Seksui 5760 Tape

My northbridge heatsink has fallen off and in thinking how to reattach it - the clip on the motherboard is gone - I came up with the idea of sticking it on. This seems attractive to me because there is evidence of it being stuck on originally, as well as held by the spring clips.

I have refined this idea to using thermally conductive adhesive tape (there is evidence of this being used originally), and looking around at likely candidates (I have no idea of what thickness was originally used, except that it appears sub-mm) I found out something about Sekisui 5760, which is sometimes used for attaching heatsinks to RAM and other purposes. Apparently it has excellent adhesion but unlike arctic thermal epoxy is not permanent (somehow...), and also apparently it has good thermal properties...

does anyone know where I can find a comprehensive comprehensible data sheet for this stuff ? (important parameters being thickness in situ, and thermal resistance per unit area, thermal conductivity if not multi-latered etc)

Dumb_Question
28.July.2012
Compaq Presario S5160UK DT261A under XP/SP3
Processor - Celeron 2.7 GHz
Motherboard - MSI MS-6577 v2.1
RAM - 1GB + 512MB (1GB +1GB max) DDR PC2700
PSU - Octigen 300W model 10270PSOTG ('upgraded' from original Bestec 250W PSU [in 2011?])
Nvidia GeForce 6200 graphics card in AGP slot.

 

I would not use that stuff. It is tape - double-sided tape. That means it is thick. Since the most efficient heat transfer occurs with direct metal to metal contact, something thick between the chip and heatsink does not appeal to me.

I recommend you use a product designed for this purpose. Why are you concerned about the bond being permanent? I would use this and leave it.

That said - so far - I have not been unable to remove a glued heatsink. It usually requires heating up the bond, so the machine needs to work. Once warmed up, slight twisting back and forth eventually breaks the bond. Pulling on the heatsink breaks the chip, socket, motherboard, or all three.

You say there is evidence it was glued or had some paste on it before. You MUST ensure the mating surfaces are thoroughly clean of old paste before reattaching the heatsink. This may include careful rubbing with a clean clothe wet with 91 - 93% Isopropyl Alcohol (the ~75% stuff can leave a film).

Be sure to take the necessary ESD precautions.

 

Hi Digerati, and thanks for your comments and advice.

Yes, Sekisui 5760 it's double sided tape, but it's very good (strong bond I mean). How thick is it ? Including the two layers of adhesive, 0.14 mm, in situ, having been compressed to some degree, probably a bit less. Is 140 microns thick ? There seems to some thermal conductivity data around, but I m unable to interpret it ["Thermal Resistance @ 45C Rja (oC/W)" ] The only stuff significantly thinner than this is FCTL from AmecThermasol which claims 0.05 mm, and indeed has 3 x the thermal conductvity per unit area of the rest, in line line with it being a third as thick as the rest. But do I want a 25m reel ? That's why I'm trying to find a sensible datasheet for this Sekisui stuff, but I'll probaly end up buying some Akasa AK-TT12-8 (0.3 mm thick, 0.3 W/(cm^2.K) ). Other users have reported on Seksui 5760 favourably (I have references if you would like them, mostly hearsay from other forums).

I have read about AS Thermal Epoxy (Arctic website) and it warns you that it is permanent. JB Weld is also said to good for this purpose, though I suspect its thermal properties are unknown. I fear doing it wrong, thus I fear making a permanent mistake ! In my experience I make a lot of mistakes. This why I don't want an irreversible solution.

I have 99.9% IPA.

The heatsink was held on by a clip (to m/b) firstly, and also by various sticky pads, I don't know how sticky because they are old and have become dusty as well. There is one around the outside, about 1mm thick), then there a region inside that looks like sticking plaster about 0.2 mm thick I guess, and right in the middle of that there is a white area with a correspondingly shaped discolouration on the top of the chip package, the thickness of I'm not even going to guess - might the same as the pink stuff changed colour by temperature effects ?? I attach a picture, the heatsink is about 2 - 2.5 cm square. The thicknessses don not show on the picture because of the angle of the shot and their small sizes.

Dumb_Question
29.July.2012
Compaq Presario S5160UK DT261A under XP/SP3
Processor - Celeron 2.7 GHz
Motherboard - MSI MS-6577 v2.1
RAM - 1GB + 512MB (1GB +1GB max) DDR PC2700
PSU - Octigen 300W model 10270PSOTG ('upgraded' from original Bestec 250W PSU [in 2011?])
Nvidia GeForce 6200 graphics card in AGP slot.

 

Compared to 0 microns, it is very thick.

Thermal interface materials (TIM) is ONLY needed to fill the microscopic pits and valleys in the mating surfaces to push out any insulating air that might be trapped. Any excess TIM is in the way.

Since it is the responsibility of the case to provide adequate air flow through the case, I am sure that tape is better than no TIM.