Looking for a small inexpensive cpu cooler with secure mounting for i5-2500k at stock

[MauveCloud]

In 2011, I bought an i5-2500k, but I don't think my ASRock H61M-VS motherboard will actually let me overclock it significantly (despite having an "OC Tweaker" section in the BIOS setup), and I've decided not to bother with that anyway.

Now I feel like my Xigmatek Loki cpu cooler is bigger than I need, and the rubber pieces for mounting the fan are hard to get in correctly. Also, when I first installed it, I mounted the fan with the airflow in the wrong direction, and I wasn't able to get all the rubber pieces out to remount the fan on the other side of the heatsink. I bent a few fins trying, and some of the rubber pieces I used to mount it on what is now the bottom side didn't go in far enough or straight enough. The window of my current case makes it easy to see this (when I first got the Xigmatek Loki, I had a case without a window)

What I'm looking for is a heatsink that meets the following requirements:
1. Smaller than a Xigmatek Loki
2. Does not rely on push-pins or clips to hold the heatsink in place (so the stock heatsink is out, even if I still had it)
3. Isn't much hotter or louder than the stock cooler (cooler and quieter would be bonuses, not requirements)
4. Costs less than US$24.99 (which is the amount I paid for the Xigmatek Loki)

I found the following cpu coolers at pcpartpicker that seem to fit my requirements (with the possible exception of #3, since that's harder to find out):
http://pcpartpicker.com/part/masscool-cpu-cooler-8w2002f1m4
http://pcpartpicker.com/part/thermaltake-cpu-cooler-clp0556
http://pcpartpicker.com/part/logisys-cpu-cooler-ic231pwm
http://pcpartpicker.com/part/titan-cpu-cooler-dc155a915zrpw

Are there alternatives I missed, reasons to rule out one or more of those, or reasons to favor one of them over the others (aside from price)?

 

[bgunner]

I'm curious, how come a tower cooler wont work for you? They have some in the price range you specified and cool better.

 

[MauveCloud]

It's not so much that a tower cooler won't work, but it feels bigger than I need, given that I'm not overclocking. If you can provide a link to a test that shows how much cooler and/or quieter one of the tower coolers in that price range is compared to one of the non-tower coolers I listed, I might consider it.

 

[MauveCloud]

After some more thought, I've realized that it's more the fan mounting that bothers me with my Xigmatek Loki rather than the size. Therefore, now I'm thinking of getting a Hyper 212 Plus - yes, it's bigger than the Xigmatek Loki and a little more expensive, but it looks like it has good mounting mechanisms, both for mounting the heatsink onto the cpu and for mounting the fan on the heatsink, and it comes with a reclosable tube of thermal compound - I think the Xigmatek Loki came with a tear-open pouch of thermal compound, and I apparently haven't kept any other thermal grease, so with some of those other options I might have to buy some separately. It would also give me an excuse for some extra exercise, since there's a physical store that carries it about an hour's walk from where I live. I might wait until Friday for that, though, because the ForecastFox plugin is showing rain predicted on Thursday.

However, I'd still like to get some links to measurements of temperature and noise for the smaller and cheaper cpu coolers if possible.

 

[ihog]

I don't know why your board has an OC Tweaker; H6x chipsets don't support overclocking.

And just get the CM Hyper 212+. You won't get better performance at that price with anything else. As for thermal paste, I always go with AC MX-4.

 

[bgunner]

just beware that the 212+ needs more care with applying thermal paste because of the gaps between the heat pipes and the mount. the 212 EVO eliminated the gaps on the 212. yes its a few more dollars but IMHO it is well worth the ease of installation. Get what fits your needs just be aware of this minor application issue. Good luck

 

[MauveCloud]

Out of curiosity, are there any circumstances under which you'd recommend a cpu cooler cheaper than the Hyper 212 evo/plus? (assuming that the stock cooler is somehow not an option)

 

[MauveCloud]

Okay, I've purchased and installed the Hyper 212 EVO, but I have to say, mounting this cooler is more difficult for me than a number of others I've installed (including the Cooler Master V8). I had to take the motherboard out, because when mounting the backplate, the nuts for the top holes were bumping into the edge of the cpu cutout that my case has. Also, as far as I could tell, there was no way for that adjustable "X" shaped bracket to be secured to the heatsink before placing the heatsink down on the cpu; I had to line it up as I was attaching the spring screws. Even after I tightened the spring screws, the heatsink could still be twisted a little.

The fan, however, is much easier to mount and unmount.

Now I think I should double-check that my temps are within normal ranges for this cooler, given my difficulties with installing.

