Exploring Below Ambient Water Cooling

[4Ryan6]

I'm back online guys, about 120+ threads to answer/respond/thank but I'm back
I'll read through and catch up here as I get time to but I'm just happy to be alive again hehe
Moto

Glad you're back!

 

[Motopsychojdn]

**We could start our on water cooling testing club!
"Idiots R Us"
or "The Cooling Geeks" **
The first rule of TEC Club......
😛
I'm liking the res Ry and the design you've used mounting it through the roof, I've started reading the link but I'm not good for time today, lot of math in there though 😛
Moto

 

[Lutfij]

:lol:

We could start our on water cooling testing club!

"Idiots R Us" :lol:

or "The Cooling Geeks" :lol:

It's a sub minority group interested in this cooling anyway.

But as Moto has said many times, there are few situations that some kind of modding isn't required, that there are very few out of the box solutions that fit perfectly, and the introduction to water cooling is also the introduction into modding.

So once one gets past the initial fear of modding a brand new case, they've already crossed the line to all other modding possibilities, but for the largest majority it ends right there, wiping their sweat off their brow they somehow pulled it off!

Whereas this modding goes beyond any boxed solution, we're writing a new story in modding and cooling altogether.

I concur mate! On a personal level I can vouch for the on hand modding experience needed to tackle each and every lil speed bump I encounter...which is nice to have. Seems like a boring day when a problem doesn't present itself [:lutfij:3]

 

[nathann92]

That looks like the AMD mounting bracket for the XSPC Raystorm CPU water block?

It also looks stressed pretty bad and you're only using 2 screws, and the plexiglass is bent, the design uses the water block itself as leverage points to decrease the clamping stress bend, the mounting plate itself was never designed to be used without the block, be careful you do not break it!

Is the Raystorm your intended cold pickup?

Yea it was stressed but was only clamped down to test the cold plate and heatsink combo before I put the raystorm block on. Then I might change the way it clamps and use all 4 screws. When I started testing heatsinks I used all 4 screws but then after having to pull it apart and change things around it was just easier to only use the 2 for testing.

That's only a 60w I have 2 100ws coming any day now so I didn't want to get to far into clamping it.

 

[4Ryan6]

Yea it was stressed but was only clamped down to test the cold plate and heatsink combo before I put the raystorm block on. Then I might change the way it clamps and use all 4 screws. When I started testing heatsinks I used all 4 screws but then after having to pull it apart and change things around it was just easier to only use the 2 for testing.

That's only a 60w I have 2 100ws coming any day now so I didn't want to get to far into clamping it.

Clamping pressure is very important that must be something you've overlooked in your information gathering.

I see by your picture your springs are not even maxed out, and there's a lot of stress to the mounting plate, I'd estimate 40lbs of pressure at the most if you still have spring tension.

You need a minimum of 100lbs of clamping pressure which is past maxing out the springs, so I've long since left the springs out of the clamping equation, and I'm not too sure that plexiglass plate is going to take the pressure you need?

When you conduct your tests and see how fast the temperature drops and frost forms on the cold side of the peltier, you need to realize the same thing will happen inside your water block

Are you going to modify the flow rate of the RayStorm or use it as is?

If you do not modify the flow rate of the water block it will possibly freeze up on you, expand and crack the block, so be careful and mindful of that.

I'm not referencing varying the pump flow rate if your pump has that capability, I'm referencing how much restriction the block itself has, because if it is too high, the water will not flow through the block fast enough.

If ice is allowed to form inside the water block, that's bad, you need a high enough water flow through the block to prevent ice from forming, that's very important, and something you do not want to learn the hard way.

It's important enough to share with you that I have 2 brand new in the box spare water blocks on my shelf and it is something I did learn the hard way.

Of course in your experimentation to cool a chair, there's no computer loss on the line.

However your discoveries will be interesting, I'm extremely curious as to how that 92mm Thermaltake heat sink performs removing the heat from the hot side?

 

[nathann92]

Clamping pressure is very important that must be something you've overlooked in your information gathering.

I see by your picture your springs are not even maxed out, and there's a lot of stress to the mounting plate, I'd estimate 40lbs of pressure at the most if you still have spring tension.

You need a minimum of 100lbs of clamping pressure which is past maxing out the springs, so I've long since left the springs out of the clamping equation, and I'm not too sure that plexiglass plate is going to take the pressure you need?

When you conduct your tests and see how fast the temperature drops and frost forms on the cold side of the peltier, you need to realize the same thing will happen inside your water block

Are you going to modify the flow rate of the RayStorm or use it as is?

If you do not modify the flow rate of the water block it will possibly freeze up on you, expand and crack the block, so be careful and mindful of that.

I'm not referencing varying the pump flow rate if your pump has that capability, I'm referencing how much restriction the block itself has, because if it is too high, the water will not flow through the block fast enough.

