Sanded down the IHS of a Q6600 to the silicone.

MrRuslan · 09-05-08, 11:02 AM

I wanted to do this for a while and finally got the patience to do it. It took 6+ hours of sanding by hand and here is the result:

s you can see the chip somehow survived. I took a screen shot in the morning and I was in a rush. My temperature before this procedure was 75-80 with both the cooler and the IHS lapped. After it will not go above 56 at 3.2 Ghz and that is with an additional +.1 to the vcore on top of the minimal needed to run 3.2 stable. this is with 1.375 at full load and I need 1.275 for that speed to be stable. I want a Solid 3.6 GHZ out of this. The Q6600 I have has a vid of 1.2875 are reported by coretemp and the cooler I'm using is the Tuniq. By the way a lot of people seem to think that the core and the IHS are soldered and this is simply not true. They use some hind of very strong silver adhesive that is no fully solid. No matter how good of a cooling system someone has, if the IHS does not have good contact between the die and the bottom of the IHS the temps will not be the best they can be. I did this ot get the most out of my gear.

I used 100 grit wet-dry and I was lapping without water because water killed the sand paper way to early. You have to completely shave off the top part of the IHS and this will take 6+ hours. If you rub too fast it will get very hot so when it starts to heat up pause and continue. after you the the whole top of proceed with caution. When i was doing it one part of the IHS was taller than the other so i almost panicked and gave up. you can pretty much see it in the first picture that it was not leveled. Keep going with a 400 grit until all the silver paste is gone and the cpu looks roughly like the second picture. then finish it off with 800 and stop. This is not a beauty pageant so there will be scratches on the silicone. they look deep but they are nothing to worry about. The goal is to have the dies leveled with the outer part of the partly decapitated IHS so whatever will be rolling your monster will not crush them into tiny bits. You have to carefully remove the cpu clamp from the motherboard. You cannot use a pushpin cooler. In my case since 2MM disappeared from the CPU my cooler was not holding it very tight so I used washers beetween the spring and the mounting plate to give it some extra hold. I picked it up by the cooler and it was fine, I also rotated the motherboard in the air by the cooler and it did not budge. I even picked it up by the cooler while the power is on to make sure there is good contact. There will very little clearance between the the cpu socket and the bottom of the cooler after this so make sure you do not shave off more than absolutely necessary. Make sure your cooler is completely flat at the base. Proceed with caution. I will follow this up with some pics later.

How to clean it: I used the arctic clean solution. Fill a small glass to a point where the cpu is submerged and dip the cpu and then use a toothbrush to make sure all that copper dust/paste is off of everything u can let the cpu soak in it but do not leave it there for more than 5 minutes. do not use a plastic or Styrofoam cup because the arctic clean will dissolve it and it will not be cool. After step 1 use the step 2 bottle and dunk it in, wait and wipe off. Use a hairdryer to completely dry and enjoy.

OldSkool · 09-05-08, 11:08 AM

Very very nice work, and a great write-up

MrRuslan · 09-05-08, 11:13 AM

I will run this test for 24hours to make sure nothing is wrong with the temps.
Looks like there is an option for all those less than perfect cpus.

ghost_recon88 · 09-05-08, 11:14 AM

Very nice. Reminds me of when I took the dremel to mine. Now I wannah see someone do this with a 45nm quad, but no one around here has the balls to

And don't look at me, I don't even own a 45nm quad

MrRuslan · 09-05-08, 11:19 AM

Quote:

Originally Posted by ghost_recon88

Very nice. Reminds me of when I took the dremel to mine. Now I wannah see someone do this with a 45nm quad, but no one around here has the balls to

And don't look at me, I don't even own a 45nm quad

I would do it, but i don't have a 45nm quad. I do have a 45 dual and after I test it and see that it has some potential then I would do this to it.

nd4spdbh2 · 09-05-08, 11:25 AM

damn thats alota work to expose the cores.... i woulda done the knife blade and blowtortch trick to pop the entire ihs

MrRuslan · 09-05-08, 11:27 AM

Quote:

Originally Posted by nd4spdbh2

damn thats alota work to expose the cores.... i woulda done the knife blade and blowtortch trick to pop the entire ihs

I wanted to keep the outer edges of the IHS to prevent the cooler from crushing the cores.

madhatter256 · 09-05-08, 11:27 AM

The first picture was scary! When I first saw it, I thought you were going to say how you failed at it.

