Lapped the i7

Q

quakefiend420

Limp Gawd
Joined
Sep 22, 2004
Messages
224
took a few hours to finish, i ended with 1500 grit and it's pretty shiny now :D

10C drop in load temps immediately, and the AS5 hasn't even cured yet :eek:

i think by the time it cures i'll have dropped my load temps by 12C or so

anyone else here had such a significant temp drop just from a good lap?
 
Impressive. I don't have it in me to lap my cpu. Heatsink yes, but cpu... I just can't afford to replace it if I mess something up.
Anyway, nice job. What kind of temps are you getting, and what is your cooling setup?
 
transformer 4 w/ 2 120mms push/pull, few 120mm case fans, nothing special

went from [email protected] when running IntelBurnTest to 72C

it was prettu badly out of true, it took me over an hour just to get all the nickel off the heatspreader

i think i'm gonna pull the heatsink and give it a good lapping as well, although i'm not sure how much material i have to play with on the heatpipes, they directly contact the cpu core
 
I noted a 14c difference on one core of my Q6600 - and about 8 on average for the others.

Lapping is the 2nd thing I do in a PC ;)
 
I wish I could lap my cpu =/. Those temperature drops are worth it though.
 
So no before and after pictures? What heatsink are you using and did you lap it also? Pretty good drop on temp for a simple lap though. Enjoy it.
 
I was considering lapping my CPU. I kinda wish I did, since I do see a lot of stories about how much good it did to people. I already saw a huge reduction in temps when I moved to water cooling though, lol.

Maybe if I get bored one day... haha.
 
I'd think with the large heat spreaders on these Intel chips it would be hard to mess it up (unlike the old days of exposed ceramic cores). With temp drops like that I may have to think about lapping my 920 :)
 
10C drop?

I think maybe I saw a 2-3C drop when I lapped my TRUE. Maybe.

Need some screenshots to verify this. Sorry for the skepticism but I've lapped my fair share of heatinsks and cpus and have NEVER seen a 10C drop in temps.
 
BlackDragon24 said:
10C drop?

I think maybe I saw a 2-3C drop when I lapped my TRUE. Maybe.

Need some screenshots to verify this. Sorry for the skepticism but I've lapped my fair share of heatinsks and cpus and have NEVER seen a 10C drop in temps.
Click to expand...

depends on how far from flat the IHS is. If it's severely concaved then I could see a 10c drop.
 
BlackDragon24 said:
10C drop?

I think maybe I saw a 2-3C drop when I lapped my TRUE. Maybe.

Need some screenshots to verify this. Sorry for the skepticism but I've lapped my fair share of heatinsks and cpus and have NEVER seen a 10C drop in temps.
Click to expand...

i don't have any before screenshots, i didn't expect to have to prove anything to anyone about it :/

besides that, if i wanted to lie about it, i could keep everything the same(voltage, fan speeds, etc) and drop ambient temps, or add more fans, or change heatsinks, or point a box fan at the side of the open case, etc etc screenshots don't prove much of anything really, and i don't have any reason to lie about it anyway

i did snap a quick pic of the cpu though...in this pic i had moved up to 1000 grit, almost all the nickel was gone at this point...that one side closest to the camera was quite stubborn

2010-03-11233401.jpg
 
HydroBudz said:
depends on how far from flat the IHS is. If it's severely concaved then I could see a 10c drop.
Click to expand...

it was a wavy surface, the outside on opposite sides started coming off first, then slowly started getting into the middle, and i thought the last side that you can see in the pic was never going to come off...i was getting nervous at this point, afraid that it may had just been so badly out of true that i was going to damage the cpu trying to get rid of that last little bit of nickel..came out ok though :D
 
mrmylanman said:
I was considering lapping my CPU. I kinda wish I did, since I do see a lot of stories about how much good it did to people. I already saw a huge reduction in temps when I moved to water cooling though, lol.

Maybe if I get bored one day... haha.
Click to expand...

liquid cooling is next on the list...i'm still working on maxing out air first
 
I don't see the point of a before shot anyway == any i7? And, after == very shiny mirror-like little square/rectangle? meh
 
I meant screenshots of desktop monitoring software reporting temps.

