P8P67 Pro and Hyper 212+, Horizontal or Vertical?

I recalled having some difficulty originally installing this for an i5 2500k and thinking I could only get it to fit aiming vertically up.
Im using a Coolermaster 912 and I put 2 intake fans on front, 1 side intake, 1 rear exhaust and the 2 stock fans upper exhaust.
My idle temps at 4.2ghz 53-55º and prime95 in 1 minute hits 73/77/80/75
Im guessing the thermal paste isnt on too well, the orientation isnt optimal, etc.

Also by design it seems the 212 doesnt come down as far as it can and it can be wiggled left and right, which is the loosest ive ever seen any heatsink perform.

I have some of the stock thermal and some AS5 so debating what I should do as I dont think these are proper temps.
32 answers Last reply
More about p8p67 hyper horizontal vertical
  1. Sounds like it isn't on right. Definitely shouldn't wiggle. You can aim up or to the back, which ever suits your setup best. Have you got the back plate on the right way around? Check these:

    Hyper 212+ installation:


    Thermal paste application:

  2. The thing is Im fairly certain I remember reading that theres a little bit of wiggle, id say a few mm left/right rotation.
    When I check the newegg comments now, under the p8p67 pro and the 212, I dont see any mention of it.
  3. Hmmm. With very minimal force you can 'shift' the position of the heat sink slightly. Is that what you mean? I wouldn't really say it wiggles :) Anyhow I'd have another go a putting it on, those temps are huge. Also with your OC, perhaps try moving the multiplier up incrementally at 1st while you are figuring out whether you've got a good CPU Cooler mount or not. See below:

  4. I realized the heatsink fan is rated to go to 2000rpm but was only sitting at 1300. So I installed the Asus Ai Suite app and have it more aggressive, at 1600-2000. So my temps are idle at 50.

    But I took some pics to better show that I can rotate the heatsink

    I would guess my accidental bumping it, etc and causing those rotations would be eventually shifting the thermal paste around, ruining it?
  5. Just to double check, you did NOT use the pea size method as the 212 requires the Line method. With the 212 and @ 4.2 you should be lower, I'm quessing mid 60's. I'm using the 9900max and hit mid to upper 50's running prime 95 (I5-2500k @ 4.2)
  6. Yea used the line, but not on the heatsink i think, just 1 on the cpu. Ill redo it tomorrow. My pc keeps randomly crashing when playing WoW and its driving me nuts. For over 2 weeks. totally random. i run gpu stress tests, works fine. evga gtx 260 superclocked at stock speed. when I tried sli it crashed, so I thought was my 750w psu because i got it from my bro. so I try my 650w, still crash. i put back to 1 card, crashed, 650 and 750 w. i tried other card alone, same thing.
    And since im not playing anything else atm, I cant see if its just warcraft. And I can play fine 2-3 days and then it starts. I dont think its temps but im willing to investigate anything.
  7. It sounds more like the Heat Sink isnt seated correctly ontop of the CPU which isn't allowing the heat to transfer properly. Make sure the Heat Sink is really tightly fastned down with the correct Thermal Paste underneath. The tighther the fit the better the heat will transfer.

    You shouldnt be seeing anymore than about 35 degree's idle and 60 degrees under load. Personally im using Arctic Freezer 13 (92mm) CPU fan and thats keeping me at 30 idle, 60 load so you should be seeing better performance with the Hyper 212+
  8. Sounds like you are having fun. I'd be suspicious that it could be your temps causing problems. If Prime gets to your core temps to 80 after a minute.. imagine after a few hours on WOW. Try running everything a stock for a bit after reinstalling the 212+ and see if you still get the crashes.
  9. Did you watch this video on installation.
  10. Yea I saw the video prior to hooking it up. I realize now what makes it rotate. the top area of the contact on the heatsink has a hole to let the clamp in, and a little peg that is what lets you align it properly. the notch for the peg is just a bit bigger than the peg hence it has some wiggle room.
    So I redid it and have it blowing backwards instead of up. Also set the cpu down to 3.8ghz, so lower temps may be just because of that.
    When I booted it was at 40º and now after an hour its idling at 43-45, at least an improvement. the very second I hit start on prime95 it jumps to 60 but now after a few mins its only at 65-70. Still a bit better. Currently have the side off because im transferring data between drives, trying to replace an older 6 year hdd that sometimes spins off.
    As for WoW, temps never went past 60 or so, it stayed around idle temps.
    And when I mean randomly, I mean I could play for 6 hours no problems.
    And then later after watching tv, load it up, 2 mins in, crash.

