CryoStar Sli and CPU evaporators by Gosmeyer

[runmc]

Eric Gosmeyer and I were talking about Single Stage coolers via emails and one thing led to another.
After sharing a few emails we went to phone conversations. The conversations soon turned to evaporators and Eric was wanting to know what I thought would be a good evaporator. Immediately I thought of Eleven's design.
I sent PMs to Unseen, esdee, Gray Mole, Jinu, hipro and even Eleven himself asking all kinds of questions about size and design. Everyone was awesome and gave us enough information to get us headed in the right direction.
Between Eric and myself and Gray we added different designs and ideas.

The CryoStar Maze design is what we came up with. Click on CryoStar to see more pictures.

Just so happens Eric works for a company that makes jewelery molds. He has access to all the machine needed to do the job. Eric is also a pro in design and programing. I was very fortunate to meet him.

Here are some pictures. If you have any questions post them here and Eric will answer them if he can. Do not ask about - sales or pricing on these open forum. Discussion of services, pricing and sales on the open phase forums is not allowed and is coming to an end.

[n00b 0f l337]

Perfect, refrigerant evaporation close to the base where its best, and a good balance of surface area. Maybe get it sand blasted?

[runmc]

Perfect, refrigerant evaporation close to the base where its best, and a good balance of surface area. Maybe get it sand blasted?

No problem NOL - Eric has all that covered Nothing will be overlooked

[Stapler]

Looks great, have you actually tested on out yet? Reminds me of Eleven's evap, but looks like it should have more surfacce area.

[Supertim0r]

wow, that's a small evap

[EnJoY]

That is very sexy. Nice work.

[TheBender]

Wow that looks like it would even work well for tight places like on a video card. Now where do I get one

[runmc]

Looks great, have you actually tested on out yet? Reminds me of Eleven's evap, but looks like it should have more surfacce area.

Eric just finished this one today - not tested yet. As you can see in the first post, this evap was patterned with the Eleven evap in mind.

[Duh]

looks very promising

give us some numbers ASAP

hope it performs as nice as it looks

J

[stuntingalone]

Give us some results!

[Gray Mole]

Turned out great Ron!

Hopefully all of the theory of what makes a good evap translates into good results too.

I see you were able to get the channels to both come into the center that should make your mounting possiblities much easier to deal with.

I also like the fact that you've given Eleven credit where it's due

Your friend has done a great job, and if it's even 1/2 as good as I think it will be you should have some really nice performance

Cheers

Gray

[sacha35]

Good job, looks very good, can't wait to see some results.

[Brettbeck]

Great job! The small circle cut-outs will definitely help the surface area.

I assume you used a CNC milling machine? If I get on a course i'm looking at, in september I will have access to quite a lot of those milling machines .

[LittleDevil]

Really nice evap


regards

[Unseen]

Nice job ron

Take a look some pics. I hope they will help you with the enclosure.

[runmc]

All credit goes to Eleven - Unseen - esdee - Jinu - Gray Mole for design. Eric Gosmeyer is the master behind programing and machining. Eric has done a great job. Hopefully he will chime into this thread and give you some specifics on the evap itself.

Our goal is to have an evaporator that will hold the load of the Quads and cores to come. It won't get any colder. Just suck the heat up and keep the multi-cores cooler

The evaporator will be tested in the next week or two. Due to the background of this design, we think it's going to do well.

BTW N.O.L. - the evaporator was blasted and then the surface was machined for a smooth fit between bottom and cap.

[esdee]

wow runmc! it turned out great!

[gosmeyer]

This ships to Ron tomorrow a.m. Hopefully we will be able to see some good numbers.
There were a lot of people who contributed to this and I personally want to thank them all.
The only thing that I can take credit for is adding the radius cut outs on the I.D. and O.D of the slots,
as well as machining it.
I made 4 models in solidworks before we settled on this one. Then it was time to make some chips.
40,000 rpm spindle speed at 40 i.p.m. .010 depth of cut per pass.
The part you see was the first part that ran, program had about 47,500 lines. (I’m lucky, I can create the solid model-
As well as drawings-create the tool paths –and machine the part , don’t have to rely on anyone)
Slots are same width as walls 2.37mm/.0937”
Depth of slots are 7.9mm/.312”
Gosmeyer on cap is etched with laser .
Next cap will have counter bore for suc line stop. I just forgot it on the first one.

[Unknown_road]

although I much prefer maze evaps over spirals, I still have a comment. when the suction is placed directly above the core and the cap tube at the side it's possible that the refrigerant above the core is already superheated especially under full load. If there is liquid refrigerant above the core it would very likely be overcharged.

[runmc]

although I much prefer maze evaps over spirals, I still have a comment. when the suction is placed directly above the core and the cap tube at the side it's possible that the refrigerant above the core is already superheated especially under full load. If there is liquid refrigerant above the core it would very likely be overcharged.

Sorry - I don't understand what your saying. Please explain and give a solution to the problem if there is one.

[Unknown_road]

because the suction is in the center of the evap, the refrigerant might have boiled off when it reaches the center of the evap. If it hasn't boiled off at the center much of the liquid refrigerant will enter the suction which suggests the system is overcharged.

If the liquid refrigerant has boiled off before the center it can't extract heat anymore so the center won't be cooled actively. Because of this the heat has to travel a greater distance. longer distance means higher core temp.

[[XC] gomeler]

I'd think with proper tuning it'll be possible to balance the liquid refrigerant flow within the block so that the maximium amount of heat can be pulled off of the evap. Can't wait to see how this thing performs, will this be an experimental design or will we potentially see a commercial run of these?

[gosmeyer]

Guess that's why they need to be tuned and see what the numbers are.
I can say that the wall thickness between the captube inlet and suc line
is .76 mm/ .03" so more or less the captube is at the core.

[Unknown_road]

that doesn't matter much gosmeyer, the most important surface area of the evap should be flooded at all time imo.

[gosmeyer]

so are you saying captub in I.D.
suc from O.D.