|
Post by ozz on Jul 24, 2016 6:12:33 GMT -5
i been playing around and delided 2 pentium 4 cpus...1st was a SL7BK, 3.OO gig, under the ihs plate was soldered but it came off with no heat at all and looks like little bubbles of solder, it just came away easily, just popped off ?, next was a 3.00 gig SL7MP, no solder at all on the chip itself but it has a little copper heatsink in the middle of the ihs plate and thats soldered to the ihs plate but the ihs plate is not soldered to the chip ? fascinating to me
|
|
|
Post by Vinster on Jul 24, 2016 6:19:03 GMT -5
that's weird. I've never played with deliding. don't think I ever will. but those are interesting findings.
Vin
|
|
|
Post by ozz on Jul 24, 2016 6:43:45 GMT -5
that's weird. I've never played with deliding. don't think I ever will. but those are interesting findings. Vin ive delidded a few vinnie, 90% i kill lol, but never seen a copper heatsink exactly the size of the chip in anything b4
|
|
|
Post by Mr.Scott on Jul 24, 2016 6:54:35 GMT -5
Interesting. I delid a lot and I've never come across anything like that.
|
|
|
Post by ozz on Jul 24, 2016 7:02:20 GMT -5
Interesting. I delid a lot and I've never come across anything like that. ive never seen it scotty both in AMD or intel, and im an amatuer at it, i just thought i'll delid this one, see how it goes, but seems its not goin the first one i got to boot then shut off at 75c so it said in bios temps lol, dunno why, water pump goin and had a lidded one on b4 it , intel and i just dont seem to get along with anything
|
|
|
Post by ozz on Jul 24, 2016 7:14:28 GMT -5
|
|
|
Post by Vinster on Jul 24, 2016 7:15:27 GMT -5
with a delided CPU i'd think you need to close the gap, I wouldn't expect the clamp to clamp right and the cooler wouldn't sit right.
if I delided a CPU I'd take the IHS and sand down the top of it, leaving just the lower step and re-install that on the CPU. then the CPU would clamp properly, then look at closing the gap on the cooler to the die.
I've never don't the process, but that's the thinking I'd apply to using it. otherwise I'd expect it to hover in the socket and not get the proper cooling.
Vin
|
|
|
Post by ozz on Jul 24, 2016 7:24:17 GMT -5
this is what it looks like delidded
|
|
|
Post by Vinster on Jul 24, 2016 7:24:20 GMT -5
I just delided a socket 478 P4 (2.4Ghz/512/400mhz/SL6S9 Philippines). it was just TIM on the inside. I don't have a 478 board anymore so no loss to me.
what socket P4's are you playing with there Ozz? 478 or 775?
I'd be more brave with 478's and AMD's as the socket is more positive with the pins on the CPU. all other's I'd just destroy the MB I think (which is why I wouldn't get in to it)
Vin
|
|
|
Post by ozz on Jul 24, 2016 7:27:01 GMT -5
I just delided a socket 478 P4 (2.4Ghz/512/400mhz/SL6S9 Philippines). it was just TIM on the inside. I don't have a 478 board anymore so no loss to me. what socket P4's are you playing with there Ozz? 478 or 775? I'd be more brave with 478's and AMD's as the socket is more positive with the pins on the CPU. all other's I'd just destroy the MB I think (which is why I wouldn't get in to it) Vin 478 vin, when delidded the foam insulation takes up the ihs plate gap, but you do have to make sure the heights are right in the foam to seal properly otherwise youll get a bad seating and condensation in the socket, more than likely if it isnt right the cpu wont boot anyway, im tryin to get somethin goin for the 478 section in the team cup
|
|
|
Post by Vinster on Jul 24, 2016 7:31:03 GMT -5
ah, ok. also with 478's the socket grips the pins so there is no top clamp (I was thinking you were doing 775's) and your pic makes sense to me now too. you're cooler would just sit on the die directly.
I get it now.