    For thermal I used the AS5 and on the outer pipes I applied a line about half the length of the cpu, and then one more slightly shorter down the middle.
    The bracket didnt align right first time so I had to reseat it again, but as I didnt do anything accept a brief touch, I didnt reapply. I know AS5 has some adjustment time so we'll give it a whirl and see whats what.

    Heatsink is as tight as itll get. Im some what thinking if I had a faulty 212 because the back plate and screws are designed to not spin, yet they would. When unscrewing the clamp, 2 of the nuts on the back fell off. If youve used a 212, the screws are flattened on 1 side to align only 1 way on the back plate, so they were already against that. Also the little screw in the middle of the bracket to apply the pressure on heatsink doesnt go down that much.

    edit: after playing with the fans and xfering data around. Loaded up wow and crashed about 12 mins in.
    Im really stumped. I just clocked cpu down to stock 3.4, which is slightly less since the turbo mode supports around 3.8ghz.
    Even ran memtest86 over night for 7 hours and no errors.
  11. It sounds like theres is a problem with the heatsink or you are using the wrong mouting bracket and backplate holes. The heatsink should not move at all, if you wiggle it the only flex you should get is from the durability of the motherboard which is very very minor.

    Idle temps should be around 25-30 degrees, even up to 4.5GHz on the I5-2500K idle temps are around the very low 30's.

    Id suggest attempting to get a replacement/refund on your current 212+ or really do make sure 100% your using the right mounting points and bracket holes for your motherboard.

    Also, what PSU are you using? It could be a problem with the PSU not having enough power for your OC voltage. Generally your PC shouldnt crash with temps of anything below 80-90 degrees.
  12. Not saying I'm a pro, but I've been putting my own pcs together since an AMD 1.0ghz a good 10+ years ago and this heatsink just seemed odd.
    All the right holes and settings, so I know its not me picking wrong ones.
    I was using a Corsair tx650w originally with no problems. When I decided to try the sli mode, I switched to an Antec truepower 750w. Which is when the crashing started.
    Later as a test I put the 650w back with the sli and still had a crash.
    So other thing im thinking is maybe the 750w is faulty and is acting up, period.
    And the 650 I tried to confirm with, cant handle SLI so its a 2 edged problem.
    I was meaning to put the 650back and then try just 1 card.
    But also it just might be the damn game itself. *shrug*

    My last windows install just wouldnt let me share files over the network. Im net+ and couldnt figure out and even had my friend who has mcsa couldnt figure it out.
    Im always cursed with odd problems.
  13. Strange? Those temps are a little high, but your right it shouldn't crash it at those temps. What are you checking the temps with?

    If it is only happening during WoW... might need a reinstall (horrible process I know... keep those patch files:) Have you adjusted anything else, voltages etc or just your multiplier. Anyhow if your running stock or below then shouldn't be the problem...
  14. Use prime95 to test memory/stability. I find Prime 95 better at detecting memory errors.
    Monitor both voltages and temps. On voltages note how much the +12 V drops ( should not dropp more than a couple of tenths). Specs are a min of 11.4 V, but I use 11.6 as my min.

    Also run Furmark in a window whilel montoring Voltages (again primarily the +12 V rail) and GPU temps.

    I use CPUZ HWMonitor normailly, but not compatable with all new MBs. So may need to use MB supplied monitoring program.
  15. Temps I was using Core temp
    And actually I did do a reinstall of WoW. Even went to the hassle of letting it dl the WHOLE client from the site which took a good 8-10 hours I would say. I normally just copy the whole folder around between my system swaps so i thought maybe residual errors. I made it a whole 2 days on SLI before the booboo. Im thinking maybe its a bad addon so Ive been toggling a few off at a time since again it can work for days without problems, so its a hassle to play totally stock without addons for too long and not lose my mind.

