Exploring Below Ambient Water Cooling

Page 37 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
[/quote]
Your buried copper idea I had also been seriously considering myself, but in South Carolina I would have had to go down 10' to get a constant temperature to rely on, and I had invested too much work getting grass to grow in my back yard to dig a hole to accomplish it.

Plus my wife already thinks I'm nuts anyway!

I could just imagine me telling her I was going to dig down 10' in the back yard to lay copper tubing in the ground to cool my computer? :lol:[/quotemsg]

I understand your wife's concern! Fortunately mine is an engineer, we both work in the energy efficiency industry so she not only understands the mechanical 'beauty' of some of the things I build but also sees the practical benefits they can provide to the environment and economy.

I wish I could find the pictures of the subterranean heat sink I built but they seem to have vanished so I'll have to give a try at explaining the design so you don't think you need to dig up your backyard. I used a roto hammer to remove a section of the basement floor slightly larger than the opening of a 5 gallon bucket. Once this circular section of concrete was removed I then took a 5 gallon bucket with the end cut off, stuffed it into the hole and started excavating the dirt in the center of the bucket - the deeper I dug the deeper I pushed the bucket to make sure the dirt did not collapse into the hole and undercut my foundation.

I used this method to go down about 3.5 to 4' under the foundation using a shop vac to remove the sand and dirt as I went. Once the the depth was right I used a 4' piece of PVC pipe and the vacuum to drill the 4 probes at an outward angle to the bottom of the hole. These probes are important because they get deeper into the ground and give access to a much larger thermal mass. Inside of each of the probe holes I set a 5' long 3/4 diameter piece of copper pipe filled with water and capped. These 4 pipes protrude into the cavity about a foot, after they were set I poured concrete into the probe holes being sure to keep the copper pipe centered in the shaft. The PVC was pulled out afterward leaving the concrete and copper free to transfer heat to the earth without the insulating properties of the PVC.

Once the 4 probes were set I coiled the 100' of 1/2 copper tubing so it was about 2" from the inside edge of the bucket and made a second coil loop a foot in diameter that coils up from the bottom of the hole about 2 inches from the outer coil. I run hot water down the center of the loop and back up the outer coil to the top.

I wired the coils, which fit right over the 3/4 probe tubing in the center of the chamber, together at a few points to keep them from moving when the concrete was poured. The chamber was slowly filled with concrete, the bucket pulled out and leveled even with the floor. I brazed a couple quick connects to the 1/2" coil and filled it with distilled water

My thinking is the hotter water going down the center fills the concrete mass with energy, the cooler water coming up the outside loses heat more readily to the larger exterior surface. I thought I'd try reversing the flow at some point but it is just so cool there has never been a need. The 4 probes transfer more heat down their length because they are thermally tied to the larger mass and loop coil; they protrude into the center of the loop so the heat can move through the concrete and copper down their length... no need to dig a 6 or 10' hole to lay the loop at that depth.

I put the whole thing under the house because the foundation insulates the earth under it from the sun and keeps its temperature lower. This way I can go down vertically with the heat instead of digging deep trenches in the yard and running more water in the loop to dissipate the heat because the earth is being warmed by the solar heat gain from above. In essence the earth under the house has a much more stable thermal mass regardless of season because the house keeps it insulated. Another plus, the computer sits above the heat exchanger so there is far less head pressure to overcome as well as better pumping efficiencies.

If you ever decide to give it a try I'd suggest using PEX instead of copper for the loop, that would be sooo much cheaper. You would loose a little thermal efficiency but there is plenty of 'cool' under your house so I doubt you would even notice.

And... don't tell your wife!

LO3
 
And... don't tell your wife!

We don't have a basement, so there's no way she wouldn't know and as far as cooling my computer, I have some of the best performance cooling you could ask for now, granted it's not classed as energy efficient, but no constantly running below ambient cooling solutions are.

Still the ground burial of copper pipe is interesting and I have considered it.
 
