Anyone ever thought to put there chip in a mill or cnc and perfectly remove the ihs without hitting the core? Expensive probably, but could give a great oc!
Printable View
Anyone ever thought to put there chip in a mill or cnc and perfectly remove the ihs without hitting the core? Expensive probably, but could give a great oc!
I know a guy that removed the IHS on a celeron D for lga775. CrazyXP1700 did it
You mean grind the IHS down until you hit the epoxy? Take a look at AMD IHS, unless it had REALLLLLL bad factory contact you gain a few C and maybe 50mhz give or take a bit, bar having horrid IHS to HSF contact like I always do. Not worth it.
Right, but amd does it, maybe we should try it. Because cutting the epoxy always ends up in a dead chip. Why not just mill off the top!
it wasn't an LGA775 ZX... it was a Celly D 478... none the less it was soldered to the IHS...
basically i cut around the epoxy with a real thin razor blade... basically removing the black epoxy around it... that gave it a little play...
after that i took the cpu... put in in the mobo without the heatsink... basically ran it till she was nice and toasty... then i took a thin flat screwdriver... and stuck it in a corner... between the IHS and the PCB... and GENTLY inserted it... most screwdrivers progressivly get thicker (kinda like a V shape from the side...) the farther you insert them the farther the screwdriver... the farther the IHS and PCB get away from each other (it did make a noise when the solder seporate)... BUT dont sporate them any more than 1mm apart at a time.. put the CPU back in the board... fired it up to get hot again... do another corner...
basically i did that until i got all 4 corners raised up... then i put it back in... warmed it up real good and raised off the PCB... and simply twisted it off...
basically if it's not warmed up enough... you'll crack the die... dead CPU...
but if you get it off... all you've got to do is clear off the solder... that stuff you can scrape off with your fingernail if you'd like... till you get down to the die... then just make sure you get it all off
but yeah... i got it off my celly D... i've been meaning to take pics... just havn't got around to it...
Yeah I know you can pop the 478.s But I have not seen a naked lga775.
i tried this same thing with my P4 520 awhile back... basically the IHS of my Celly D is secured the EXACT same way... with solder... soooooo... i guess this same method would work for the LGA775's... it's just i screwed up my 520
i mean i know what i did wrong with my 520 about 8 months ago... when i put the screwdriver inbetween the PCB and the IHS... i thought it was just epoxied on... not soldered... epoxy pops off easily normally... solder on the other hand is MUCH harder...
basically i had the IHS and PCB about 4mm apart and i heard the DIE crack...
maybe if i can get my hands on another 775 i bet $100 i can get the IHS off without cracking the die :D
Wait wouldn't the lga bracket lock thing be higher than the CPU die with no IHS?
Yeah. You'd have to pad it.
Pad what?
Yould have to put a shim like thing on the processor so when the lock bracket type thing comes done it would lock not just little it slid uselessly in it.
I think it would make an improvement but not a worthy one for the risk. I know my FX-57 I pretty much got no omprovement when removing it. Then again it was cold bugged and I was using LN2 so was able to hit optimum temps. ;)
I think it may be almost worth the experiment. You just gotta have a lathe take it off insanely slowly and very little at a time.
huh?Quote:
Originally Posted by afireinside
Ive had a video where they straighten p4 ihs with cnc-mill. It dropped temps few degrees if i remember correctly. I think mill would suit better for the job than lathe.
im sorry, but thats absolouty retarded....
if the lga775 locking clip is higher than the core, how the hell are you going to mount a heatsink on it.. that's what he means... :rolleyes:Quote:
Originally Posted by ingentingmendeg
Look who posted it...Quote:
Originally Posted by IYP
Quote:
Originally Posted by n00b 0f l337
Never lathe a CPU. No good can come of it.
I use a die grinder before I'd use a lathe or mill ;)
The heat generated by grinding or CNC'ing could be a problem. It would hard to find a way to mount it for precision cutting without causing damage.
