Core running HOT, but only 'sometimes'

buononutbuononut Michigan
edited June 2007 in Hardware
Hi all. Is there any explaination for inconsistant CPU temperatures? It appears completely RANDOM (as impossible as that is for computers) just how hot my CPU runs any given time I fire up my rig.

AMD Athlon XP 2600+
I upgraded to a new heatsink (aluminium but has copper core) w/70mm fan, plus that silver 5 goo on the back. I figured it to be more than sufficient to help cool a bit better. I'm not over-clocking it.

Normal conditions (browsing, playing music) can be 40-58*C, with very little fluxuation.
When gaming (say, Star Wars: Empire at War) it runs around 55-59 on a good day 60-64 on a bad day. Yesterday, chart shows it hit 68 !... :sad2: thus, I had to write this thread.

Did I install fan and/or paste incorrectly? The Athlon has that small, silver, rectangular raised area in the middle... is that the only spot that should get the paste? Either way, how much?

THANKS.
(Sorry I couldn't make the Memorial day bash.)

Comments

  • DonutDonut Maine New
    edited June 2007
    Try it with the side of the case removed. This may (hopefully) drop your temps. My Athlon chips I tried to keep them below 50c.

    What are you using to monitor your system temps?
    Could you list your system specs including the case and any case fans?

    Here's a link on how to apply Arctic silver.http://www.arcticsilver.com/ins_route_step2amdas5.html
  • buononutbuononut Michigan
    edited June 2007
    This link definately helps. I have both ArticSilver and AS Cermaique, so now I have instructions for both and will try a few tests (i'm sure I've got too much compound in there). Hopefully, that will explain the eratic temps.

    After, I will try to take the side panel off (if still no luck) but I've tried to ensure good flow through.

    Thanks much.
    buononut
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited June 2007
    What exactly is your heatsink?
  • LeonardoLeonardo Wake up and smell the glaciers Eagle River, Alaska Icrontian
    edited June 2007
    i'm sure I've got too much compound in there
    I very common user mistake. You just need a very thin patina, just enough to fill microscopic irregularities on the CPU core and bottom of the heatsink. Arctic Silver's site has an excellent guide (with pictures) on applying their thermal paste correctly.
  • buononutbuononut Michigan
    edited June 2007
    What exactly is your heatsink?

    *********
    The heatsink is called the TR2 M3 by tr2tt (ThermalTake).
    I can't yet post links so you'll have to google "tr2tt TR2 M3".

    I'm trying not to dump $$$ into this machine as a whole new PC is pending. Again, I have a Athlon XP 2600+ (Socket A) so i have little/no room to upgrade.

    It has the full copper bottom but was not very smooth, so I sanded and buffed the contact side to a shine with just water and fine sandpaper (no compounds, etc.) and did the same to the back side of copper and the aluminium portion to ensure good contact.
  • buononutbuononut Michigan
    edited June 2007
    I just read around a few other forums on the topic.

    My unit is 4+ years old:
    * When might damage occur to the CPU after time?
    * Also, what TEMPERATURE will cause instant damage?
    * Lastly, what symptoms might a heat-damaged CPU exhibit?
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited June 2007
    Heat damage will be manifested in computer seizing, failure for the PC to boot, or failure for the PC to post. CPUs are generally rated to run at less than 60*C for prolonged periods of time -- years -- and running them at 50-55C isn't a big deal at all. When you start getting into 70-80*C, you run the risk of instant damage.
  • buononutbuononut Michigan
    edited June 2007
    Not much luck experienced, even after following the instructions properly. The PC will start up around 42-45 for a while, but once it heats up after doing a few tasks, it stays around 52-55 instead. Bummer.
    I think the unit just need more space inside (HP a220n is the smaller ATX size).
    Is water cooling really the best way to go, if I'm going to get a new PC in the future?
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited June 2007
    No, not really. Good air-cooling is just as good as inexpensive water cooling these days.
  • buononutbuononut Michigan
    edited June 2007
    Thanks for that. I'm going to try and flip the CPU fan around next (removing the side door did little to help).
    Maybe the problem is that I have a 'trifecta' of sucking going on right near my CPU: my PS is pulling air up and out from just above my CPU, and the rear exhaust does that same just below that. The CPU fan may have no air to blow on heatsink.
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited June 2007
    Nah, it's a good thing to have the fan blowing directly onto the heatsink. It disperses the hot air off of the fins and into the case, thereby letting your exhaust fans take it out. :)
  • buononutbuononut Michigan
    edited June 2007
    Thanks for all the help, Thrax. After a quick test, you are definitely right.

    So, now I flipped the exhaust fan in the back of the PC around so it is brining in more air for the CPU fan to blow onto the heatsink (I may put a new hole and a fan in the side panel instead for this purpose in the near future). The PS fans are strong enough to handle exhaust duties back there, I think.

    My BIOS hardware monitor shows ~1600 RPM for the CPU fan but once Windows and "SpeedFan" loads, the RPM comes up to ~2600 RPM and the temps slowly dropped from 55* to 41*, where I am now. =)

    Also, I noticed my GPA card is dumping lots of hot air in the bottom front of case (near HDD). I had an intake fan there, but I flipped it to pull that hot air right out.

    Summary: I sort of have 2 air zones now: (1) CPU, PS (2) GPA, HDD. This may not be kosher, but it seems to be working.
  • Your-Amish-DaddyYour-Amish-Daddy The heart of Texas
    edited June 2007
    Yeah it's always good to keep heat sources separate, because unlike cold, hot temperatures stack to almost 1.5:1. Cold barely makes 1.1:1, which sucks but when you think about it physically, cold air falls because hot air is lighter and tries to get above the slower particles......Yeah, whatever.
Sign In or Register to comment.