Removed IHS Results thread

Now, this thread is to remain free of debates about whether this is a wise or safe decision. We want results and posts relevent to non-ihs issues or problems.

here's a link to the thread with some of CrazpXp's results and how remove a soldered on IHS
http://xtremesystems.org/forums/showthread.php?t=81209

here's FreeCableGuy's guide to preparing your mobo for ihs-less cooling.
http://xtremesystems.org/forums/showthread.php?t=115252.

A link provided my kocsonya of another person's removed core 2 duo.
http://modbox.hu/forums/index.php?ac...&st=0#entry384

pics of my removed e6400
http://webpages.charter.net/aari/641.JPG
http://webpages.charter.net/aari/642.JPG
http://webpages.charter.net/aari/643.JPG
http://webpages.charter.net/aari/644.JPG
http://webpages.charter.net/aari/645.JPG


Some work of CrazyXp's

Quote:

---

Originally Posted by CrazyXP1700

CONGRATS MENLATIN!!!!!
did you use my method from the other threads?

BTW... i did this a few weeks ago... (i ended up having a bent pin in my socket says ASUS... i checked for that before RMA... but ok ASUS!)

http://img.photobucket.com/albums/v3...p/DSCN0919.jpg

http://img.photobucket.com/albums/v3...e/DSCN0928.jpg

http://img.photobucket.com/albums/v3.../Guide/WIP.jpg

i joined in the thread after a buddy (ZX) rememberd i removed the IHS off a Prescott Celly D a year ago... and the thread was going on in the LN2, and Dry Ice, Intel Heatspreader Removal thread

my method outlined above and in the other thread is for SMD's cpu's mostly... can be applied to all cpu's without SMD's ofcourse since you dont have to worry about how far the razors stick under the core (just dont go too far or apply pressure)

and to everybody wondering in the other threads... it's solder, or some kind of alloy! ive removed IHS's a few times (2 successful, 3 unsuccessful) eventually i cracked the cpus mounting large heatsinks... so be careful

i am trying to sell an old PC so i can finally get my core2duo and take off the IHS

---

some pics of 4-layer electric tape feet to protect the core from being crushed by uneven mounting pressure

http://webpages.charter.net/aari/tape1.JPG
http://webpages.charter.net/aari/tape2.JPG
http://webpages.charter.net/aari/tape3.JPG


My results so far...

Here's some temps and clocks i've compiled...

Old ds3
Stock IHS
Apogee
3440mhz
1.36v actual, 1.4v bios
62c easy Tuner
66c Core temp
28c Ambient

Old ds3
Stock IHS
Apogee
3328mhz
1.3v actual, 1.325v bios
55c east Tuner
58c Coretemp
20c Ambient

Old ds3
Lapped IHS
Apogee
3608mhz
1.46v Actual, 1.5v bios
No easy tuner results
59c Coretemp (skewed because of super pi simultaneously running, IIRC it was 66c prime)
20c Ambient

New ds3 (much less vdroop)
Lapped IHS
Apogee
3600mhz
1.49v actual, 1.5v bios
75c Easy Tuner
74c Coretemp
28c ambient

New ds3 (much less vdroop)
Removed IHS
Apogee
3520mhz
1.41v actual, 1.425 bios
58c Speed Fan (same as easy tuner reading)
69c Coretemp
28c Ambient

Notice the ambient temps though. Makes a huge difference.

Screen shot of 3520mhz result
http://webpages.charter.net/aari/3520.JPG

CEDAR MILL:


http://www.jpop.com/fcg/CPU3.JPG


http://www.jpop.com/fcg/CPU2.JPG


http://www.jpop.com/fcg/CPU1.JPG


PRESLER:


http://www.jpop.com/fcg/CPU4.JPG


CONROE\ALLENDALE:


HTTP://www.jpop.com/fcg/12.JPG


HTTP://www.jpop.com/fcg/13.JPG


-FCG

http://members.cox.net/kjboughton/ttr.png

will be interesting to see comparisons with same hardware/same ambient/cooling and rest

mentalin there's just a lot of variation there with your testing but certainly looks like you've dropped the temps........has your max OC increased and how much

With what stuff do you guys use to remove that black (silicone) crap around CPU?
How about the stuff on core (alu)?
:)

FCG, how do you clean the die from the leftover messy solder ? Lap it ? or use some chemical ?

acetone maybe?