If ice is allowed to form inside the water block, that's bad, you need a high enough water flow through the block to prevent ice from forming, that's very important, and something you do not want to learn the hard way.

It's important enough to share with you that I have 2 brand new in the box spare water blocks on my shelf and it is something I did learn the hard way.

Of course in your experimentation to cool a chair, there's no computer loss on the line.

However your discoveries will be interesting, I'm extremely curious as to how that 92mm Thermaltake heat sink performs removing the heat from the hot side?

I did read up on the clamping pressure but its hard with that BTX mounting style heatsink. On the other hand it does drop the temp pretty fast i would say frost forms within 15 secs ( never timed it so only a guess) and i have never really measured the hot side but i can touch the heat sink and cold plate comfortably.

I have purchased a second hand Noctua NH U9B heatsink so im going to try with that over the next couple of days and use 4 bolts & nuts with the intel raystorm mount and get the full clamping pressure it needs. Pretty much like what you have done with yours.

I purchased a XSPC raystorm cooling kit and just sold off the parts i don't need it seemed the cheapest way to get the block, pump reservoir and all the fittings and tubing.

I am running a fair amount of anti-freeze with the water that should stop the ice inside the water block right or will i still have to mod it?

 

[cbrunnem]

I did read up on the clamping pressure but its hard with that BTX mounting style heatsink. On the other hand it does drop the temp pretty fast i would say frost forms within 15 secs ( never timed it so only a guess) and i have never really measured the hot side but i can touch the heat sink and cold plate comfortably.

I have purchased a second hand Noctua NH U9B heatsink so im going to try with that over the next couple of days and use 4 bolts & nuts with the intel raystorm mount and get the full clamping pressure it needs. Pretty much like what you have done with yours.

I purchased a XSPC raystorm cooling kit and just sold off the parts i don't need it seemed the cheapest way to get the block, pump reservoir and all the fittings and tubing.

I am running a fair amount of anti-freeze with the water that should stop the ice inside the water block right or will i still have to mod it?

from what i just read a 50/50 mix with antifreeze will freeze at -30*f and with a 75/25 mix it will freeze at 10*f. so unless you are going below freezing you should be good.

 

[4Ryan6]

I am running a fair amount of anti-freeze with the water that should stop the ice inside the water block right or will i still have to mod it?

It should theoretically and by all anti-freeze claims?

But there are no spaces within an automobile engines water circulation path as tight as what's inside your water block, so good luck!

Let us know what happens?

Additionally: Anti-Freeze is a wonderful product for automobiles and water cooled motorcycles, but not so much when actually brought into the home to cool something.

A leak with a vehicle is an inconvenience, it may leak in your yard, carport, garage, etc. all basically outside the home, but inside your home cooling a chair as you intend, right?

Anti-Freeze will ruin even hardwood floors, laminate floors, carpet, and a host of other stuff, and can kill any animals that lick it up if they consume enough of it.

So make sure that wherever you put this chair, you assume full responsibility of the consequences of a leak.

This warning was brought to you by 4Ryan6!

I'm 4Ryan6 and I approved this message! ROFLMAO :lol:

I'm so glad the US election crap is over!!!!!!!!!!!!

 

[nathann92]

It should theoretically and by all anti-freeze claims?

But there are no spaces withing an automobile engines water circulation path as tight as what's inside your water block, so good luck!

Let us know what happens?

That's something that crossed my mind a couple of weeks ago.

If I open it up and don't think I can mod it would you be interested in doing it for me? I'm happy to pay for your time and of course shipping both ways?

 

[4Ryan6]

Well the TecBox is almost completed but I ran out of wire ties!, (I hate when that happens!), however is is 100% fully operational, and looks pretty good, if I may say so myself!

The TecBox is housed inside a AZZA Solano 1000 full tower case, it has 10 5 1/2" drive bays fully open from top to bottom, I started this mod by completely gutting the entire case down to the bare frame.

This was almost like building a computer even though it was all about cooling, mounting the power supply into the case brings all the wire handling to the table.

 

[4Ryan6]

Originally planned on painting the reservoir black to match the case, but then discovered the top of the case could be used to conceal the reservoir so I just thoroughly insulated it and found the perfect home for it inside the case, the pump is mounted directly to the reservoir.

You can see the mounted fan controller behind the reservoir.

Reservoir in it's mounted position.

TEC Assemblies mounted in 5 1/2" bay area, set exhausting their heat out the front of the case.

 

[4Ryan6]

Insulation begins.

Two full independent wrapped layers of insulation, the insulation comes in a roll 2" wide 1/4" thick foam with an aluminum face, the contact side is adhesive covered and sticks very well.