I don't mean to be the asshole but I just don't believe every claim I read on the internet. Coming on and saying 10C drop would mean a lot more to me if there was a screenshot under load before and after with a temp monitoring program like realtemp running to show the temp differences.

And at the same time, if the 10C drop is real, congrats.
 
@ BlackDragon ... of course, brain turned off here. And... I tend to agree. Although, "who knew" if I were the OP as I would not have expected a significant change.

Then would have come the questions about the original mount install.
 
BlackDragon24 said:
I meant screenshots of desktop monitoring software reporting temps.

I don't mean to be the asshole but I just don't believe every claim I read on the internet. Coming on and saying 10C drop would mean a lot more to me if there was a screenshot under load before and after with a temp monitoring program like realtemp running to show the temp differences.

And at the same time, if the 10C drop is real, congrats.
Click to expand...

I know what you meant, reread my reply. If I wanted to lie about it I could easily change conditions the screenshot doesn't reveal. And I honestly wasn't expecting a drop of that magnitude either, maybe 5C tops...and as the poster below me said, it's also possible that my original heatsink mount wasn't as good as it could have been the first time around as well. Regardless, I'm quite pleased with the results :)
 
quakefiend420 said:
I know what you meant, reread my reply. If I wanted to lie about it I could easily change conditions the screenshot doesn't reveal. And I honestly wasn't expecting a drop of that magnitude either, maybe 5C tops...and as the poster below me said, it's also possible that my original heatsink mount wasn't as good as it could have been the first time around as well. Regardless, I'm quite pleased with the results :)
Click to expand...

I'll be honest. This thread almost made me yank my TRUE off today and touch it up again... and then lap the damn cpu :). Almost. Maybe when I've got 4 spare hours. She definitely runs hot.
 
We call this kind of reporting "anecdotal evidence". Assuming OP is as honest as he can be, lapping an i7 can only be beat out with several HSF mountings & re-mountings (sounds like fun) before & after. I have read a lot of conjecture about how lapping is not necessary ... as I said "conjecture". Yes, the net is full of experts that have done what they are preaching.:eek:

It can only be known by *doing* & being as careful & consistent as possible. In other words report how you do what you do so that others can at least come close to copying.

Just my $.02
 
i lapped my old i7 965xe , the temp drop is not that significant. but once i have time i will lapped my W3520 and report back
 
Psi* said:
We call this kind of reporting "anecdotal evidence". Assuming OP is as honest as he can be, lapping an i7 can only be beat out with several HSF mountings & re-mountings (sounds like fun) before & after. I have read a lot of conjecture about how lapping is not necessary ... as I said "conjecture". Yes, the net is full of experts that have done what they are preaching.:eek:

It can only be known by *doing* & being as careful & consistent as possible. In other words report how you do what you do so that others can at least come close to copying.

Just my $.02
Click to expand...

Very true...I wasn't going for any sort of scientific result, and I can't be 100% sure that just the lapping was responsible for the temp decrease.
 
BlackDragon24 said:
I'll be honest. This thread almost made me yank my TRUE off today and touch it up again... and then lap the damn cpu :). Almost. Maybe when I've got 4 spare hours. She definitely runs hot.
Click to expand...

I'm going to go after the heatsink sometime this week...anyone have any experience with lapping a direct contact heatpipe cooler?
 
I have an E5200 in my system and they are pretty darn cheap.... and I am planning on upgrading to a q9400 or q9550 sometime in the near future so I may just have to lap my E5200 for the fun of it. I have always wanted to but never had a disposable CPU in case things went horribly wrong! :D
 
I would not lap your heatsink, you would be very likely to punch a hole through the direct contact heatpipes and let the working fluid out
 
jeffline said:
I would not lap your heatsink, you would be very likely to punch a hole through the direct contact heatpipes and let the working fluid out
Click to expand...

well, maybe i won't then

i did get to 4ghz tonight though :D

think that's about all she has in her on air though
 
Maybe I should pull my 920 out and lap it some more... I just did a quick lap on mine... no copper showing whatsoever. I do get a pretty good contact patch though, not perfect, but pretty good.