Quote:

---

Originally Posted by Revv23

acetone maybe?

---

LOL, that solder is made from lead or metal, acetone can't solve it ! :stick:

Well, prolly works for the left over gunk or epoxy, but I doubt it too since acetone might dissolve that green pcb too ? Dunno, c'mon guys, share your experience.

@bing: the solder is extremely soft and most of it can be scraped off with a finger nail.

FCG: any tips on getting all solder and rubber stuff off, to get such a clean shiny finish as yours?



:toast:

metal polish can be used after most of the solder is removed. Brasso, or mothers, or Blue magic, whatever you can find. I used a paper towel, but you might want to use a lint free cloth, as i think that has less chance of producing static.

yeah FCG, how did you remove the silicone?

I wish i could get more results, but my p5b's have been total whores, and do no OC at all. Neither of them even detected both cores. I keep saying that i'll wait for my ds3, but good lord, i can't wait that long!!

But with it one core @ stock everything, i get a 4-5C delta detween load an idle. IIRC, with the ihs and apogee, i was getting 11c delta

The silicone on my now-slaugtered E6400 wore off quite easily even by rubbing it with a cotton cloth. Isopropyl alcohol makes this stuff softer thus easier to get rid of.

The stuff on the core is a bit trickier, though.
The TIM on Intel CPUs is an alloy consisting mainly of gallium with some tin and possibly other elements in it to make the alloy solid below ~80-90ºC.

Few facts about this stuff:
- organic solvents (acetone, isopropyl alcohol, etc.) don't readily dissolve it
- the stuff remains soft for couple of minutes after IHS removal meaning most of it can be scraped off with a piece of plastic eg.
- once cooled becomes about as solid as pure lead

I've no idea though how FCG managed to get his CPUs so clean. And I fathom he won't be revealing his technique any time soon...

Hmm..so FCG is making business out of it, doubt he will share it.

Crazy idea, use a temp controlled heat plate and set it to 80ºC and start lapping the die on that ! Anyway, its only near throtlling temp ! LOL

Well, one thing for sure, with mercury, this liquid metal can solve most metals like gold/platinum/iron/copper easily like acetone on plastic, dunno if it can solve that galium/tin solder and without affecting the silicon core.

lapping is NOT necessary, and will only worsen the surface of the die. All you need is a little metal polish and about 20 mins of polishing. We're all computer geeks here, i know you can polish for 20 mins.

That is, of course, after you scrape the solder off, wihch can be done with any hard plastic edge... like a credit card. Thats how i did it.

I starting to worry about the left over tin/galium in micro sized that can't be seen thru naked eyes, it will actually become a perfect solder/glue to your cooper heatsink/water block once they reach their "soft" point and glued the die at the block like soldering it at IHS.

My worry is if that small left over tin is so perfectly melted and stick to hsf, and imagine when you pull that HSF when cooled ...chipped or lifted core ... yikes !

Prolly, I'm worrying too much, better start practicing on my cheapo smithfield D805 toaster. :D

Anyway, menlatin thanks for the encouraging jobs there ! :D

@crazyxp 1700 , great stuff.Love article like the one you have shown.

Wow.. Conroes run really hot. D:

Quote:

---

Originally Posted by bing

I starting to worry about the left over tin/galium in micro sized that can't be seen thru naked eyes, it will actually become a perfect solder/glue to your cooper heatsink/water block once they reach their "soft" point and glued the die at the block like soldering it at IHS.

My worry is if that small left over tin is so perfectly melted and stick to hsf, and imagine when you pull that HSF when cooled ...chipped or lifted core ... yikes !

Prolly, I'm worrying too much, better start practicing on my cheapo smithfield D805 toaster. :D

Anyway, menlatin thanks for the encouraging jobs there ! :D

---

Hopefully, your CPU won't get to those temps while running.

Quote:

---

Originally Posted by DragonOrta

Hopefully, your CPU won't get to those temps while running.

---

With D805 toaster ? Even my grannie could easily do that, give her a stock cooler with high VCore ! :D

For those who haven't read yet, i think my e6400 is half dead. IT seems that when windows see's both cores, (or windows installation) it blue screens, but if botting to an existing single core installation, it boots once, then after windows "installs" the second core and reboots, it blue screens all over again.

