topless Northwood. NICE!

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
well, tonight i took the leap and took the heatspreader off my 1.6a. i had previously lapped the heatspreader to a mirror sheen, and i didn't know what to expect. well, i took it off by slowly slicing into the glue between the spreader and the chip using an exacto knife. 3 sides were easy, but one was a real pain in the ass. eventually (and after a lot of sweat, that chip is tiny!) i got the spreader off.

i had to take a set of washers off my alpha, since the die of the chip is a lot lower on the socket than the heatspreader, but that is no big deal. a little bit of AS3 on the core, and voila. how does a 7 degree reduction of idle temps sound? i haven't tested it for max temps, but i'll run Prime95 over night to give it a work out. it was a stressful operation, but worth it. and the best thing is, the heatspreader already has a hole in it for a keychain ;). cheers!
 

paco83

Member
Nov 4, 2002
170
0
0
Intriguing. So the fact that the heatspreader makes more contact with the heatsink then the chip doesn't make up for the fact that the heatspreader is shi**y at conducting heat.

Kudos on the guts to take that sucker off. :cool:
 

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
well, i damaged the chip. it won't stably go through more than a couple minutes of prime95 at 2.4 gHz even at 1.75 volts. it runs fine at 2.13, at 1.55 volts, though. oh well. 2.133 will be fast enough until i can get a new chip. not like i'll really notice anyway. you pays your money and you takes your chances. i have a hope that this problem is being caused by some AS3 that is smeared on the chip around the die (there are traces on the chip top). i bet that i damaged a trace, but i'll pull the chip out and clean it tomorrow. however, even if it will only run at 2.133, it is almost worth the reduction in speed for the 10 degree+ lower temps.
 

paco83

Member
Nov 4, 2002
170
0
0
Oh well in that case the fact that your running slower then before probally has quite a bit to do with the reduction in temps. The reduction from 2.4ghz
-> 2.13 would result in about 12% less power to dissipate by the hs&f. This accounts for most of the drop in temp, the rest is probally from the AS3 or sumtin.
 

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
i can't post a pic because i don't have a digital camera. however, i'll try to borrow one.

i cleaned off the chip and applied some 2 year old radio shack paste, and what do ya know? so far so good. 2.4 gHz/1.65 volts (i can live with that), and it has survived 10 minutes of prime95 (it didn't go through more than 10 seconds, before), with a max temp of 44*C. before i took the heatspreader off, the temps hit 56*C sometimes during Prime95, at 2.4 gHz/1.55 volts. assuming this is stable, i would consider that a major difference in temp. the little drop from 2.4 to 2.133 before definitely didn't explain the temp diff. it was over 10 degrees. i'm not a superstitious person, but i'm crossing my fingers. :)
 

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
45 minutes of Prime 95, load temp of 45*C. looks to me like cleaning the chip worked. w00t.
 

Blain

Lifer
Oct 9, 1999
23,643
3
81
I popped the heatspreader of an old school Celeron before. It's not a pretty sight. I ended up epoxying a HS onto it... but it was never the same. :Q
 

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
hopefully i'll be able to borrow a digi cam in the next couple days. you'll get your damn pictures! :)
 

Lizardman

Golden Member
Jul 23, 2001
1,990
0
0
sounds tempting, but my better judgement tells me not to do it since my temps are stable and I would be royally fuc%ed if I broke my computer during school.
 

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
well, i think i can safely say that the chip is perfectly stable at 2.4 gHz/1.65 volts. low temps, too. all in all, i'd say that this was a success. would i do it again? probably not. :)
 

EdipisReks

Platinum Member
Sep 30, 2000
2,722
0
0
it looks exactly like that, though most of the adhesive came off with the heatspreader. i think that 9 months of use baked the thermal adhesive on my CPU, which made taking the spreader off harder and messier.

[edit: well, it didn't have that big crack in it ;)]
 

RalfHutter

Diamond Member
Dec 29, 2000
3,202
0
76
If people want to see pictures of a P4 w/o the IHS go over to OCforums. There's a thread in the "Intel CPUs" forum titled something like "So I removed the IHS from my P4" and it's got a dozen pictures of an IHS-less P4. Lots of good pics of the die as well as the IHS.