Good job on it. Hope you wore a mask and did it where the copper and zinc will not get circulated in your home. It can cause health problems and the metal can build up on your motherboard and short it out.

xtkxhom3r · 09-05-08, 11:32 AM

thats awesome bro i would do it to my q6600 but i dont really need too lol.... at 3.6 my max temp with prime is 52....so yea.....

good write up your the 2nd person i have ever seen do this

ratbuddy · 09-05-08, 11:32 AM

Quote:

Originally Posted by MrRuslan

I wanted to keep the outer edges of the IHS to prevent the cooler from crushing the cores.

Awesome work, and

nd4spdbh2 · 09-05-08, 11:35 AM

Quote:

Originally Posted by ratbuddy

Awesome work, and

SERIOUS, that is some friggin AWESOME work!

dicko555 · 09-05-08, 02:37 PM

Quote:

Originally Posted by nd4spdbh2

SERIOUS, that is some friggin AWESOME work!

+1 on that.

I would love to try that but I got nervous just lapping the CPU down to a nice flat semi mirror finish.
OOOooHHhh bye the way that really is a fantastic job and very nice review of the process.

Joe Camel · 09-05-08, 03:39 PM

wow... just wow

major props

Ashura · 09-05-08, 04:33 PM

Quote:

Originally Posted by Joe Camel

wow... just wow

major props

What he and everyone else said!

MrRuslan · 09-05-08, 05:18 PM

Thanks for all the props and the warm welcome

Here is an update:
I left it to prime for 16 hours and it passed.

I turned off the AC and close all the windows before going to work. I came back and took this screen shot. The ambient temp was 27C at the time of the screen shot.

Sam__ · 09-05-08, 05:29 PM

Lord, you have some major balls. props.

petteyg359 · 09-05-08, 05:37 PM

Wish I was brave enough to do that to my Q9450...

MrRuslan · 09-05-08, 06:22 PM

Quote:

Originally Posted by petteyg359

Wish I was brave enough to do that to my Q9450...

Practice on a Celeron DC

MrRuslan · 09-05-08, 07:24 PM

This is how much you need to take off. I have a celeron that i might make a ginnypig and take it to a power tool carefully...

SolidxSnake · 09-05-08, 07:41 PM

That looks absolutely AWESOME. Excellent work, and major props to you.

What I would have done was gotten it to the point in your first picture, separated the inner IHS pieces stuck to the die and then just torched those off instead of the whole IHS (so you still have the borders). That way you don't risk sanding the dies too far.

09-05-08, 11:02 AM	Thread Starter #1
MrRuslan New Member Join Date: Sep 2008	Sanded down the IHS of a Q6600 to the silicone. I wanted to do this for a while and finally got the patience to do it. It took 6+ hours of sanding by hand and here is the result: s you can see the chip somehow survived. I took a screen shot in the morning and I was in a rush. My temperature before this procedure was 75-80 with both the cooler and the IHS lapped. After it will not go above 56 at 3.2 Ghz and that is with an additional +.1 to the vcore on top of the minimal needed to run 3.2 stable. this is with 1.375 at full load and I need 1.275 for that speed to be stable. I want a Solid 3.6 GHZ out of this. The Q6600 I have has a vid of 1.2875 are reported by coretemp and the cooler I'm using is the Tuniq. By the way a lot of people seem to think that the core and the IHS are soldered and this is simply not true. They use some hind of very strong silver adhesive that is no fully solid. No matter how good of a cooling system someone has, if the IHS does not have good contact between the die and the bottom of the IHS the temps will not be the best they can be. I did this ot get the most out of my gear. I used 100 grit wet-dry and I was lapping without water because water killed the sand paper way to early. You have to completely shave off the top part of the IHS and this will take 6+ hours. If you rub too fast it will get very hot so when it starts to heat up pause and continue. after you the the whole top of proceed with caution. When i was doing it one part of the IHS was taller than the other so i almost panicked and gave up. you can pretty much see it in the first picture that it was not leveled. Keep going with a 400 grit until all the silver paste is gone and the cpu looks roughly like the second picture. then finish it off with 800 and stop. This is not a beauty pageant so there will be scratches on the silicone. they look deep but they are nothing to worry about. The goal is to have the dies leveled with the outer part of the partly decapitated IHS so whatever will be rolling your monster will not crush them into tiny bits. You have to carefully remove the cpu clamp from the motherboard. You cannot use a pushpin cooler. In my case since 2MM disappeared from the CPU my cooler was not holding it very tight so I used washers beetween the spring and the mounting plate to give it some extra hold. I picked it up by the cooler and it was fine, I also rotated the motherboard in the air by the cooler and it did not budge. I even picked it up by the cooler while the power is on to make sure there is good contact. There will very little clearance between the the cpu socket and the bottom of the cooler after this so make sure you do not shave off more than absolutely necessary. Make sure your cooler is completely flat at the base. Proceed with caution. I will follow this up with some pics later. How to clean it: I used the arctic clean solution. Fill a small glass to a point where the cpu is submerged and dip the cpu and then use a toothbrush to make sure all that copper dust/paste is off of everything u can let the cpu soak in it but do not leave it there for more than 5 minutes. do not use a plastic or Styrofoam cup because the arctic clean will dissolve it and it will not be cool. After step 1 use the step 2 bottle and dunk it in, wait and wipe off. Use a hairdryer to completely dry and enjoy.
	QUOTE Thanks