Tried this with XP and 2k with the same results. Also, NT4 setup would only last about 30 seconds before reboot.

The last 2 p5b's i've had haven't even recognized the second core, and this 3rd ds3 does the blue screening. Chances of all 3 boards having a problem like this seems very unlikely, so my thoughts are pretty much stuck @ the cpu being the problem. Give me a little time, maybe the end of the month, and i'll get another e6400 and see what i can do.

The cpu looks fine though. The contact pads are still all there and have not ben worn through, the core looks just a tad scratched up but nothing a good 30 second polish can't take care of, and the core is not chiped anywhere. Its just weird. Maybe ESD? Some suggesting that through my ounting trials, un-even pressure could have dont damage to the inside of the core without showing outside damage ie., a crack or chip. I dont know...

But again, give a litte time, and i'll risk another e6400 for the cause.

Concerning removing the solder: try different methods trying to remove the solder left on the IHS before experimenting on the core.

The metal alloy TIM can be removed easily with some metal/jewelry polishing paste and a Q-tip.

I'd love to see someone leave the sodder on the die and somehow melt it onto the HS base :D. Then you could fill the gap between the chip and the HS (with caulk or something lol) to take the weight off the core completely.

--mds

Nice Work!

Wish I could do this to mine, I just havent got the guts to do it! :(

:( Why'd intel make it so hard :rolleyes: :p:

Hmm I wonder if a heatgun would get it hot enough to remove the IHS...

no, i've used a lighter in the middle of the Ihs about 40 sec with Cpu upside down, and a glass full of water under, maintain the cpu by the blades, when the ihs reach the right temp the ihs and cpu fall in water in 2 pieces.

Succesfully tested on a 331 D, i will try soon on 6700.

with an heatgun , the heat is not enough concentrated , the pcb and core will be too warm.

Come on folks. Someone else needs to post some results. There has to be more ppl that have tried their, becuase i know there are others who have done this... I'm at a halt until 15th. Can't afford a new cpu from my Oct 1st check, so i musta wait for a new 6400.

Failed and killed one E6600 and one E6700 , pcb of 6600 was probably too much bended while warming, and core cracked with 6700. Hard loss.

Avoid removing IHS.

Maybe i will try another method anyday (Cabriolet IHS by lapping).

Blagh. I only have a heatgun. I don't wanna have to go and buy a torch. :(
Gonna just lap the IHS then. ;x

Quote:

---

Originally Posted by vudoodoodoo

Hmm I wonder if a heatgun would get it hot enough to remove the IHS...

---

That's a big Negative vudoo. :nono:

Your standard torch (propane gas powered) blows a flame in excess of 5,500F. And the heat at the tip is very concentrated, most heat-guns you will find, only blow at 400F by heating atmospheric gas with coils of wire. :D

Quote:

---

Originally Posted by misteroadster

Failed and killed one E6600 and one E6700 , pcb of 6600 was probably too much bended while warming, and core cracked with 6700. Hard loss.

Avoid removing IHS.

Maybe i will try another method anyday (Cabriolet IHS by lapping).

---

I'm sad to hear, but honestly, the top of the lighter flame is about 600 degrees C and the core will be much warmer than the rubber by the time it melts. I think it would be better to just cut the rubber all around the edge and then hold the CPU around the bottom pcb and warm the middle of the IHS with a lighter for a shorter period of time and then twist the IHS of if it doesnt fall off by itself.

If I saw your post earlier I would've warned you..

Quote:

---

Originally Posted by misteroadster

Failed and killed one E6600 and one E6700 , pcb of 6600 was probably too much bended while warming, and core cracked with 6700. Hard loss.

Avoid removing IHS.

Maybe i will try another method anyday (Cabriolet IHS by lapping).

---

Wow, that sucks. I think maybe that glass of water might have had something to do with that. Or did you crack the core during mounting? In that case, did you use and shim/spacer/feet to prevent your cooling method of rocking back and forth?

If you have learned anything in High School class is that if you do a high delta in temperature on a substance it will crack and bend.

Please dont drop hot things in water.

Quote:

---

Originally Posted by corsair8575

If you have learned anything in High School class is that if you do a high delta in temperature on a substance it will crack and bend.

Please dont drop hot things in water.