Hi Really creative set up for both cooling your pc and warming your house/room. Where is the thumbs up button.


Sorry for my Disappearance Ryan, School and stuff. Ill have the money next month to revamp my whole computer. so expect a lot of new question once i jump in to this TEC stuff in the middle of month.
 


I'm personally not interested as I'm running above the condensation forming threshold so I don't have to insulate the motherboard, but I'm sure someone out there may be interested.

 
Hello Ryan,I wanted to share something I recently noticed while running this pseuexperimental cooling system was that if I lowered the total amount of water and added Three 2 liter bottled frozen water is that it could last up to 14 hours Game before it hit room temp assuming of coarse that the water was already cold at the start point and didn’t have to waste much on looking the water.


This is far from science fact, Due to not acutely testing but just observation. The more water in the (Res) was less time total cooling because I had less (Ice) in there. Now I am adjusting the pumps to 3/2 depending on how much cooling I need on the CPU.

As i see it now i want the most bottles of ice i can fit with the lowest amount of water In addition I need the lowest pump speed's so i don't freeze the CPU outright.
 


All makes perfect sense!

The higher water volume would tend itself to bring the ice to it's temperature range, so a lower water volume allowed the ice mass to reverse the effect, more ice slowed the cold transfer, and slowing the pump speed extended the cold even further.

Good discovery to share! :)

I was doing all sorts of testing when I was ice water cooling a lot focused on various water mass levels from 10 gallons down to 2 gallons, but I was no longer testing ice amounts below 2 gallons of water, because once I had gone below 2 gallons I was using the peltier cooling.

Your discoveries will help others using or considering ice water cooling!

 
When i first found this thread more than a year ago it was very interesting, something id definitely want to try my hand at if i ever were in a position where i could spend the money. From my understanding the restriction of using chilled coolant in a loop is condensation, the fact that it would form and get on components frying things. Just now though i had a thought, Cooling a system that's suspended in a vacuum! I saw a video on youtube, something random and interesting, about Adam Savage making a vacuum chamber for a friend of his. So i wonder, what would the challenges be of placing a system inside one such chamber and running a chilled coolant through a loop inside?

Apparently condensation will only form inside a vacuum if there is already moisture present. If your system is inside, would it be possible for you to remove the moisture with something like a pile of dry rice in the bottom to absorb the ambient moisture? If rice didnt do the job then maybe you could replace the air inside the chamber with a different gas of some kind such as nitrogen. This is where im utterly ignorant as to what, how and where to look. Many questions pop up such as, how would a nitrogen filled container act? Would bottled nitrogen carry with it moisture? (i should hope not if its 100% pure nitrogen) and would replacing the air in that vacuum chamber with nitrogen effect the components in a bad way? Also how would the thermal paste react to both a vacuum and a pure nitrogen environment?

If it were effective then maybe it would allow you to go so far as to refrigerate your coolant as opposed to simply chilling it.

Any who, just a passing thought i had and felt the need to share.
 
I generally run a 10c actual water temperature which is 15c below my 25c ambient room temperature and there is no condensation at that temperature level.

Many think that if you go below ambient at all there is condensation forming instantly but that's not true, you can go below ambient and operate condensation free down to a certain temperature which needs to be discovered in your local environment.

I simply run my water temperatures above the point condensation even starts to form which gives a major overclocking edge over traditional air or water cooling.

Click on my CPU-Z validation and look at the idle CPU temperature at the time of the validation. (All done with zero condensation)

Running everything in a sealed vacuum would be a seriously expensive undertaking you would need a vacuum chamber to handle the pressure of the internal vacuum and the outside air pressure, and the evacuating equipment to create the vacuum in the first place.

It would be much cheaper to just insulate the motherboard and then condensation wouldn't be a problem down to ice forming at sub zero temperatures, for a far less monetary investment.

