Closed Solved

Just upgraded CPU cooler: No temperature difference...

Hey everyone,

I just upgraded the CPU cooler from an Arctic Cooling Freezer 7 Pro R1 to a Noctua NHU12P

This should give me a pretty significant temperature drop right? I am getting exactly the same idle temps as i was with the freezer 7.

I used Arctic silver 5, made sure i followed the instructions from the arctic silver website: I used the "1mm line across the width of the CPU" method which is recommended for core 2 quads. The heatsink is securely mounted, and i cleaned both surfaces with IPA, and then primed them (applied AS5 and then rubbed it off with a lint free cloth) before applying the AS5 to the CPU.

Is there anything else worth checking for that i might have done wrong? Obviously id rather not take it all apart again because it was a pretty big effort taking the motherboard out to fit the new cooler, plus i seem to have disassembled my PC a million times in the last month. I havent lapped either surface, and id rather not, but surely this wont make a massive difference. (the freezer 7 pro wasnt lapped either)

I am running a Q6600 G0 Stepping at 3.2Ghz in a lian li v1000. Idle temps are 39/36/36/35 for each core respectively.

Thanks so much
18 answers Last reply Best Answer
More about just upgraded cooler temperature difference
  1. Best answer
    Hello rollojarvis;

    A bigger drop at the top end. Maybe not much difference at the low end.

    You might not see too much change at idle temps if you were getting good results with the AC Freezer 7 Pro.
    It will be the top end where the bigger sink and fans of the Noctua NHU12P should make a noticeable difference.

    What are the under load temps like?
  2. WR2 said:
    Hello rollojarvis;
    What are the under load temps like?


    Ah okay, didnt realise that, Im running P95 now and ill get back to you. Also i know arctic silver 5 has a fairly long break in period, will this affect temps much?
  3. First of all you need to make sure you have heatsink across all of your processor, Using the credit card or ziplock method is fine as long as it is coated across the whole CPU. Even bolted down it will spead evenly. But if you use the dot in the middle, you might have problems. Just make sure it is coated all over the CPU and you will be OK. A little extra will not hurt like some say on this forum. Complete coverage is the big deal.
  4. Load temps seem to be stable at 52/47/47/48.

    I cant remember what i had before but that seems pretty decent. Its weird how core 1 is always hotter - it was with the previous cooler as well, could it be a stuck sensor?
  5. I use Artic silver and there is no break in period. It cools right off or it does not. You should have in the mid 30* to begin with and if it goes down the better. Mine is OC'ed to 4.0 and runs in the low 40's*, but never goes above 52* on anything i throw at it. The breakin therory is crap. It cools or it does'nt. I have been through several materials to gap the space and silver is about the best beside the compound that comes with the Notcha NH-D14.
  6. pjmoses said:
    First of all you need to make sure you have heatsink across all of your processor, Using the credit card or ziplock method is fine as long as it is coated across the whole CPU. Even bolted down it will spead evenly. But if you use the dot in the middle, you might have problems. Just make sure it is coated all over the CPU and you will be OK. A little extra will not hurt like some say on this forum. Complete coverage is the big deal.


    I've heard so many methods from different people, so i chose to follow the instructions from the thermal paste manufacturers as i assumed they'd know what they were talking about. Their reasoning seems to be that the cores are only in the middle of the headspreader, so the thermal paste only needs to be in the middle.

    For the surface spread method, AS5 is bloody hard to spread around, is it alright to use your finger wrapped in cling film?
  7. Good results there for an OC'd Q6600. Time for 3.6Ghz?

    A stuck sensor would be - stuck. You wouldn't see it change all that much or it would give values totally out of scale with the rest.
    What you're seeing with that type of temp variation is pretty common.
  8. Cling film, or even easier, a plastic baggie, works fine.

    But I think you can run for a while without making any changes. Those are pretty good temps.
  9. haha wasnt planning on going higher than 3.2... but if i get bored...

