Zero Condensation Sub Zero Cooling Concept.

[Aquados]

The air in a given volume of space has approximately 80/ Nitrogen, 20/Oxygen  and Other.   Imagine the Pc rig sitting in a 100% Nitrogen environment with the Sub-zero cooling hardware outside of it.

A box that is big enough for the Pc, will have 3 holes.  2 will be input and out valves for the gas. The third will be cable and tubing.  The C&T hole will be filled with silicon for air tight seal.  Inject the Nitrogen gas with the output valve open until all of the non-nitrogen gas are pushed out. Close valves without building pressure.  There should be no moister in the box. 

Nitrogen is lighter then oxygen and water. Output would be on the bottom and input on top.    LTT has a cnc, laser machine, and maybe a 3D printer. They got the resources to make it.  

I think it is time they get an engineer or a person who loves and know how to build stuff while using tools correctly.

 

Imagine a world with sub-zero cooling with no threat of moister. 

[Enderman]

Or just hook up the sealed PC case to a dehumidifier to remove moisture.

[Jurrunio]

just cover the motherboard with paper towel to suck the moisture

[Amazonsucks]

IBM has been doing refrigerant cooled systems for decades. They are large so the cold plste is just housed in an air tight enclosure with a rubber gasket to prevent moisture, but there are silica dessicant packs inside just in case moisture gets in.

 

Might like thus article.

 

https://www.electronics-cooling.com/2011/03/reasons-to-use-two-phase-refrigerant-cooling/

[orbitalbuzzsaw]

4 minutes ago, Jurrunio said:

just cover the motherboard with paper towel to suck the moisture

z u c c

[W-L]

21 minutes ago, Aquados said:

-SNIP-

That'a very doable as long as you can feed in coolant lines, there was a user that did something like this on the forums by displacing the air with argon to prevent condensation of any sort. I recall something about a yearly top up to ensure it was completely filled up. 

[Aquados]

14 minutes ago, Amazonsucks said:

IBM has been doing refrigerant cooled systems for decades. They are large so the cold plste is just housed in an air tight enclosure with a rubber gasket to prevent moisture, but there are silica dessicant packs inside just in case moisture gets in.

That is industrial use.  I am talking about at home.

 

19 minutes ago, Enderman said:

Or just hook up the sealed PC case to a dehumidifier to remove moisture.

The dehumidifier needs to suck air in. It would need an inlet.    The concept is for 100% no moister or anything else.  Sub-zero hardware plus a dehumidifier  would make it more complicated and less efficient.  

 

5 minutes ago, W-L said:

That'a very doable as long as you can feed in coolant lines, there was a user that did something like this on the forums by displacing the air with argon to prevent condensation of any sort. I recall something about a yearly top up to ensure it was completely topped up. 

What is the price for argon and nitrogen. LTT got an CNC machine. They should attempt.

 

18 minutes ago, Jurrunio said:

just cover the motherboard with paper towel to suck the moisture

the idea is to eliminate moister from the environment. 

[W-L]

Just now, Aquados said:

What is the price for argon and nitrogen. LTT got an CNC machine. They should attempt.

Just straight Argon will do, and it's relatively cheap any welding supply house can provide you a tank or fill one if you bring an approved vessel in. 

[Enderman]

46 minutes ago, Aquados said:

The dehumidifier needs to suck air in. It would need an inlet.    The concept is for 100% no moister or anything else.  Sub-zero hardware plus a dehumidifier  would make it more complicated and less efficient.  

Put a dehumidifier outside the PC.

It takes air in from the case and outputs the dehumidified air back into the case.

Once the humidity is out it can be shut off and disconnected.

[manikyath]

or you make a mineral oil submerged PC that's cooled by an AC condenser, rather than just radiators, and keep it close to freezing point, but still far enough above the oil's freezing point.

 

[Amazonsucks]

11 hours ago, Aquados said:

That is industrial use.  I am talking about at home.

 

The dehumidifier needs to suck air in. It would need an inlet.    The concept is for 100% no moister or anything else.  Sub-zero hardware plus a dehumidifier  would make it more complicated and less efficient.  

 

What is the price for argon and nitrogen. LTT got an CNC machine. They should attempt.

 

the idea is to eliminate moister from the environment. 

I know its in a datacenter and not home but the same basic principles apply. You wouldnt want to do a PC in a sealed box with nitrogen inside, unless the nitrogen itself was also being circulated. The rest of the components that get hot would need to have the atmosphere in the case do the cooling like any other case, even if the CPU and GPU had refrigerant cold plates on them.

 

In refrigerant or air cooled servers, there still has to be airflow throughout the enclosure so that VRMs, PSUs, and all the orther stuff that gets hot inside a computer can be cooled in addition to the chips themselves.

 

The way im envisioning your setup is a refrigerant loop with cold plates going into a sealed case filled with nitrogen.

 

Unless your setup was to be a sealed case with a ducted blower that circulates chilled nitrogen as well as direct to chip refrigerant cold plate cooling.

 

Though you could just do a refrigerant loop with standard humidity control modifications instead of needing to purge the sealed case every time you work on it.

