Printable View
I have a E4400 that I have removed the IHS from, really wanna try lapping the core a little bit.:p:
The core is really concave and has a minimal contact surface, so I think the gain would be substancial.
Anyone seen this been done, and how thick is the protectionsurface on the core?
Lapping the core?? is that posible
noQuote:
Originally Posted by GAR
not if you want your cpu to live.
Does that mean you have tried it?Quote:
Originally Posted by Lestat
It can be done on a Athlon/Opteron...
I got a piece of glass from work with about 2x the area of a sheet of sandpaper (hehe). I started with 220 grit, then 800, then 1500. Went very fast. Copper is so easy to sand. I rotated the CPU with each change of grit. This makes your progress very obvious.
It is a major pain to sand a Scythe Ninja btw. The Al fins are like razor blades, so you have to grasp it at the bottom, which is difficult because of how wide the top is and how little room there is under it. I ended up grasping the heatpipes right at the base, which is risky because they are hollow, soft copper. Took a lot more effort and time than the CPU. Wasn't even really worth it IMO because it had a very smooth finish to start with and was impressively flat.
I got at least a few degree C improvement. Maybe around 5. Hard to determine without controlled testing.
@ Alik4041
you have to use less thermal paste because after lapping you will have shallower valleys and pits, hence the good reflection. remember the paste is only there to fill those tiny air gaps. i used AS5 and it says about 1 to 1 1/2 grain of rice in size. well i used about half of a grain of rice and got nice results, except when i tried mounting the heatsink the wrong way as an airflow experiment. upon lifting the heatsink i could see the AS5 had spread biassed to one side of the middle of the chip
follow the guide on the arctic silver website, the blob on the chip, pushing down and rotating will eliminate chances of air being trapped.
as for your temps, touch your heatsink with no fan blowing on it and see how warm it is. if its cool and the chip is 50 odd degrees then you have a contact issue. but if the temperature of the heatsink is fairly warm and so is the processor then contact should be good.
other possibility is that the cpu or heatsink is concave or convex. pull both out and put a stanley / razor or something deffinately flat across and see if there is a gap. another possibility is that if its in a tower with you, and being top heavy, when you have put it upright it has sagged and seperated a little by the laws of gravity. myself and others have combatted this issue with the use of zip ties off the case or PSU above to give it something to "hang" onto.
also i had to take my board out of the case when re-mounting because i wasnt satisfied i could do a good enough job whilst still installed - not enough room to work. also after pushing & rotating the heatsink down onto the chip i held it in place firmly and painstakenly put the screws in with one hand, unwilling to remove the pressure. i suppose ideally you would be better with a 2nd pair of hands to hold it once its on.
a few things to try, let us know how you get on ... also which method of lapping did you use?
[QUOTE=swaaye;2452908] blah blah QUOTE]
know what you mean. the thermal right's fins are closer together so less chance of digging in but still a pain. i wadded kitchen roll around the heatsink fins but ended up with 3 fingers on the top of the block. took a while to find right amount of pressure and right amount of speed but the heatpipers make it awkward to hold when you rotate.
some people have lapped in a straight line and get as good a finish. someone should get 3 of a particular heatsink and lap each differently; circular, straight & figure of 8 then post results. will be nice to know.
didnt think soQuote:
Originally Posted by Lestat
I lapped my Q6600/Tuniq Tower and got -10C drops across the board, definitely worth the 2 hours it took.
A few thoughts.
* Perhaps one reason temps also get better after lapping is copper is a better conductor than aluminum and the lapping gets rid of the AL. Funny, but my temps actually rise after a lap dance, but that may be an overshare.
* Overclocking doesn't void your warranty, utter nonsense. Not to mention, Intel was running a 5GHz+ overclocking demo by none other than XS owner FUGGER at IDF. Lapping would likely do it though. :ROTF:
* There are other TIM's than Arctic Silver 5 and Ceramique and some work better, like MX-2.