Vin
|
|
|
Post by ozz on Jul 24, 2016 7:38:51 GMT -5
ah, ok. also with 478's the socket grips the pins so there is no top clamp (I was thinking you were doing 775's) and your pic makes sense to me now too. you're cooler would just sit on the die directly. I get it now. Vin yeah no top clamp on the socket, even with 775 when you delid if your cooler base is too big (say 60mm x 60) like mine is you have to take the top clamp off otherwise the cooler wont sit low enough to make contact with the chip face, (50x 50mm ) will as it sits inside the main socket surround but i do have 50 x 50 mm on this one, its not as good as the 60 x 60, the copper plates are 2 x 5 mm thick each, this 1 is only 1 copper nickel plate x 3mm and i re-drilled the top plate and cut off the ends to fit where it doesnt hit anything to fit the 478 socket coz i didnt have a water block for 478 socket
|
|
|
Post by georgekokovinis on Jul 24, 2016 7:38:54 GMT -5
Correct de-lidding procedure is not easy. It is a delicate job and needs patience.
Here is how it goes :
1) Only use razor method. carefully start from the corners. SOFT BUT DECISIVE MOVEMENTS. 2) As soon as 4 corners give in and the razor has enter about 1cm under the IHS, start working on the 4 sides. 3) Soon the IHS will be off. 4) You will need :
a) Arctic clean 1+2 b) Pure alcohol c) Liquid metal paste d) Sand paper n.320 and n.800 e) High temp black rubber sealant
5) Clean thoroughly both the IHS and the cpu. Remove any debris of sealing material. 6) Sand the sides of the IHS with n.320 and bring it down about 1mm 7) sand the center part of the IHS with n.320 and then n.800 until you see bronze. This is approximately 0.5mm 8) Clean everything with pure alcohol and dry. 9) Reclean with Arctic n.1 - dry - and then n.2 10) CAREFULLY - EXTREMELY CAREFYLLY - apply with the provided tool am even layer of Coolaboratory liquid metal paste ONLY on the processor. This material is electrically conductive. 11) Apply an extremely thin layer of rubber sealant on the two vertical sides of the IHS. 12) Install only the cpu without the IHS in the board socket. 13) Aligh and firmly attach the IHS on the CPU. 14) Holding down the IHS firmly with one hand, lower and secure the clamp. 15) Let it stay 24 hours.
Enjoy -8 to -12C.
If you are unsure ask. If you are afraid DONT DO IT.
De-lidding the cpu invalidates any warranty.
|
|
|
Post by ozz on Jul 24, 2016 7:48:19 GMT -5
when i delid george the cooler whatever it is goes straight on top of the die, just with tim, but you have to get your heights right if youre cooling with either chilled water or dice or ln2 to stop the condensation, i havent done dice or LN2 but i have chilled water and i know if the insulation heights arent right, youre in trouble, same as if it isnt delidded
|
|
|
Post by Vinster on Jul 24, 2016 7:51:48 GMT -5
Correct de-lidding procedure is not easy. It is a delicate job and needs patience. Here is how it goes : 1) Only use razor method. carefully start from the corners. SOFT BUT DECISIVE MOVEMENTS. 2) As soon as 4 corners give in and the razor has enter about 1cm under the IHS, start working on the 4 sides. 3) Soon the IHS will be off. 4) You will need : a) Arctic clean 1+2 b) Pure alcohol c) Liquid metal paste d) Sand paper n.320 and n.800 e) High temp black rubber sealant 5) Clean thoroughly both the IHS and the cpu. Remove any debris of sealing material. 6) Sand the sides of the IHS with n.320 and bring it down about 1mm 7) sand the center part of the IHS with n.320 and then n.800 until you see bronze. This is approximately 0.5mm 8) Clean everything with pure alcohol and dry. 9) Reclean with Arctic n.1 - dry - and then n.2 10) CAREFULLY - EXTREMELY CAREFYLLY - apply with the provided tool am even layer of Coolaboratory liquid metal paste ONLY on the processor. This material is electrically conductive. 11) Apply an extremely thin layer of rubber sealant on the two vertical sides of the IHS. 12) Install only the cpu without the IHS in the board socket. 13) Aligh and firmly attach the IHS on the CPU. 14) Holding down the IHS firmly with one hand, lower and secure the clamp. 15) Let it stay 24 hours. Enjoy -8 to -12C. If you are unsure ask. If you are afraid DONT DO IT. De-lidding the cpu invalidates any warranty. great info, better than slamming a CPU in a vise and hitting it with a hammer... that vid scared the shit out of me... never expected it to boot. Vin
|
|
|
Post by georgekokovinis on Jul 24, 2016 8:08:58 GMT -5
when i delid george the cooler whatever it is goes straight on top of the die, just with tim, but you have to get your heights right if youre cooling with either chilled water or dice or ln2 to stop the condensation, i havent done dice or LN2 but i have chilled water and i know if the insulation heights arent right, youre in trouble, same as if it isnt delidded With all respect, I am extremely opposed to that.