LGA775 IHS removal is no problem, just cut the RTV sealant stuff under the edges of the ihs and hold a small torch on the spreader for a few seconds while applying a light twisting pressure to the IHS and next thing you have a naked chip! Also, the solder that intel use to hold the ihs onto the core melts at around 98*c so the heat from the torch won't cause any damage if you aren't reckless.
wasnt it about 70°C?? thought someone (bbmods??) postet it some time ago over at hwa)Quote:
Originally Posted by MeltedDuron
:toast:
maybe it was :) I just remember it being below 100, which a non-running chip can take no problem :toast: anyone wanna send me a dead conroe? :fact:
sure, even 100°C shouldnt be a problem for the cpu :)Quote:
Originally Posted by MeltedDuron
dead conroe... try it with a working one! no risk - no fun! ;) :eek:
i saw someware someone used a cigarette lighter to heat the ihs and as long as the flame doesnt touch the chip youre fine, then use a soldering iron and solder wick to clean the solder up, then use the outer edge of the ihs as a shim and the rest as a keychain
can't do it.
i tried once and CPU = died!!!!!!!!
it is totally different then AMD.....
core is like super glue with IHS
very few chips are superglued... you probably didn't heat up the IHS enough.
:nono: :nono: That would be wrong. Imagine if the IHS is secured in a different way and he'd kill a good Conroe CPU. Imagine the waste!! :DQuote:
Originally Posted by 2fink
Why not hold the chip up side down in a liter of hot water. This way you can control the water temp, and have enough heat capacity to melt whatever solder is in there. Though you'll need steady hands, or risk using distilled water (my wc leaked and the distilled water didn't kill it, I didn't realize that till I saw water on the mobo haha.
Yep BBmods got confirmation from intel that the solder melts at 77C The core can take upto 140C though so it's safe. Although the performance benefit of removing the IHS. over just lapping it is far out weighed by the risk for me. :toast:Quote:
Originally Posted by 2fink
Just lap it man :)
Or you wanna try removing it on x6800? :D
I did it with my Celeron D 341 before, I got the IHS off with the core still intact.
But I must have cut some traces on the chip with the razor blade I used because its dead now.
I now have a Celeron D keychain.
When the contact between the ihs and the core is that good, removal will gain 0% higher clock's imho.
for the people who want it still, why not grid it off, take some time but the heatspreader is not that massive, maybe start with a file.
Look @ VR-Zone, there is an article about IHS removal ;)
heh looks like shamino f-ed up two cpus. Looks like it pretty much cant be done.
http://www.vr-zone.com/?i=3878
Actually it can. You need to heat it up to 77c and then you should be able to get it off.
screw temps, you heat it up until its hot enough! shamino was too paranoid, holding a small butane torch or heat gun to the ihs with very gentle prying will get the ihs off no problem, as long as you dont heat it for too long and melt the solder on the other side of the die.
I am just saying thats when it melts. The cores can take up to 140C.....
shamino pried upwards, wich lifted the core off, if he had twisted, he would have had no prob
You can do it with a larger tipped soldering iron, I know this for a fact.
How good a thermal paste can solder be?? I mean, I can almost completely say that there would have to be a decrease in temps if it was done right. How could there not?? Unless the solder is like >80% silver (which it cant if it melts @ 77c) then I think there would have to be a decrease in temps (unless the die is completely ragged and jagged or something).
i heated it up to 95C but still can't remove it... haha
golden E6700 S5
OK anyone saying this is doable I suggest you post a pic of one you have done yourself.