* If you use alcohol, use 91% or better. 70% is about 30% water and can rust metal. Don't believe me? Put a pair of cheap nail clippers in a cup of 70% isopropyl for a few days and see what happens.
* A nice thing to have on hand for cleaning off wayward TIM and any insulation materials (for those of you who chill water or phase change) is some nice spray-on contact cleaner. Works like a charm.
It's nickel actually. It's an awfully thin layer. It is there to prevent copper oxidation. Nickel is much worse at heat conduction than Aluminum even, apparently.
http://www.engineersedge.com/properties_of_metals.htm
I'd like to know what the solder they use to connect the heatspreader to the CPU is made up of. Lead sucks as a conductor.
Why wouldn't overclocking void your warranty? Of course it does. Whether or not they can tell you overclocked is the question. If they knew you did, they would never honor your warranty. I wouldn't be at all surprised if the chip had something like the ECU of a car, in that it records and measures various data. There is extra hardware in the CPU for debugging and testing.
Alcohol won't get paste out of the metal as well as a Xylene or Acetone cleaner. Or, even better, ArctiClean. I love that stuff. I'm more than happy to pay for a cleaner that smells like oranges and doesn't wreck havok on my hands or lungs, while also cleaning as well or better. I bought the big bottles of Articlean and they've lasted me years. Alcohol is useful for lighter hardware cleaning tho. Cuz it dries so fast and leaves no residue (other than that 9% of water for a bit). I have Articlean, Oops, and Isopropyl 91% in my "lab" lol. The Oops rarely gets use because it is so nasty.
though i read that mx-2 performed no better than AS5? thought i read it on these forums too by someone? spawne32 maybe?
Its possible, used to be done to Coppermine pentium 3's to get rid of the thin layer of blue plastic protecting the core.Quote:
Originally Posted by Lestat
I wouldn't do it on any cpu now though, as the cores are really flat anyway, and its not worth doing anything to make them more suseptible to chipping.
I just got some MX-2 in today and put some on my lapped Q6600 / Scythe Ninja pair. The stuff is much thinner than AS5, that's for sure. I used ArcticSilver's application method, where you apply a line of paste across the CPU and put the cooler on top without spreading it yourself. The MX-2 spread out across most of the heatspreader in a nice thin, translucent layer when I took it off to see how it looked.
It is cooling better than AS5 was, too. I'm seeing 3-4 C cooler at least. The AS5 would probably improve over time though because I'm sure pressure on it thins it out over time. That's what I've seen in AS5 applications left untouched for months. But this stuff was nice and thin immediately.
Nonsense. First of all, the Intel warranty says nothing about overclocking. It does mention the vague misuse, neglect, etc. Nowhere in the Core 2 Quad warranty booklet I am holding does it say "DANGER! ONLY RUN AT SPEC SPEEDS AND VOLTAGE OR YOUR WARRANTY IS VIOLATED!"Quote:
Originally Posted by swaaye
Now, I'm not sure what law school you went to, but the one I went to made it pretty clear that contracts - especially provisions which affect your rights - must be very specific and clear, especially those in contracts of adhesion (pre-printed, offered in a take-it-or-leave-it basis), preferably written in big, bold letters. No such language exists. And contrary to what endless retards say on Internet forums, it is not up to Intel to say whether or not a warranty is violated. That is a legal matter. They can fight you on it, but ultimately that would be a decision for smalls claims court in Santa Clara, CA, to make (choice of forum clause in Intel's contract).
Secondly, there is the fact that Intel does everything short of bragging how overclockable it's chips are, offers its top chip with unlocked multipliers, and just invited Charles Fugger to overclock one to 5.3 GHz at their official developer conference (and you can freaking bet I'd be in court with that video, starring Charles Fugger and Intel playing on my laptop. Nice job BTW FUGGER, Jesus H 8.4 SuperPi I just noticed that, ha ha). Intel is actively marketing their Bad Axe boards to overclockers. Have you not seen the front page of this Web site?
