S939 A64 IHS Removal

Jokester · 1 Dec 2005 at 19:28

I got about 100MHz on phase, so it should be of a similar improvement I would have thought.

Jokester

Jokester · 3 Dec 2005 at 14:02

Looking at the photos of it's base I'm not convinced it will fit.

Jokester

Jokester · 3 Dec 2005 at 14:13

Though the bit in contact with the core is small it's got stupid "wing" bits that curve out so that at the edge I would think they would hit the back of the socket. They might miss it, but then again they might hit them.

Jokester

Jokester · 6 Dec 2005 at 18:07

You do need to apply a fair bit of force, the nearest I can come to describe it is if you were cutting a thick, hard bit of cheddar. Because of the force you need to be careful with using the knife obviously and also from damaging the pins or internal components if the knife slipped.

Jokester

Jokester · 8 Dec 2005 at 07:39

If the block covers any portion of the rear of the socket you might not be able to get full contact with the core as the block can hit the socket.

Jokester

Jokester · 17 Dec 2005 at 09:53

MikeTimbers said:
I will do mine but I want to get the before overclock and temperatures first. Has anyone else got any hard data to share on the difference it made on both overclock and temperature? Also for those who have done this, how was the quality of the thermal bond between the die and the IHS?

I found on phase, with an X2 that temps dropped about 10C and got a coresponding increase in stable speed by about 100MHz. All the tops I've taken off look like they had decent 100% contact with the core.

Jokester

Jokester · 20 Dec 2005 at 18:39

NEW TIP!

Spray your blade with IPA to lubricate it, makes a lot of difference.

Jokester

Jokester · 20 Dec 2005 at 23:30

Yep, isopropyl alcohol.

Jokester

Jokester · 23 Dec 2005 at 14:39

Success then?

Certainly sounded like poor contact but there was nothing obvious when I took the top off.

Jokester

Jokester · 2 Jan 2006 at 14:20

From the photo I think you're problem is pin melt to be honest.

Jokester

Jokester · 2 Jan 2006 at 14:54

Sorry, I thought you weren't serious.

If you're using a Vapochill you must increase the reach of the head to ensure that it makes full contact with the core otherwise it will overheat quickly.

Also, what on earth have you coated the surface mount components with? It also looks as if you've also burnt the PCB with something?

Jokester

Jokester · 2 Jan 2006 at 15:32

Yeah I would clean it off and try again.

Also refer to post 41 as to what to do to get full contact with a Vapochill (assuming you're using a clamshell).

Jokester

Jokester · 2 Jan 2006 at 17:21

Was that grease to protect from condensation?

Jokester

Jokester · 6 Jan 2006 at 12:18

A quick check of my DFI Lanparty UT nF4 SLI-D shows that the socket height is less than that of the core so even if the HS overhangs it will make good contact.

Jokester

Jokester · 8 Jan 2006 at 13:25

Yeah swapped over to the Big Typhoon, seems to cope a lot better with the extra heat load that the X2 pumps out.

Jokester

Jokester · 8 Jan 2006 at 18:03

Mr Mister said:
jokester im getting a mach 2 gt for my ihsless 4000, do i need to put that black stuff over the small resistors like on your fx57? What is it and what is it for?

Thanks

The black stuff is Liquid Electrical tape, I'm not totally convinced if it's needed, as I never had problem with condensation forming on the top of the surface components.

Di-electric grease would also do the same job.

Jokester

Jokester · 11 Jan 2006 at 19:45

Yeah I don't think that'll work with current generation P4s, they're in effect glued on and the only way to remove them is to use precision machine tools by all accounts and it's extremely risky.

Jokester

Jokester · 13 Jan 2006 at 18:07

Yeah, you should be safe enough without the pads but just to check how does the Zalman 9500 heatsink mount?

Does it just screw down vertically?

By the way what sort of string did you end up using?

Jokester

Jokester · 13 Jan 2006 at 18:28

Could well be, an alternative to supporting the HS is to use string or rubber and to tie it to your case above and try to balance it with the weight.

The other thing to check though is that the HS is making good, even contact with decent pressure on the core.

Jokester

Jokester · 13 Jan 2006 at 18:42

The core should leave a clear impression on the base of the HS, with very little TIM where the contact is and a rim round that where excess has been squeezed out the sides.

Jokester