Naked 631!
done with the torch method...
i took 4 razor blades in each corner...and placed it in a clamp suspended... I set it so the IHS was above the blades (so the cpu would drop when it let go)
i heated up the top until it just popped off
http://img.photobucket.com/albums/v3...p/DSCN0899.jpg
http://img.photobucket.com/albums/v3...p/DSCN0907.jpg
Not 100% sure if my cpu survived... but it sure looks like it...
i've got to get a memory stick soon to throw it in and see if she'll boot
prayin like hell that it does
(5Ghz on air? hopefully)
77с?
hm, some people say it has indium solder, which may be true then. its melting point is a bit above 60c, cant remember now for sure ;)
crazyxp that looks very promising, please keep us updated as to if it works or not.
well... i polished up the cpu abit...
i went to install my Scythe Ninja... and it's not making contact with the die...
http://img.photobucket.com/albums/v3...p/DSCN0913.jpg
no not a kingston plug :p: (i dont think i even have any kingston sticks)
so definetly going to have to modify some coolers to go IHS less thats for sure
i think my old chilly1 tube might have cleared it... but i dont really think so...
so some modifications might be needed...
http://img.photobucket.com/albums/v3...p/DSCN0914.jpg
so fun!
gonna break out the dremel and shave it down about 1mm all the way around the green and see if it'll make contact
would be nice if i had the proper equipment
http://img.photobucket.com/albums/v3...p/DSCN0919.jpg
tada!
looks like hell... but it's making contact...
now i just gotta lap the base of the heatsink with some 2000grit paper and i should have the best contact possible :toast:
would suck if this cpu was actually dead
i need to aquire some money to go pick up some new sticks...
i killed my D9's because my stupid multimeter wasn't reading properly
im gonna go and take some pics of how i actually heated up my 631... so you people can see... anyone brave enough should be using my method...
shimano's is horrible
he would have had better luck inverting his cpu so that the IHS was locked in the vice... and twisted the PCB when it was all warmed up... thats how i did my first successul IHS removal on my Celeron D
im tryin to get out and get a new memory stick from the local store (staples... and they have 512 sticks with D9's! not fattys though... their 1GB PNY looks to be fatty's though, but their under heatspreaders)
but my brother just stole my car (he blew up my jeep a few months ago too!)
http://img.photobucket.com/albums/v3...e/DSCN0925.jpg
go around the IHS with the razor blade... some cpu's have resistor banks (like my 805) so only go in around .5cm... you just want to get the RVT silicone off
http://img.photobucket.com/albums/v3...e/DSCN0926.jpg
Put a razorblade in all 4 corners of the IHS...
now this is going to be used to do 2 things
1. it's going to suspend the PCB so that it will drop right out once the solder reaches it's melting point..
2. the razors are actually going to be applying all the pressure that you need to seporate the IHS from the PCB... so you shouldn't have to twist or apply any pressure to it... once it's warmed up... all the presure is just gonna shoot that CPU down... and the razors should be holding up the IHS still
http://img.photobucket.com/albums/v3...e/DSCN0928.jpg
place the razors holding the CPU over a vice, or something so the cpu can drop freely... now you'll want to put some foam or something like that below because it will literally POP off!
Core 2 duo's may be a little tricky... basically you've got a couple of these you've got to pop... i think warming up faster would be the best option... that way one core doesn't pop before the other... circular motions around the IHS would be the most ideal, the heat should transfer throughout... but still a fration of a second later could send one core popping off while the other is still firmly attached... (i havn't done one yet!)
http://img.photobucket.com/albums/v3...e/DSCN0929.jpg
I used a big ass propane torch, a fire maker, 4 razor blades, and a CPU to risk!
Core 2 Duos are real dualcore = Two cores, one die :)
Kentsfields are two C2D dies on one package. They may be tricky.
I'm gonna do this with a $1600 Kentsfield when I get it :D
Good job XP, do you know if it's a working chip yet?
nope... im having the same problem with these new memory sticks as i did with the board after i did the volt mod... and (i think anyways, i fried out my fatty D9's)
the board isn't reading any voltages
i think im gonna have to go up to raleigh and stop by tiger direct in the next few days and get a board to slap it into... while i try and fix this one
i might get a 975X board if i can spring it... any donations would be appriciated :p:
ugh... im wore out by this stupid thing...
it's the same problem as before... so i think my cpu is fine...
