Quiet CPU Cooling

roskov

Honorable
Apr 18, 2013
20
0
10,510
Hello! I have an FX-8350 and I am looking for a QUIET cooling system for it. Right now, I have the stock CPU heatsink and I do not mind the performance I get from it. However, I need something quieter. If you could recommend anything that would be great! I am looking for low sound emission over performance (although I hope the performance won't be too bad). Thanks
 

roskov

Honorable
Apr 18, 2013
20
0
10,510
I have an ASRock 970 Extreme 3 mobo (http://www.newegg.com/Product/Product.aspx?Item=N82E16813157280) and the Rosewill challenger case (http://www.newegg.com/Product/Product.aspx?Item=N82E16811147153).

Any idea if the Hyper 212 EVO will fit?
 

clutchc

Titan
Ambassador
With that case, I have my doubts. The EVO measures around 6.25" (159mm) tall from the base. Take a measurement to see. If it is too large, look into the TX-3. It has a slightly smaller fan, but is only about 5.5" (139mm) tall. Still a quiet cooler. And an easy fit for an AM3+ socket.
http://www.newegg.com/Product/Product.aspx?Item=N82E16835103064

I can't vouch for this one, because I've never used one. But it is low profile with a large fan: http://www.newegg.com/Product/Product.aspx?Item=N82E16835103177
 

roskov

Honorable
Apr 18, 2013
20
0
10,510
Yeah I will definitely have to do some measuring (I really like the EVO but it might be too big)...In the meantime, are these the quietest options I have? Is there anything else out there that would fit my pc that is not ludicrously expensive? Any help is much appreciated.
 

Applesniper

Honorable
Apr 11, 2013
8
0
10,510


well all CPU fan cooling is pretty loud, i am not sure about hyper 212 evo since i have hyper 212+ the stocked fan is so loud i am going to replace with another fan. try water cooling since i heard they are pretty quiet. otherwise i suggest once u get your hyper 212 replace the stocked fan with a some other fans like sp120, af120 or Fractal Design FD-FAN-SSR2-120
 

roskov

Honorable
Apr 18, 2013
20
0
10,510
Wait water cooling is quieter? I was not aware of this...I had been told that the right air cooling was the way to go. I just needed help finding it
 

clutchc

Titan
Ambassador
You don't need water cooling. Besides, they have a fan too. So you're still going to have fan noise. Your problem is the inexpensive, noisy stock fan that ships with the FX-8350. A good quality aftermarket air cooler will not be that noisy.

This is a good site to get a review of various coolers: http://www.frostytech.com/top5heatsinks.cfm

There are 2 main things to look out for when sizing a cooler. 1) is it too tall for my case. 2) does it clear my RAM (or anything else under it).
 

roskov

Honorable
Apr 18, 2013
20
0
10,510


Alright sweet that's a good resource...quick question: when I compare the sound ratings, the lowest dbs come at around 35. I'm guessing that number is when the cpu is under a heavy load and thats the max db? Any way to find out what the db level is for the heatsink that comes with the Fx-8350?
 

roskov

Honorable
Apr 18, 2013
20
0
10,510
Hey so I went ahead w/ the Evo and it ended fitting in my case (barely). I have had an issue arise: according to coretemp, the evo is showing HIGHER temp than what I had with the stock heatsink...any idea what might be causing this?
 

clutchc

Titan
Ambassador
Yeah, that is definitely not right. The EVO should be keeping the FX-8350 at considerably lower temps than the stock cooler. And doing it quieter. Usually that would indicate a poor surface contact between the cooler's base and the CPU's heat spreader. Or it might be a poor TIM (thermal paste) application. Did you clean the surfaces of both the cooler and CPU with good cleaner?

