Hello friends,
 
        I think the hardware Gods are still mad at me for my juvenile manslaughter of my Athlon with liquid nitrogen.  Something seems to be up with my new rig's CPU temps.
 
        I built this new machine (around air cooling), precisely to be cool.  The case is an Antec 900:  It has 5 120mm fans on full blast, and a 200mm fan on full blast.  The CPU is an Core 2 duo 6700.
 
        The first cooler I installed was a Thermaltake Big Typhoon with AS5.  BIOS reported temps of about 22 degrees.  When XP was up and running, however, temps averaged between 40-44 degrees without any programs running.
 
         I've never fully trusted Thermaltake's aircoolers, and I'm really not a fan of AS5 (I did like AS3 though).  Yes, I know its the best, but I just don't feel comfortable with the new application instructions (ie, a subjectively thin line which is to be smeared on by the heatsink).
 
        So I get a Zalman CNPS9700 LED.  I didn't use the AS5, I used the supplied Hg-based STG1 thermal grease (which I hear is just as good as AS5 -- and a hell of a lot easier to apply).  
 
        Yet, core temp still posts an average of 38-41 degrees!  An improvement over Thermaltake, to be sure, but not the target 26 degrees necessary for O/Cing.  I'm afraid to bring CPU to load.  What gives?  BIOS posts between 19 and 24 degrees, but not Coretemp.  I've reseated the cooler 3 times (using Arctic Silver's cleaning products each time), and coretemp posts nowhere near 26 degrees with XP booted up.  What the hell?  Any advice??

you got the asus stuff off in bois and have you installed the asus oc stuff in windows
my bother had same issues until he uninstalled it all and turn stuff off in bois
 
plus i belive CD2 reads temps of die not heatsink so i will read higher(might be wrong havent got my 6700 yet)

Are you referring to the mobo drivers?  I tried uninstalling them and reinstalling them to no avail.  I even did a fresh format and XP install.  no change.  To what ASUS OC windows stuff are you referring?  Didn't come with the hardware and isn't on ASUS' website.

That's not the coolers I'm sure. I have the 840 and Big Typhoon and I get about 35c idle. Must be reading the wrong temp. You should install back the Big Typhoon if you still have it along with a thin layer of AS5. One way you can test it is run some intensive applications and games. Then touch the cooler's base plate. That's one of the way you can measure the heat. I don't really trust bios reading as it gives wrong voltage and temp reading. At least my mobo does, giving me 199c idle.    
 
Well, check bios and if you can adjust the cpu voltage to the minimum operational voltage requirement. Sometimes the motherboard just put out over the minimum to make sure the cpu would run fine. And you can even adjust the voltage lower than the it's cpu's minimum voltage requirement. Just do some test after wards to make sure it's stable at full load. Use Orthos for testing.  
 
Also the motherboard's temp sensor is right under the cpu, not on the cpu heatsink so it does measure the temperature of the space between the cpu and socket.  
 
But simple way like I mentioned, test the cpu with the most intensive application or gaming you run on that pc, and touch the coolers base plate. If it's not hot, but warm then you're good to go. You can use the Orthos for further testing and stability.

I'll load up Flight Simulator X and give it a go.  Too tired for that now.  I would expect the 840 to run hotter than the c2d.. you're not using liquid cooling anymore?  Off topic, by the way, but your icon looks like my girlfriend's little bastard maltese that chewed up my 600w power supply's 12v connector.  shame it wasn't on j/k
 
EDIT:  At load, I'm at 53 degrees.  Not very safe for overclocking it would seem.  dammit!

no imeant all that ai nos stuff and cool and queit etc etc turn it all off so your cooling is just running stock speeds no interferace etc
 
ps make sure they are uninstalled and also off in bois as well
 
see how you go then my brothers droped 10 deg after he turned all that off.

First I was using the Big Typhoon on the 840. Now I have to go water cooling at 4Ghz overclock, the Big Typhoon can only cool about 150 watts of heat output. The 840 at 4Ghz @ 1.55vcore puts out over 200 watts of heat at full load.  
 
Again don't really much on the reading of the bios and softwares, it might be wrong.  
 
How much AS5 are you using? Just asking.  
 
At 53c at load well it's okay temp and as long as it doesn't go over 60c-65c then your cpu is on the safe side.

As for the AS5, I applied about a 15mm long line down the center of the heat spreader just under the width of a toothpick.  Presently, however, I'm using the Zalman compound, which is very easy to apply in thin layers (it comes with a brush ala wite out).
 