I machined an Athlon down to get the IHS for a load tester I'm building and I found that the small piece of circuit board was not parallel with the face of the IHS or the chip. I was able to machine it off very slowly using a milling machine and a fly cutter, taking 0.001 inches per pass, until I had the chip exposed. The very next pass just shattered the chip. My point is that it could be done, but I don't know how you would get the chip flat in the vise if you have to hold it from the pin side, I held it from the IHS side which was flat.
looks like you have copper shavings all over the motherboard, is that just dust?Quote:
Originally Posted by CrazyXP1700
it's a combo
i hit it up with my compressor before i fired it all up
i just took the pic in a hurry
HahahQuote:
Originally Posted by CrazyXP1700
I like this mans style.
Keep us posted on the 631
Dude looking good CRazy! First time I've seen this done. Your method is a bit rogue, but I like the style.
from what i can tell... this is the most promising way possible... it makes the most sense aswell... it SHOULD be successful...
i mean it makes complete sense... with the razorblades under the IHS like that... it's putting enough pressure so when the solder melts, it's going to seportate itself... you dont have to worry about it overheating the core though cause once it's at that melting point... it's gonna pop right off... im a genius! :p:
lack of hardware kinda leaves me hangin though... i dont know if my board has gone out... or my cpu... (if it was my cpu... it could have been before i removed the IHS... cause i attempted removal because i was bored!)
i might try and get my asus to crank up again... i've got some business to take care of first... but hopefully i can take another crack at firing it up tonight
http://img.photobucket.com/albums/v3.../IDIOThaha.jpg
thats my old 520 that i removed this IHS off over a year ago.. . (january 05 i believe)
with that chip... i was heating it up and prying it up one corner at at time... the razorblades is a much simpler method to say the least
needless to say... too much pressure on one side will crack the core... non the less i never heated up the core nearly hot enough
love the filename on that one :D
you do have a sweet method there.
well i fired it up... and i got nothing...
im not getting any vcore readings out of my P5WD2... so that kinda makes me wonder
i even took the heatsink off and put my finger on the core... and it did nothing didn't get scorching hot real quick like my naked celly D did...
so that makes me wonder is it the board or the cpu...
id love to blame it on the board... cause i was adjusting my mods when it last messed up
now i've got to find someone who has a 775 board or cpu in my area to test things out...
(my local compy shop doesn't even sell LGA775 gear... like it's the dead socket :rolleyes: )
OK I gave into temptation and tried your method on my celeron D :p:.Quote:
Originally Posted by CrazyXP1700
Attachment 50278
Attachment 50280
My method was basically to cut around like a normal 939/754 IHS to cut the rubber stuff and then to wedge 2 blades under oposing corners of the cpu. I then thought how would be best to hold the cpu and how to heat it up, I eventually decided to hold one of the blades with a wet teatowel and heat the IHS with a lighter. I held the cpu upside down (pads up) and wore a fireproof glove while using the lighter and just kept heating it until it dropped off. The entire process took about 2 minutes, not including thinking time and I found it easier than 939/754 IHS' thanks to it having no smd's. It fits in the socket fine thanks to the way that the bracket dips in the middle although a bit of rubber or something may help hold it in better. The IHS also cleared the top of my bracket so that's fine too. I can not show you it working as I think the m-boards dead but it fails at error code "45" on my diagnostic card which is good as if it's a cpu fault it's code "00". :)
The next time I do it though I think It would be better to put something to catch the cpu with rather than just letting it drop to the floor :rolleyes: and it may also be usefull to have something you can use to quickly whipe the solder off the core because mines cooled down and is solid now :D :toast:
looks good...
not all cpu's are free of SMD's under the IHS... my smithfield core has them.... (i believe presler, smithfield, and kentsfield have them... possibly more)
clear off that left over solder and take a looksee (pics!!)
i used a guitar pic... or some kind of rounded plastic...
i just scraped it off until i got close to the bottom and had a thin layer of solder on there... (heck you can even use your fingernail) then i hit it up with a small amount of polish to clear everything else off and remove that last bit of solder
you need to find someone else with a board and check that cpu... (same here!)
CrazyXP1700 your username really suits you :p: Got to say nice work to the guys who have attempted this. Anyone get their cpu tested as working yet though?