Check the mounting of the cooler to be sure it is tight and flat against the CPU. Core Temp has returned some odd results for me sometimes. Check it against another temp pgm like HWMonitor.
 

roskov

Honorable
Apr 18, 2013
20
0
10,510


Yeah I will have to check the contact, but the other bases should be covered. I used arctic silver thermal compound and cleaned with their 2 step solution beforehand. I know there is a break-in period but I expected reduced temps right off the bat. Could it be fan placement? I can't really think of any other reasons. Guess I will also see if the issue is coretemp
 

clutchc

Titan
Ambassador
Even with the break-in time, it should be nearly the same as it will be after. I used to use it all the time until I found something a bit better; Arctic MX-4. No break-in time, not electrically conductive, and results in slightly cooler temps than when I used AS5. But either one is a good TIM. I would lean towards an installation problem keeping the surfaces from good contact.

Did you follow the Arctic clean/application instructions?
http://www.arcticsilver.com/amd_application_method.html#
 

roskov

Honorable
Apr 18, 2013
20
0
10,510


Yeah I followed the instructions and made sure not to overdo the compound because I knew there could be an issue with the compound overflowing onto the mobo around the chip. I will have to check the installation again...would you recommmend just re-installing the heatsink from step 1( i.e. taking off the heatsink entirely, re-applying the solutions to clean the compound off the chip and then re-applying the AS5 and putting the heatsink back on)?

I went into BIOS and sat watching the H/W Monitor...the temp went up to 60 deg C. That's not normal for an aftermarket heatsink when my pc is not even under load, right? When I booted up and went into coretemp while idling I was ranging from 25-40...
 

clutchc

Titan
Ambassador
Right, that is too high fore idle... even in BIOS where the CPU is always at full speed. Do as you suggest. Start over with the installation. You may want to take the MB out of the case for the cleaning/installation of the cooler. I always do mine outside the case so I can see what is happening to the components as I assemble them.
 

roskov

Honorable
Apr 18, 2013
20
0
10,510


Hey just an update: I re-installed the heatsink, making sure to make the connection right and apply the right amount of solution. The result: Haven't seen temps go over 40. Also, quieter then the stock heatsink. Pretty happy so far, thanks for all the help!
 

roskov

Honorable
Apr 18, 2013
20
0
10,510
Tried out IBT...don't really know what to make of the output numbers, but according to CoreTemp under full load my CPU temp was 44 degrees.
 

clutchc

Titan
Ambassador


If your processor only got to 44C after a 10-run of IBT, you have it well cooled. Granted, you probably ran at stock speed, but still good. Basically, all you are looking for with IBT is that it passed. At stock you'd expect it to of course. But as you go thru the O/C process, it will let you know if the settings are unstable or not.
 

roskov

Honorable
Apr 18, 2013
20
0
10,510
Yeah I'm going to take the OC process slow...I know this is not the right sub-forum to ask it in, but what is a good starting point to OC my FX 8350? Any input much appreciated!
 

clutchc

Titan
Ambassador
You're correct to head over to the 'overclocking' section for the best advice. But the place to start with a Black Edition processor like yours is always with the CPU multiplier. Turn off Turbo first. Then take the multi up .5 at a time and test for stability. I can usually get good O/Cs with just that increase alone. My other PC has a 3.5GHZ FX-6300 in it that I have at a stable 4.0GZ with no other changes than the multiplier. Personally, I don't like to increase voltage if I don't have to because that can shorten longevity of the cores if not done carefully. The front side buss is another place to increase the O/C, but remember that increasing that will increase RAM speed as well. So a crash is harder to determine if it is the RAM or the processor that was the culprit.
 

roskov

Honorable
Apr 18, 2013
20
0
10,510


Hey thanks a lot for that! I will definitely give that a whirl. Quick side question: you mentioned testing for stability, would that be through programs like IBT or Prime95? Or do I just use my PC and see if there are any issues? I know I would need to monitor the temps, but as far as performance I'm not too sure.

 

clutchc

Titan
Ambassador
Yes, IBT is my usual stability test at 10 runs, default. Lately, that's all I use. Although more runs (or Prime95 running for a couple hours) will be a slightly more thorough test if you are going for the ultimate max. But I've found so far, that if it passes IBT/10 runs, it is stable for every game I've played.
 

TRENDING THREADS