Submersion : Thinktank Revision 2.0

B

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,211
It's probably time I started this thread. :D

I am about 18 months into the evolution of the Thinktank


-----

I am designing the thinktank 2.0 to be a submersion rig , with pure alcohol (anhydrous isopropanol ) as the liquid.

This IPA will be pumped from a large vacuum insulated container of my own design (yes it's a keg).The active coolant will be dry ice . The entire system has been designed and built with a spec of -80 C in mind.

I have rebuilt the thinktank 'tank'.I have taken the original glass tank , and built and inner glass tank inside it , using Borofloat 33 siliconed together using RTV 157 aerospace grade silicone...The two tanks are designed to be separated by an active nitrogen blaket .
Here's a shot of the new tank :

tt2_112.jpg


---

The original lian-li pc65 has become part of something bigger :

tt2_48.jpg


Here's a shot of the back with the vents :

tt2_68.jpg


-----

So.....

I have every major component I need to build this.....I am doing it now.

The first rig that will go into it :

Asus p5k-se
Intel e6600
Mushkin D9 ram , 2x 512Mb
Video card yet-to-be-determined.

The psu will be external....

-----

Cooling systems specs :

The IPA will be pumped using this pump :

tt2_113.jpg


That's a 316 stainless steel ,self priming pump , 3/4 inch I/O....

The coolant lines are all "Vacuum Insulated".

The pump will be housed in a positive pressure N2 system......

I expect the worst temps I will get will be below -60 C .

------

I'm not far off the first test runs......I actually have every major part in my possesion...just gotta put it together.

Next pics: the rig running.

cheers !

:D
 
I hope that you are not gonna be submersing electrolitic capacitors. Correct me if I am wrong but they don't like this level of cold.
 
Yes in fact I am.

But only the new polymer capacitors.No 'old ones'....
also , I'm making the battery external as well.

;)
 
I had this moved here (thanks FLECOM) , so I can add more pics etc...

I'll add some later today.

:D
 
Suscribed! This is going to be so awesome when you get it done. Hope it lives up to your expectations. :)
 
So how "sustainable" will this thing be while running? The first thing I think of when I hear someone mention dry ice cooling is having to constantly add more dry ice, which would get real old after about...two fill ups.
 
Very ambitious! I can't wait for the outcome! Good luck on the World Record!
 
:D

----------
Harb said:
So how "sustainable" will this thing be while running? The first thing I think of when I hear someone mention dry ice cooling is having to constantly add more dry ice, which would get real old after about...two fill ups.
Click to expand...

It's not designed to be sustainable.

Each 'run' will consume at least one "K" type nitrogen tank , and probably at least 50 pounds of DI , so that's a couple hundred bux a run ( a run being 4-8 hours...)

---------

I'll take this time to explain why Isopropanol ,and, it's 'safe' use.

Cost , Operating temp and materials compatibility.

Out of the fluids capable of operating at -60c and colder , (fluorinert , silicone oils etc) IPA is the cheapest , and as a bonus , Isopropanol is compatable with 99% of all electronic components.

After I did the thinktank 1 , I discovered the biggest pain of an oil based submersion system is cleaning the stuff off components .After cleaning that rig in boiling IPA about 4 times , It donned on me , why not just use Isopropanol....

This does however lead to other issues , extreme flammability being one ...

90% of the work on this rig is a result of either a) providing a coolant loop designed for dry ice/-80c ( I.e. insulation etc ) and b) safety revolving around the use of a flammable coolant.

One of the main safety items ,is a positive pressure nitrogen system.This is there for 2 reasons.

1) To provide inert 'fireproof' gas zones where required.This feature is strictly to save money.Technically , this rig should be build with 'explosionproof' motors , switches/computers/sensors etc......However , I can save a lot of cash by just putting these components in a positive nitrogen atmosphere.

2) To provide an inert (read;fireproof again...) gas for the evaporation component of the cooling system.This rig won't just be about dry ice and alcohol...I intend to try and evaporate the alcohol from both tanks as fast as I can.This will increase cooling.....(Michael Farady used a similar method to achieve -110C)

----
Between the co2 output of the dry ice , and the n2 output of the fire safety system , there is another safety hazard : suffocation from displaced o2.As a result of this , the offgases are going to be fully vented with solvent recovery.

----------------

:D
 
The original Thinktank inspired me to get into water and dice cooling, and projects in general that are absurdly over the top and beyond what most sane people will ever attempt. 4 years since you posted the original's awesomeness, and now we get rev 2.0 to drool over too.

I know you have a few surprises up your sleeve, but how exactly are you cooling the iso with dry ice? Evaporation inside a container cooled on the outside, or evaporation plus dry ice submersion? If it matters, isopropanol that's been filled with CO2 from dice is a lot more difficult to ignite than straight iso, and does build up a decent of pressure. Stay careful and keep us updated :D
 
The big keg gets filled with 30 litres of Iso , and 35 pounds of dry ice.

:D

-----

Here's a labled photo and some more info ,and a few more pics..

tt2_48b.jpg



Here is the breakdown ;

A : This is the thinktank proper.It's the original thinktank with a new 'inner' tank'

-.The space between the two glass tanks will be filled with nitrogen.