09-05-08, 11:08 AM	#2
OldSkool Rest In Peace Join Date: Feb 2008 Location: Lewisville, TX	Very very nice work, and a great write-up __________________ No Rig currently :( New Rig to be built in coming months......i7? Maybe :) "I Reject Your Reality, and Substitute My Own" E8400(CO) @ 4.2Ghz with 1.312v/4.5Ghz with 1.38v - Retired Heatware[/SIZE]
	QUOTE Thanks

09-05-08, 11:13 AM	Thread Starter #3
MrRuslan New Member Join Date: Sep 2008	I will run this test for 24hours to make sure nothing is wrong with the temps. Looks like there is an option for all those less than perfect cpus.
	QUOTE Thanks

09-05-08, 11:14 AM	#4
ghost_recon88 Senior Jedi Master Premium Member #9 Join Date: Feb 2005 Location: Michigan	Very nice. Reminds me of when I took the dremel to mine. Now I wannah see someone do this with a 45nm quad, but no one around here has the balls to And don't look at me, I don't even own a 45nm quad __________________ The will to live will always outweigh the ability to die
	QUOTE Thanks

09-05-08, 11:25 AM	#6
nd4spdbh2 Member Join Date: Nov 2005 Location: Camarillo, CA!	damn thats alota work to expose the cores.... i woulda done the knife blade and blowtortch trick to pop the entire ihs __________________ #12 in the 4ghz core 2 duo club baby! - 52 inches of CRYSIS! - Black Belt Ubercloxx0r -- Heatware Main -- Lian Li V2000b -- i7 2700k -- MSI Z68A-GD80 -- 12gb 1600 -- GTX560Ti --Adata 128gb S510 & 2x500gb -- Seasonic S12 600w -- Dell 2209WA HTPC -- Silverstone GD05B -- E7200 @ 3.2ghz -- G41-M7 -- 4GB 800mhz -- g210 passive -- 320gb 2.5in WD -- Antec EA 430w -- 52inch Samsung LCD Server -- Antec NSK4480B -- E3300 @ 2.5ghz -- ECS G33T-M2 -- 3gb 667mhz -- Onboard Vid -- 500Gb + 1TB hatachi + 1tb WD -- Zalman 400w Netbook - Dell Mini 9 - Atom 1.6ghz - 8GB ssd - 2GB pc6400 - WIN 7 PRO! Need 4 Speed - BH
	QUOTE Thanks

Sanded down the IHS of a Q6600 to the silicone.

Internet Services

Website Development

Tech Hardware

09-05-08, 11:27 AM	#8
madhatter256 Member Join Date: Jul 2008 Location: CFL	The first picture was scary! When I first saw it, I thought you were going to say how you failed at it. Good job on it. Hope you wore a mask and did it where the copper and zinc will not get circulated in your home. It can cause health problems and the metal can build up on your motherboard and short it out.
	QUOTE Thanks

09-05-08, 11:32 AM	#9
xtkxhom3r Join Date: May 2008 Location: San Antonio,TX	thats awesome bro i would do it to my q6600 but i dont really need too lol.... at 3.6 my max temp with prime is 52....so yea..... good write up your the 2nd person i have ever seen do this __________________ [Xeon 3520 (I7 920 D0) - 4.5GHZ \| Swiftech GTZ- MCR 120.3 - 6 Fans] [Rampage Extreme 2 \| 3x2 GB Corsair 1600 ram @ 1800 7-7-7-20] [x2 XFX 4890's \| x2 MCW60's\| MCR120.3 - 3 Fans \| X-Fi Titanium Fatal1ty] [x2 OCZ Vertex raid 0 \| x3 WD 2TB \| Coarsair 1000HX \| 55" LED Samsung] [Lian Li V2000 \| Logitech diNovo Edge \| Logitech G7 \| Scythe Kaze Master] Laptop Service Manuals
	QUOTE Thanks