    How would I check voltages though?
    My gpu idles at 50 and I believe when gaming only up to 65 or so. When I had the sli then the top one would go a lil more, 70 maybe.
    I tried furmark and it scared me because it jumped the temps up to 90. I ran it 5 mins before I turned off, but again, no issues there.
    I need a damn diff game to play a lot of and crash it. maybe grab street fighter 4 AE.
    If this crashes again I may drop down to the corsair 650 and single card, because that setup was working 2 months no problems.
    Maybe combination of 650 cant handle sli, even though it says its rated for sli, and the 750 is faulty.
    I wouldnt be surprised. Bro in law is in the next room and his stuff is always dying and breaking. he has an aura of technical failure. Only man who I know had a dreamcast break in 1 month. Doesnt mistreat it, stuff just doesnt like him and rather commit suicide.
  16. On both my Gigabyte P55 MB and the ASRock Z68 MB both temps and voltages are displayed on the same page in the respective MB supplied utilities (Both in side the OC app.

    You can try HWMonitor. It displays everyting fine for the P55, but voltages are not displayed correctly for the ASRock Z68 (I believe my version is 1.17 and the latest is 1.18) - worth a shot.

    Yes, Furmark will give your GPU a workout, But if GPU cooling is good it shoud stay below max - Need to check but max for GPUs is close to 100C. More important is to use it to check your +12 V rail, and that only takes about 5 Min.

    Link for prime 95: http://www.mersenne.org/freesoft/

    If you can run furmark for 30 Min, I would not suspect the PSU or the GPU (excluding the specific Game). Prime 95 will normally find errors in 2 hours, many also leave this running for upto 8Hrs. With Prime 95 you will come close to the max temperature that it will generate with in 30 Minutes. So for the first 15 Mins monitor temp closely, then can relax and watch TVor play with spouse, but check periodically.
  17. Ill take a look at Furmark for those other settings and see what safe max is on my card.
    Prime95 i think I didnt run for 2 hours or so when I first got it.
    Old pc used to shutdown with heat warnings after a bit.
    My pc is in the living room type area and when it gets hot, you can burn yourself on the outside of my old case..
    hence this one i put the 6 fans. It may be hot to touch sometimes but ive had it run fine like that, then crash at night when its kool..
    I guess tonight ill try some more stress tests.
    ah, hwmonitor doesnt have any of those voltage fields, it has
    CPU Vcore
    and VIN6

    I see theres a section under my vidcard, VIN0, min 1.05v and max 1.18v

    but the AI suite crap for asus does.
    Currently says +12v, 12.192
    Other area just has a slider, 10.200 to 13.801, and its at middle. shrug, dunno that one.
  18. I found something similar to hwmonitor, hwinfo64, and it shows all the fields.
    right now after about a 30 mins of prime95, temps are at 70/72/76/76, also note I put the cpu back up to 4.2ghz so that alone brought my idles back up to 45.
    the 12v is still just sitting at 12.192.
    Ok after 4 mins of so of furmark, heat shot up to 96º so I killed it.
    the 12v did move though, min 12.096, max 12.192.
    And I only saw it flicker under 12.192 2 or 3 times, just quick flicker, current stayed at 12.192 majority of the time.

    Ill try again with the sli later.

    Well, have some "good" news.
    I ran prime95 again while watching tv. Just came by and saw nothing was opened. I checked, pc crashed and rebooted, so I guess we can finally rule out the game and video card ideas.
    I had it logging to a csv file so I know exactly when it crashed.
    If anyone can take a look.
    So now what the hell do you guys think it might be? Ive gotten it prime95 before and no probs and here we have a crash after 9 mins.
    I looked in eventvwr and like usual it doesnt tell me a whole lot, but the errors look the same as before
    "The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly."

    Any ideas ?

