drawing air from between the duct plate and motherboard and from underneath the motherboard above the duct box:

Ultra%20120%20duct%20and%20mobo%20duct%20panel%20together.jpg


To build the duct-box for the motherboard, we need to do a little drafting just to get a foam-board prototype:

The%20P-I-T-A%20of%20Ducting.jpg


upper%20mobo%20duct%20to%20exhaust.jpg


Upper%20mobo%20duct%20to%20fan.jpg


CPU%20and%20mobo%20ducts%201.jpg


At this point, we wanted to add ducting for the VGA card, which had been fitted with a ThermalRight heatpipe cooler of its own. The pipes and fins hang down the rear side of the VGA card nearest the CPU cooler.

VGA%20Duct%20Extension.jpg


Later, a face-plate was added which covers the front side of the VGA card and includes a 92x15mm fan forcing air up and over the VGA card, to be drawn across the motherboard plate and out the rear exhaust. [I have the pictures, I just need to locate them.]

Here:

VGA%20extension%20installed.jpg


Eventually, these prototypes were either polished up a bit and painted with silver enamel, or whole flat panels were cut out of the prototype and replaced by clear Lexan plate. Like I said -- the pictures are somewhere -- I just need to locate them.

Truth be told, that project crowded a huge case with ducts that could've been smaller. Smaller means more work, but that's the only problem with it.

AIGOMORLA says he doesn't like solutions that require you to remove something to get to the essential components. I agree, and that was the problem with this "chrome window" project. But it did the trick. Especially, the motherboard was equipped with a blingy copper heatpipe necklace, so the attention to motherboard cooling really helped.
However, there are other ways to skin that cat. ASUS did it with their Sabertooth boards:

http://www.asus.com/us/Motherboards/SABERTOOTH_Z77/gallery/

The "TUF Armor" or duct-plate is fitted with two 40mm fans, both blowing into the crevice between the plate and board. If you wanted to devote maintenance time to dust buildup, you might have an intake fan and an exhaust fan with refinements to channel air over the northbridge heatsink and the VRMs.

And you can build a "TUF Armor" of your own -- much simpler than the duct box in Chrome Window. You just need an engineering drawing of your motherboard from the top in 1:1 scale, to use as a template or stencil to cut a foam-board plate that fits on the board before you add the VGA and other cards.

This could give you some extra mileage and would be less imposing than the other project -- with water-cooling. You only need to accommodate the rubber hoses and fittings of a waterblock to the duct-plate.

CONTINUED . .
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
Moving on, some pictures from my acoustic duct for an AP-30 exhaust fan ported to the rear of an NH-D14 with a single Akasa Viper 140R ("round") fan between the towers as both pusher and puller:

duct%20box%20and%20noise-cone%20front%20view.JPG


And it's not all that ugly:

duct%20box%20complete%20installed.JPG


This also resolves the problem of accessibility. The duct box is a close fit to the fan and cooler. It simply rests on the upper edge of the fan and cooler, and can be wiggled free by pulling sideways and out.

So you could build a simpler duct for the CPU cooler, and add a duct plate like the Sabertooth board.

But all this has limited application. You can get maximum efficiency from your heatpipe cooler, but then you have to pay attention to acoustic considerations -- which can be done or so I assure here. Part of that efficiency arises from allowing air to enter the front of the cooler and exhaust immediately out the case-rear: this would improve CFM airflow through the cooler by itself. But the other part of it comes from using a stronger fan as dual exhaust and puller. Without the duct, the strong exhaust fan just doesn't pull so much CFM through the cooler; without the fan, the CFM will also be reduced somewhat.

This all seems like attention to fly-shit. But it does improve things. Even so, I must make a grandiose pronouncement: These new processors need better cooling, and the odds of getting it from a heatpipe are less; the odds of getting it from an AiO are greater but limited; and the odds of getting it from custom-water or an exotic combination of it are much, much better.

You get a better payoff fiddling with hoses and fittings, radiators and reservoirs -- for some essential cooling priorities. There's going to be tedium and trouble either way. You can choose to use both water-cooling and motherboard ducting -- to go after "the last grain of rice."