When I left it on the BIOS H/W Monitor menu for 10 minutes or so, I saw these results:
CPU Temp: 35.5 C
M/B Temp: 31.0 C
Ambient (from my clock with temperature display on the wall above the computer, not from BIOS): 74.4 F / 23.6 C
CPU Fan Speed: fluctuated between 643 and 653 rpm
Chassis Fan Speed: fluctuated all over the place, possibly because I've got a total of 3 fans connected to the 2 motherboard headers (top fan on one, 2 bottom fans on fan header splitter on the other)
Power Fan Speed: stayed at 904 rpm with only brief dips/spikes

After about 20 minutes light load, RealTemp 3.70 showed the following min-max temps:
31-36, 31-35, 35-39, 21-26

I have Prime95 already installed in case I should run that.

 

[bgunner]

The temps will depend on the case and air flow that it has. My temps @ 68F - 20C are 77F - 25C Idle. Now before you condemn the cooler I have 2x 120mm front fans 1x 120mm rear fan and 1x 200mm top fan and have added a Thermaltake AF0026 fan in push with the stock fan in pull. The stock fan is at max RPM and the thermaltake has a variable switch that I can adjust as needed for cooling.

This is also on a different CPU so temps will be different and the temps should go down after a day or two for the TIM to set By about 1-4 degree's.

I agree the "X" mount can be a bit tricky but no more so than most other tower mounts.

Try Hardware monitor from CPUID (the makers of CPU-Z) to test your temps instead of the Bios. The reason I say this is on "my" PC's (yes more than one) the fans will adjust themselves better with the OS running than in the Bios.

your temps look fine considering the ambient temp. Wait a day or two of heavy use or use "Intel burn test" to set the Thermal Insulating Material (TIM) to get a final temp reading. Run IBT for at least 2 hours before calling it good for setting the thermal paste (TIM).

HW Monitor is found here http://www.cpuid.com/softwares/hwmonitor.html
Intel Burn Test is found here http://www.softpedia.com/get/System/Benchmarks/IntelBurnTest.shtml

good luck and let us know what your temps turn out to be

 

[MauveCloud]

I don't trust CPUID hardware monitor - I tried it before, and one of the system temperatures was reading a max of several hundred degrees celsius, and more importantly, remained even when I told it to "clear min/max values". I'll run IBT tomorrow morning. I've got the default fans in my Raven 3 case, and the bottom fans are set to low speed. I haven't added a second fan to the Evo. I ran the Prime95 blend test for over 30 minutes, and the max among the cores was about 62 C, so I don't think I completely botched installing the cooler.

BTW, I believe TIM stands for "Thermal Interface Material". Thermal Insulating Material would not be a good thing to have between the cpu and the heatsink. :lol:

Edit: Perhaps I should also mention that I put the motherboard back in the case after mounting the backplate but before mounting the heatsink. For future reference, do you think it would have been easier if I'd mounted the heatsink as well before putting the motherboard in the case?

 

[bgunner]

I mounted my cooler out side of the case then installed the motherboard but my case does not have the cut out for cooler replacement. I was forced to remove my board to install the back plate so I figured, as long as I can get to the screws to mount the board, why not install the cooler while I have the extra room.

Woops your right, Insulation between the CPU and heat sink could be major issue.

Yes HW monitor can give false readings but usually on sensors that are not actually on your equipment. This is also a board to board issue as not all will have far out readings. As for CPU, system and GPU temps I personally have had good luck with the program and have found many more that are not as accurate. In fact the OC program that came with my board is wrong when it comes to low temps. It will only read 32C and no lower for my CPU when I know for a fact that it is lower. But any ways find a program you feel comfortable with reading the proper temps and run the IBT and watch the temps. IBT will usually give the highest temps out of the stability test programs I have used, including Prime95.

 

[MauveCloud]

At this point I'm fairly sure installing the heatsink outside the case would have been a better idea, so if I ever have to replace this case, I'll consider having a cutout behind the cpu to be an insignificant feature.

As far as HWMonitor, I realize it's for a sensor that isn't available on my system, but that doesn't excuse failing to clear the max value when I tell it to. Also, isn't there a way for HWMonitor to tell that the sensor isn't supported?

I went ahead and reinstalled HWMonitor, since it shows a lot more than just the cpu temperatures, and am now running IBT at Maximum with 40 iterations (10 iterations only took 38 minutes). I'll edit this post and add a link to an image of the results after it's done.