Really nice work! If your method really works i'm sure all other (crazy :P) people will do that!
Well I can give an update to mine, the cpu is dead :( but it was dead before I removed the IHS. I have tested it in a friends 925x board and had no go and 45 code still. Also you can remove the solder with your finger nail as crazyxp said so that's one problem solved. I'm trying to get another celery now and I will try again :toast:
i havn't been able to test my cpu yet...
i'll probably be picking up a mobo in the next few days... (possibly today)
HOPEFULLY my cpu is not dead... if so im screwed cause i have a memory review to write...
here you go guys i tried it on a d830. It would have turned out ok if i knew there were SMD components under the IHS:D ohh well i will do it right next time if i decide to give it another go.
I just did this for :banana::banana::banana::banana:s and giggles i dont think removing the IHS will prove to be a worthy risk.
http://img244.imageshack.us/img244/20/1001788mn7.jpg
This is really promising, it basically shows that removing the IHS is almost as simple as with AMD correct?
Is it a worthy risk, I think it is, Intel CPU's do tend to get hot and with the recent complaints about concave and otherwise bent Conroe IHS's this might be a blessing.
And even though it didnt do much for the A64's I reamoved the IHS from I still did it anyway. :D
outcast... i said in a couple posts the cpu's that have SMD's under the IHS... smithfield being one cause i killed my 805 because of the SMDS :rolleyes:
http://img.photobucket.com/albums/v3...p/DSCN0898.jpg
Celly D Prescott ZX was talking about earlier in the thead (it lived till i crushed the core mounting a heatsink months later) Smithfield, and the 631!
basically if you dont know... google the core name in google images... and there should be PR shot of the cpu without an IHS...
tommorow i find out if the 631 lives :toast:
(DFI Infinity 975X)
aleast we know this method can get the IHS off without cracking the core :D
thats more success than shimano had
I did it before you said that:D ohh well i was thinking it was going to look like the 631.
yeah... i posted it in the OTHER thread in the intel conroe section (conroes DONT have SMD's!)
preslers, Kentsfield and smithfields... i think thats all that i was able to find
but you did successfully get the IHS off without cracking the core... thats the hardest part!
Yes! Conroe's don't have SMD's...
*censored*
these may not be 100% accurate... (because a few pictures showed smithfields without SMD's... and we know for a fact 2 of them have them) but
http://img.photobucket.com/albums/v3...de/Presler.jpg
http://img.photobucket.com/albums/v3...ieldconroe.jpg
http://img.photobucket.com/albums/v3...l_IMG_1264.jpg
Just a few more hours and we'll know if i have a working naked 631 :D
it would be a shame... cause this thing was at 6ghz... and 32M stable under phase (maybe cascade... would be a nice LN2 chip)
631 = Dead...
i think it was my vdroop mod that just pumped a ton of volts into it or something
must have took out the board and the cpu...
Core2Duo here i come!
(if you want to donate, cause im broke at the moment... hit up my paypal mike_mackenzie18 at hotmail.com)
Too bad I was hoping to see some good results because I have a 631 aswell. Better luck next time
Oh yeah... there will be a next time for sure too...
im 100% confident that this method works...
i just need to get my hands on a Celly D or something just to prove the removal process doesn't kill the chips
IT CAN BE DONE!!!
I went and put some tape on my CPU... to help hold it in place... and it booted!
my heatsink was making horrible contact so i modified it by grinding it lower... and i cracked the core while remounting it...
IT CAN BE DONE!
their super fragile...
i need to get something to help spread the weight over these tiny cores...
http://img.photobucket.com/albums/v3...cracked631.jpg
I have a couple of IHS-less northwoods, and I have made shims using this. It seems to do the job.Quote:
Originally Posted by CrazyXP1700
Nice thread.
