View Full Version : Whats the Highest "Grit" sandpaper you have ever lapped you HSF with
I finished off with 4000 Grit jewelers sandpaper.
Warlord2
06-18-01, 03:30 PM
2000 grit
I finshed my waterblock with 3000 grit wet dry used dry.
anarki_gr
06-18-01, 03:32 PM
2000 and I think that's more than enough.
I started out with 800 and finished off with 1500 W/D,left a pretty mirrored surface on the hs.Use water with the sp also.
coldmop
06-18-01, 03:48 PM
toothpaste, it's pretty fine
harderclock
06-18-01, 05:24 PM
i have lapped some no name brand hsf with 300 and finished with 600 i didnt spend more than 5 minits doing it and i brought down the temps of a t-bird 900 from 38 idle 31 and under full load it never goes over 48 so it made a big difference and i didnt have to work too hard
cold mop (Jun 18, 2001 03:48 p.m.):
toothpaste, it's pretty fine
Have you actually tryed that or are you just presuming that because it's an abrasive it would work as I have tryed regular toothpaste and it didn't get my heatsink white enough :D
No I'm kidding, I tryed it and it just didn't work as it was really gunky. Smokers tooth paste powder may work which I'll try tomorrow.
Fess_ter
06-18-01, 06:05 PM
I thought it was discussed somewhere that too smooth of a finish degrades heat transfer.
A less than perfect finish /\/\/\/\/\/\/\/\/\/\/\ gives thermal paste a place to flow and allows the points to make contact with the surface to be cooled.
A more perfect smooth finish ========== will float on the thermal paste and have less contact with the surface to be cooled.
I may have read that in one of the articles posted here at Overclockers.com.
jeff_harrison_344
06-18-01, 06:12 PM
Yeah but its only gonna make maybe a 1C difference if that.
Plus when I take my heatsink off most of the Artic Silver has been pushed out and there is only small amounts in very thin patches. No thermal compound offers as much heat transfer abilities as a heatsink so the less you have to use the better which is why I disagree for leaving troffs for it to go into. With the small size of the die on coppermines and durons/tbirds the excess is squished out quite easily and with a bit of wiggling and then running without the fan on for a short period of time which seemed to lower my temps, the get a very good heat transfer.
m_kelder
06-18-01, 06:58 PM
corcus cloth...
corcus cloth... ????? care to elaborate?
Cut him some slack Phil. We all know he meant Crocus cloth. For those of you who are unfamiliar with the term, it is a sheet of stiff cloth with ruby dust bonded to it. As an abrasive, it's useless for removing large amounts of metal, as it clogs too fast, but it is an excellent polishing material. Kind of like dried Brasso on a cloth. Unlike Brasso, the harder the material being polished, the better it does. This stuff will polish glass, chromium, virtually anything. I get mine from the local hardware store for $1.00 a sheet. Some carry it, some don't.
Personally, I feel anything past 1000 is overkill, especially if that 1000 was administered with soapy water as the wetting medium.
Hoot
m_kelder
06-18-01, 08:21 PM
:)
CrystalMethod
06-18-01, 08:30 PM
2000 grit, although I had accsess to a perfectly flat surface on which to do it on, and the measuring tools to verify that the heatsink was still flat to 100 000 of and inch, afterwards. I also used rubbing compound (3M finishing compound) after sanding to further polish it. I might take the HS off and re-lap it, with a heavy grade (1000 grit), to see if it really makes a difference, going past 1000. If and when I do, I'll post my results.
coldmop
06-18-01, 08:32 PM
hey where is my toothpaste response..posted a while ago
yea I use it all the time for stuff. used it yesterday for taking the oxidation off a waterblock. Best is saving scratched CD's with it though
inertia
06-19-01, 05:27 PM
220 -> 320 -> 400 -> 600 -> done. not polished, but definitely flat. No temps yet as I've not turned it on yet.
k_lined4lyf
06-19-01, 05:35 PM
Yup.
Mirror reflection - not necessary, and definitely not the objective.
You just want it flat.
I used 2000 as my final grit when I lapped my both of my HSFs, but I've come to realize it's not really necessary.