    Wow, was in middle of posting before. Stuck my 650w in and started prime95 and it crashed within 3 minutes.
    *sigh* So we rule out power I guess. What other methods can I use to troubleshoot this issue?
    I think its also safe to say temps shouldnt be an issue because they really dont seem to get that hot.
    So its only crashed when gaming or running prime95.. Memory? Should I start testing em out in sets of 2? I have 4x4 g skills. But again 6 hour memtest was fine.
  19. Ran OCCT CPU linpack stress set to 90% memory, and crashed in 1 minute.
    I pulled 2 sticks out and left banks A1/B1 full. Ran again. Came back and saw was still on. Oh wait a min, nope, froze the screen. I had the WoW 3 or 4 times just freeze, not autorestart so I have a feeling were on the right track. Clock said 5:13 and it was 5:25 when I found it. Chcked logs, crashed at 6 mins.
    Removed those 2 sticks and put other 2 into A2/B2 as per manual preferred slots and it ran for 8 mins, and then it stopped saying sensor failure. I happened to sit at it at the time and tried opening apps so not sure if thats a reason. I dont know how to read the charts though, but didnt outright kill the pc.

    Before all this I minimized the system to the point of all case fans unplugged and surplus hdds. so its pretty naked.
  20. I think its slowly coming down to a faulty motherboard? Either that or a faulty CPU. Even with a stock cooler i dont think you should be seeing temps that high.
  21. I think its the memory. I just did 6 hour of prime95 over night and no hitches. Also I got a good 6 hours of wow in no probs, even put the sli cards back in.
    My temps sadly were always a hint higher than the norm, i think its because the room has poor ventilation itself and what not.
    Gskill says lifetime warranty on mem. Just wondering if I should still put the other 2 in and narrow down the actual bad one or just ask for RMA now.
    Normally I like to ride something all the way through.
  22. You've done well. Did you buy the memory in a kit? I'd send it all back a get a new kit. Don't like to mix up memory even when it's the same model number.

    Did you say back there your GPU hit 96 degrees in furmark? Might be getting hot in WoW too. Sounds like it's putting out a lot of heat. That wouldn't be helping your CPU temps either... lot of hot air blowing around.
  23. Yea furmark stressed it up to 96, but in game ive checked, doesnt go past 70 normally, and with the 2 big azz cards in sli, i saw it was 75 and /65 or so. top one a lil hotter cuz the bottom blows up onto it. Additionally they dont seem to really affect the cpu either. Hence my original dilemma, the temps cant be that bad. My previous build was horrible with circulation and what not. I opened the side and burned myself on the case metal where you screw in the hdds. only thing in contact with it was the hdds, and it was just that hot.

    Also side note, the heatsink fan supplied has the following specs,

    heatsink COOLER MASTER R4-BMBS-20PK-R0 Blade Master 120mm
    600 - 2000 RPM
    Air Flow:
    21.2 - 76.8 CFM
    Noise Level:
    13 – 32 dBA

    I remember pulling some fans out of a crazy dell xps I wanted to use but couldnt due to hardware fitting issues.

    Nidec fan
    . Speed: 3600 RPM
    . Air Flow: 130 CFM
    . Noise Level: 50 dBA

    These things blow air hard. Most of the fans I have like the supplied coolermaster case fans blow so weak, I have to put my hand right next to it to make sure its even doing anything. These nidecs are so rough, the second I plugged one in to test, it was standing on its side and knocked itself over. I can feel the wind from 4 feet away.
    Now that I think ive nailed the mystery down, I was thinking maybe mounting one of these as a victory dance to the heatsink... Also that I found out how to set up cpu fan speeds. I was afraid of this thing blowing max pressure 24/7

    Oh one turns out diff model, looks same, this claims
    3900 RPM

    Air Flow
    148.34 CFM

    Noise Level
    51 dBA

    Yeaaa,.... ill try this for a little bit.

    As for the memory, yea it was 2 kits
    G.SKILL Ripjaws X Series 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800)
    I never planned to go 16gb but it was on sale at the time and figured why not. Figures that would bite me in the ass down the road.

    I went to lay down again after previous post and set the OCCT to linpack mem stress again and fine after an hour.
    For giggles Im going to reboot with the other pieces mixed in and see that it is indeed one of these 2, or with horrible luck a bad mem slot on mobo.
  24. On GPU temp (looked back, didn't see the specific card).
    Max themps are generally above 90C, I've heard of some that are 120 C (Ouch).
    EX Found this for the 5870.
    "Talking to AMD about GPU temperature it was related to us that the GPU should start to throttle itself at close to 95C degrees."