Abandoning overclocking for many years Intel started to move forward for enthusiasts desires. Starting with Bad Axe and advancing with Bad Axe 2 overclocking community finally received many tweaking options featured motherboards. Finally as a token of appreciation for trustful and experienced overclockers and as final touch to fully insure the community with the demanded hardware and it's abilities Intel presented it's new generation of upcoming platform based on innovative X38 chipset and astonishing quad-core processor, built using the 45nm manufacturing process - the pioneer in this microelectronics segment presented by one of the famous overclockers and holder of the worldwide community forum - Charles Wirth a.k.a FUGGER. Short performance show shocked the public and community worldwide with fascinating 7 World Records in most popular benchmarks and achievements.
New tweaking utility allows control all major settings within Windows thus giving new, never seen before perspective of Intel motherboards.
And now Intel wants to get up and claim overclocking voids its CPU warranties? Yeah, just try that and see how fast 1) Intel gets hammered by class action suits and 2) AMD uses it against them in marketing.
Finally, there is the element of proof. How, again, would Intel prove this? You cannot just pass over proof. It is an essential element of all legal claims. And no, they do not have a black box in their chips. This urban legend has been thoroughly debunked, including in this thread. :rolleyes:
So there it is. That's my story and I'm sticking to it, here, and in my business law courses that I teach. :p:
Standard disclaimer, IAALBNYL (I Am A Lawyer But Not Your Lawyer). This is not legal advice, so do not rely on it. If I were your lawyer, you would have signed a RETAINER AGREEMENT and paid me a RETAINER FEE. You did not, so we are merely two guys bullsh_tting on the Internet. Nobody in their right mind would believe anything they read written by some anonymous guy on the Internet claiming to be a lawyer, let alone rely on it to their legal detriment. If you do this, despite this clear disclaimer, no jury would be stupid enough to reward you for such folly, should such unjustifiable reliance lead to your detriment. In other words, DO NOT RELY ON THIS AS LEGAL ADVICE BECAUSE IT IS NOT. After all, I am also a brain surgeon, an astronaut, and the President of the United States.
Search google for the reviews and see. MX-2 usually beats AS5 by ~1-4C. Even MX-1 beat AS5 in tests. See this post for more.
The reason I use Ceramique is FUGGER himself said it's what he uses for phase change. I meant to do a head-to-head AS/MX phase test, but I am still getting around to that one. ;) But as for air/water, the results speak for themselves in the reviews.
Quote:
Originally Posted by it_burns_when_i_pee
http://www.intel.com/support/process.../cs-020033.htmQuote:
Originally Posted by uclajd
http://www.intel.com/support/process.../CS-009862.htm
How about some choice quotes from their warranty:
Quote:
Intel warrants to the purchaser of the Product...if the Product is properly used and installed, it will be free from defects in material and workmanship, and will substantially conform to Intel’s publicly available specifications for a period of three (3) years...
That seems to cover their asses pretty well, I'd say. Of course if they can't tell how it was used, the only thing blocking a person from collecting on that warranty is whether they are willing to lie for it. So, a court will judge in my favor if I try to go after a brand new free Intel CPU while informing everyone I blasted my busted one with more volts, clock rate, and heat than it was specified for? Uh huh.Quote:
Further, this Limited Warranty does NOT cover:
- damage to the Product due to external causes, including accident, problems with electrical power, abnormal electrical, mechanical or environmental conditions, usage not in accordance with product instructions, misuse, neglect, alteration, repair, improper installation, or improper testing;
- any Product which has been modified or operated outside of Intel’s publicly available specifications or where the original identification markings (trademark or serial number) has been removed, altered or obliterated from the Product.
Until someone calls them up and gets the go ahead for a free replacement of a baked, overclocked, overvolted chip I'll stick to my belief that they aren't going to give me a new chip. Just because they want to attract overclockers to their product doesn't mean they're going to back them up with a free replacement.
