Sustainable subzero overclocking

[Danioki]

Alcohol (isopropyl alcohol, ethanol and isopropanol) is a polar solvent (very conductive) and is potentially corrosive (contains water).

nope.

 

it will have to be pure ethanol

[Cavalry Canuck]

5 minutes ago, Danioki said:

Alcohol (isopropyl alcohol, ethanol and isopropanol) is a polar solvent (very conductive) and is potentially corrosive (contains water).

 

nope

Thanks for the google copy-paste.

 

Why wouldn’t you quote the post where I brought it up?

[Danioki]

4 minutes ago, Cavalry Canuck said:

Thanks for the google copy-paste.

 

Why wouldn’t you quote the post where I brought it up?

i did it, but then it changed to another post, i don't know what i did if i'm honest.

 

maybe use some mix of mineral oil and non conductive-antifreeze.

[Jurrunio]

24 minutes ago, Cavalry Canuck said:

You’d get sufficient cooling with the stock mobo heatsinks if you use a circulated fluid.

Not enough to keep it entirely forzen still

 

24 minutes ago, Cavalry Canuck said:

Your ice would have to be pretty damn cold for there to be no melt in contact with one of these uncooled heat-producing components/areas.

So is the CPU, just port some coolant over to the waterblock for the VRM

[Cavalry Canuck]

12 minutes ago, Danioki said:

i did it, but then it changed to another post, i don't know what i did if i'm honest.

 

maybe use some mix of mineral oil and non conductive-antifreeze.

What about Kerosene?

[Danioki]

[Cavalry Canuck]

3 minutes ago, Danioki said:

Need oxygen for that. If this is a sealed system, this wont be a problem.

[Danioki]

Just now, Cavalry Canuck said:

Need oxygen for that. If this is a sealed system, this wont be a problem.

too dangerous, imagine a fire in the house. you have a bomb waiting in your room.

 

i think mineral oil with sufficient fluid movement will be enough.

[Cavalry Canuck]

1 minute ago, Danioki said:

too dangerous, imagine a fire in the house. you have a bomb waiting in your room.

No more so than the natural gas lines that service your furnace, hot water, stove, dryer, etc. 

 

3 minutes ago, Danioki said:

 

i think mineral oil with sufficient fluid movement will be enough.

That’s what I was originally thinking as well. But somehow that wasn’t enough, so I began listing alternatives. Admittedly it’s  rather fun to consider what else *could* be used.

[Danioki]

1 minute ago, Cavalry Canuck said:

No more so than the natural gas lines that service your furnace, hot water, stove, dryer, etc. 

yeah but natural gas lines are installed by professionals and the certified. we are talking about some DIY thingy.

 

https://www.engineeringtoolbox.com/freezing-points-liquids-d_1261.html

 

so we can take many liquids from here and then check for conductivity.

 

so Nitrogen has a freezing point of -210°F -346°C, and is an electric insulator, so it's the best alternative

[Cavalry Canuck]

5 hours ago, Danioki said:

yeah but natural gas lines are installed by professionals and the certified. we are talking about some DIY thingy.

 

https://www.engineeringtoolbox.com/freezing-points-liquids-d_1261.html

 

so we can take many liquids from here and then check for conductivity.

 

so Nitrogen has a freezing point of -210°F -346°C, and is an electric insulator, so it's the best alternative

You can install your own and only require an inspection, depending on where you live. In my place, they are all threaded cast iron pipes with only thread tape to seal the connections.

 

I think there comes a point where the vapour/boiling points of a liquid become a concern. Nitrogen wont stay liquid at room temp and 1 atmosphere of pressure. And if you want to talk about explosive, liquid nitrogen is VERY explosive due do it’s expansion ratio. Your case would need to be pressure rated, not just sealed.

[SolarNova]

15 hours ago, FratChad said:

I have an idea for constant subzero overclock. As we all know the biggest problem with going sub is the buildup of moisture. Why not just remove the air and water out of the equation by submerging the system in a non-conductive liquid. With this the temp can be dropped far below zero and maintained. What are yalls thoughts?  

Wouldnt work well at all.

 

however the work around is to instead of using a liquid to dunk the components in to avoid moisture, to just remove the moisture from the air the components are in.

 

A airtight case filled with any inert dry gas would solve the condensation issue. Nitrogen, Helium, Argon ,etc

 

Removing condensation from such a build is the easy part tbh.

