TechSpot

Temp E6600

By Post
Apr 30, 2007
  1. Just got a new system and I am curious if the tempatures are good.

    System is

    -E6600 processor (not overclocked).
    -Asus P5B-E
    -2 GB Kingston RAM
    -Antec 6000 case with 120 mm fan outtake (back) and 1 92 mm fan intake (front).
    - Artic Freezer Pro cooler for CPU using thermal path

    Idle, running at
    - System : 39 Celsius
    - CPU : 47 Celsius
    - AUX (?) : 119 Celsius
    - Core 0 : 27 Celsius
    - Core 1 : 27 Celsius

    Stressed with
    - System : 39 Celsius
    - CPU : 63 Celsius
    - AUX (?) : 127 Celsius
    - Core 0 : 45 Celsius
    - Core 1 : 45 Celsius

    I think the difference is to big for CPU temps.
    Anyone ideas ? Should Artic Silver do the job ?

    Btw temps are measured with Speedfan and CPU stressed with Orthos
    Also checked the Bios and they gave a 52 degrees after the test when I restarted the system.
     
  2. raybay

    raybay TS Evangelist Posts: 7,241   +9

    Idle looks about right to us, based on shop experience, but stressed readings are too high. You might want to try checking it with a thermal temperature gun. As the stressed 63 CPU is a bit on the edge for a cpu that is not over clocked. We sometimes find as much as 9 degrees off with the motherboard equipment, so it could be a machine error.
    A good thermal temperature gauge gun can cost $60 to $90, so many you can get a shop to check it.
    You won't get much change with Artic Silver, but it is worth a try as you begin to rule out things.
     
  3. ceegee

    ceegee TS Rookie

    try core temp version .95, some other monitors do not read core 2 duo temps correctly, also try intel's TAT (thermal assement tool) it will read cpu temps and stess at same time. more stressful than Orthos
     
  4. CMH

    CMH TechSpot Chancellor Posts: 2,039   +9

    More stressful than Orthos? o.O

    What, it runs the processor at 110%? :D:D

    Thermal temp guns only measure surface temp right? In that case, it would always be cooler than the actual temp.... nothing beats a calibrated probe right beside the CPU, which is obviously not very practical for shop use (gotta remove heatsink to remove probe).

    I'm not too sure, but the core temps are quite alright. I don't see why there should be such a huge difference between CPU temp and core temp, the cores make up the CPU, am I right?

    Btw, normal operating temps for the 6600 goes up to about 66C according to Intel. So you're fine, but threading a pretty thin line. You can, and should try Arctic Silver 5, especially if you didn't install that cooler with any thermalpaste at all.
     
  5. Post

    Post TS Rookie Topic Starter Posts: 20

    Thanks for the replies :D

    Installed the cooler with the thermal pad (pasta) wich is already on the cooler.

    Could try core temp version .95 to see what the temps are.
    Already tried Intel TAT, but Vista 64x doesn't run the tool :)

    Also is it correct that (Thermal Pad) pasta needs between 100 / 200 hours of work before it becomes optimal effective (bit weird sentence, sorry me is dutch :) )
     
  6. CMH

    CMH TechSpot Chancellor Posts: 2,039   +9

    when it is optimally effective, the temps would be lowered by a small amount, say, up to 5C lower.

    I wouldn't be worried really, other than the fact that your heatsink isn't installed properly (is it tight enough?). Load up the CPU, and then touch the heatsink where you think it would be the hottest. If it is somewhat warm, then the heatsink is installed properly (somewhat warm is subjective, can't really say more about that). If it is cold, as cold as the metal parts of your case, then you've not installed it properly, and should get AS5 and reinstall the cooler.
     
  7. Post

    Post TS Rookie Topic Starter Posts: 20

    Heatsink is to my opinion tight enough.
    For the touch test, I have to wait until to tonight after work.

    Also I read on other forums that the Core(s) temps should be higher then the CPU temp. Is this correct ?
     
  8. Tedster

    Tedster Techspot old timer..... Posts: 6,000   +15

    thermal grease tends to be more effective than pads as it covers more surface area, fills in micro-gaps, and bonds better. I would go with arctic silver.
     
  9. Post

    Post TS Rookie Topic Starter Posts: 20

    For some visuals, these are my temps

    IDLE > Speedfan (QFan in Bios enabled)
    [​IMG]

    STRESSED > Speedfan (QFan in Bios enabled)
    [​IMG]

    IDLE > Speedfan (QFan in Bios NOT enabled)
    [img=http://img22.imagevenue.com/loc773/th_54201_speedfan_0stress_1_122_773lo.jpg[/img]

    STRESSED > Speedfan (QFan in Bios NOT enabled)
    [img=http://img162.imagevenue.com/loc578/th_54205_speedfan_100stress_1_122_578lo.jpg[/img]

    IDLE > Everest (QFan in Bios NOT enabled)
    [img=http://img22.imagevenue.com/loc623/th_54210_everest_0stress_122_623lo.jpg[/img]

    STRESSED > Everest (QFan in Bios NOT enabled)
    [img=http://img141.imagevenue.com/loc591/th_54214_everest_100stress_122_591lo.jpg[/img]

    Didn't had time to feel the heatsink, will do asap (girlfriend was visiting :) ).
    Planning to order Arctic Silver 5 :D, see what difference we can get.
    And more solutions / tips are certain welcome :)

    btw : read something about the IHS problem with some Conroes ?

    btw2 : TAT and CoreTemp .95 are not running in Vista 64x
     
  10. Post

    Post TS Rookie Topic Starter Posts: 20

    Mhhh re-applied thermal pasta (had some Spire) and now the IDLE temps rised with 5 degrees celcius and load made it to 72 degrees celcius. :confused:

    It was a very thin layer of thermal compound (you could see thru), so it looks like the cooler is not making good contact ?!
     
  11. LinkedKube

    LinkedKube TechSpot Project Baby Posts: 3,481   +44

    well you dont want too thin of a layer, you should apply an amount about the size of a dime. even with a stock fan you should be that high of a temp. Definately a heatsink problem by lookin at the rest of your temps, case temp is fine. Hope it works out for you.
     
  12. Post

    Post TS Rookie Topic Starter Posts: 20

    You mean that the cooler doesn't make good contact or that the IHS is not flat (checked that and the IHS was flat, didn't see any weird inconsistness).
     
  13. LinkedKube

    LinkedKube TechSpot Project Baby Posts: 3,481   +44

    yeah, that's what i mean, was talkin about the AS5.
     
  14. CMH

    CMH TechSpot Chancellor Posts: 2,039   +9

    There is no such thing as too thin a layer. The thinner the better when it comes to thermal paste.

    However, there is such thing as too little, if it doesn't spread out to make full contact with the whole chip, its too little. If it does, its good enough.

    You don't need it to cover the whole cover for the chip (in chips with covers, like Intel chips), just the chip that lies below it (you can check out how that looks like in some review sites, or the AS5 installation guide).

    This is because the AS5 does NOT dissipate heat, it just bridges that gap, which would otherwise be filled with air (which is a really crap thermal conductor). AS5 on its own is a crap conductor, although better than any other thermalpastes around. But compared to pure silver, or pure copper/aluminum/any metal, it conducts heat pretty crap.

    Which is why you want as little of it as possible.
     
Topic Status:
Not open for further replies.

Similar Topics

Add New Comment

You need to be a member to leave a comment. Join thousands of tech enthusiasts and participate.
TechSpot Account You may also...