Hell, AMD used to void warranties if they found Arctic Silver on the chip. They required the use of Shin Etsu or a thermal pad along with a stock cooler, I believe. That was years ago though.
http://www.xtremetek.com/info/index.php?id=14
Some companies are starting to officially say that you are allowed to overclock and retain the warranty, however.
classy.Quote:
Originally Posted by uclajd
Well, I guess your expert legal opinion carries the day. :rolleyes:
Did you not even listen? AMD cannot "void" a warranty, nor can any other contractor. Only a court can do that.
You want to not overclock your processor, that is of course your right. What you are doing on XS, I am less sure of. Bottom line, I teach contracts and this one DOES NOT SAY ANYTHING ABOUT OVERCLOCKING. And no, in a pre-printed form contract, Intel has not, in fact, covered their asses from a contractual standpoint. You don't want to test this, fine. But I read the same contract you did before I posted, and I stand by my opinion.
I just love people on the Internet who think they are experts on everything. A lawyer gives his opinion and you just had to come on and argue the law. Amazing. :rolleyes:
And you want to make personal attacks, take it to PM sparky.
Quote:
Originally Posted by swaaye
So, I go to court and tell them that I pounded said fried CPU way out of spec. They are going to reward me with a free one? What exactly would motivate them to do that? What would entitle me to that nice replacement?
Even if the warranty is supposedly worthless to Intel, which I would say is rather strange to say the least, would my ability to go to court matter? Convince me that somehow I can get a free CPU after I knowingly cook the old one. How does that seem right?
Is this how you treat your supposed students too? Heh.Quote:
Originally Posted by uclajd
I've got much better things to do than try to convince you of something. It is obvious that you have no desire to learn anything new, and that you are more interested in winning your argument than arriving at the truth.
So I yield the floor to Swaaye, legal expert.
Quote:
Originally Posted by swaaye
Yeah, ok. From my end it seems more like you just want to be accepted as legal genius (as an anonymous poster, of course) and not questioned. Especially with each post informing me of my closed-minded retardation. I had no idea I was so unwilling to learn.Quote:
Originally Posted by uclajd
I just see it as unlikely that even if the warranty is useless to Intel that a court would let me have a new CPU after I knowingly used it improperly. Of course, I suppose what is proper usage is up for debate.
So, by what you say, warranties don't really exist for any purpose other than as a consumer scare tool? I can do whatever I want with a product as long as said use is not explicity forbidden within a contract. And if I'm willing to go to court over it after the company denies me, I can get a free replacement .
Take it to another thread and stop the hijacking :down:Quote:
Originally Posted by swaaye
Yeppo Moderator better remove off topic posts.... go PM lads and let us just know the final verdict...
I also thought ceramique was preferred over AS5 when applying thermal paste on a lapped CPU... correct me if I'm wrong lads
I think the Ceramique preference is for more extreme cooling situations. I remember reading that AS5's properties change when it is chilled to a point. Not exactly positive on that though.
Ceramique's performance is very close to AS5 so really you wouldn't lose much at all by just going with it. It's cheaper I believe, and you lose that electrical capacitance that scares some people.
i already had AS5 for when i applied mine. but upon building a machine i checked google to see if something at the time (beginning of the year) was better and there wasnt.recently i saw this MX stuff and googled it and the only recommendation i saw was a cheaper price tag for same performance as AS5. i cant speak from experience so i am left relying on reviews but even with reviews you have to take a pinch of salt sometimes because the "laws of averages" where someone will make a better fit than someone else. method of installing / slightly too much or too little paste etc
also in one review the authoer stated that he cleaned the surface with some toilet roll before applying a different paste for his testing. yeah thats good enough! (google sarcasm if you dont get me)
as for AS5 curing time .... on the last 2 machines i didnt notice any drops or increases of temps after the specified time. anyone else found that? not tried ceramique or know if AS5 performs poorly when below a certain temp. im reliant on online info for that.