Edit: here's the link:
http://www.mauvecloud.net/hwmonitor.jpg

 

[bgunner]

WOW 84C The max temps of that CPU should be 72C. Are you sure you seated the cooler correctly?

here is a link to the specs of that CPU and is where I got the max temp from. http://www.intel.com/content/www/us/en/processor-comparison/processor-specifications.html?proc=52210

EDIT: I just noticed that the fan speed did not rise with the heat of the CPU. Make sure your fan adjusts itself properly. The max rpm for that fan is 2000 RPM +/- 10%

 

[MauveCloud]

I think the motherboard temperature sensors are a little flaky. I cleared min/max, and CPUTIN showed a max of 57 C pretty much as soon as the numbers were showing again, even though it's staying at about 28 C with low cpu load.

My BIOS settings were as follows:
CPU Fan Setting: Automatic Mode
Target CPU Temperature: 45 C / 113 F (lowest available on the menu)
Target Fan Speed: Level 1
Chassis Fan Setting: Automatic Mode
Target CPU Temperature: 45 C / 113 F
Target Fan Speed: Level 1

I changed both Target Fan Speeds (which aren't well explained in the BIOS menu, and the printed manual has no explanation of any BIOS settings) to Level 9, and I'll try IBT again.

Edit: CPUTIN climbed to 77 C before IBT has even finished the first iteration, though the max "Package" temperature read from the cpu is only 59 C. I pulled out the motherboard, and the cpu cooler was seated firmly as far as I could tell.

 

[bgunner]

We both have an AsRock board so I know exactly what the target fan speed means. It means the lowest RPM the fan speed will drop to when the CPU cools down. Example level 1 would lets say allow the CPU fan to drop to 700 RPM where as level 5 the CPU fan would only drop to 1400 RPM.

when the temp climes do you physically hear the the fan speed raise? This would only happen if the target fan speed was set low like 3. I'm just curious if the fan is speeding up properly.

Have you installed AXTU from the Disk or web site? It has a monitoring program and also you can adjust the target fan speed from with in the program to hear if the fan speeds up or not.

 

[MauveCloud]

I had too much other background noise before to tell by ear, and I hadn't installed AXTU, but I have now, and it seems to be treating the "Target Fan Speed" as a fixed speed, not a minimum. If I set it high, it stays high even when there's only a small load and the cpu temperature is below the target. If I set it low, it stays low even when stress-testing the computer with cpu temperatures reading well over the target temp.

 

[bgunner]

Something is stopping the the fan from auto adjusting and is most likely in the Bios. There is one option I can think of in the OS that could cause this issue. In Power options under advanced settings then Processor Power management there is a setting for System Cooling Policy. This should be set to active and if not could cause the issue.

EDIT: as for the target fan speed settings, if it is set high the fan will not go under that setting. same thing if it is set to low. I will re edit this post with the minimum fan speeds i get at each setting so you will have a better understanding of what i am talking about

EDIT again: ok this is my numbers. keep in mind I have a fan pushing air through my cooler and it is keeping my fan rpms higher than what you have. The fan will not go below, +/- a few rpms, these numbers at that level. This sets the minimum fan speed. The rpm that the fan will not go below. As the CPU heats up the auto adjustment is supposed to raise the fan speed but when it cools back off it wont drop below the set value of target fan speed. I hope i'm not misunderstanding you and this helps.
level 1 =788 rpm
level 2 =843 rpm
level 3 =1014 rpm
level 4 =1210 rpm
level 5 =1378 rpm
level 6 =1513 rpm
level 7 =1661 rpm
level 8 =1822 rpm
level 9 =1904 rpm

 

[MauveCloud]

When I ran IBT for posting the image with the temps, I had the power plan set to Power Saver, with cooling policy set to passive, but before installing AXTU, I had set the power plan to High Performance, which has the cooling policy set to active.

Edit: as far as CPUTIN, I thought maybe the reading was occassionally spiking, but I did some Googling, and found OCCT which allowed me to get a graph of cputin vs. time, and it wasn't showing anything like that. However, when I ran OCCT's built-in stress test, I noticed that the max cputin shown in CPUID's Hardware Monitor program was rising to match exactly double the Celsius reading I was seeing in OCCT (which went up in increments of 0.5 C)

Edit #2: I let OCCT run its test for a little over 10 minutes, with several programs monitoring - OCCT and Open Hardware Monitor agree that CPUTIN had a max of 41 C, but now CPUID HWMonitor says CPUTIN hit a maximum of 437 C, which is not believable given that my computer is still functional, so I'm going to consider the cputin readings from that program void, and say the heatsink is properly installed. I think the failure of the fan to spin faster is because that's triggered by the motherboard CPUTIN reading, which only really got up to about 41 or 42 C, which wasn't high enough to require spinning up the cpu fan.