Since schedule 40 PVC pipe with it's thick walls was used to construct the reservoir, the schedule 40 pipe also has a certain amount of insulation factor by itself, and so far the the cold retention is very good.

It is actually a little better than the Coleman or Rubbermaid coolers, of course there's much less air space left inside the reservoir compared to those standard coolers.

Insulation Complete Top of case is on.

Front of case is on.

Side View.

And Wallah! One TecBox in operation!

Wow Batman! There is a desktop after all!

So many changes have been implemented since the beginning of this thread all aimed towards below ambient cooling, TEC cooling has replaced the 54qt Coleman Cooler and freezing and swapping out blocks of ice every single day is history.

The TEC cooling has been downsized to fit inside a single case and it's cooling performance is amazing.

 

[4Ryan6]

That's something that crossed my mind a couple of weeks ago.

If I open it up and don't think I can mod it would you be interested in doing it for me? I'm happy to pay for your time and of course shipping both ways?

Nathann92, I live in the United States if I remember correctly you live in Australia right?

I wouldn't mind doing it for you, but you are talking a lot of money in shipping, I don't think you may realize how much it will cost you, you can do all you've done to this point but you cannot use a Dremel tool?

Modifying the block is not that difficult.

 

[4Ryan6]

Question? Why would anyone want to even consider this type of CPU cooling solution?

Answer! For overclocked load temperatures lower than the greater majority of your idle temperatures!

Quote Below is from the beginning of this thread.

CPU Temperature Results:

These comparative tests below were run with my 2500K overclocked to 4500mhz @ 1.325v with one 580GTX at 23c ambient.
The 4 core temperatures were averaged together for one total score.

Air Cooling;
Noctua NH-D14, Idle = 32.75c, Load = 54.75c
Thermalright 120 Extreme 2 fans in push/pull, Idle = 32c, Load = 51.75c

Standard Water Cooling Closed Loop;
XSPC Rasa RS240 kit, Idle = 30.75c, Load = 50c
XSPC Rasa with a Black Ice 240 Radiator, Idle = 30.25c, Load = 48.75c

2700K at same 4500mhz clock @ 1.305v except with Hyper Threading Enabled (Which brings more heat to the table!) and running 2 580GTX in SLI, (Which also adds more CPU processing heat), at a 25c ambient, 2c higher than the original tests were run at.

If you have the ability to control your actual water temperature you are cooling with, then ambient room temperature is actually totally irrelevant.

But this isn't, (I forgot to mention I was downloading in the background at the same time I ran the tests! :lol: ).

The same benchmark tests that were run above using an 8c water temperature, Idle = 11c, Load = 30c.

8c is still not in the condensation problem zone, where I live that is! 😀

 

[Lutfij]

I'm still fixated on this initiative of going close to ambient before I go sub-ambient

Keep up the great work mate! Oh yeah I also would like to thank you for the Dremel tool link, though not for me - nifty tool to have should I setup my workshop.

[:lutfij:3]

 

[4Ryan6]

I'm still fixated on this initiative of going close to ambient before I go sub-ambient

Keep up the great work mate! Oh yeah I also would like to thank you for the Dremel tool link, though not for me - nifty tool to have should I setup my workshop.

[:lutfij:3]

I'm just sharing my thoughts below, not criticizing, or anything like that!

Close to ambient is a challenge in itself, if there's not enough radiator and the right cooling fan airflow balance, fluid flow rate, case cooling performance, etc., you can end up seriously frustrated!

Cases themselves have a lot to do with what can and cannot be accomplished inside of them, and even if they're close to whats needed they always fall short somewhere and we're forced to mod to pull off our intentions.

Cases themselves are sometimes the biggest obstacles to overcome, when it's all crammed inside the internal airflow becomes impeded sometimes to the point of radiator stall, to where the cooling performance seems to seriously suffer, IMO the overall best solution is the radiator assembly mounted outside the case or completely in it's own RadBox.

It would be nice if case designers would sit down with us, and get our input on designing a water cooling case for today, and then put a price on it that doesn't break the bank to attain, and comes with all the needed option panels, and availability without having to purchase additional mounting panels adding price on price. (Referencing the Mountain Mods cases and even some of them would not work for my needs)

It's very important in a water cooling setup that airflow across the motherboards voltage regulators is good, and too many today overlook that additional heat load has to be accounted for inside the case, or motherboard failure is just a matter of time, it will happen.

Most don't take into consideration that overclocking adds heat to the loop, and that amount of additional heat is not always 100% calculable across all brand platforms, so what's calculated out for AMD may not even work for the particular setup, and the same for Intel setups.

Most times the unaccounted overclocking heat load is what turns the frustration corner, when one has invested a certain level of money in their water cooling setup and it doesn't deliver the performance hoped for., there's a big difference between settling for what the end results are, and being 100% completely satisfied with the cooling performance.