    What kinda voltage increase would i be looking at from 3.2 to 3.6? (currently at 1.24 in cpuz) Also, i currently have the option on (C1e? - cant remember the name) that drops the multiplier off when not under load. Is this alright for an OC'd cpu? Sorry, but just thought of a few more questions, 1) Does overclocking significantly increase the power consumption of the computer, im thinking about electricity bills... 2) Is it alright to leave the northbridge voltage at auto when overclocking?
  10. No, and No! your cpu ws not made to run at those low of voltages and if overclocking is will not preform right at all. Yes increasing the clock of the CPU wil increase the voltagek but only if you let if. What CPU do you have and why such a low voltage>
  11. The temps are fine, but you need to stablize the voltage to get a stable overclock. Is you CPU rated at 125w? If so you can increase voltage to 1.35 without any problems and run a cleaner overclock.
  12. I just got it prime stable at 3.6GHz (400*9) with vCore of 1.4125 in bios, LLC enabled, VID 1.25, but the temps are a bit high for my liking, it idles at about 35, but load with small FFT goes to about 70 degrees max.

    Im gunna drop back down to 3.2 for 24/7 use.

    CPU is Q6600, also the low voltage i said earlier was due to a large vdroop, and a vdrop on load, it was set to 1.325 or something in the bios
  13. pjmoses said:
    I use Artic silver and there is no break in period.


    Artic Silver's web site says different.

    http://www.arcticsilver.com/as5.htm

    Quote:
    Important Reminder:
    Due to the unique shape and sizes of the particles in Arctic Silver 5's conductive matrix, it will take a up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink or with a low speed fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop 2C to 5C over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired.


    At least one reviewer puts that thermal cycling requirement equal to "one year of normal usage"

    Shin Etsu matched AS5's thermal performance w/ no curing issues.
  14. rollojarvis said:
    Hey everyone,

    I just upgraded the CPU cooler from an Arctic Cooling Freezer 7 Pro R1 to a Noctua NHU12P

    This should give me a pretty significant temperature drop right? I am getting exactly the same idle temps as i was with the freezer 7.


    What did you base your selection on ? Make sure the testing on whatever site you reviewed uses a test system w/ a CPU with an identical or at least similar heat signature. For example, using a LGA 775 test system to base cooler selection on would not be representative of performance on a 1155/1366 CPU.

    Also note that while lapping may increase performance on a poorly finished cooler, it will void the warranty and actually reduce performance on many coolers as a slight curve is part of the design and performance level.
  15. Ill remember that for the future but ive already bought the AS5 and dont really want to buy something else
  16. JackNaylorPE said:
    Artic Silver's web site says different.

    http://www.arcticsilver.com/as5.htm

    Quote:
    Important Reminder:
    Due to the unique shape and sizes of the particles in Arctic Silver 5's conductive matrix, it will take a up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink or with a low speed fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop 2C to 5C over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired.


    At least one reviewer puts that thermal cycling requirement equal to "one year of normal usage"

    Shin Etsu matched AS5's thermal performance w/ no curing issues.


    That is what I am saying, a year breakin cycle. Most of us do not have a cooler on there or the same cpu that long as an enthusiast, I am ramping up my CPU when something with a higher clock rate comes out with AM3 platform, if it does. I made the mistake of ordering a T1100 6 core and found out it is usless in my configuration. But back to Artic silver. I have a Noctua NH-D14 and have to remove it to do anything from PSU removal to RAM replacement, and I have to clean off the compound and start over. So I do not think I will ever see the breakin period fruitition. That is why I said there is no breakin period like the compound that came with the Noctua that I went through after 3 installs. But you are right if you build one and never do anything else to it you might get the breakin temps, finally.
  17. Best answer selected by Proximon.
  18. This topic has been closed by Proximon
Ask a new question

Read More

CPUs Cooling Temperature Arctic