dasparx

Thanks Mate, im just trying to max that sucker out on a budget board.(proving that chipsets with intergrated gfx pwn ^^ )

__________________
Got a free PCI-E x1 Slot, and want to do more folding? Click here!

psykoon

I just removed the IHS off of my e2180, and I was wondering if I could get some ideas on how to mount the heat sink without cracking the core?

I have an Arctic Freezer 7 pro with a bolt through kit instead of push pins.

My initial thought was to get some foam or electrical tape and make a square around the core. The thickness should be slightly higher than the core so it supports some of the weight of the heat sink. Then I was going to make some standoffs for the 4 feet of the heatsink so that I can screw the bolts in all the way without having to worry about applying to much pressure to the core.

Any thoughts and other ideas are welcomed...thanks

SRV

I have very similar situation. E2160 with IHS removed, cooler - AC Freezer 7 Pro.

Mounting system is push-pin, but I'm willing to make my own mounting system, the main problem is how to avoid core cracking.

I considered puting some foam between cpu and cooler base, just like you, I believe that would help to distribute pressure evenly on CPU, especially because there isn't anything that is holding cpu in the socket except pressure made with cooler.

I believe some foam is enough to prevent core cracking, and just like you said, foam needs to be a bit higher than cpu.

My first post.

GigaByte

Slightly snug is best, start loose and slowly tighten, you should come to a point quickly where making it slightly tighter makes the temp worse or temps do not change. That is a good indicator that you hit the sweet spot for tightness. In general its hard to crack a core unless you screw down the cooler so tight screws start stripping or something, majority crack their cores during the removal of the IHS on soldered CPUs.

@SRV Welcome to OCN, please go here to fill out your system specs http://www.overclock.net/specs.php?do=addsystem

__________________

SRV

Thanks. In meantime I already did this, i got ~8C lower temps, but with less noise. With same amount of noise, I guess difference would be bigger.

In begining:


After 20min:

You could notice a little scratch in the upper right corner, fortunately there isn't anything. I was very lucky, CPU is still 100% stable.

Very important thing is to remove this:

as stated in this thread:
http://thetechrepository.com/showthread.php?t=23

That was first attempt, because I didn't have cooler with bolt-thru mechanism, base didn't touch core, so i had to wait 'till I manage to work this out.

2 days ago, I attempted to finish this mod:

First I isolated mbo from getting scrached.

Further:

How it looks like from above.

Now, some foam to protect core and to distribute force evenly:


It was necessary to tighten the upper two screws much more than lower two, because cooler was hanging a bit and contact was bad, temps rose up to 70C.

Unfortunately, I don't have pictures with cooler instaled, I was pretty nervous to get this done.

In the end, CPU needs slightly less voltage for 3200MHz: instead 1.384V it is enough 1.368V. Also, it is stable on 3350MHz @1.44V, bit it's too much for 24/7, altough temps are 55-56C, just like 3200MHz @1.384V prior to IHS removal.

GigaByte

Very nice! As you noticed once the CPU is cooler and the heat is able to get removed faster it won't need as much voltage for the same speeds, what is also neat is you can have 3.3Ghz on stock cooler at lets say 1.36v and temps at 66c and it refuses to overclock anymore, then use an after market cooler and have the same temps on the same volts you will notice the test using the after market cooler will tend to overclock more even if the temps are the same, this is because once more heat is introduced the cooler is able to remove it much faster than the stock one can. Its hard to explain but once you see it in action you will understand it better.

Also you have an E2160 which is a 65nm CPU, meaning you can go up to 1.50v (says processorfinder.intel.com) safely, 1.44v won't harm it at all if you're worried. In the long run nice job there

__________________

SRV

Thanks. I understand what you mean.

I know, 1.44V is ok, but it is 3350MHz, compared to 3200MHz with just 1.368V, that's much cooler/quieter (CPU fan speed depends on temperature) - 1800rpm vs. 2300, and MBO components are cooler too.

JEK3

I screwed up. I tried removing my IHS, but I brain farted and didn't keep the razor blade flat to the board. I cut through the green coating to the copper traces on the board, and now I get no response when I try to boot.

Just airing my dirty laundry so no one repeats my mistakes.

Fortunately it was a Celeron 1.5GHz socket 370. Gives me an excuse to upgrade to a Q9300.

__________________
Rig complete, and folding for 37726...

Brutuz

I tried this, couldn't get my razor under the IHS, to cut the adhesive, I'm going to try again now though, will say if it works!

__________________

Azazel

I took the ihs off my old athlon x2 and I just used a little plastic razor holder from walmart. It cost like 5 bucks for the holder with 5-10 razors in the package. Then you can just put the razor on the part to be cut, put the other side on the cpu up to something hard(tile/countertop/table) and push in. The plastic will hit the side of the thing before the razor goes far enough to damage the stuff inside. Just do it once per side and pop the ihs off.

__________________
X-fi digital out -> Buffalo DAC ES9008 -> Firstwatt F1 -> AKG K701(balanced stock cable)/AKG K1000/Victor HP-DX1000(balanced stock cable)