That was my frustration and what started this thread, was seeking a cooling solution that literally broke the limitation boundaries, so the calculations became irrelevant, if I choose to run close to ambient, I can, if I choose ambient, I can, if I choose below ambient, I can, even if I choose to insulate the motherboard add anti-freeze and go sub zero, I can.

So as you are watching what's going on in this thread, I'll be continuing to test the boundaries of this cooling solution and changing unknown to known, so others can duplicate and use it successfully themselves, if they so choose to do so.

It's finally gotten Moto's curiosity so it will be interesting to see where he goes with his ChillBox discoveries?

You are welcome for the Dremel tool link!

Also your build thread is most impressive!

Ry

 

[Lutfij]

...my frustration and what started this thread, was seeking a cooling solution that literally broke the limitation boundaries, so the calculations became irrelevant, if I choose to run close to ambient, I can, if I choose ambient, I can, if I choose below ambient, I can, even if I choose to insulate the motherboard add anti-freeze and go sub zero, I can.

Ry

now that is the stuff I look for, something I can do whether it be going up or down.

Also your build thread is most impressive!

Ry

I need to be on my A-game to stay within our circle of trust , but thanks man

 

[Motopsychojdn]

Duct tape will make the res black Ryan
looks shiny though, I'm itching to crack on with mine but times a little thin lately in the Psycho household
Moto

 

[4Ryan6]

Duct tape will make the res black Ryan
looks shiny though, I'm itching to crack on with mine but times a little thin lately in the Psycho household
Moto

Or Red, Green, Blue, Purple, Camoflage, White, etc., I didn't realize duct tape came in so many colors!

The bad thing about duct tape is sometimes it leaves it's glue behind when it is removed, and this res/cooler is in it's testing phase anyway.

So far it has performed very well!

Just completed replacing the switch for the 2nd TEC assembly, ran longer wire and mounted it under the lip of my desk, approximately 6" away from my mouse.

Now that's convenience!

 

[Lutfij]

:lol: convenience it is! If left on for a long period of say more than a few days, yeah the adhesive will have a residue.

 

[nathann92]

I'm stuck on my project. I have everything assembled I'm using 3 meters of copper pipe and it has around 1 little of coolant in the loop. When the pump is off the peltier & water block freezes over but as soon as I turn the pump on it defrosts and warms up. I'm using a different heatsink and a push and pull fan setup with a 100w peltier I can attach photos if needed.

Any ideas why it's not staying frozen the water in the loop is only at room temp so there isn't a massive heat load on it.

 

[4Ryan6]

I'm stuck on my project. I have everything assembled I'm using 3 meters of copper pipe and it has around 1 little of coolant in the loop. When the pump is off the peltier & water block freezes over but as soon as I turn the pump on it defrosts and warms up. I'm using a different heatsink and a push and pull fan setup with a 100w peltier I can attach photos if needed.

Any ideas why it's not staying frozen the water in the loop is only at room temp so there isn't a massive heat load on it.

Regarding the water block freezing be careful how long you play with that, if you allow it too long it will split your block open.

Did you modify the water block?

Is all that copper pipe insulated to retain the cold transfer?

Lets see some pictures of the setup?

 

[nathann92]

I opened the water block and it just looked different to what you had explained so i left it as it was.
http://i1326.photobucket.com/albums/u653/nathan_nelson1/IMG_1763_zps0a7df1d7.jpg
http://i1326.photobucket.com/albums/u653/nathan_nelson1/IMG_1764_zps50082972.jpg

Only the plastic tubing is insulated, I need the cooper pipe exposed since its the cooling source (Don't now what you would call it) that goes in the chair.

http://i1326.photobucket.com/albums/u653/nathan_nelson1/IMG_1801_zps3a016956.jpg

Will i need a more powerful TEC?

This is what happens when the pump is unplugged.
http://i1326.photobucket.com/albums/u653/nathan_nelson1/IMG_1797_zpsb064c921.jpg

As soon as its plugged in.
http://i1326.photobucket.com/albums/u653/nathan_nelson1/IMG_1800_zps98385f33.jpg

 

[nathann92]

Just playing around before i leave for work 🙁. I turned the TEC on without the pump to freeze then i switched the pump on which defrosted it but slowly after around 8min its slowly freezing the waterblock again. I guess it just takes time to cool the water.

 

[4Ryan6]

Regarding the water block you may get the flow you need by removing the jet plate and the small rubber ring, and reassembling the block without them.

Even though you do not see the frost when the pump is running the TEC assembly is producing cold, you have to allow it to build up the cold storing it in an insulated reservoir.

I'm running 2 245w peltiers and they do not show any frost when in operation.

I'm also running a 1/2 gallon or 2 litre capacity insulated Cooler/Reservoir.