Printable View
And who forced you to? You decided to waste your time. So what? :D Some poeple around here are really funny.
You should try to understand that there are two thermal junctions. One is between the heatsink and the IHS, and the other is between the IHS and the actual core. If you don't remove the IHS, you'll always be limited by the junction between core and IHS. That is reality and you'll have to live with it. ;)
Unless he got one of those new Core2Duo's that dont need to be lapped.
http://www.xtremesystems.org/forums/...=140701&page=6
I'm WC'ing my chip, and using a bowed GTX that hopefully will enable me to avoid lapping. I change chips too fast to sacrifice the ability to sell it at a good price...
Quote:
Originally Posted by celemine1Gig
As I'm inclined to agree with celemine1Gig, I also have to remember you guys, especially Gam3Ra, that thermal paste/grease does have a curing period. If you have the same temperatures right after installing the cooler, chances are in a couple of days those temps are going down a couple of degrees, after the thermal paste has settled. Also, remember you may have put too much paste, or some air got caught between the IHS and the cooler, etc. There are too many variables to linearly discard lapping right away...Quote:
Originally Posted by Gam3Ra
And, if nothing else, you might as well have learned how to lap. That's always something...
Might also be true. In fact, I think I might be one of those... My E6400 is capable of insanely low voltages. Through RMClock, I can go as low as 1.163v (if I remember correctly) without any problems (stock speed) and my load temps don't even cross 50ºC, with semi-passive cooling (only the PSU's 12cm low speed fan moves air). Probably lapping wouldn't do me any good, as idle temps are usually on the 30ºC mark, which is probably 8ºC above room temperature...Quote:
Originally Posted by Kurz
Cheers.
Miguel
Yes, I know all that ;)
When I lapp my P4 540J cpu at load loose some 8-12C, gives me more OC. The IHS was perfect.
That fools me to lapp the C2D.
Well, at least you didn't get worse temps after lapping. No, seriously speaking: You didn't damage anything, that is one positive point. You may have ruined your warranty, but who cares as long as the chip still works. AND, you still have the option to fully remove the IHS later. So, IMHO, the lack of temp decrease is rather a minor problem. OK, you lost one hour of your life, but it's the same when watching TV nowadays. :D
The fact that you lapped the CPU in the first place shows some xtremesystems spirit and that's what counts. ;)
Well, a Netburst-based 90nm 3.4GHz CPU is bound to be VERY HOT, and benefit from lapping much more than a C2D... You have to factor in the law of diminushing results. It is much easier to shave some degrees from a hot thing than from a cold one...Quote:
Originally Posted by Gam3Ra
But I guess you also know that :p:
Nice one! :lol2: :rofl:Quote:
Originally Posted by celemine1Gig
Very true. X2 that.Quote:
Originally Posted by celemine1Gig
Cheers.
Miguel
Actualy is 3.2ghz ;) And LOAD at that frequency with 1.15V was at 30-32CQuote:
Originally Posted by __Miguel_
C2D @ 3600 1.33V shoud be hotter, but no changes after the lapping.
Lapping the P4 gives me stability with lower voltage before that, I will test the C2D :p:
EDIT
3200Mhz 1.175V good
Is it possible not having any drop on temp after lapping? Eventho it's my 1st time lapping a cpu and i managed to pull out 7-9 degress with AS5.
You did read all postings on this page, did you?Quote:
Originally Posted by laziepig
A proper lapping job should produce better temperature results.Quote:
Originally Posted by __Miguel_
It's not so much dependent on the processor type, rather on the lapping job and the block used to cool the cpu.
The Noctua (heavy) is keen to flex the board.
Take a look at these results, extracted from the review Cooling-Masters web site made of the Apogee GT.
http://www.cooling-masters.com/image...es/fullc2d.png
:eek2:, that's a cool Pentium, nice one! Don't know why I thought it was a 3.4, though... Maybe I'm confusing with the 9xx series. Or I'm getting old... :rotf:Quote:
Originally Posted by Gam3Ra
Oh, one thing I remembered: try to lap the cooler too, as I've seen in the thread some people have had better results after lapping both things.
One little question, though... Doesn't lapping may eventually lead to loss of connectivity between the IHS and the cooler? I mean, LGA 775 assumes the CPU (IHS included) has a determined height. Lapping changes that, and may in fact widen (spelling?) or create a gap between those two components. That would hinder cooling performance, wouldn't it? Or am I just being silly?
Cheers.
Miguel
[EDIT]Fiber9, those are very interesting results. It seems there are coolers that gain more efficiency with high water flow, while others see the biggest drops in temperature with low flows... Just to show not every water block is as efficient as its cousin... Two other findings that seem interesting to me are the overall "flatness" of the temperature lines, which means a better thermal performance; and that MP-05 SP, which has an impressive drop in temperatures... One of the worst with the stock IHS, and clearly the best up till 250L/h with lapping... That is really amazing...[/EDIT]
Well... Gam3Ra is the only person complaining! hehehQuote:
Originally Posted by celemine1Gig
I think Intel put a enough space for any change up to a few mm.Quote:
Originally Posted by __Miguel_
Unless you take off the IHS you wont have a problem.
Thanks, that's good to hear. I never actually made any measurements, but it would be rather weird it that was not possible...Quote:
Originally Posted by Kurz
Cheers.
Miguel
Very nice guide, Eastcoasthandle. And with pictures, too :D Cool! Way to go!