Yeah, 53 degrees is safe, but that's only at the stock speeds of 2.66 Ghz.  I'd be really dissapointed if I couldn't overclock a 6700 with a beefy aircooler like the Zalman or typhoon .
 
the lower temps are reported by BIOS, but the 41-54 degree temps are reported by Coretemp.  I just really feel like I'm missing something here because Wusy's c2d OC guide says I should be idling at 26 degrees.

I'm sure it's more likely a misread. I don't believe that Core 2 would run that hot at idle. By touching it would be a good test. If it 65c and over it is hot to the touch and if not then your cpu is not cooking.  
 
How about try the paper thin AS5 application method.  
 
Here.

 
This will give you a very thin uniform layer.

I really hope it's a misread.  I never trusted my mom when I was a kid and she would feel my forehead to ascertain whether or not I had a fever, and many people in the forums absolutley swear by what coretemp says.
 
What could possibly cause a misread?  Is this common?

My motherboard for one reads 199c idle at stock and sometimes it reads fine like 28c idle. I don't think it's common, not that I know of but it is not to trusted 100% either.

53C is cool, you're 32C away from the throttling point.

Hi Richard.
 
Let me clear this up for you since I have just gone thru this headache myself.
 
Here are the facts:
The newer processors, Core 2 Duo and last few, no longer read the temperatures of the bottom of the package.  By the time this temperature signals a disaster nothing can be done.  Furthermore reading the temps that way used a system that ran to a chip on the motherboard that then converted an analog reading into a temperature.  The circuit path amongst other things often rendered that value off by 10's of C.   This means that the values two different identical computers might read could be off by more than 10C.
 
Intel decided that they were sick of getting complaints about dead and damaged CPUs that users swore they kept below THE maximum so they changed everything.
 
They decided they would flat out prevent damage to the CPUs by the method of throttling.  When the CPU reaches a certain temperature they just start blocking machine cycles from occuring.  They essentially control the CPU temperature directly via throughput limiting.
 
To accomplish this they could not use the old slow inaccurate temperature of the bottom-of-the-case method that also depended on the mutherboard designers.  They came up with DTC or Digital Temperature Control.  They put all the measurement circuitry directly onto the CPU die. The sensor and the analog converter.  Then they even calibrate it!  This all hooks to the internal throttling circuitry which protects the CPU and allows extremely fast nearly instant throttling.
 
What does this mean to you? Well your motherboard does not understand this.  What you are seeing is either utter nonsense or the actual internal die temperature which is always far higher than case temperatures.  
 
As a CPU owner this is vastly superior from the previous methods because there are; no die-to-case variations, no heatsink-to-case variations, no temperature sensor-to-case variations, and no temperature sensor-to-analog to digital converter variations.   The DTC value is THE actual calibrated temperature of THE actual point you really care about.
 
Wanna OC?  The DTC value is THE absolute best measure of how far you are pushing the CPU in the thermal realm.  (I say thermal realm because they haven't got any protection scheme in place for detecting overvoltaging the core which causes actual punch-thru damage of the die circuitry if pushed too far.)
 
So quit worrying about the temperature you are seeing with your present method and embrace the superior temp data you now have access too.  How you ask?  By gabbing a copy of CoreTemp presently Beta 0.94.
 
It understands all this AND actually knows what the allowed die temperature is!  When you run it it shows the actual core temp of each processor and the allowed maximum at which throttling will occur and you are at the treshold of damage.
 
While I type this my;
 
ASUS P5W64 WS
Core 2 Duo 6700
Zalman 9500 LED
2G DDR2-800
eVGA 8800GTX
Dual 10,000RPM WD 74Gig SATA Raptors striped in a RAID 0
Running XP64
 
Is running:
Core #0 at 39C
Core #1 at 37C
The maximum allowable die temperature, called Tjunction, is 85C
 
Meanwhile the motherboard is reporting an absurd yikes! 93C earlier it was reporting 63C so it's all over the map stoopid!
 
So don't worry. Be happy!
 
CoreTemp source:
http://www.thecoolest.zerobrains.com/CoreTemp/
 
Tell me what you get on yours.

 
Are you sure you put the line the right way? You may have put it across the heat spreader rather than down (or vice versa), so that the line straddles the two cores rather than covering them. Hmm, that is difficult to explain in english, I'll try and draw what I mean:
 
If the cores are like this:
 
*
*
 
You may have done this:
 
--
 
Rather than this:
 
!
!

Why don't you OC and see what your temps are?  As long as your under 60 at load you should be fine.