Sign in with
Sign up | Sign in
Your question

CPU Lapping

Tags:
Last response: in CPUs
Share
June 30, 2009 6:25:20 PM

So.. i've been trying to lap Pentium 4 (as i havent lapped before), before i proceed with my i7..

For some reason, there is a circle, where the copper just wont show..

the rest of the processor IHS removal is going very smooth, but there seems to be a spot that I just cant remove..

My grits were: (wet, and tried dry after):

400
600
800
1000
1200

i didnt find it necessary to move to 1500/2000 yet, since there is still IHS on the surface.

I tried going backk and forth on the grit sizes to remove it, but it just wont come off..

Of course i followed the steps as told, applying no pressure, etc etc..

any ideas?

ive attached some pics (from my mob phone, so excuse the quality)






More about : cpu lapping

July 1, 2009 12:14:47 AM

bump? any1 can please help me out here?
a b à CPUs
July 1, 2009 12:48:20 AM

Keep going, it will get down to the copper eventually.
Related resources
July 1, 2009 12:54:18 AM

i turned back to the 400.. been doing it for like 45min.. nearly dissapeared..

shud i still continue with the 400, and as soon as it dissapears, go to 600, smooth it out, 800, smooth it out, etc?

i've been using up/down motion and turn 90 degrees after lets say 30 times..

is that a good method? or should i use another? if yes whch would you recommend?
a b à CPUs
July 1, 2009 3:07:26 AM

Hmmm, spending the evening lapping your processor.
SURELY you have got better things to do in your spare time???
Stick a good heatsink on that I7 and forget about it!
It will be just fine......
July 1, 2009 3:29:50 AM

ombadboy said:
i turned back to the 400.. been doing it for like 45min.. nearly dissapeared..

shud i still continue with the 400, and as soon as it dissapears, go to 600, smooth it out, 800, smooth it out, etc?

i've been using up/down motion and turn 90 degrees after lets say 30 times..

is that a good method? or should i use another? if yes whch would you recommend?

The main issue to keep in mind is if the IHS is FLAT, not smooth. Have you checked the surface with a straight edge or by using the water & glass method? Even if the IHS has a polished finish, temperatures will not drop unless you were able to flatten the surface almost perfectly.

As for the nickel plating, removing it entirely will only give a negligible reduction in temperatures, but it doesn't hurt to lap the entire nickel layer off if your willing to spend more time doing that.
July 1, 2009 10:04:24 AM

i've read loadsa tutorials on lapping etc, and they mention a razor to check for flatness, although they do not go into any detail on how to go on about it..

how would i proceed with your 2 flatness tests? ill give it a go, and see how 'flat'' it is..
July 1, 2009 12:29:08 PM

after a few more hours of work.. here's what i got after 1500 grit..

dont want to use my 2000 grit sandpapers, as i want to save em for my 'real' lapping of the i7..

what do you think about the result?





July 1, 2009 1:59:44 PM

ive placed the razor on top, seems that there is contact throughout the razor's edge, except the edges which are desgined in a weird way on the P4
July 1, 2009 2:03:26 PM

btw.. as my english aint very good.. when we say razor blade.. does the tool use to cut paper count as razor blade?

the one u can retract etc etc..

edit: went and bought a razor blade.. cant see any light peeking through.. so i guess all is ok..
July 1, 2009 8:59:28 PM

well my P4 aint working.. so i just used its processor to practise a bit..

i've just finished lapping my i7.. i tried the razor blade test.. i cant see any light seeping thru (as in reflecting on the other side), but when i place my head/eye on the same line with the razor blade edge touching the processor, i can very vaguely see the light on the other side..

i've tried testing flatness using a graph paper, and the reflection the CPU makes..

the reflection of the lines are pretty straight, besides the very edges of the processor (i find the corners of the i7 a bit curved, and only the 400 and 600 grit were able to make an impression on them)..



if thats what has to be tested, i find it kinda impossible to prevent this, as its light and ull be able to see it even in the smallest space..
July 2, 2009 12:07:03 AM

pics of the lap.. what do you think?







July 2, 2009 12:12:58 AM


Nice! I suggest you pop it in and try it out But before you do, be sure the contacts on the bottom of the chip are clean; the copper dust can cause a "jumper" between two contact areas, leading to a short. I've found that using a lint free cloth or q-tips with rubbing alcohol can clean off any dust you find (if any).
July 2, 2009 12:33:39 AM

aye.. ofc course ill clean it up..

some pics from the graph paper test (excuse the dirty surface.. didnt clean it well yet with alcohol):




July 2, 2009 2:43:44 PM

lapping it as we speak.. and its takin aaaages :p  on 220 grit and it just wont flatten.. the edges are very low
July 2, 2009 10:44:16 PM

just finished lappingmy hfs.. and started relapping my CPU..

should i get it down to the level of the corners? the corners are pretty damn low (done an X mark.. sanded like for 30x8 rounds on the 220 grit.. and the marker on the corner is still visible).. the difference of the level of the corners is even visible by the naked eye..

or should i give up with the edges and just move on the 400 flatten the middle even further and polish with 800 1000 1500 and 2000?
!