The die is so sensitive, that it can give in and crack the resistors under the cpu. Not even 1c is gained, if the procedure I suggested is applied, compared to remove IHS fully. Tested on multiple cpu's from S775 and onwards. Many risks involved by applying the cooling device right on the die.
No risk in my suggested procedure. Cpu is sealed, no condensation can enter the cpu, no special precautions needed when securing the cooling device.
By the way, here is an example.
Motherboard Asrock M-OCF. Processor Intel I7-6700K retail. Cooling method EKWB AIO.
Before de-lidding - 49x100 - 1.45Vcore - CBR15 - temps hitting 85c
After de-lidding 50x100 - 1.44Vcore - CBR15 - temps hitting 72c max.
Not bad.
|
|
|
Post by ozz on Jul 24, 2016 8:19:46 GMT -5
pete cam up with this foam idea ages ago and it works like a beauty on chilled water, dice and LN2, i use it and ive seen pete do it on dice and now on LN2, as long as you cant blow through the foam it will seal and if you have the foam pad heights and downward clamp pressure right, it takes a bit to get organised wit it but when you do its there always for the next time, different boards different size foam pads for different cooler bases, delidded and not delidded, intel, amd, delidded, non delidded and a base foam for the board, this case its for the maximus v1 hero, whch i killed lol
|
|
|
Post by ozz on Jul 24, 2016 8:43:21 GMT -5
i agree and disagree geroge, you cant say not even 1 c is gained , ive proved that to be wrong in my efforts ive done here, each has their own method i guess with both positive and negative results at times, sure the die is very sensitive, too much pressure and youll crack it, for sure ive done it myself, ive also done the refit of the ihs plate with no gain at all with temps, ihs plate off ive gained up to 10c temp reduction, here is the different foam thicknesses i use
|
|
|
Post by georgekokovinis on Jul 24, 2016 8:49:50 GMT -5
Cool. Whatever suits you best Shane. Different approaches. I like to be on the safer side
|
|
|
Post by ozz on Jul 24, 2016 9:03:15 GMT -5
both approaches work george, but as you know , sometimes either approach doesnt get the hoped for results every time
|
|
|
Post by Macsbeach98 on Jul 24, 2016 10:45:03 GMT -5
Everyone has different ideas for delidding Shane I want to see this CPU and IHS so dont throw them away are you sure thats copper on the IHS plate and not the bottom of the Die. I have got 30 to 40 478 delidded CPU's here and none of them are like that. George you are right it is hard to run late intel with no IHS plate on without cracking the die the wafer is real thin and they have no strength. With my 3770k I delidded it the day I got it without running it (Scotty would remember that) and I had a heatkiller 3 block that I used to run on it naked. The standard Heatkiller mount was the exact height that it would squeeze any excess paste out from between the die and the Block I just used to tighten the screws down tight. I removed that CPU from the socket umpteen times and never had a problem with it. But in saying that I wouldnt do it with any other Block because the minute you put too much pressure on the die they will crack.
George have you tried the Cool Laboratory Pro under cold it doesnt go well once frozen but its great at atmospheric temps.
Shane if you want to try some I have got a syringe of it here I wont use it because its no good frozen.