But lately, I've taken the pulse of the membership, and I think people want quick solutions. You buy the AiO cooler and bolt it in. You can make improvements -- more powerful fans with acoustic considerations and other possibilities. So the question is simply this: do you want to take the time and trouble for any of it?
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
The duck delivers :thumbsup: :thumbsup:

Well, you only get so much mileage with this business.

I also remarked elsewhere: with a 4790K or Haswell E processor, the die-size is smaller. So thermal wattage and TDPs don't offer a basis of comparison with my Sandy Bridge results. Heat transfer RATE depends on area of contact. So like WGusler shows with his C70/H60 configuration, the temperatures are still going to be higher. But they'll be lower than you get with your standard 212 EVO, no duct and limp-o fans.

I just flabbergasted myself!! Who gets IBT "Maximum" temperatures below 70C with an air cooler on ~140W overclocked Sandy? Without water-cooling? Without AiO? I don't even think an AiO will beat it. We'd had a discussion of the Swiftech H2O 240X, and I looked at the reviews. Unless you modify the Swiftech and use more powerful fans, it's still a dead heat.

Now just as a footnote, I measure "average-of-maximums" recorded for at least an hour -- ~77+F room-ambient. The lead core sensor may bounce up to 73C or so. But that's not the best measure. It only applies if something trips OCCT or another program to "stop" because the lead core went to high. And it's NOT going "too high" with this setup. If my OCCT quits at 87C now, I've got maybe 14C of wiggle room with this.

And as a footnote to that, Average-of-Maximums -- maybe like "Package" temperature -- is a conservative measure. Take a log file over an hour from your IBT run; load it into a spread-sheet; take the average for each core (including only high-load portion of the test); then take an average of the averages for all four cores.

We're lookin' at something maybe 5C lower than 69C here. But as I said -- you'll have higher temperatures for IB and Haswell with identical thermal wattage. Of that -- I'm sure.
 

ClockHound

Golden Member
Nov 27, 2007
1,108
214
106
Thanks for the pics, BD! That's just ducty.

Definitely like version 2 better. Seems like something I could hack together using only blunt kitchen utensils and a bad attitude. ;-)
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
Thanks for the pics, BD! That's just ducty.

Definitely like version 2 better. Seems like something I could hack together using only blunt kitchen utensils and a bad attitude. ;-)

Xacto knife. And frankly -- I wish you could purchase Xacto saw blades. Do they do that? Did I miss something?

But yes -- I agree. No need to completely enclose the cooler: one only needs to assure that every molecule of air exhausted from the case has passed through the cooler on its way. With the enclosure, you only have air intake from one side of the fins. The more minimal approach has intake from three sides of any heatpipe tower.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
Implements:

http://www.dickblick.com/products/foamboard-rabbet-cutter/

http://www.dickblick.com/products/logan-701-1-straight-cutter-elite/

the nickels and dimes can add up.

I made my last big "buy" at Michael's Arts and Crafts Store. It was a three-panel "science project" display-stand, with black backing paper. I still have plenty of it. I get a secure feeling having it handy.

It's good for blocking off vents you choose to seal. You can secure it with wisely-chosen glues, cable ties, Nylon nuts and bolts you'd find at an electronics jobber warehouse or store. I could even imagine making a little drive cage for a handful of SSDs, secured somewhere in a case with Velcro.
 
Dec 30, 2004
12,554
2
76
Moving on, some pictures from my acoustic duct for an AP-30 exhaust fan ported to the rear of an NH-D14 with a single Akasa Viper 140R ("round") fan between the towers as both pusher and puller:

duct%20box%20and%20noise-cone%20front%20view.JPG


And it's not all that ugly:

duct%20box%20complete%20installed.JPG


This also resolves the problem of accessibility. The duct box is a close fit to the fan and cooler. It simply rests on the upper edge of the fan and cooler, and can be wiggled free by pulling sideways and out.

So you could build a simpler duct for the CPU cooler, and add a duct plate like the Sabertooth board.