- The motherboard (p5k-se,Intel e6600,mushkin d9gmh ((2x512 Gb)),video yet-to-be-determined) goes into this glass tank.The whole works will be fully submerged : The glass tank is full of Isopropanol when running.

- The upper section of the lian-li pc-65 is where the PSU,Single WD raptor HDD and Matrix Orbital GX Typhoon LCD will be housed.

-There will also be both N2 input lines , and vent lines , but I havn't made em yet so......

B : This is the coolant resevoir.It's the keg/stainless pipe assembly shown above.
-This is just a fancy insulated container.It's just like a 'thermos'.Vacuum is the best Insulator there is.Period.....so you put 1 container inside another , and establish a vacuum between the two....


C : Related to 'B' these are just the lines that circulate the iso from the res to the computer.They are vacuum insulated lines.

D : This is the solvent recovery tank.This is actually a 'vert' keg from WildRose breweries here in Calgary.
-When I get the vent lines made , they feed into this first.It gets filled with dry ice as well , to condense the iso out of the vented air/gas.Waste not , want not.

E: This is the monitoring computer.
-This rig runs the MO MX-222 Display ( G ) which will take care of all temperature monitoring above -50 C.......

-My USB Instruments Digital 'scope will also be running from this rig.....

F: Vacuum manifold.Basic valve control of each vacuum bit ; tank and pipes.

H: future home of the n2 manifold.

-----

Here's a shot of the tank top , vent lines tacked in.....

tt2_88.jpg


---

The coolant lines connect via 7/8" flare unions....The males will attack to the tank back :

tt2_106.jpg


---

Here's a couple of shots showing the n2 lines installed during tank construction..

tt2_95.jpg


tt2_110.jpg


-------

:D
 
I discovered the biggest pain of an oil based submersion system is cleaning the stuff off components
Click to expand...
Tell me about it, still trying to get the oil off my 7950 GX2s.

Good luck with this project, I'll definitely be watching for updates.
 
The very best part about this is the Wild Rose vert ;)

This is an incredible project: Ambitious, experimental, and potentially explodey!

Next time you're in the brewery, your first pint's on me mate.
 
Here's a nice little tip for cleaning the oil off your components in a few seconds flat:
3M Novec Electronic Degreaser

We use it at work and you can literally spray it on electronic motherboards and components, and the oil just runs right off. Then it dries clean with no residue left over.
 
heh.

Definitely a good product.I've used it.

Not *quite* as good as boiling IPA tho....although if I need really clean ,a fast wash in hot Toluene+MEK (laquer thinner) does the miracles...

:D
 
w1retap said:
Here's a nice little tip for cleaning the oil off your components in a few seconds flat:
3M Novec Electronic Degreaser

We use it at work and you can literally spray it on electronic motherboards and components, and the oil just runs right off. Then it dries clean with no residue left over.
Click to expand...


You're just chock full of informaiton, aren't you. Are you going for your EE? no one knows this shit!
 
EE/CE.. senior year at Umich. It's the gayness.

But the degreaser does work wonders, at least the times I've used it. Saying it was such a pain to get the components cleaned, I wasn't sure anyone here has ever used it, else there would be no problems.
 
It's an awesome degreaser for-sure....I needed something I could do 'baths' with (mobo size) ...but for small stuff I have used it.

Technically , something like an 'Ether' would be best , but I havnt tried that kinda solvent on a mobo before , might melt the crap outta it...

:D
 
BrainEater said:
heh.

Definitely a good product.I've used it.

Not *quite* as good as boiling IPA tho....although if I need really clean ,a fast wash in hot Toluene+MEK (laquer thinner) does the miracles...

:D
Click to expand...

Off topic and nogaf but:
I thought Methyl-Ethyl-Ketone had been banned? Or maybe the EPA just made it very restricted to use? :confused: Either way that will actually take powder coat off of parts as well LOL
 
I will also say, for your solvent trap good choice on using dry ice. In my lab I use LN2 for my solvent trap, on my schlenk line, but the solvents are always under a pretty anaerobic atmosphere. Still LN2 is cold enough to condense out any residual oxygen (resulting in liquid oxygen)and you can generate explosive situations by having LO2 and solid or liquid solvent left there. I always keep heat away when I am venting my solvent trap, it is under vacuum (that's why it is a solvent trap, it traps solvent before it reaches my vacuum pump and destroys the pump oil or worse).
 
Sorry everyone ,I'm on it....

Host's fine , but there is some issue ,I'm workin on it
:D
 
lazz said:
The very best part about this is the Wild Rose vert ;)

This is an incredible project: Ambitious, experimental, and potentially explodey!

Next time you're in the brewery, your first pint's on me mate.
Click to expand...

:D

Cheers !

For another pint , we'll re-label the big keg. :p

------
It'll probably be a week or so before I get anymore updates up , just finishing up some work type "work".....

I did order my lighting......3x Luxeon 45 Lumen 'cyan' leds....best 'cold' color I could get.

:D
 
You must log in or register to reply here.
Back
Top