3770k Discussion thread - Heat Posible improvement!

Discussion in 'Processors and motherboards Intel' started by Bloodduty, Dec 20, 2012.

  1. thatguy91

    thatguy91 Ancient Guru

    Messages:
    6,451
    Likes Received:
    0
    The Ivy Bridges are assembled in a couple of different regions of the world, I guess maybe the one you had used a different crappy TIM...
     
  2. hallryu

    hallryu Don Altobello

    Messages:
    11,379
    Likes Received:
    0
    Well having given advice to be patient I decide to delid my CPU at midnight last night. The only thing I think I succeeded in doing is turning my CPU in to a very poor paperweight. Having nearly completed the process I noticed a very small nick in the PCB when removing the black glue. It must be fatal as there is exposed copper visible.

    I think I realise how I did it. During the delidding I fumbled the razor blade and created a minute nick in the blade, I think this was enough to cause the later nick in the PCB.

    I'm pretty pissed off at the moment. I got impatient and went for it when I should have waited for a more appropriate time.

    Lesson learned I suppose. No photos as I was too disgusted with myself.
     
  3. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    When trying the first corner i thought i done the same thing, the point of the stanley blade even snapped off.
     
  4. Agent-A01

    Agent-A01 Ancient Guru

    Messages:
    9,605
    Likes Received:
    0
    RMA it to intel, tell them it died for some unknown reason, after you glue it back together though.
     

  5. ---TK---

    ---TK--- Ancient Guru

    Messages:
    22,135
    Likes Received:
    0
    Hope this crappy TIM junk is just relegated to IB and not a sign of things to come from Intel.
     
  6. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    Hence why thread was created, to deal with such problem and not take every opportunity to just bitch about it.
    One of us has suffered a massive loss in the process, i knew when doing it i was looking at three weeks without my pc if anything went wrong, what does this mean for you, hallryu?
     
  7. hallryu

    hallryu Don Altobello

    Messages:
    11,379
    Likes Received:
    0
    I do some more overtime and buy another! :D I'm down, but not out baby!
     
  8. cowie

    cowie Ancient Guru

    Messages:
    12,797
    Likes Received:
    0
    Stuff happens Hal dont rush the next one and you will be good to go.

    I still rather just mount a dice pot on my 3570k then delid.
    The 3770k i have runs hot so i may delid that one of these days since i can go to microcenter and get another for cheap if it dies.
     
  9. hallryu

    hallryu Don Altobello

    Messages:
    11,379
    Likes Received:
    0
    Yeah **** happens. I'll take my time with the next one! Just don't tell the wife :puke2:
     
  10. chanw4

    chanw4 Ancient Guru

    Messages:
    2,297
    Likes Received:
    0

  11. PhazeDelta1

    PhazeDelta1 Moderator Staff Member

    Messages:
    15,611
    Likes Received:
    0
  12. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    6 days ago, page 2 of this thread, after reading hallryu's link:

     
  13. macdaddy

    macdaddy Ancient Guru

    Messages:
    2,388
    Likes Received:
    0
    I have removed the IHS on my cpu. I can tell you that there was certainly a gap. Also that when i was slicing it open I noticed that one the top side of the cpu was easy to slip the razor in and the other was tight. Seems like when Intel glued the IHS unevenly on mine. As soon as i remounted with new TIM idle temps were all even on the 4 cores with a gain of maybe 2c-3c lower. Load wise with OC of 4.9ghz i just manage to keep around 95c with LinX. Prime 95 low 80c. Maybe shaved 6c-7c from load before IHS removal. Only weird thing is the difference of lowest core to highest and the other 2 cores are in between, for me its close to 20c apart. One core won't move above 66c while im hitting 95c on the other. Maybe have to reset waterblock.
     
  14. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    Best results will be seen after an uber cooler has been installed, now with the tim change the cooler will be able to do it's job more effectively, i'm maybe in the market for one next week, noctua n-14 or corsair h80i, and a new tube of ic diamond. Sucks to have to remove mobo to get cooler backplate on but what the hell, if it'll let me get towards 4.8ghz stable then it's worth it.
    *currently with zalman performa x10 with x2 noctua f-12 fans.
     
  15. macdaddy

    macdaddy Ancient Guru

    Messages:
    2,388
    Likes Received:
    0
    Just spent last hour playing with the cpu. I think the sweet spot with my cpu is 4.5ghz @ 1.18 vcore. I have also undervoltted all the voltage settings for my mobo. Temps are very very good. Ran IBT a few times and it seems fine for now I'll test some more. If interested i keep you guys posted.
     
    Last edited: Jan 5, 2013

  16. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    I just tried installing cpu cooler directly on top of chip, system started up for a few secs then shut down.
    It was a half-hearted attempt though i did manage to get cpu cooler tight down on the chip.
    Anyways one thing i did notice this time and that was the thickness of the IHS, it's a horrible chunky piece of metal if ever there was one, why the hell is it so thick?
    I had no spare TIM for this job but my temps seem fine with the tiny amount i managed to scrape together for the cpu cooler and cpu chip, it'll be that damn IHS that needs taken care of next.
    I'm tempted to get another cooler but i'll give this application time and see how far temps rise.
     
  17. macdaddy

    macdaddy Ancient Guru

    Messages:
    2,388
    Likes Received:
    0
    Always could try to sand down the IHS. Might be tempted to do it myself.
     
  18. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    To recap what i've done with cpu and results:

    1
    Ok i removed ihs (it was a rushed job .. or better said a job i was unprepared for) and applied the tiny amount of ic diamond i had left to the cooler and some old antec formula5 to the cpu die.

    RESULTS :
    the idle temps dropped but load temps remained in the 70s

    2
    i removed ihs again, i had no spare thermal paste this time, i tried a direct-die mount* of the cpu cooler but it wasn't uber tight like it has to be so it booted for only a few secs.
    Now my cpu cooler is mounted without thermal paste pretty much, the cpu die + IHS has a tiny amount of second hand ic diamond /antec formula5

    *a cpu with cooler mounted to the die but not attached properly especially a highly OCed one will die very fast in this state, make sure there's tension always when powering up.

    RESULTS :
    Currently my idle temps are not too shabby in general the cores are at say ...35C idling, however i'm in the eighties pretty quick these days i'll say that much.

    Last words till next post: :p
    Today i hope to get some paste and try do the job well, i'll forget about the direct die mounting but it's very easy just make extra space with tape inbetween back of mobu + cpu cooler mount that attaches to rear so the screws can be tightened down tighter.
     
    Last edited: Jan 12, 2013
  19. macdaddy

    macdaddy Ancient Guru

    Messages:
    2,388
    Likes Received:
    0


    How did you check if there was no gap when mounting cooling directly onto the cpu die ?. Also i feel that the surface area of the cpu die is insufficient to transfer large amount of heat. Gets to a point where i cannot be efficient .
     
  20. SLI-756

    SLI-756 Banned

    Messages:
    7,604
    Likes Received:
    0
    I gave the cpu cooler a shoogle. :)
     

Share This Page