I can believe that a 10c drop is possible. Haven't seen that much drop from lapping in quite a while, but it is pretty close to what I got when lapping my Opty 170. Then I ended up removing the IHS and ended up with over a 20c drop in load temps vs unlapped.
 
10C that is impressive! A little off topic, I lapped my amd x2 3800+(no way would I lap my q9450) as an experiment. I only saw 2C off but then I didn't lapped it correctly and only spent 30min-45min :/. OP, do you mind telling me what method you used to lap? Thanks in advance!
 
quakefiend420 said:
I'm going to go after the heatsink sometime this week...anyone have any experience with lapping a direct contact heatpipe cooler?
Click to expand...

Yes. I lapped my Xiggy S1283. Best thing to do is take a sharpie and put criss-cross lines on the heatsink base then start lapping. If they all come off after 10-20 passes on 6-800 grit then don't lap any further. Only remove material until the lines you drew are gone.

You'd have to lap for quite some time with low grit paper to get through the heatpipe.

Aside from that my base was quite flat to begin with. Didn't notice any significant drop in temps, but it did make me feel better knowing my IHS and heatsink base were as flat as I could possibly get them. My Q6600 @ 3.6 1.41v never got over 55c at load with dual medium speed yates in push-pull.
 
hmm judging from some of the earlier posts apparently the standard [H]ardcore practice of at least checking your IHS and or Heatsink for flatness seems to be a lost art. 5 seconds with a razor blade.

Those are not a good idea to lap ... doesn't make sense actually ... does it?
Click to expand...
In the sense that much more care and much less material can be removed or the risk of sanding through a heatpipe and destroying the heatsink goes through the roof, yes. But I would not hesitate to at least check what is going on, I might touch it up, carefully, very carefully.

We call this kind of reporting "anecdotal evidence". Assuming OP is as honest as he can be, lapping an i7 can only be beat out with several HSF mountings & re-mountings (sounds like fun) before & after. I have read a lot of conjecture about how lapping is not necessary ... as I said "conjecture". Yes, the net is full of experts that have done what they are preaching.
Click to expand...

We have years and years of posts with before and after data, there is no question. Additionally the thermodynamics and physics are pretty clear cut. We do not require peer reviewed posts. You are free to discount the advantages of lapping an uneven IHS but I dont think you will get many followers here.
 
BillParrish said:
We have years and years of posts with before and after data, there is no question. Additionally the thermodynamics and physics are pretty clear cut. We do not require peer reviewed posts. You are free to discount the advantages of lapping an uneven IHS but I dont think you will get many followers here.
Click to expand...
WELL!!! We will just see what my minions have to say about this!!!

Aren't all forums pretty powered by opinions Bill? What exactly is the point of your post?
 
Last edited:
BillParrish said:
hmm judging from some of the earlier posts apparently the standard [H]ardcore practice of at least checking your IHS and or Heatsink for flatness seems to be a lost art. 5 seconds with a razor blade.


In the sense that much more care and much less material can be removed or the risk of sanding through a heatpipe and destroying the heatsink goes through the roof, yes. But I would not hesitate to at least check what is going on, I might touch it up, carefully, very carefully.



We have years and years of posts with before and after data, there is no question. Additionally the thermodynamics and physics are pretty clear cut. We do not require peer reviewed posts. You are free to discount the advantages of lapping an uneven IHS but I dont think you will get many followers here.
Click to expand...

I'm thinking the power trip isn't necessary. The thread is quite civil.
 
BillParrish said:
hmm judging from some of the earlier posts apparently the standard [H]ardcore practice of at least checking your IHS and or Heatsink for flatness seems to be a lost art. 5 seconds with a razor blade.


In the sense that much more care and much less material can be removed or the risk of sanding through a heatpipe and destroying the heatsink goes through the roof, yes. But I would not hesitate to at least check what is going on, I might touch it up, carefully, very carefully.