    On Memory, As one of the posters indicated - keep in original pairs, and RMA as a pair (but this is up to the company).
  25. Nod. I dunno, lets name them rams 1,2,3,4 and banks a,b,c,d
    I pulled 2 and 4 from b and d, crashed.
    Ok so i tag 1 and 3 as bad. I leave 2 and 4 in b and d, all good, prime95 over night.
    Ok, so I lets find the badguy. I swap 4 out for 1, no probs after 30 mins or so. I swap in 3, still runnin now prob. Ok maybe its the slots. So I put the "good" 2 and 4 into banks a and c, runs fine. OK, i put the "bad" 1 and 3 into banks b and d, runs fine.
    So im wondering maybe its an issue when all 4 present, however I did have that 1 freeze when 1 and 3 were in banks a and c.
  26. Are 1 and 3 an originally purchased together pair? Maybe the different sets/kits are upsetting each other. I have heard of this happening even with RAM with the same model number. When testing, experiment with it in the pairs you bought it in staying together (maybe that's what you've done.. dunno just a thought). You're getting to them bottom of this... keep going :)
  27. Yea 2 kits of 2 bought exact same time. And since identical, I have no way of knowing which is which. I put the 2 "bad" ones in and ran OCCT linpack test again. Woke up and pc was frozen and nothing coming on screen when I turned monitor back on. Logs said it died after 20 mins.
    So know I know it has to be one of these. Marked them with tape and have one "good" with a "bad" and let that ride today. Itll take a while i guess but im gonna get to the bottom of this.
  28. As of now, no more crashes. SLI cards in, extra hdds added in for testing.
    those 2 "bad" sticks, 1 made pc crash again under an hour of occt linpack testing. the other lasted a bit but while browsing web and lipacking, screen would go off and give nvidia driver fails. So something was up.
    After I pulled them I saw the 2 suspect pieces had sequential serials, so maybe bad batch. I req an rma for both of them. So far so good.
    Also req rma for parts for the hyper 212+, because whether its working as well as it should, the screws are clearly jacked up.
  29. Sounds like you've got it covered. Good luck with some speedy RMAs. Are you getting a whole new Hyper 212+ or just the screws? It can't be working right at those temps.
  30. Coolermaster just now replied saying their shipping my rma request. I would assume just bracket n screws. the rest looks fine to me. And i cleaned off previous thermal, etc.
    But yea, 2 weeks or so, not a single crash. and I have the slis in their blowin heat all over.
    Idling now at 45-47. Ill that that for now, back up to a 4.2 oc, and its hot in the room in general.
    Im going to continue consolidating these hdds to cut down on crap inside.
    I tried them big ass fans. If I plug it into the cpu fan spot, I lose all control, runs max rpm nonstop. But if I plug into a chassis pins I can use the bios/asus app to control it. shrug.
  31. Hopefully the new screws sort it out. Based on your original description sounds like they might have been the problem. As described you can shift the position of the 212+ with very minimal force (similar to the position of pics you linked), but I wouldn't describe it as being 'loose' or able to wiggle. I'm not sure if it's posted in this thread somewhere or not but you can even attach another 120mm fan to the back of the cooler to get a nice push/pull flow going. Some people highly recommend it. But really, even with 1 fan, if your temps under load in Prime95 go over mid 60's at 4.2ghz, something aint right and I'd have another go at putting it on.

    Fastest warranty replacement I ever had was earlier this year... got a Mobo back from ASUS in 6 working days.
  32. The memory arrived there Monday so no update on when I'm supposed to receive the new ones. Coolermaster sent me an email rma ready,t hen rma shipped. Gave me an rma # yet the website says doesnt exist when I try to check status.
    And when I log in and check all rmas for me, nothing pops up. So that will get here when it gets here.
    But again, not a single hang up or crash since, so Im fairly happy about that.
Ask a new question

Read More

Heatsinks Overclocking