 
I thought it would be pricey, likely in the low thousands just for a finished unit let alone the cost of developing the methodology for its construction. The goal would be to get a coolant temp of 2.5c, Just above freezing, to begin with and then with more research and testing of available liquids aiming for even lower. Its not something you'd do lightly by any means!
 


http://www.tomshardware.com/forum/282844-29-peltier-water-cooling#t2033960

With the Peltier/TEC chilled water cooling I am presently running I can go down to 2.5c with steam distilled water, but I would need to insulate the motherboard for the condensation as condensation begins forming very fine looking like frost at an 8c water temperature, that's 2c below my 10c target run temperature.

As I go lower condensation begins forming more clearly distinct water droplets and by 5c water droplets are clearly formed, at 2.5c water temperature condensation would definitely be trouble! :)

The lowest I've taken it without M/B insulation was 4c, with high speed fans blowing air over the CPU water block and tubing, drying the water before it was able to fully form, but the fans I was running are too loud to run daily and it was just a test anyway.

 
Having read the original post by 4Ryan6, i was Hooked to get into peltier chilling 🙂

Im gathering parts at the moment, but i have a question regarding the waterblock on the cold side of the TEC.

I know he modded the blocks interior to avoid it freezing up and wanted to avoid glycol in case of leak 🙂

I am getting a tec1-12730 and a ekwb supremacy evo, and i cant help but wonder if this TEC wil not be to powerful to avoid freezing the block.

Id like your opinion on keeping the evos internals and adding glycol to the MiX to avoid freezing

Do you think it would work ?

My reservoir will be made of sewer pipe ø110mm, PU foam as insulation and a ø160 vent pipe as exterior shell w top and bottom ends.

Ill be using the thermalright true spirit 140 to cool the hot side of the TEC.

Cheers from Denmark
Toe
 
I still have my dewar,it may have cobwebs in the bottom.
I really don't want to be fighting -100c coldbugs on miniscule architecture improvements.
Here's to 2017! May it be good!
Happy Thanksgiving,even though Ryan and Saint19 aren't American.
 
For the record I still run chilled water cooling to this day.

I did make some changes though, I originally thought I could run safely at 10c coolant temperature but 2 years ago I discovered I couldn't, because condensation actually built up over time underneath the water block edges. It was not visable to the eye upon inspections, so I had no clue it was even there, I now run my coolant temperature at 13c and have had no more problems.

Running the coolant temperature at 13c is still 10c below ambient room temperature, and there is a lot of freedom to experiment with at that temperature area.
Since in one form or another I have been running some type of chilled water cooling since 2011, I have had many discoveries along the way. Meaning if I was not still learning, and benefiting from it, 11 years later, I would not be doing it!

I have worn out at least three D5 water pumps but all of my peltiers believe it or not, are still 100% operational.

The reason I bring that up is most look at any kind of serious water cooling as a burdensome one time monetary purchase, not really taking into consideration how long these components actually last, including radiators and such as that, and can be transferred to a new setup.

Things have to be cleaned occasionally and sometimes water blocks for CPU and GPU, bought for transfer to different hardware, but it is actually for the most part an investment, that is upgrade-able.

Hope you all are doing well, and in good health! :) Ry
 
I think all coolers, water air and AIO coolers can be swapped around from setup to setup. Intel and AMD have done a great job keeping the mounting holes standard for a really long time now. You might need a shim here or there but they do deserve some praise for that.

My biggest issue with water is the chance for leaks. Which goes up with a setup like yours. AIOs are generally ok with only the odd issue with leaks, but building your own does introduce that risk. You said the parts other than pumps haven't broke so your doing well, I just question what happens when someone lacks the experience and has something break. I'm sure even you have had issues with water on parts before. More so if condensation which you didn't see was building up. (I also admit that yours is a more extreme setup. Not only water cooling but phase change as well.) I loved reading this thread and it helped me learn even more about phase change. It's a welcome change of pace from "will X work with Y".