Good air cooling heatsinks?

[Tech Tweaker]

I've been using the Thermalright Ultra 120 Extreme (TRUE) basically forever, but recently I've began to wonder what other good heatsinks are out there for air cooling.

I guess you could say I'm a bit biased when it comes to heatsinks, as I haven't really tried many others. Tried the Zalman 9500 and 9700, but both gave consistently higher temps than a TRUE in my PC, and their stock fans are rather loud at full speed (and are not able to be swapped out for quieter fans). Also tried a Tuniq Tower 120, which was an abomination to mount, and gave worse temps (quite disappointed in that one).

So, I'm wondering what heatsinks are out there that can match or beat the performance of the TRUE without breaking the bank (sub $50 let's say).

I've heard a lot about the Prolimatech Megahalems being a good heatsink, but it's also quite expensive when bought new.

Hear of the CM Hyper 212+/212 Evo fairly often, but they seem to not score as well in most reviews as the TRUE. So, even though they are relatively cheap, they aren't the best option for replacing a TRUE.

Someone recommended the Xigmatek SD1283 Dark Knight to me a while back, and it does look kind of neat.

I've tried a Thermalright Silver Arrow SB-E Extreme, but it honestly showed no difference from my TRUE performance-wise in my test system. This was probably because the CPU's don't produce much of a heatload being Core 2 Duo's and Core 2 Quad's, so not enough heat produced to show a difference in performance. And with the stock fans it was a good deal louder for no performance gain.

 

[PolRoger]

I've used quite a few air coolers... TRUE, Megahalems, Noctua, Xigmatek, HR-02 but I think my favorite is the Archon. I've never tried any of the double tower coolers like the Silver Arrow, NH-D14 or the Phanteks. The original Archon was priced over your budget cap and I don't really know if they have come up f/s or have been offered used for ~$50 or less in the classifieds or not.

 

['Cuda340]

Worth at least looking into: Thermalright TRUE Spirit 140 Power

 

[Mjolnir]

Is your true lapped?

You'll find most air coolers won't gain you a huge margin from your true. Especially if you already tried the silver arrow. There's the Noctua d14 and u14s as well as phantecs dual tower model. But I can't see these allowing a huge improvement over your current cooler.

I'd say you'd be looking at true water cooling if you want a huge upgrade...

 

[NiHaoMike]

If possible, swap the fan on it for something more powerful.

 

[Mandrake4565]

Tech Tweaker, I'll speak for the 3 different heat sinks I have, all we're mentioned. True Ultra Ex 120 lapped, Prolimatech Megahalems and CM 212+. I did an unscientific test on my 2500k oc'ed to 4.7. Using the same fans on each test, all the tests were done in the same room within a 2 hour period but I didn't monitor the room temp, though it was in the winter so it was a pretty steady 70f give or take a degree. I ran prime blend for each test for 20 minutes. What I saw was the Prolimatech outperformed the True by 1-2 c and the CM 212+ by 9-10c. There are certainly variables that could account for the difference in temps between the Megahalems and true such as Tim application. Though, I think it was pretty conclusive they both op the 212+. I picked up both the True and Megahalems used for less then $20.00 with s/h and from what I've seen they still keep up with the newer high end coolers!

 

[Mr.Scott]

The only good thing about the 212+ is the bang for the buck. It is NOT considered a high end cooler.

 

[doyll]

The Silver Arrow SB-E Extreme is only as loud as you want it to be. If CPU fan temp/rpm curve is set right it is near silent. no louder than any cooler running 1300rpm fans. I'm running TY-143s on my 980 and can only barely hear the fans when CPU is running 98-100% load for extended periods.. Fans never go over 1050rpm and are still quieter than HDD is with CPU always below 50c. I've used TY-143s on TC14PE, R1 Ultimate, Silver Arrow, SB-E & IB-E, TRUE Spirit 140 Power, Macho HR-02, and even played with them as case fans. If they are loud it's because the fan curve needs setup or the CPU is using mega-watts and making some very serious heat.