K-Lined4lyf
*spazzed*
06-19-01, 07:49 PM
Sony (Jun 18, 2001 02:06 p.m.):
I finished off with 4000 Grit jewelers sandpaper.
What temps did you get?
any increase/decrease in performance?
*spazzed* (Jun 19, 2001 07:49 p.m.):
Sony (Jun 18, 2001 02:06 p.m.):
I finished off with 4000 Grit jewelers sandpaper.
What temps did you get?
any increase/decrease in performance?
It's hard to tell the temperature in my area flucuates so much it not really posible to get an acurrate reading but I say It went down a substancal amount. Right now i'm at 35C all I have running is Winamp and IE ever since I put in my GF3 my temps have gone from 27C idle to 32C idle which I don't think should have happened because I have a exaust blower blowing out the heat right under it 1 PCI slot down leaving one slot inbetween them blank. So I have no clue what happened
Hoot (Jun 18, 2001 07:53 p.m.):
Cut him some slack Phil. We all know he meant Crocus cloth. For those of you who are unfamiliar with the term, it is a sheet of stiff cloth with ruby dust bonded to it. As an abrasive, it's useless for removing large amounts of metal, as it clogs too fast, but it is an excellent polishing material. Kind of like dried Brasso on a cloth. Unlike Brasso, the harder the material being polished, the better it does. This stuff will polish glass, chromium, virtually anything. I get mine from the local hardware store for $1.00 a sheet. Some carry it, some don't.
Personally, I feel anything past 1000 is overkill, especially if that 1000 was administered with soapy water as the wetting medium.
Hoot
No I was asking what it was the smiley face was an accident.
outhouse
06-20-01, 12:11 AM
1500 and soapy water worked my way up with different grades of sandpaper. did not want to use any kind of polish as it may leave residue in the pores allot of polish has wax in it and if touching your CPU or the bottom of your HSF with a your finger can have bad results then a waxy residue may too besides 1500 was more then smooth enough.
FizzledFiend
06-20-01, 04:46 AM
does it do any good to lap the heatsink to the fan i did this myself and gained 3-4 degrees. I asume it was the lapping as I only had they system running for a few hours before reading some more post and articles and added more fans... BTW I read another article by HOOT on his squirl cage wich i mentiond I was think ing of useing in another post of mine.................I always get off subject sheesh!
Blue Jester_2112
06-20-01, 08:25 AM
2000
Ok,for you automotive buffs,how about valve grinding compound? now thats fine! Only thing is im not sure how ez it would be to work with,lol.
2000 grit, but I am coming around to the theory that Fester mentioned. When I get my ThermalRight SK6 all copper, unless it's really poorly finished, I'm gonna try without lapping first. After recording a few days temps I may lap it to 800 and test a few more. Then 2000.
CrystalMethod
06-20-01, 10:32 PM
J.S. (Jun 20, 2001 07:42 p.m.):
Ok,for you automotive buffs,how about valve grinding compound? now thats fine! Only thing is im not sure how ez it would be to work with,lol.
Actually, I was a mechanic till about 3 months ago, and I'd have to advise against using valve grinding compound. It's more abrasive than you think, it's used to grind into hardened steel. Heatsinks are usually made of aluminum, or copper (in the rare instance). The valve grinding compoung will eat into it with the approximate grade of 180 grit sand paper.
Kingslayer
06-21-01, 10:02 AM
I finish with 800. I saw a few reviews that said polising is a bad thing. I tried it on one of my other machines. One was just finished with 800, the other was finished with 800 then polished with a mirror finish with my Dremel™. These were older Global Win heatsinks. The ones they made before the FOP-32. Big black thing with a FOP-32 fan. Can't remember the name of it.
I didn't see any differnce in temps. But it was done on an older AT/ATX motherboard and a Pentium 233 MMX, and back then the on board temp probes werent the greatest. But I put them both in my wife's dually with Celeron 500's and they both showed the same temp. Usually one runs a few degrees hotter than the other because one CPU is right by an exhaust fan. So maybe there is a difference.
You would think that polishing it would be great because you increase surface area by removing all the pits. That's why you put paste on. Who knows. To each his own.
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.