• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Sandpaper trick worked for me!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

ElectricMan

Registered
Joined
Dec 17, 2000
Location
Northeast Tennessee
I tried that sandpaper trick And It worked!!!!!!! I am proud to say that I am at 1010 mhz . My temp is 43c idle and 47c under load. I don't know who started that trick but it does work, at least it did for me! Thanks to the unknown trick man:)
 

Allan Nielsen

Senior Member
Joined
Dec 18, 2000
Location
Denmark, Europe
Yeah, lapping works for most... unless they just can't stop lapping at a certain point! :)
Good, old SurlyJoe taught me the trick. Btw., where is he now? Doesn't see him post anymore?
 

Murphy

Senior Member
Joined
Dec 18, 2000
Location
Amsterdam, the Netherlands
where to get it?

o.k. I'm convinced now. But where can a find a good article/link about how to do it and what I need for it. p.s. what is "sandpaper", the same thing as used to do wood? No, I'm not a native english speaker, but I think I can figure out how to do it!
 

Allan Nielsen

Senior Member
Joined
Dec 18, 2000
Location
Denmark, Europe
Yes, you can use the same sandpaper as you use for wood! Make sure to use some VERY fine sandpaper when you lap your cooler, and be sure to keep it level, to improve cooling performance.
 
OP
ElectricMan

ElectricMan

Registered
Joined
Dec 17, 2000
Location
Northeast Tennessee
h2k (Dec 20, 2000 12:16 a.m.):
I think I'm missing something...
what exaclty does this trick involve?
Is there a url?


well what I did was take out the cpu and clean off the heatsink paste from both the cpu and heatsink ( I used rubbing alochol with a qtip to clean paste off) Then I took some 600 grit emery cloth sandpaper ( The black sandpaper) and very carefully polished the core until it looks like dark chrome, and used 400 grit sandpaper and did the same to the heatsink, cleaned them with rubbing alochol and reapplied paste and was 5c cooler all the time!
 

markedmundb

Senior Member
Joined
Dec 18, 2000
Location
Kent, UK
Allan Nielsen (Dec 18, 2000 05:13 p.m.):
Yeah, lapping works for most... unless they just can't stop lapping at a certain point! :)
Good, old SurlyJoe taught me the trick. Btw., where is he now? Doesn't see him post anymore?

I was pondering the same thing...
 

Han Solo

Registered
Joined
Dec 18, 2000
Then I took some 600 grit emery cloth sandpaper ( The black sandpaper) and very carefully polished the core until it looks like dark chrome, and used 400 grit sandpaper and did the same to the heatsink, cleaned them with rubbing alochol and reapplied paste and was 5c cooler all the time![/quote]

starting at 600 grit is find but you really should end up w/ 1200 or 1500 also you should use glass or some other surface you know to be absolutly flat to put the sand paper on. For me I started at 600 like you then went to 900 and finally 1500. My heat sink bottom is like a mirror now. just doing 600 won't hurt but leave really small lines going to 1500 youll end up haveing very small lines. in the area of 2.5 times smaller...it might help lower temps another degree or 2
 

pestul

Registered
Joined
Dec 19, 2000
I don't see the trick. Isn't it obvious that the manufacturers are craptastic in the first place. I did it to my Geforce DDR, but that was just to put a new hs/fan on it.
 

amd

New Member
Joined
Dec 23, 2000
Not sure I agree with Han Solo.
I would have thought the theory behind this technique is the increase in surface area provided by indendations/scratches; and therefore the increase in heat transfer.
Polishing it smooth only serves to minimise surface area.

It would seem to me to be taking the concept of the large surface areas of the coolers and applying it on all levels.
 

CFusion

Registered
Joined
Dec 17, 2000
amd (Dec 23, 2000 08:45 a.m.):
Not sure I agree with Han Solo.
I would have thought the theory behind this technique is the increase in surface area provided by indendations/scratches; and therefore the increase in heat transfer.
Polishing it smooth only serves to minimise surface area.

It would seem to me to be taking the concept of the large surface areas of the coolers and applying it on all levels.

The smoothness has a decrease in surface area that is negligible, it's nothing. But the actual surface-to-surface contact is increased enough to greatly improve heat tranfer between core and heatsink. (must use a super thin [talking paper thin] layer of thermal goop to effectivly pull this off.
 

cjtune

Member
Joined
Dec 18, 2000
CFusion (Dec 23, 2000 12:33 p.m.):
amd (Dec 23, 2000 08:45 a.m.):
Not sure I agree with Han Solo.
I would have thought the theory behind this technique is the increase in surface area provided by indendations/scratches; and therefore the increase in heat transfer.
Polishing it smooth only serves to minimise surface area.