 

[Tech Tweaker]

Is your true lapped?

You'll find most air coolers won't gain you a huge margin from your true. Especially if you already tried the silver arrow. There's the Noctua d14 and u14s as well as phantecs dual tower model. But I can't see these allowing a huge improvement over your current cooler.

I'd say you'd be looking at true water cooling if you want a huge upgrade...

No, I've never bothered with lapping either my TRUE or TRUE Black.

If possible, swap the fan on it for something more powerful.

I'm running Delta EFB1212SHE fans when I'm benching and need massive airflow to keep temps in check, otherwise I run Cooler Master SickleFlow fans or Antec TriCool's (except mine have black frames and no LED's).

The Silver Arrow SB-E Extreme is only as loud as you want it to be. If CPU fan temp/rpm curve is set right it is near silent. no louder than any cooler running 1300rpm fans. I'm running TY-143s on my 980 and can only barely hear the fans when CPU is running 98-100% load for extended periods.. Fans never go over 1050rpm and are still quieter than HDD is with CPU always below 50c. I've used TY-143s on TC14PE, R1 Ultimate, Silver Arrow, SB-E & IB-E, TRUE Spirit 140 Power, Macho HR-02, and even played with them as case fans. If they are loud it's because the fan curve needs setup or the CPU is using mega-watts and making some very serious heat.

I don't have the slightest idea how to set a CPU fan temp/rpm curve.

I just ran mine at 12v straight off the motherboard connectors, or via 4-pin to 3-pin adapter connected to the PSU.

 

['Cuda340]

FYI

Courtesy johan851

 

[doyll]

EasyTune6 on Gigabyte boards and Fan Xpert on Asus boards make it easy. I'm familiar with ET6 if you need a helping hand. I set minimum at about 20C@20% and max abut 70c@100% as a starting point. Than watch temps, rpm and listen. If it's too loud and temps are not high I lower the rpm by upping the temperature to rpm ratio. Higher temp at same rpm % means fans spin slower.

 

[NiHaoMike]

If you get one of the variable speed Deltas, you can use it all the time.

 

[Bobnova]

Sure, if you wear earplugs or listen to loud music at all times, or if you live in Schmitty Lane Magical Fun Land.

If you live in the real world and like a quiet system, big fat fans aren't going to do it for you no matter what control method you use. I've tested many.

Very few heatsinks actually beat a TRUE by a respectable margin. The Dark Knight falls short. If you want to beat it by a meaningful amount you need a massive dual tower thing, the Phanteks unit is my favorite personally.

 

[RnRollie]

or one could opt for top down cooling to ensure the VRMs & RAM is cooled also....

http://www.coolermaster.com/cooling/geminii-series/geminii/

 

[doyll]

That CM cooler looks nice but I see lots of issues with heated cooler exhaust mixing and heating up the intake air going to cooler. Much easier to keep cooler intake seperarted from cooler heated exhaust with tower coolers.

Most GPU cooler are terrible at keeping heated cooler exhaust away from cooler intake.

 

[Bobnova]

Turns out to make less of a difference than you'd think, based on my testing with a Tt top down piece.

 

[doyll]

I think it makes more difference than you think.
Having tested a fair few top-down and tower coolers I've found 6-10c difference using tower design vs pancake with both having same number / size pipes, fin pack area, and using a reference fan.
But of course more often than not systems do not perform the same and it may not make a difference in your system where it does in mine.. and several others I've suggested to try it.
It's all about how well system removes heated cooler exhaust and supplies nice cool air to coolers.
Some see one way, some both, some see another.
. . .

 

[Bobnova]

Yeah you do need to make sure you have decent case airflow through the CPU area, that's for sure.
If it's a tight case, the results will be poor.
What I did was use the same Tt heatsink in a case and then with a cold-air-intake on it so it couldn't get case air. There was a difference, but not a huge one.