AMD doesnt have any hole in the middle of most of their sockets so the wafer is fully supported you can clamp them down with pressure delidded no worries.
When I delid soldered ones I cut around the edge as usual then leave 2 blades in there at opposite sides and support them so the cpu is 5mm off the surface I am working on then I hit it with the hot air tool flat out in the middle of the IHS and in about 30 secs it drops down. To clean up the Die I drag a single sided razor blade across a sheet of 1200 a couple of times on both sides to ensure there are no burrs. Then I scrape the rest of the solder off the die I put a bit of pressure on it to get it all off it doesnt scratch the top of die at all. Then for the magic trick I put a bit of car cutting compound on my finger and rub over the top of the die it gets the residue off and they come up with a mirror finish just like a die that has never had solder on it at all.
|
|
|
Post by Mr.Scott on Jul 24, 2016 12:03:34 GMT -5
After looking at those pics closer, it does appear to be the die stuck to the lid.
|
|
|
Post by georgekokovinis on Jul 24, 2016 12:39:58 GMT -5
George you are right it is hard to run late intel with no IHS plate on without cracking the die the wafer is real thin and they have no strength. With my 3770k I delidded it the day I got it without running it (Scotty would remember that) and I had a heatkiller 3 block that I used to run on it naked. The standard Heatkiller mount was the exact height that it would squeeze any excess paste out from between the die and the Block I just used to tighten the screws down tight. I removed that CPU from the socket umpteen times and never had a problem with it. But in saying that I wouldnt do it with any other Block because the minute you put too much pressure on the die they will crack. George have you tried the Cool Laboratory Pro under cold it doesnt go well once frozen but its great at atmospheric temps.
No Peter, I have tried it till -70 with Dice and for one hour it did not crack.
But, on LN2 I simply can't know ( - 180 to -190 ). In any case I have plenty of syringes of Coolaboratory, Thermal grizzly Kryonaut and Gelid Extreme. If I were to go real cold though, I would use Thermal Grizzly. Seems all champs have a good word for it ( besides advertising Roman ).
|
|
|
Post by Macsbeach98 on Jul 24, 2016 17:39:12 GMT -5
I use Gelid almost exclusively for cold. Thermal Grizzly is too dear here I would like to try it though.
|
|
|
Post by georgekokovinis on Jul 24, 2016 19:00:11 GMT -5
I use Gelid almost exclusively for cold. Thermal Grizzly is too dear here I would like to try it though. I can check it out and see if I can buy them for you. Interested ?
|
|
|
Post by ozz on Jul 24, 2016 19:27:08 GMT -5
the copper square in the middle of the ihs plate is soldered onto the ihs plate, i can see the solder squeezed out the sides of the copper square that exactly the same size as the die size
|
|
|
Post by ozz on Jul 24, 2016 19:33:27 GMT -5
where as this one the SL7KB COSTA RICA has solder on the ihs plate and it just popped off when i cut around the black silicone stuff and need no heat at all to come off and those scrape marks are from my finger nail,the solder was coming off easily
|
|
|
Post by Mr.Scott on Jul 24, 2016 19:45:46 GMT -5
the copper square in the middle of the ihs plate is soldered onto the ihs plate, i can see the solder squeezed out the sides of the copper square that exactly the same size as the die size View AttachmentYeah...that's not copper buddy. That's the whole die. I can see the circuits right in the pic. They have a holographic like reflection. That's why that CPU doesn't work.
|
|
|
Post by ozz on Jul 24, 2016 20:00:49 GMT -5
yes scotty i see now, the die has come away altogether off the cpu , silly me i also saw the holographic pattern in that pic where as i didnt see it b4 and said that has to be the chip itself thats come away from the cpu , no wonder it wouldnt boot, it had nothing to boot from now i know the pentium 4 in those models are soldered to the ihs plate but with the other cpu the ihs plate just came off easily with no heat, just popped off
|
|
|
Post by Macsbeach98 on Jul 24, 2016 22:51:58 GMT -5
Yep some of the top end P4 478 are soldered
|
|