Tip: Use Teflon on HS or WB mounting posts

I don't know if anyone has suggested this before, but it helped me so I though I'd mention it.

I recently setup a P4 2.4C system with a White Water waterblock. I had removed the IHS from the CPU to improve my cooling.

I got the block mounted, but in the process of mounting it, it felt like the block was hanging up on the threads of the mounting posts. Because the heat spreader wasn't protecting the CPU, I didn't apply much force to ensure the CPU was making good contact. But I wiggled the block around until I thought it was seated okay and bolted it down.

I started the machine up and was getting load temp measurements around 42C when I had been getting 65C with stock air. I thought my cooling was good. But, I was not able to get the CPU above 3.06 GHz.

After weeks of being frustrated about the relatively low overclock I was getting, I decided I had to try reseating the waterblock. This time, before I put the block back on, I sprayed dry teflon spray lubricant into a cup and used a cotton swab to spread it around the mounting posts and the mounting holes of the block. This time the block slid smoothly over the mounting posts, and my overclock went up to 3.3 GHz. The load temp reported by my mobo only dropped 1C.

So, my advice to anyone who's block doesn't slide smoothly over the mounting posts: get some dry teflon spray and see if it helps.

By mounting posts, you mean the allthread that sticks through the four corners of the heatsink/mounting plate, right?

Really wierd that it would have such an impact on your overclock... any theory as to why?

Lithan said:

By mounting posts, you mean the allthread that sticks through the four corners of the heatsink/mounting plate, right?

Really wierd that it would have such an impact on your overclock... any theory as to why?

Click to expand...

Yes, the allthread.

My best guess is that the block wasn't laying flat on the die. It may have been making ok contact with the part of the die where the thermal diode was. This would explain the temperature measurement. But, some other part of the die was not getting cooled well. Looking at the imprint in the TIM, it seemed that in one corner, the paste may not have been squeezed out as well.