---

Exactly what I wanted to add. ;) I can only agree with corsair8575. Letting the CPU cool slowly would be much better for the material. I even heated my CPU to ~150°C or higher accidently (hadn't fully cut through the black rubber), but let it cool slowly afterwards. The CPU is still working fine.

Yes, i'm thinking that was the mistake on the 6700. But now i'm not really confident to try a third time :/

What up folkity folks. Well, i finally got my e6400 working again... except only 1 core is functioning. Kinda sucks, but i can limp for the time being. Turns out i wont be able to get a new cpu this month, so is going to have to last me.

Anywayses. Here's some results. I could only get the cpu to work with the Ds3 and not my P5b-deluxe. Sucks, yes, but atleast it works and overclocks. It gets to 3.6ghz on 1.425 bios volts, 1.41v windows, 1.39v load.

Ambient 25C
Core Temp 54C idle / 64C load
Speed Fan 33C idle / 45C load

The cpu overclocks and works, but it does some weird things too. cpu-z is missing stats, cpuid is missing stats, and PC wizard is alittle off. I'll get screen shot later.

There is far too much risk involved for me to try it. Nice job!

i have never had the ballz to remove the ihs, but thank you guys for sharing such detailed info, and pics.

got a new e6400 from The Nemesis today. Its nice. after booting into windows @ 4ghz, with high temps, i decided to lap. temps still high, but i got 3997mhz 1m pi stable. Open the windows to the cold outside, and got 4005mhz 1m psi stable. Pulled the IHS off, and now i'm @ 4000mhz dual 32m pi stable with normal (68*F)
Pics
http://i89.photobucket.com/albums/k235/menlatin/ihs.jpg

http://i89.photobucket.com/albums/k2...atin/4g32m.jpg

http://i89.photobucket.com/albums/k2...latin/4100.jpg

From my experience I always found a benefit removing the IHS from the cpu, a better cooling for sure. Anyway the naked core is very fragile and I'm often worried to break it during cooling device mounting procedure. Now I use this rubber frame to prevent craking and I sleep better tham before

Wohooo, the price of these foam pads is hilarious. :eek: I'd bet that their profit should be around 100-200% with these things, if not more. Still, it's one of the best ways to protect your CPU-DIE from broken edges etc. .

Or you could just get a hold of 1" of electric tape and some scissors. 5 layers of tape is pretty darn close to the height of the core. 4 layers on A64's and 5 on Core2. I've done 5 bare cpu's like that so far.

EDIT: I guess while im at it i should add the latest results

http://i89.photobucket.com/albums/k2.../3880prime.jpg

http://i89.photobucket.com/albums/k2...in/13scpuz.jpg

I went Ihs-less last weekend with my E6700. Temperature gain is about 6-7C. Used a lighter and a box cutter. You'll probably need a glove too.

DISCLAIMER: You do this at your own risk!

Quick guide:
1. Cut sides with the box cutter or razor blade. Conroe cores are located in the center of the pcb so unless you're determined to cut into the core, you probably won't unless you're stinking rich and won't mind. Don't do this while you're high, or drunk, or both :).
2. After cutting sides, fire the lighter (I used a stove lighter, bigger flame more heat). Invert CPU and hold lighter so that the flame is about an inch below the surface of ihs. You will need to wear gloves.
3. Tricky part. In order to ensure that you don't burn the core, while holding the flame to the CPU, insert the box cutter between pcb and ihs and pry gently. When the TIM hits the right temperature, the ihs will come flying off. That's it.

PS: PCB (CPU) will most likely bend a little bit (this may increase with the application of too much heat). But you can remedy this by straightening it as much as you can while still warm. Don't apply too much pressure, it'll straighten out completely when you install it (heatsink pressure+heat from core would do the job).

PPS: Another tricky part is the mounting. My waterblock is sitting right on the core with AS5 in-between them, and a lot of good pressure for my own good. How you approach yours is up to you. There are a few guides around. (Check here: www.thetechrepository.com).

Have fun, and goodluck.

Oh, before I forget, I lapped the core with a very fine grit foam and cleaned the whole topside of the pcb with isopropyl alcohol.

I REALLY wanna unlid my E6400... I may just do this... hopefully it'll help me gain a little bit...

I killed my E6600. http://www.xtremesystems.org/forums/...d.php?t=145723

here's IHS removed on a kentsfield
http://xtremesystems.org/forums/showthread.php?t=126707