1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Cooling Sudden change in temps.

Discussion in 'Hardware' started by yougotkicked, 13 Apr 2014.

  1. yougotkicked

    yougotkicked A.K.A. YGKtech

    Joined:
    3 Jan 2010
    Posts:
    251
    Likes Received:
    9
    Hey all, it's been a while since I've posted here, but I could use some help figuring this out.


    A few days ago I noticed my CPU (machine in sig) was suddenly running a bit warmer; I'm not certain of the numbers but if I recall correctly they went from low to mid 30's at idle and 40-50's under load, to high 30's idle and 60-70's under load. I figured it had been a while since I re-seated my heatsink so I reached for a new-ish tube of IC diamond. I put down an application (with great diffuculty, that stuff is hard to apply) and booted up.

    To my horror my boot-up/load temps rose into the high 90's, and the CPU began to throttle. I immediately assumed a bad mount because of how hard it was to manipulate the IC diamond.

    I pulled the heatsink off, grabbed a ~6 year old tube of AS Ceramique, painted both contact plates to a mirror like surface, and re-mounted. I booted up to the same 90+ load temps.

    2-3 more attempts with the ceramique and I was seeing little if any improvement. At this point I knew something else must be wrong, but to be certain I grabbed a new-ish tube of AS5 and mounted one more time, no change.

    Now that I have ranted a bit about having to re-mount my heatsink half a dozen times, here's the details of my cooling system.

    I'm using a Tuniq Tower-120 from 7-8 years back with a modified retention plate, if you aren't familiar with this cooler, here's a link: http://www.bit-tech.net/hardware/cooling/2007/11/01/tunic_tower_120/1

    Since purchasing it, I've replaced the fan with some high-power scythe fan that has not trouble moving air through the dense cooler, it also sit's in the prime location for air exhaust in my case, with two 120 mm exhaust fans above it, and one behind it, matching airflow with the heatsink fan. So airflow is excellent.

    My CPU is one of the 4.5Ghz 2500k's that were so popular a few years back, the core voltage top's out at just over 1.4v according to software readings, I don't recall the exact setting I have in bios, but that should be pretty close. I honestly have no idea what the normal voltage people use at these speeds is, though I always feel like I'm running a bit high.

    While idling my CPU clocks down to 1.6 Ghz and about 1.0v core voltage, temperatures tend to rest in the high 30's right now.

    So, with all that said, I would like some help sorting out what could be causing these super high temp's all of a sudden, I have a few idea's myself but I would like to hear some outside opinions on the matter. Let me know what you're thinking, I'd really like to understand the cause of the problem, even if the solution just ends up being a new heatsink.

    Thanks in advance for your help, and sorry for the long post, I just like to be thorough, but I end up being long-winded.
     
  2. Guest-44432

    Guest-44432 Guest

    I guess your CPU is on it's last legs running at 4.5GHz with 1.4v vcore. It is known to happen that when running an overclock that in time the CPU will degrade...

    Only other option is to de-lid the CPU and apply some new paste to that too. But that is down to you, and I take no responsibility if you kill your CPU in the process.
     
  3. lancer778544

    lancer778544 Well-Known Member

    Joined:
    5 Jan 2011
    Posts:
    2,871
    Likes Received:
    368
    OP has a 2500k T_G which are soldered on I believe so attempting a de-lid would kill the CPU.
     
  4. yougotkicked

    yougotkicked A.K.A. YGKtech

    Joined:
    3 Jan 2010
    Posts:
    251
    Likes Received:
    9
    That could be the case but I would expect to see the system crash more often as the CPU degrades rather than a temperature change.

    Does anyone happen to know if the i5-2xxx series CPU's use TIM or solder under the heatspreader? I seem to recall hearing that they used solder and thus were very difficult to delid. I'll poke around a bit later to find out, but if someone knows off hand please let me know.
     
  5. Shirty

    Shirty W*nker! Super Moderator

    Joined:
    18 Apr 1982
    Posts:
    12,474
    Likes Received:
    1,694
    My money's on a duff heatpipe. Get hold of a cheap/free hsf and see if that helps.
     
  6. TheMadDutchDude

    TheMadDutchDude The Flying Dutchman

    Joined:
    23 Aug 2013
    Posts:
    4,745
    Likes Received:
    522
    Sounds like a cooler problem to me.

    Running an overclock will not suddenly cause your temperatures to rise. Those 2500Ks were strong and could take one hell of a beating. 1.4 volts is nothing. I had mine at 1.5 volts for months and it didn't even go over 72c under max load.

    For your next mount, only apply thermal paste to the CPU rather than to both the cooler and the CPU. Too much thermal paste is as bad as too little. I always only apply it to the CPU and I've never had an issue.
     
  7. Shirty

    Shirty W*nker! Super Moderator

    Joined:
    18 Apr 1982
    Posts:
    12,474
    Likes Received:
    1,694
    If a heatpipe has malfunctioned (rare but it does happen) you will need a new cooler, no fix for that sadly.

    Do you have a stock cooler you can test with? Or even an incompatible one that makers good contact, which you can just hold down whilst temp testing?
     
  8. Guest-44432

    Guest-44432 Guest

    Oops, teach me to have two windows open at once, for some reason I assumed he was running Ivybridge...

    They are soldered on, for some reason I assumed it was a Ivybridge chip. I was inbetween replying to another post.

    CPU's do degrade and produce more heat over time. More often this is seen in GPU's.
     
  9. yougotkicked

    yougotkicked A.K.A. YGKtech

    Joined:
    3 Jan 2010
    Posts:
    251
    Likes Received:
    9
    Thanks for the input everyone,

    The more I think about it the more I think a failure in the heatsink makes for the best explanation. I've never had a heatpipe go bad on me before, so I don't know exactly what can cause them to fail, but the scenario makes sense: A pipe starts to go out causing the increase in temps which prompted me to re-seat, then the stress of being unmounted and remounted pushed it over the edge. I actually have the contact plate mounted with so much pressure the 3mm steel retention bracket bows slightly, so if a pipe was starting to loose contact with the plate, suddenly removing all that pressure could cause it to separate completely.

    I have a handful of stock intel coolers that might fit, but they are the lower-profile ones which come with i3's (I think, got them from a surplus store) they really aren't cut out for the thermal load of a 4.5Ghz processor, but if the mounting gear lines up I may strap a high power fan to it and see if the temp's are more stable with a different cooler.

    typically I would, and I tried a few methods over the course of my multiple re-mounts. I decided on the double application because both the cpu and heatsink contact plates are a little worn out, so I wanted to be sure it worked TIM into the imperfections on both sides, plus my CPU is a victim of intel quality control, and has a concave heat spreader, so a little extra TIM helps bridge the gap.
     
  10. Teelzebub

    Teelzebub Up yours GOD,Whats best served cold

    Joined:
    27 Nov 2009
    Posts:
    15,796
    Likes Received:
    4,484
    I use the diamond ic paste a lot the best way is to warm it first and apply a large grain of rice sized blob to the centre of the CPU only and let the pressure of the heat sink spread it
     
    Last edited: 14 Apr 2014

Share This Page