But all this has limited application. You can get maximum efficiency from your heatpipe cooler, but then you have to pay attention to acoustic considerations -- which can be done or so I assure here. Part of that efficiency arises from allowing air to enter the front of the cooler and exhaust immediately out the case-rear: this would improve CFM airflow through the cooler by itself. But the other part of it comes from using a stronger fan as dual exhaust and puller. Without the duct, the strong exhaust fan just doesn't pull so much CFM through the cooler; without the fan, the CFM will also be reduced somewhat.

This all seems like attention to fly-shit. But it does improve things. Even so, I must make a grandiose pronouncement: These new processors need better cooling, and the odds of getting it from a heatpipe are less; the odds of getting it from an AiO are greater but limited; and the odds of getting it from custom-water or an exotic combination of it are much, much better.

You get a better payoff fiddling with hoses and fittings, radiators and reservoirs -- for some essential cooling priorities. There's going to be tedium and trouble either way. You can choose to use both water-cooling and motherboard ducting -- to go after "the last grain of rice."

But lately, I've taken the pulse of the membership, and I think people want quick solutions. You buy the AiO cooler and bolt it in. You can make improvements -- more powerful fans with acoustic considerations and other possibilities. So the question is simply this: do you want to take the time and trouble for any of it?

that's nice but you need to put some shipping tape over the sides of the fins of the cooler.

and that's the only thing that I've done to mine.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,727
1,456
126
that's nice but you need to put some shipping tape over the sides of the fins of the cooler.

and that's the only thing that I've done to mine.

"Shipping tape" is useful for this if you don't care about the aesthetics.

Here are my thoughts on this.

Some of the heatpipe coolers have closed sides: you can find this in a cooler like the NH-U14S -- possibly the D15 cooler; most certainly, this is the design of my EVGA ACX single-tower cooler. No air enters from the sides of the cooler.

The D14 has the fins open on the sides. If the case has pressurizing intake, more air flowing through the cooler fins from three sides is better than airflow exclusively from front to back.

What do you do about a cooler with closed sides? You can (a) either leave it alone, or (b) build a "leaky duct" which draws air through 1/4" slots extending beyond either side of the cooler. These sides of the duct will be extended to the forward side of the cooler, and admit air through a 1/4" to 1/8" slot, pulling the air from fore to aft of each side. This adds more airflow contacting both surfaces of the cooler on each closed side.

Of course, there's the tedium of fiddling with such types of refinements . . .
 

DesiPower

Lifer
Nov 22, 2008
15,366
740
126
Few days back I upgraded to i7-4790k, I again ran some mkv to avi conversion using DVDFab, just to test, left it running for about 3.5 hours, using the same 212 Plus and the highest temperature was 60 degrees. This cooler just keeps delivering. However the CPU usage was around 85% compare to 95% + that I used to see on i5-2500k. Maybe its the software's imitation, need to try a different software to encode.

I have not overclocked yet, will do soon and see how it performs.

*EDIT*
Got handbreak to utilize 100% CPU

c5Yxg1Y.png


Notice how it goes trubo and runs at almost 4.2 mhz!
 
Last edited:

jihe

Senior member
Nov 6, 2009
747
97
91
Few days back I upgraded to i7-4790k, I again ran some mkv to avi conversion using DVDFab, just to test, left it running for about 3.5 hours, using the same 212 Plus and the highest temperature was 60 degrees. This cooler just keeps delivering. However the CPU usage was around 85% compare to 95% + that I used to see on i5-2500k. Maybe its the software's imitation, need to try a different software to encode.

I have not overclocked yet, will do soon and see how it performs.

*EDIT*
Got handbreak to utilize 100% CPU

c5Yxg1Y.png


Notice how it goes trubo and runs at almost 4.2 mhz!

If you don't overclock the stock cooler is meant to be sufficient, so 60 degrees is what any after market cooler should get you.
 

DesiPower

Lifer
Nov 22, 2008
15,366
740
126
If you don't overclock the stock cooler is meant to be sufficient, so 60 degrees is what any after market cooler should get you.

I did try the stock cooler before I dismantled the previous build, it would easily reach close to 80 degrees under full load
 

jihe

Senior member
Nov 6, 2009
747
97
91
I did try the stock cooler before I dismantled the previous build, it would easily reach close to 80 degrees under full load



Don't worry about it. As long as it doesn't reach 100 and throttles it's fine.