But this is so hopeful you could do the work! :woot:Quote:
Originally Posted by CrazyXP1700
Btw i'm sorry for the dead chip... When you'll try with a celly D be sure it is a 65nm :toast:
Hehe, nice, well doneQuote:
Originally Posted by CrazyXP1700
Are Allendales free of SMD's under the lid as well? I'm really not liking my 62c load temps @ stock voltage.
nothing shows up... so id use extreme caution...
dont stick the razor blade in like a cpu without SMD's
Hmmm I might try this on my celery D 331 (2.6ghz)... can do 4.2ghz on the stock air cooler :)
If it goes well, might try on my pent d 805... gimme a day or 2 fellas, got an exam in 2 days!
Hi,
I just tried CrazyXP1700's IHS removal method today on my Pentium D 820 SL88T and it seems as if it worked fine. (didn't test the CPU again, but it looks promising.
Just after removal of the IHS:
http://www.xtremesystems.org/forums/...1&d=1157750344
After cleaning and lapping of both, the CPU core and the IHS:
http://www.xtremesystems.org/forums/...1&d=1157750344
@CrazyXP1700:
Defintaley BIG THX for the info man! Your detailed pics saved my ass, err CPU, I think. :D
Here are the original pics (not resized; size: 2048x1536 !). BEWARE, these are pretty big (over 300KB each)!
http://www.fh-augsburg.de/~madocer/P...0(cleaned).JPG
http://www.fh-augsburg.de/~madocer/P...20removed).JPG
Quote:
Originally Posted by CrazyXP1700
Now, you still think I flamed earlier? :rolleyes:
Looks real good :D
pop it in and give it ago
im still tryin to come up with a better way of mounting heatsinks without modding them or ripping apart the socket... :rolleyes:
Small price to pay for being one of the first... and coming up with the best/less risk method :up:Quote:
Originally Posted by ewsforos
i mean it's a cpu... it's not like i wrecked my car... easily replaceable with something faster... and naked :toast:
@celemine1Gig: n1 ... hope your CPU doesn't work up a sweat again :stick:
Agreed, I was just told off on the intel thread because I just stated the risks involved cause there were like X new into overclocking people that were like "oh! It's easy, I'll do it!".
Posted on the ln2 forum, that's fine since there are more seasoned members here that already know the risks, no need to tell them. However, you see how many dead cpus have pilled up already :)
Either through cracked cores, damaged SMDs, etc.
All I was trying to say there was that although it is good, and "less risky than other methods" (no disagreements there), it still is not the best thing since sliced bread as it was portrayed to be ;)
Good luck with the next cpu though, I'd really like to see it working. Seasoned OCers will definately put it to good use under phase/LN2 :)
Today I had time to mount the CPU without IHS...
http://img171.imageshack.us/img171/4365/dscn1462dg1.jpg
It's an E6600 and I've removed the IHS more than 1 week before...
And yes, it's still working ...
Wow nice. Congratz.
Id be too scared to do this to a cpu worth over £5 lol
but seeing as my only income is £10 randomly every 2-4 months then you can see where im coming from :P
Crushed cores are easy to do btw, i did that to my duron 650 accidentally, but it still worked/works :D
Intel Celeron D 356 w/IHS removed and completely cleaned:
http://www.jpop.com/fcg/CPU3.JPG
http://www.jpop.com/fcg/CPU2.JPG
http://www.jpop.com/fcg/CPU1.JPG
wow the black pcb colour looks like teh sex
FCG.. your cleaning skills!:eek:
Boy, if you ever need a job as a house-cleaner, give me a pm! :stick:
A steel bowl in a pot of boiling water could heat a cpu nice, just like chocolate moose... Max temp 100. :)
its nice to see a few people having successful removals... and no more cracking cores...
guess i should be a little more delicate with my next one...
FreeCableGuy cleans up pretty good :p:
id like to hear if anyones picked up a few more mhz... or dropped a bunch of temps...
:woot: have you tryed to boot the CPU? Still it works?? Temperatures?Quote:
Originally Posted by Entsafter
I've written: "And yes, it's still working..."
The CPU runs great... after the IHS removing I used the CPU with remounted ihs... and now it's running under a waterblock...
temps are useless for you -> room temp, watertemp and so is different...
But temps are much better without ihs than before!