I think the trick with the indium is the same with LM. You have to spread it out with rubbing to get it to adhere (wet) to the surface. If you notice when you apply LM you can't just pt a bead in the center and put the HS back on. It won't spread out evenly. I think that's because there is a microscopic film of left over greese unless you cleaned THOROUGHLY. And if you did, then there defintely is a microscopic film of oxides that form on metal surfaces. So when you spread out the LM and rub it along the surface, your scratching through the oxidation and bringing the LM into contact with the metal surface. It needs to be a liquid for that to work. Which is why the Indium foil didn't work well on the IHS/HS interface. The die is silicon, which has a high affinity for oxygen and will form a film too.
Under a powerful electronic microscope the metal surfaces are like mountain rifts, so there is always a bit of air trapped in between the rifts. The silicon of the DIE doesnt have that rifts, his internal structure is like "glass" (or diamond) where the atoms are very well aligned to each other in crystaline structures, so the flatness of the DIE surface is near perfect
The "wetting" term you used fits very well btw, the liquid metal needs to be "rubbed" on top of the surfaces to either push some atoms inside the rifts or/and top remove the air (and break that "surface tension" property of the atoms of the fluid)
Btw, for the lapping there are a couple of tips worthy to be mentioned, first thing is to "scratch" the original heatsink surface with your finger nail... if you feel like several "tac" "tac" "tac" is very bad, ive seen some PS3 heatsinks that had horrible "scratches", in the PS3 slims that scratches are paralell (because the surface was flattened with low quality machines), in some of the PS3 fat heatsinks that scratches are concentrical circles (this is a different machine, bit better, but not so perfect), if your heatsink is one of those yeah... a lapping is going to help a lot with your experiments
Btw another trick to identify the bending on surfaces, is to use a razor blade placed vertically on top of the surface, in front of the bulblight of a lamp... and try to see if there is some light rays that passes the "gap" in between the edge of the razor blade and the surface
Im mentioning this because the IHS uses to have some bending... is the consequence of that metal piece to be "forged" individually... there is a bigger amount of metal in some areas and the atoms have different "tensions" so this results in a small bending when it gets cold, this could be easylly fixed by "mechanizing" the surface but it seems they was not doing it with the IHS's
If your IHS's have an small bending (and you remove it by lapping them) yeah, this is another thing that is going to help too with your experiments
Are minor details, but everything counts, for every small thing maybe the improvement is just -1ºC but is the sum of everything that makes the total... -2ºC here, -1ºC there, another -2ºC... etc...
Btw, have you tryed to melt by heat the old/used indium "flakes" you removed from your previous test ?
Im wondering if it could be used to "wet" the surfaces before applying a new indium foil... you know like when we do a "pre-tin" with a solder iron in a standard solder job
Maybe this could help to achieve that "wet" adhesion to remove the microscopical air bubbles
I think I have a solution to this problem. If I place the HS on my IR preheater set to 80C, it should get hot enough to melt indium (57C). I'm not worried about the heat pipes bursting because they just contain Deionized water under partial vacuum. 80C is surely enough to volitilize it, but the pressure won't be enough to break the seal and It's not hot enough to weaken the copper/brazing. Then I should be able to get the indium to melt and spread it with a cotton swab. Of course lapping the aluminum first would be better.
I was about to mention something about the copper pipees in my previous post, as far i understood (from some schematics, never seen with my own eyes) inside the pippes there is also some kind of "porous" material intended to create a circuit where the gas/fluid moves in 2 different directions
The theory is... are like 2 ducts, concentrical... the extternal duct is used to move the liquid toward the hot spot... in the hot spot the liquid becomes gas... then it moves away (to the extreme of the pipe) in the internal duct
The reason why im mentionin this is because that porous material (i guess is some syntethic plastic, dunno) could be sensitive to the overheats
You know... the gas/fluid doesnt have much problem with the overheats because the pipes are sealed so is not going to escape or degrade (unless the pipe explodes, lol), but that material with the pores could melt or crack (ruining that concept of the 2 ducts)
Btw, another detail that worths to be mentioned from that pipes of the PS3 heatsinks is the glue used to attach them to the heatsink sometimes is bad quality or have "gaps" where it was not applyed enought glue
Check that too, in the PS3 fat heatsinks that glue (or better said, thermal glue) is like candy and can be "cracked" with your finger nail... it have the consistency (and the color) like melted sugar, personally i dont like it at all, lets say... is too much "industrial" (a.k.a. cheap)
If i was you, i would try to see how many of that crappy thermal glue can be removed... and i would use a decent thermal glue to replace it and to fill all the "gaps" in between the pipes and the heatsink... you know, the goal is to maximize the area of contact in between the copper pipe and the alluminun of the heatsink
About the graphite/LM combo. I minored in chemistry so I know that graphite is just carbon-carbon bonds. It's very hard to form new bonds with carbon. So I'm not worried about LM/Indium reacting with it at all. But it does have microscopic pores filled with air (terrible heat conductor). These pores "might" allow the LM to seep in and displace some of the air, greatly improving the thermal pads performance. On the other hand they could be so small that they act like a barrier to the LM. LM has high cohesison to other LM molecules and if there is no cohesion to the carbon, then it will not wick into the pores (capillary action), but instead be repelled (surface tension). So it's pretty iffy. Again, an untested hypothesis.
I'm curious to try it, but these are expensive materials (graphite thermal pads, LM, Indium foil). 1 failed test could be $50 in the garbage! I think the the idea of melting it on and spreading makes more sense than adding a layer of graphite to the mix. So I want to try and get the Indium foil to work. It's the better solution overall. It's just a matter of getting it to adhere properly to the entire mating surface.
Im not sure, but i use to think in this considering how it would look under a very powerful microscope and also the posible chemical reactions that could happen (somethig that usually nobody knows, because we dont have a laboratory to do tenths of weird tests)
The graphite pads are a nice idea, the performance is not impressive but are failproof for long term, mixing them together with liquid metal doesnt makes me much confident, maybe there is some way to "force" the liquid metal to enter in the pores (under pressure ?) to "wet" it properly but dunno... i guess someone else tryed this kind of experiments before... maybe google have some answers for this combo
I would use the graphite pad for the EE/GS, maybe is not needed but is located near CELL/RSX so there is some heat generated by CELL/RSX that is transfered to the EE/GS ---> to the metal shield with the graphite pad