 

The two hardest parts are building a subzero chiller powerful and efficient enough to chill coolant sufficiently below zero to be worth while. Your not going to get away with using a r134a chiller, nor using a basic DIY AC to chiller conversion. You will need to use a dedicated scratch built subzero chiller using a refrigerant at least below -40c. it would have to be designed to chill a liquid, so most likely a plate heat exchange would be used.

 

The 2nd hardest part is dealing with the PC side loop and modifying it to handle subzero temps. Normal o-rings wont do a very good job at -30c and acrylic topped blocks and tubing will become extremely brittle and develop cracks. As such ur going to have to either replace o-rings with a more suitable material type and go with glass tubing, or build a complexity braised copper loop (basic plumbing), which would involve buying full metal blocks, removing the o-rings and braising them closed.

 

The mineral oil submersion cooling is fine for sub ambient chilling, but not subzero.

[Danioki]

12 hours ago, Cavalry Canuck said:

You can install your own and only require an inspection, depending on where you live. In my place, they are all threaded cast iron pipes with only thread tape to seal the connections.

oh, here you are required to do it with a certified professional and then pay again to get the thing certified for use.

 

Quote

I think there comes a point where the vapour/boiling points of a liquid become a concern. Nitrogen wont stay liquid at room temp and 1 atmosphere of pressure. And if you want to talk about explosive, liquid nitrogen is VERY explosive due do it’s expansion ratio. Your case would need to be pressure rated, not just sealed.

well, we will all die anyway, so why not go in a blaze of glory.

[Cavalry Canuck]

5 minutes ago, Danioki said:

oh, here you are required to do it with a certified professional and then pay again to get the thing certified for use.

Ooof

5 minutes ago, Danioki said:

well, we will all die anyway, so why not go in a blaze of glory.

It wouldn’t be so much a blaze as it would be a loud thud. If the shrapnel and over pressure don’t  kill you, asphyxiation will. No blaze, so no epic burning to death. 

 

Quite lame.

[Danioki]

1 minute ago, Cavalry Canuck said:

It wouldn’t be so much a blaze as it would be a loud thud. If the shrapnel and over pressure don’t  kill you, asphyxiation will. No blaze, so no epic burning to death. 

 

Quite lame.

that is why you keep a candle lit at all times beside the PC. no one want a lame ending.

[Cavalry Canuck]

11 minutes ago, Danioki said:

that is why you keep a candle lit at all times beside the PC. no one want a lame ending.

Nitrogen wont burn.

 

I’m being a total buzzkill today.

[Danioki]

1 minute ago, Cavalry Canuck said:

Nitrogen wont burn.

 

I’m being a total buzzkill today.

ahahah you are right. well the we go back and use ethanol or kerosene.

 

i need an explosion, mythbusters kind of explosion ahahaha.

[PowerBaller]

Alphacool has a solution like this.

By theory could be any simple refridgeration loop.  Intergrate the block as the evaporator, then a compressor, then condenser, than the PWM expansion valve. 

But honest i dont expect the result to be fancy for daily use. Conventional watercooling take more advantage of the total heat capacity of water, instead of merely the temperature difference. Since phase change always has to be involved in refridgeration cycle, you cant have more than 200 grams of it in your loop. Plus, you have to weld, instead of using fittings to connect your loop.  

 

 

 

 

[SolarNova]

8 hours ago, PowerBaller said:

Alphacool has a solution like this.

By theory could be any simple refridgeration loop.  Intergrate the block as the evaporator, then a compressor, then condenser, than the PWM expansion valve. 

But honest i dont expect the result to be fancy for daily use. Conventional watercooling take more advantage of the total heat capacity of water, instead of merely the temperature difference. Since phase change always has to be involved in refridgeration cycle, you cant have more than 200 grams of it in your loop. Plus, you have to weld, instead of using fittings to connect your loop.  

 

 

 

 

 

Aye but its not very good and is overpriced for what it is.

 

At stock it only chills down to 20c, worse than most aquarium chillers which can go down to 4c. It also uses r134a refrigerant, which means even if u bypass the thermocouple , it wont get very far below subzero.

 

With the limitations it has ur far better of getting a cheap aquarium chiller like a Hailea unit, similar to the one LTT used in a vid once, but would recommend a far more powerful version than they used.

They are cheaper and work down to 4c, if ur lucky enough to find an older model like the HC1000B, BH, and HC2200BH, running r22 refrigerant u can even bypass the thermocouple and get usable subzero performance out of them.