 

[RnRollie]

Yeah you do need to make sure you have decent case airflow through the CPU area, that's for sure.
If it's a tight case, the results will be poor.
What I did was use the same Tt heatsink in a case and then with a cold-air-intake on it so it couldn't get case air. There was a difference, but not a huge one.

Cool.. A difference where? How much? On what?


My take is if there is not a huge difference with or without intake shroud/funnel on CPU temps, its because the max dissipation performance/capacity of the cooler has been reached.

From my point of view a topdown cooler using an intake shroud is almost a must on certain MBs -esspecially those which are not really designed to handle an overclocked AMD for example- because they will provide more airflow/cooling over VRMs and nearby components like RAM or "northbridge" compared to a tower cooler.

Look at it like this: most MB makers must have some idea of what they are doing, so you end up with the CPU sitting in the middle of a square of (VRM) Heatsinks (in case of Gigabyte ludicrous big sometimes)... probably based on the idea that a stock CPU cooler is used and its simple design ensures extra airflow of those VRM HSs... why would one willingly break that concept/setup by installing a HUGE tower cooler which leaves a 2" thick dead air spot on/around the VRMs?

A few years back, the majority of the (cheapish) cases were designed with side intake/shroud/funnel over the CPU area... one has to assume this was not a "looks cool" design fad, but that the engineers/designers were not total morons and had a vague idea what they were doing




On a more personal note... i've always wondered what was gained by those who dumped the stock HS, replaced it with a HUGE tower with 2 or 3 fans in push/pull AND then adding an extra fan with ducktape at a weird angle hoping to get to get some directed airflow over the VRMs.... I mean, i understand the quest for that last .5 of a degree; but ..... there seems to a a lot of sensibility (& knowledge) missing in a lot of people.

.

 

[Tech Tweaker]

My take is if there is not a huge difference with or without intake shroud/funnel on CPU temps, its because the max dissipation performance/capacity of the cooler has been reached.

From my point of view a topdown cooler using an intake shroud is almost a must on certain MBs -especially those which are not really designed to handle an overclocked AMD for example- because they will provide more airflow/cooling over VRMs and nearby components like RAM or "northbridge" compared to a tower cooler.

Look at it like this: most MB makers must have some idea of what they are doing, so you end up with the CPU sitting in the middle of a square of (VRM) Heatsinks (in case of Gigabyte ludicrous big sometimes)... probably based on the idea that a stock CPU cooler is used and its simple design ensures extra airflow of those VRM HSs... why would one willingly break that concept/setup by installing a HUGE tower cooler which leaves a 2" thick dead air spot on/around the VRMs?

A few years back, the majority of the (cheapish) cases were designed with side intake/shroud/funnel over the CPU area... one has to assume this was not a "looks cool" design fad, but that the engineers/designers were not total morons and had a vague idea what they were doing

On a more personal note... i've always wondered what was gained by those who dumped the stock HS, replaced it with a HUGE tower with 2 or 3 fans in push/pull AND then adding an extra fan with ducktape at a weird angle hoping to get to get some directed airflow over the VRMs.... I mean, i understand the quest for that last .5 of a degree; but ..... there seems to a a lot of sensibility (& knowledge) missing in a lot of people.

.

I've actually owned some of those cases with the side intake funnel for the CPU fan/heatsink. My experience was it only really helped in cases that had poor airflow and couldn't get enough cool air to the CPU fast enough to help dissipate the heat produced. With cases that had this airflow problem it did help a bit, by say 2-5°C usually. I'd say those were in use primarily from ~2002-2008. I mainly saw them with AMD socket A rigs, Intel socket 478, and some Intel LGA775.

I've always seen quite a large difference in temps going from the stock cooler to an aftermarket tower cooler, usually 5-10°C cooler, sometimes even 15°C cooler or greater with some CPU's (or platforms).