I might try something like that with my Ninja... Maybe I can make it even better than it already is... hehe But first, I have an HTPC/modding project I have wanting to do for the longest time. Who knows, I might even lap something in the project... lol
Cheers.
Miguel
Thanks, I hope it's useful to everyone who might want to try lapping.Quote:
Originally Posted by __Miguel_
Hmm, i lapped my e6600 and it was my first lapping chip :D got it to a nioce mirror finish and its around 4-5C cooler. Also lapped my Scythe Ninja.
I kinda used my own technique. I just started with some 400. Go the IHS as flat as possible. Then with the 600 i started to just move up and down, rotating the chip whenever. That started to peel the tin off. Then i stopped the 600 when all of the tin was off.
Started up with some 1000 grit i got from the auto store. I used 99% alcohol to make the chip wet. Worked very well. Then i just kept on going up to 1500, then 2000.
Took a total of 2hrs or so to lap the CPU and ninja, w/ one person on the cpu and one on the HS ;)
Looks great, cept i dont have any pics. I was in too much of a hurry to put it back in my pc and check the temps... :D
Yeah know., the last cpu heat spreader I lapped was a PD 805. I have lapped a couple (ie: X2 4200+, 3000+, PD 805, P4 2.66, Celly 346) and I have NEVER got a 10c gain in temperature. I think anyone that claims they did: 1)had improperly installed their HSF to start with, 2) has bad eye sight or a temperature program or temp sensors that are very inaccurate or 3) are stone cold bullcrap artists. I have at present 4 C2D's and one on the way. I run them 24/7 folding and have checked all of them with the "ketchup glass" technigue to see if they were concave or convex (thankfully none so far are, I'm waiting on another E6600, I already have one, and I haven't had a chance to check it yet) IMO anyone that laps a cpu heat spreader without checking it first and voids the warranty is not very smart. (even if you never warranty it, the chip would be harder to sell) I've also read on many sites creating a "mirror" finish is a waste of time, it seems 600 grit paper is sufficient (I like to go to 1500 grit, I'm apearence anal), flatness is what's the most important. And another thing I read a post saying the AS "line" method of applying TIM wasn't any good, all I can say is I'll join the other millions of satisfied users and say I think the line method works great. I've done many cpu resets using the "line" method, both for me and people with first class gear I know. Evidence to support what I have I posted "you've got to be kidding me". Being a point provider in the top 20 of a folding team (pretty regularly) of over 6,000 people ought to be a little convincing even for a "moron". jws2346 , the greatest team on the "Net" Team 32 (overclockers.com)Quote:
Originally Posted by Contero
Quote:
I lapp my E6400, the temps are the same, I waste 1 hour of my life and small amount of AS Ceramique
10 degrees? you gotta be kiddin! thats insane! that is a complete lie !Quote:
Yeah know., the last cpu heat spreader I lapped was a PD 805. I have lapped a couple (ie: X2 4200+, 3000+, PD 805, P4 2.66, Celly 346) and I have NEVER got a 10c gain in temperature. I think anyone that claims they did: 1)had improperly installed their HSF to start with, 2) has bad eye sight or a temperature program or temp sensors that are very inaccurate or 3) are stone cold bullcrap artists. I have at present 4 C2D's and one on the way. I run them 24/7 folding and have checked all of them with the "ketchup glass" technigue to see if they were concave or convex (thankfully none so far are, I'm waiting on another E6600, I already have one, and I haven't had a chance to check it yet) IMO anyone that laps a cpu heat spreader without checking it first and voids the warranty is not very smart. (even if you never warranty it, the chip would be harder to sell) I've also read on many sites creating a "mirror" finish is a waste of time, it seems 600 grit paper is sufficient (I like to go to 1500 grit, I'm apearence anal), flatness is what's the most important. And another thing I read a post saying the AS "line" method of applying TIM wasn't any good, all I can say is I'll join the other millions of satisfied users and say I think the line method works great. I've done many cpu resets using the "line" method, both for me and people with first class gear I know. Evidence to support what I have I posted "you've got to be kidding me". Being a point provider in the top 20 of a folding team (pretty regularly) of over 6,000 people ought to be a little convincing even for a "moron". jws2346 , the greatest team on the "Net" Team 32 (overclockers.com)
Why should it be impossible? Do you have a clue what you are talking about? If you have a convex CPU and a concave heatsink or vice versa and than lap both, a 10°C drop is well within the possible range. I'd even go as far as to say that much higher delta T's are posible when comparing the highest overclock.
Has anyone else tried to lap their IHS or HS to see if you decrease your temps?
jws2346,
BTW... What is the "ketchup glass" technique? Googled it but didn't find anything...
13 pages? by far the best thread i ever made :lol:
temps may improve but not that drastic! where talkin about 10 degrees here, 3 is feasible. but 10? whew! besides the convex part you're talkin about is covered with thermal gunk! its job is to facilitate the transfer of heat on the not so even areas. do some reading first before you blurt out anything will ya.Quote:
Why should it be impossible? Do you have a clue what you are talking about? If you have a convex CPU and a concave heatsink or vice versa and than lap both, a 10°C drop is well within the possible range. I'd even go as far as to say that much higher delta T's are posible when comparing the highest overclock.
a 10 degree difference using a hi end hsf versus a watercooling setup is sometimes even hard to reach. if you're idea's so great howcome the boys in intel havent thought about this? lolz