Overheating? Check these first (Beginners) ?Sticky

[IrishRangerWingMedic]

I am a beginner who had serious problems with overheating in my relatively new system.

I found the following things should be checked first - Obvious to most experts here.

1) Check temperatures with software application also - eg. Speedfan or Core Temp

2) Watch the temperature for a few minutes. Does it rise steady while idle? If so then something reallly basic is wrong

3) Open the case and power on - Watch your PCs fans -
Are they all spinning? (I accidently unplugged one and whole system temp went nucleur)
Are they all clean?

4) Examine fan outputs - do the vents have space? (Make sure they aren't hitting a wall)

5) If you have moved about hardware, make sure items aren't blocking air flow.

6) If processors have been changed, make sure you clean away the old thermal paste and apply fresh thermal paste. Also make sure CPU heat sink is rammed evenly onto that processor - no movement allowed now!

Thermal Paste - Why use it? - Because the surfaces of the processor and heat sink fan are not completely flat, even when bolted together, and so, as the air between them is a bad conductor, a paste is applied.

Which one? - Everyone loves Artic 5, but I'm not sure theres not much difference between manufacturers.
A Silver based paste conducts heat well, but also conducts electrical charge, so don't short the motherboard circuits by being sloppy!

There are 2 schools of thought on application of thermal paste -
1) the spread it like butter school and 2) The squashed pea in centre school.

I go for the later as my heat sink fan has a circular copper contact, so squashing makes sense for a nice uniform circular film. Others say that spreading will introduce air pockets.

Remember - Less is more - You are only filling those microscopic surface imperfections - Only the thinest application is necessary - about the size of a very small pea (Much more comes in the tube).

Also, it takes a little time for the paste to spread and settle - I have read about an hour? I have definately noticed this, as there was no change to my processor temp immediately after application, but 3 hours later the temperature had dropped 40%

My Core2Duo at stock 2.4GHz was starting from 40C and rising to 62C idle within 6 minutes!

By fixing 2 basic problems, I got to 38C idle and 55C loaded -

Problem 1 - Fresh paste not applied after removed and installing processor
Problems 2 - System (Case) Fan was not reconnected!

Hope this helps.

 

[!gOtVolTage!]

Hey great post .Check this out , I have some info that i have discoverd for applying thermal paste to a lapped CPU / IHS with lapped Waterblock/Heatsink . Smooth ,flat and lapped/polished surfaces will get way lower temputures if you apply thermal paste as described below.

When Applying the Thermal paste AS5 ,MX-2 or whatever you have with the (Dot method) for AMD X2's or use the (Thin line Method) for Quads .Make shure to spread the PASTE PAPER THIN across the WHOLE Polished IHS One Dot of paste will cover a AMDX2 paperthin!,Do NOT just slap the waterblock or heatsink on the IHS if the paste is not spread out paper thin or it will be too thick and not lower your Coretemp as good when using a( LAPPED IHS with LAPPED WaterBlock or HeatSink). The Dot and Thin line method is for UNLAPPED IHS/Wateblocks/Heatsinks only not LAPPED POLISHED ONES

I have tested this several times with Overclocked Cpu's running @ and above 3.5ghz with upto 1.63vcore. I used a Bigtyphoon LAPPED /POLISHED first, Themalright120 ExtreamLAPPED/POLISHED second, and lastly im now using a Apogee GT non BOWED Waterblock LAPPED/POLISHED .

I recomend to anyone that has LAPPED/POLISHED the IHS and the WaterBlock/Heatsink to try this. It Really makes a differance and does mean that extra Mhz can be achieved or just piece of mind. I guess im just a lil [H]ardCore but am always looking for the next best Oc. I discoverd this meathod only after i removed my Heatsink to see how the Paste spread out when using only a DOT. When i saw that this dot was only spread the size of a nickle but still thick as crap ,I now know that installing AS5 by thier method/instructions was not right for a polished surface! I also though to myself "This is why my temps must be same after lapping the IHS and Heatsink ". You would think temps would go down a lil bit,but on a smooth surface AS5 and other compounds/paste become insulators!. In short ,i then spread the remaining AS5 that was on my IHS to paper thin , if you so much touch it it would smear paper thin and like magic it went from 36c idle to 23c idle @ 22c room temp!
This is a drastic drop over the AS5 Correct method !. I really thought i installed my Waterblock wrong the first time! So i repeated this test 5times with the Apogee, ThermalRight and BigTyphoon using AS5 DOT METHOD and my PaperThin Method.This gave me 30X times to practice propper waterblock/heatsink installs

Note: SouthernComfort Helped out ,but did not adversly affect any of the test results listed below.

Results are based on using CORETEMP.095 and SpeedFan@22c room my temp. Averages are a result of 5X each Method! I think i was sick when i decided to do this

Apogee GT LAPPED/POLISHED with mcr120 and mcr220 with 54cfm fans .

6000+3.5ghz 1.58vcore AS5 DOT method = 36c idle
6000+3.5ghz 1.58vcore AS5 PaperThin method= 23c idle

Net Temp differance= 13c

ThermalRight 120Xtream/1x110cfm LAPPED/POLISHED

6000+3.5ghz 1.58vcore AS5 DOT method = 39c idle
6000+3.5ghz 1.58vcore AS5 PaperThin method= 31c idle

Net Temp differance= 8c

BigTyPhoon 110cfm LAPPED/POLISHED

6000+3.5ghz 1.58vcore AS5 DOT method = 44c idle scary!
6000+3.5ghz 1.58vcore AS5 PaperThin method= 38c

Net Temp differance= 6c


Have fun ,Try it see what im talking about.

 

[IrishRangerWingMedic]

I'm afraid to lift off my THS again , but would be interested to see the result of my pea squash method.

Your technique certainly seems to have yielded great results - I'll definately have to remember that for again.

As a matter of interest, my AKASA silver Based Thermal Compound did actually come with a credit card sized spreader.

 

[!gOtVolTage!]

Akasa and MX2 are way better than AS5 . My results with As5 are pretty decent . I will be testing same way with the new MX2 paste but only with my Apogee water block. I takes so long swapin out coolers .So ill just stick to the one for now,but its interesting and fun


Just remember my results are for LAPPED /POLISHED IHS with a LAPPED /POLISHED WaterBlock/Heatsink. Both must be lapped/plished for the test results i posted.

We need more people to jump in on this. Hmmm the this could be the making of "The Ultimate Quik Guide to Cooling tread" for ya