Cooling my Q6600

[OldGeek]

I just received hardware to mount my Xigmatech S1283 Red Scorpion using a metal brace instead of the LGA775 pins. Xigmatech makes a package called Crossbow part# ACK-17361. All well and good. I've never trusted the pins to give me a unconditional tight seal between the CPU and the cooler. Several questions come up. They are coming up because I have to take the system apart and remove the motherboard to put on the hardware...can hardly wait.

First is the orientation of the cooler. Vertical or horizontal? Does the contents in the heat pipes have a preference, i.e. are they more effective one way or the other?

Second question. Am using Arctic Silver 5. What way is the best way to apply. Now before you send me off to read the Arctic Silver install guide , remember that guide makes or I presume (hate to assume) that they are referring to a CPU cooler that has a perfectly flat mating surface, thus the tiny bead across the CPU such that putting the cooler in place will ensure that the thermal compound will spread out due to the pressure exerted by the installation of the cooler will squeeze the compound in a manner that will spread out nicely over the cores.

Well the Xigmatech does NOT have a perfectly flat mating surface. It has three pipes that actually contact the cpu surface and strips of the cooling block between the pipes. There is a minute valley between the heat pipe and the cooling block. I would expect that valley to suck up some compound. The interface is hardly perfectly flat. So as I see it I can:
1) Put the bead across the CPU such that it crosses both cores and have the head pipes at a 90 degree angle to this bead. Down side is all three heat pipes do not cross both cores.
2) Put a rice or smaller amount of compound on each heat pipe and run the heat pipes horizontally across the CPU, i.e. 90 degrees different than my #1 and thus all three heat pipes service both cores.
3) use a baggy and lay down a extremely thin layer of compound right on the cpu and again have the heat pipes run horizontally across the CPU such that each pipe runs over both cores.

Now that I have 3 heat pipes perhaps that means they should run such that all pipes run across both cores thus answering my first question on orientation. Now...please educate me with your knowledge and experience and I shall be most grateful.

Oh, my vote is for method #3, however am not limiting the choices to my three ideas.

[Aleslammer]

I faced my 1283 to the rear of the case and it worked well, never tried it pointing to the top. The aluminum portion of my base was a little proud, so I lapped it with 1200 grit sand paper until I brought the aluminum down to the copper and used a X layout of the TIM. Did pickup a C or two having my fan in the pull postion over the push, used the pushpins, had bought the crossbow base but switched to a TRUE before trying it.

Here is a link to a site discussing TIM application.
Best Thermal Paste Application Methods | best thermal past application,thermal paste application methods,thermal compound,thermal interface material,best practices,Best Practices for Thermal Interface Material Paste Application Methods and Thermal Gr

[OldGeek]

Aleslammer...were it in my power I'd give you some ale to slam . The article you pointed me to held the exact answer I had been looking for...the correct way to apply the TIM to an HDT CPU cooler. As I had mentioned, I was concerned about the grooves between the heat pipes and the cooling block. As it turns out, I had reason to be. Although a ROYAL pain in the the backside, was able to carry out the very tedious operation.

A couple of points. I really disagree with the ambivalence of the writer regarding the Intel pin method of connecting the cooler. Using the Xigmatek metal with screws and springs is so much better and I don't need to stay up nights worrying about a pin popping as it did on another rig with a E6420. I found the loose pin before damage had taken place. No, nothing like good solid metal to be on the backside of the board to give peace of mind on keeping the CPU to cooler interface perfectly intact at all times. The bad news is I had an incident of probable... not sure exactly what happened... carelessness on my part such that I ruined the CPU connector on my EP45-UD3R (a half dozen bent pins). Great excuse to upgrade to the EP45-UD3P.

Second point...all this work did not change my temperatures one degree on any core. Core zero is still running 4 to 8 degrees warmer than the rest of the cores. In a fairly idle state the cores run 43, 35,35,37. I've heard that is a normal manufacturing variance. Decided to keep the cooler fan oriented to push air rather than pull. There is a case fan not 2 inches or so...on the other side of the cooler's fan and a huge fan above the cooler on the Antec 900 that really sucks out heat. A seasonal problem that I have been plagued with is the ambient room temperature during the long grotesquely hot days in the Sonora Desert. It's been over 100 on several days in the past couple of weeks already! A wee bit early for these parts but time to get used to it...and brace for the electric bills that go with the high temps.

Again thanks for the help!

[Aleslammer]

Your Welcome.

No bad on temps, my q6600 B3 had about 8c between the hottest and coolest core, after I lapped it that difference shrank to 2c.

Been over 100f where I'm at a couple times so far, hoping it doesn't mean a warmer July / August.

Want kind of Flight SIMs, I play with IL2 and OFF a little.