It would seem to me to be taking the concept of the large surface areas of the coolers and applying it on all levels.

The smoothness has a decrease in surface area that is negligible, it's nothing. But the actual surface-to-surface contact is increased enough to greatly improve heat tranfer between core and heatsink. (must use a super thin [talking paper thin] layer of thermal goop to effectivly pull this off.


NO NO NO It's to decrease the roughness of the surfaces in contact. Look, the main reason why thermal paste is used is to fill in the air gaps created when two rough surfaces contact. Although the thermal conductivity of most thermal pastes are leagues below that of aluminium, the thermal conductivity of air is much worse, plus the NET surfaces in contact (CPU to heatsink) is also less than if you smoothened both (yes, consider lapping the bottom of your heatsink as well) surfaces by lapping. As a bonus, there is also less thermal RESISTANCE from the CPU die as you have effectively removed some amount of them and this places the heat source (CPU core) closer to the heatsink. Go find some 'Engineering Heat Transfer' textbook or something, it's all there...look under the topic 'interface/contact resistance'. Putting pressure on the heatsink onto the CPU will smoothen (literally squash the microscopic surface imperfections) out some of the surface roughness as well, so a strong heatsink retention mechanism would be helpful as well.
 

surlyjoe

Senior Member
Joined
Dec 27, 2000
Location
far west
ya that's right ,,you want as little goop as possible ,,since it doesn't transfer heat as well as the metal does.


oh ya ..I been real busy w/real life :p
 

kn7671

Registered
Joined
Dec 19, 2000
For those who still do not get the benefit from lapping two surfaces, let think of this in a different larger way. If you take a fork, the kind you eat with, and deeply scratch a flat surface of wood with each point of the fork, say 4 points on this fork. Now that you have 4 scratches, observe their depth. Question 1: Does this increase surface area? Yes! Question 2: Does this increase surface contact? No, it reduces surface contact. Take a small piece of glass and lay it on top of the wood, this will simulate your CPU surface. Do you now see that those scratches will not contact your glass, CPU. Next, what is in those scratches, air. Air heats up very quickly and does not transfer heat well at all. Infact, air pockets will create hotspots, so where these scratches are will infact remain cooler on the heatsink and hotter on the CPU since they don't make contact. Next, heatsink compound, it will fill in these scratches or gaps, but its main purpose it to eliminate air pockets and small imperfections between the surfaces and reduce the hotspots between the contacts. So back to you model with wood and glass. Fill the scratches in the wood, slightly overfill them so you have a rise is the oil above the surface, with a light cooking oil or water and replace the glass. You should now observ the oil flatten out across the entire surface between the scratches, if you didn't you did not use enough oil or water. Now if you apply light pressure to the glass with you finger tips spread open across the glass you should notice even more speading of the oil, possibly squeeze out beyound the glass or wood. This is because all of the oil has filled the holes and now has to escape. This actual surface contact has not really increased, you still have the scratches in the wood, but the oil will transfer some of the heat to the glass now, having reduced air pockets. Apply this knowledge to CPU's and heatsinks. This concludes lesson 101. Still don't get it? Put a frying pan on a burner, put the heat on low, and rest your hand on the pans flat surface for a few seconds, probably hot, but what happens when you press down with your hand, it flattens out more, reducing air pockets, increasing heat transfer, making your hand even hotter. Now put cooking oil in the pan, this will fill in any grooves, finger prints, etc... on your hand, and press down hard, probably even hotter than without the oil. Get it yet? If not you have no reason to be touching a computer, go back to picking up garbage on the highway or robbing people and smoking crack because you are a lost soul.
 

OpenFriday

Senior Member
Joined
Jan 9, 2001
Location
Hamilton, Ontario
well this is just my opinion but if you have to lap something dosent it show problems with the product? i mean seriously if there was a way to make it 5c cooler for like 2 cents dont you think a quality company would do it
 

BlastMaster

New Member
Joined
Jan 23, 2001
Used a Dremel And polishing compound on my super orb. Yes I do use a lot of fans on my system, but even after 3hours online with quake3 I never top 125F