'Stepper' Evap design

**Gray Mole** · 09-14-2005, 12:41 PM

Rethought the design a bit, sizing a little better and a lot cheaper. One less plate, and the tube for inlet is larger.

I thought about what Soddem was saying about the liquid getting a little bit of a head start in evaporation before it hit the base, and whether using a 1/8" with cpev, or a .031" captube, just braze it close to the top of the inlet tube, and it should act like an 'accelerator' in a way, the expansion should speed up the refrigerant and possibly lead to a greater percentage actually evaporating against the base. the top is pretty thick and I'm not sure if I'll keep it that thick, but should add to the overall mass and control load well enough if it does stay that way.

inlet tubes not to scale(just a paint sketch, nothing is to scale) but that's the final idea. Will get a prototype pretty soon.

I think that's different enough I feel ok about it. As soon as it comes in I should have a mach1 or 2 here to test the design on first and see how it works both sideways and downward.

We'll see if it works, and how well soon I hope.

Gray

**jinu117** · 09-14-2005, 01:46 PM

I wonder how the ceter of big mass will behave with heatload as it is sitting on top of hottest part of heatspreader actually. Hmmm... Tell us how it works

**hatemi** · 09-14-2005, 06:34 PM

Just one tip: Put the suction in the midle of the evap. That way it will be much easier to insulate and mount.

**jinu117** · 09-16-2005, 01:31 PM

Few questions and ideas.
1) You want to have all liquid changing into vapo if possible. Wouldn't it be better to vary the size of space in steps to progressively get larger from liquid to suction portion to give it even less pressure each steps?
2) What would be real advantage of having suction at top and cap tube at bottom vs cap tube at top and suction at bottom from above? Getting curious on this. I thought for some reason having liquid coming through top and as it goes down evaporating, the place most of evaporation happens would be most heat absorbed? Probably close to cpu die is what I thought where temperature difference is greater...

Thanks in advance

**Gray Mole** · 09-16-2005, 01:47 PM

I'm working on a different configuration of the base chambers. I'm trying to get back to a conventional inlet, and using a portion of the base for pre-evaporation.

Instead of thinking in terms of 'as much liquid as possible on the base' think 'as efficient as possible evaporation at the base'

If you can get the liquid into a 'prepared' state of evaporation, you can locate the most efficient level of evaporation at the base.

When the terminator in T2 got covered in LN2(or whatever that was) he didn't freeze right away right? That's because it wasn't evaporating yet. I realize it's a movie, but the principle is entirely correct. The process of evaporation is where the latent heat is taken, and that's the key. It's why I wanted a 'prechamber' for initial bubbling to take place, so that the refrigerant is somewhat prepared when it actually contacts the base.

Watch out for the next revision, little different...nonsymetrical...but I think an improvement in this principle. Hope it's cool, I'm almost finished the drawings.

Biggest trick is guessing the internal volume to try to attain about the 15-20% preevaporative volume, though I think this one will negate too many ill effects of that if it's a little more than that.

Gray

**Gray Mole** · 09-16-2005, 02:59 PM

Changed teh base configuration in this, though without a 3D model, it's harder to visualise.

The base chambers are cut from each side, leaving a square block in the middle with a number of 1/8" holes drilled through for refrigerant to travel through, and hopefully 'tuning' the evaporation rate to cause consistent evaporation across the center of the base. The base could actually be thinner in this design, as the lower central block is solid except for the holes. Could be more like 5mm, and still maintain good load capability due to the central mass.

Just don't know how easy it is to cut swathes from the side of a cylinder. The pre-evap chamber is still at the base, so if it's too large, it still acts on the base and central mass.

Imagehosting is down, so they're attached.

Gray