We have years and years of posts with before and after data, there is no question. Additionally the thermodynamics and physics are pretty clear cut. We do not require peer reviewed posts. You are free to discount the advantages of lapping an uneven IHS but I dont think you will get many followers here.
Click to expand...

I agree with all of this. Lapping has been proven time and time again. It's not for everyone. Lappers don't give a squat about warranties and care more about raw performance I guess.

It's not rocket science. Basic thermodynamics show that two flat surfaces mated together will transfer heat more efficiently than two warped surfaces.
 
Psi* said:
WELL!!! We will just see what my minions have to say about this!!!

Aren't all forums pretty powered by opinions Bill? What exactly is the point of your post?
Click to expand...

Lapping works and anyone who thinks otherwise is uneducated on the subject. When opinion flys in the face of fact it is no longer an opinion but an untruth and I do not knowingly let untruths about technology propage here. The rest of the web is 98% crap, but not here in OC& cooling. Not if I can help it. Dont like my stand, my opinion or my methods? Feel free to take it up with the management.
 
Last edited:
The only time lapping won't benefit is if you either screw it up or they were so close already that it wasn't possible for you to make it better. Both are unlikely.

My question is will lapping for my phase change cooler benefit enough to make it worth it or is it pointless? Wondering if it's only an air / water thing.
 
BillParrish said:
Lapping works and anyone who thinks otherwise is uneducated on the subject. When opinion flys in the face of fact it is no longer an opinion but an untruth and I do not knowingly let untruths about technology propage here. The rest of the web is 98% crap, but not here in OC& cooling. Not if I can help it. Dont like my stand, my opinion or my methods? Feel free to take it up with the management.
Click to expand...

Bill:

No one is saying lapping doesn't work.

Nobody has said that in this thread, at least not the posts I can read that haven't been deleted.

The debate is about HOW beneficial it would be, you know, cuz we are deciding if we want to void the warranty on our $300-$1000 cpus. OR we want to avoid lapping right thru the heatpipes on our fancy direct contact coolers.

I completely understand your reasoning for taking stands against people propagating lies in your forum, I just feel like we are doing fine without the police getting involved.

Cheers
 
BlackDragon24 said:
Bill:

No one is saying lapping doesn't work.

Nobody has said that in this thread, at least not the posts I can read that haven't been deleted.

The debate is about HOW beneficial it would be, you know, cuz we are deciding if we want to void the warranty on our $300-$1000 cpus. OR we want to avoid lapping right thru the heatpipes on our fancy direct contact coolers.

I completely understand your reasoning for taking stands against people propagating lies in your forum, I just feel like we are doing fine without the police getting involved.

Cheers
Click to expand...

The choice to lap is up to the end user. How beneficial it will be is a crap shoot. Best to take a razor blade to the IHS surface and see if light passes through. If it does, at all, then you probably will benefit from lapping. How much is the golden question. Won't know till you try. My stance is make sure the proc works 100% before lapping at all. If the proc is 100% then a warranty is the last thing you should be worried about. Only voltage will kill it at that point. Technically if you overclock you void your warranty anyways, so what's the difference?

As for lapping a HDT cooler. Do as I said before. Use a sharpie and draw hash marks across the entire heatsink base. Then lap until the hash marks are gone. No further. You're only taking off microns worth of material. Unless your lapping with 120grit paper I wouldn't worry about lapping through the heatpipe. Start with 800 grit and end with 20 passes over 1000 grit.
 
I said nothing about lies nor called anyone a "liar". I am an engineer, I say what I mean and mean what I say. Untruths most certainly do not equal lies. There is a huge fundimental difference.

This entire subject "is it worth it" has also been beaten to death many times. But whatever, have at it. I could ask Kyle if he will change the name of the site to havenotesticalesOCP but I after seeing his method of erasing data off of old hard drives I do not think I will.

http://www.hardocp.com/news/2009/03/21/50_bmg_api_vs_18_hard_drives/
 
Last edited:
You must log in or register to reply here.
Back
Top