Direct Die cooling

[Farabomb]

Seems that there are others besides our own "Direct Die Dicki" that don't mind taking watercooling one step further.

Take a look at what Doctor Ice (http://www.3dnet.hr/dr-ice11-en-01.html) has cooked up in the lab.

Some nice work there. :)

[Brinky]

Pretty cool... i wonder if you used like a silver HS on top of the core... with small fins... maybe that would increse performance...
like a tiny HS and still use the same method... but serusly that kind of cooling is scary, leakwize...

[tripodal]

just the front page image is eneough to give me chills

[tripodal]

i bet silver would do something positive if you could get it mounted really well... hard to say for sure tho, would be more like a water cooled hs LOL

he didnt seem to get the problem dicki had.. with water going through the die

[Brinky]

I mean the same concept... like this..
maybe try it out... see if it performs any better

[tripodal]

Thats not a bad idea, except that you will lose a LOT of what you are trying to gain by using the adnesive. The best way would be to mount it onto the core w/o someting between

[Brinky]

yeah i get the whole point of the cooling technic... I just think that a flat surfice conducts less heat then an non flat because there is more friction...

[prokaryote]

Hi people, this is my first post here. Yipee!

Okay this is just some observations from an ex semiconductor package applications engineer (I think we may have discussed before Dicki on another site, I don't remember). Take them for what you will.

1. The epoxy used to fill the gaps between the die, the solderballs and the organic substrate is probably not waterproof. Many epoxies used in the semiconductor business aren't.

2. The soldermask may not be waterproof and the organic substrate itself is not waterproof.

3. Residual flux may react with moisture to form acidic compounds.

4. Moisture and heat cycling can lead to delamination of the organic substrate that the die is attached to. Though it usually requires higher temps around and above 70C.

Now, if you were to substitute the water with say the goo that was used in the famous experiment and article, "goo cooling for fun", you many be able to avoid those issues. Just got to find a suitable pump.

Okay, I'll go now.

prok

[tripodal]

Smart A$$, but its welcome here.

[evilmatt]

another advantage of using goo or goo compatible substances rather than water is they are liquid down to much lower temperatures. The stuff we used was ok down to -50 deg c.

Course its a lot thicker than water and gets to the consistency of treacle at those temps. Even at room temp the pump had some issues with shifting the stuff.

but direct die goo cooling sounds like a plan ;)

[lechumbl]

Hi prokaryote,

:D :D WELCOME :D :D to Liquid Ninjas.
Glad you came "from the darkside" to visit.
Feel free to stay and grace us with your knowledge.

Take care, and Welcome again........

[Farabomb]

:beerm8s: Welcome prokaryote :beerm8s:

Glad to see you here. :thumbsup:

Add your $.02 anywhere you want :)

[tripodal]

perhaps someone will adapt a typical automotive airpowerd fluidpump to this application.. tho the pressures usually excede 100psi then