09-05-08, 03:39 PM	#13
Joe Camel Senior Camel Kicker Join Date: Aug 2003 Location: ---> NEW HOUSE 7/17/09 !! <---	wow... just wow major props __________________ -->> Help US fight: CANCER, Alzheimer's, Parkinson's & MORE with YOUR computer <<-- Joe Camel is FOLDing for Team 32 R.I.P pete_scout I wish we could have FOLDed a cure, for you. HEAT "So you see, it's not at all pointless, you just have to look for a different point" ... hafa 5/17/04 6800 EX Conroe @ 4740 MHz (395 x 12) - - Dry-Ice in The Camel Pipe - - X1900 CrossFire @ 846/945 - - MachII Slave, D-Ice Master - - 2x gig Corsair XMS 6400 CAS3 @ 494 MHz 4-4-3-1 - - ASUS P5W-DH - - 700w "Mr Zippy" PSU
	QUOTE Thanks

09-05-08, 05:18 PM	Thread Starter #15
MrRuslan New Member Join Date: Sep 2008	Thanks for all the props and the warm welcome Here is an update: I left it to prime for 16 hours and it passed. I turned off the AC and close all the windows before going to work. I came back and took this screen shot. The ambient temp was 27C at the time of the screen shot.
	QUOTE Thanks

09-05-08, 05:29 PM	#16
Sam__ Member Join Date: Mar 2007 Location: UK	Lord, you have some major balls. props. __________________ "Computers in the future may weigh no more than 1.5 tons." - Popular Mechanics, 1949 "I think there is a world market for maybe five computers." - Thomas Watson, chairman of IBM, 1943
	QUOTE Thanks

09-05-08, 05:37 PM	#17
petteyg359 Join Date: Jul 2004 Location: University of Southern Mississippi	Wish I was brave enough to do that to my Q9450... __________________ MSI 890GXM-G65 - X6 1055T - 16GiB 10700 - 5850 1GiB - 7 Pro x64 / Gentoo amd64 - HX520 Dell XPS 15 L502x - i7 2760QM - 16GiB 12800 - GT 540M 2GiB - 7 Pro x64 / Gentoo amd64 - Agility 4 512GB [GB ≠ GiB] [MB ≠ MiB] [kB ≠ kiB] [1000 ≠ 1024] [Giga ≠ gram] [Mega ≠ milli] [Kelvin ≠ kilo] [Byte ≠ bit] "Apparently, Plaintiff believes that he could sue an egg company for fraud for labeling a carton of 12 eggs a “dozen,” because some bakers would view a “dozen” as including 13 items." - Western Digital 2006
	QUOTE Thanks

09-05-08, 07:24 PM	Thread Starter #19
MrRuslan New Member Join Date: Sep 2008	This is how much you need to take off. I have a celeron that i might make a ginnypig and take it to a power tool carefully...
	QUOTE Thanks

09-05-08, 07:41 PM	#20
SolidxSnake Member Join Date: Dec 2004 Location: Massachusetts, USA	That looks absolutely AWESOME. Excellent work, and major props to you. What I would have done was gotten it to the point in your first picture, separated the inner IHS pieces stuck to the die and then just torched those off instead of the whole IHS (so you still have the borders). That way you don't risk sanding the dies too far. __________________ Too funny. When I logon to Windows, I press Ctrl+Alt+ELITE i7-2600 4001MHz @ 39x102.6~ASRock P67 Fatal1ty Professional~MSI N580GTX Lightning 1536MB @ 1000//2400 @ 1.149v~8GB Corsair Vengeance DDR3-1600 9-9-9-24~256GB Samsung 830 SSD and 60GB OCZ Vertex~2x Western Digital Caviar Green 1TBs in RAID1 ~BFG ES-800 PSU Custom Watercooling Loop \| Apogee GT, MCW82, MCP35x + Res, Thermochill PA120.2 and 120.3 (both rev1) Yate Loon D12SL-12s pull, 7/16"x5/8" MasterKleer, 100% Distilled Water with Iodine ---All encased in a Silverstone TJ07B-W Join in on the fun! Heatware OCForums Metal Club member #13224
	QUOTE Thanks

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search