[kilgore_T]

how about throwing your pc in a bath of distilled water

bingo no more worries about condensation lol

[For Science!]

This is not a new concept and has already been prototyped.

 

 

[Amazonsucks]

11 hours ago, kilgore_T said:

how about throwing your pc in a bath of distilled water

bingo no more worries about condensation lol

Even better than water.

 

And for the condensation

 

 

[Jumper118]

On 11/11/2018 at 10:49 PM, Aquados said:

The air in a given volume of space has approximately 80/ Nitrogen, 20/Oxygen  and Other.   Imagine the Pc rig sitting in a 100% Nitrogen environment with the Sub-zero cooling hardware outside of it.

A box that is big enough for the Pc, will have 3 holes.  2 will be input and out valves for the gas. The third will be cable and tubing.  The C&T hole will be filled with silicon for air tight seal.  Inject the Nitrogen gas with the output valve open until all of the non-nitrogen gas are pushed out. Close valves without building pressure.  There should be no moister in the box. 

Nitrogen is lighter then oxygen and water. Output would be on the bottom and input on top.    LTT has a cnc, laser machine, and maybe a 3D printer. They got the resources to make it.  

I think it is time they get an engineer or a person who loves and know how to build stuff while using tools correctly.

 

Imagine a world with sub-zero cooling with no threat of moister. 

This has already been done by l and l cooling. 

[SolarNova]

On 11/11/2018 at 10:49 PM, Aquados said:

The air in a given volume of space has approximately 80/ Nitrogen, 20/Oxygen  and Other.   Imagine the Pc rig sitting in a 100% Nitrogen environment with the Sub-zero cooling hardware outside of it.

A box that is big enough for the Pc, will have 3 holes.  2 will be input and out valves for the gas. The third will be cable and tubing.  The C&T hole will be filled with silicon for air tight seal.  Inject the Nitrogen gas with the output valve open until all of the non-nitrogen gas are pushed out. Close valves without building pressure.  There should be no moister in the box. 

Nitrogen is lighter then oxygen and water. Output would be on the bottom and input on top.    LTT has a cnc, laser machine, and maybe a 3D printer. They got the resources to make it.  

I think it is time they get an engineer or a person who loves and know how to build stuff while using tools correctly.

 

Imagine a world with sub-zero cooling with no threat of moister. 

 

 

This is effectivly, well ..IS.. what i have planned for my next build. Spent a few years planning it, its now on the back burner untill other life things are sorted.

 

So what i can tell you form my extensive research and planning.

This will obviosly work. but...

You will need to ensure you dont pressurize the sealed case, but you will want to ensure that it is sealed to the best of your ability so a slightly pressurization and simple soapy water / pressure gauage test will be required.

You will want the case to servicable, so atleast one pannel will have to be removable and thus require a large oring.

Easy method for sealing wires and such will be to cut a hole in the case then use a section of hard tubing, thread the wires through then fill the section up with sealant, then seal the pipe to the hole in the case.

Cooling pipes going in and out of the case just use fittings to seal up against the inner and outer panel.

 

For subzero liquid cooling you will need to decide how far below 0 your going to go. Anything beyond -5c or so is going to likely require either replacmnet o-rings for all fittings and blocks, or custom/braised blocks. You will also have to avoid plastic tubing and go with ether glass or copper depending on if you plan to heat cycle the system.

The fluid of choice would have to be Mayhems XT-1 in a ratio solution appropriat for your temperature target, lowest would be -50c at 60/40 mix though i would test viscosity at this point as it may determin your pump choice.

 

Next choice is res. If you will be heat cycling then smaller res, single loop 1 pump. Though your going to have to wait for temps to go down before putting any significant load on a OC'ed system.

Other choice, and the 1 im going for, is a Large res and a 2 pump dual loop. The cooler (what ever u decide to do) will chill the res and stay on controlled by a temperature controller with 1 pump, keeping the coolant at target temp all the time. The second pump and loop will go from the res to the PC and will only turn on when the PC is on. All of this will have to be very significantly insulated, to avoid conesation (getting everywhere around it wet) and ICE buildup, not to mention ambient temperatures affecting prerformance.

Sensor in the case would also be good, especialy humidity, to keep an eye on things.

Your also going to want to wire up a front IO to control motherboard functions for if u need to clear the BIOS and such otherwise ur goign to have to re-gas the case every time you you need to do it.

Crucialy, if you go the route of using a dry air environment for the case, make it look nice, this is the kind of situation that allows you do avoid 'most' condensation protection thus avoiding the ugly look. I would still advise some bassic condensation protection, even though you shouldnt need it, like grease in the CPU socket to avoid pinrot, and perhaps some kind of clear conformal coating over the coldest areas.

 

 

Think the covers everything ..problably forgot somthign though

On 11/11/2018 at 10:53 PM, Enderman said:

Or just hook up the sealed PC case to a dehumidifier to remove moisture.

This wont work very well. Went through this idea myself. Most if not all dehumidifers only go down to 30%, not low enough to go subzero.