TEC/Peltier CPU Chilled Water Cooling

[4Ryan6]

So you intend water cooling the hot side of the peltiers?

It can be done with a really good air cooler, why waste all that hardware attempting to water cool them?

Don't see any peltier assembly, of any kind or configuration, in your picture?

Did you post the wrong picture?

 

[toolmaker_03]

No that is only a radiator module then there is the case module and the last will be the TEC module

the TEC module uses those little copper rads attached to the TEC's hot side and a CPU water block attached to the cold side the copper rads get hot but the water only gets to 80C I have a big 200mm fan cooling all the little copper rads too. but the system is not ready to connect and show off yet still have some control issues to work out, thinking of using a 3-way parallel system on the CPU blocks to mix the water because at idol I only need one small TEC to cool the entire system while at load I need 3 TEC's to cool the system down, let you know how it turns out, and if this will be the final configuration for this module or back to the drawing board and try something else.

 

[kauedg]

Oh man, I'm almost giving up setting this up. When I find the required hardware, it's very expensive, when I find it to buy from ebay and such, it's not delivered here or it costs the same as the products. The government has raised the tax on international buys from 0.38% to 6.38%, I have a limit of $50 dollars and if the package is held by the customs for checking, I must pay as much as 60% of it's price in taxes (and believe me, after all that, sometimes it's cheaper by 1/3 than buying it here).

I'm kind of pissed off of living in this shit pound.

 

[kauedg]

Just an example... let's say I want to buy one used TRUE heatsink:

http://www.ebay.com/itm/Thermalright-Ultra-120-Extreme-lapped-for-socket-775-CPU-Heatsink-Fan-/221277140932?pt=US_CPU_Fans_Heatsinks&hash=item33852537c4

It costs 39.99 + 43.05 shipping = $83

Add 6.38% for intl shop = $88.33

Then custom's 60% = $93.63

And convert it to BR Real = $93.63 * 2.35 = 220,05

So, for a $40 piece (roughly R$94) I get to pay R$ 220,05. This, because of our incompetent government wich does not promote science and research, but drives a consumption-based economy of internal products and taxes everything that comes from the foreign or else the local economy would ruin.

 

[4Ryan6]

@kauedg

Where do you live to face those additional prices?

 

[kauedg]

@kauedg

Where do you live to face those additional prices?

Brazil

 

[kermdawg]

Ryan I had a thought and wanted to get some opinions on it before I wen't out and spent the cash on it, and being a fan of your thread I thought you might be able to give some advice.

I've been thinking about the buried copper idea(great but alot of work) or the toilet idea(I think I could make it work but pain in the butt) or even making a slush box outta a cold water line and a drain...then I got to thinking-

It's winter time. It doesn't quite freeze where I live(Las Vegas) but it does get into the upper 30's/lower 40's every night. High of mid 60's..meaning the ambient temperature has a great average in the lower 50's or so.

I also have a swimming pool in the back yard, that doesn't freeze. I figure it holds around 10,000 gallons of about 40 degree water. I was thinking of running a supply and return to and from the pool(maybe the deep part?Might take some experimentation) and cycling that through a heat exchanger feeding my cooling loop.

I really can't see the 200-300 watts of heat my computer puts out having much of an effect on that much water. Passing it through a heat exchanger would prevent any contaminates from entering the loop, and of course no heat exchanger is 100 percent efficient so the delta will be a bit higher. Obviously would need something like a pond pump(or maybe 2 in series?) to get enough head.

Sound like a plan or am I wastin my time?(before I waste my money?)

 

[4Ryan6]

Ryan I had a thought and wanted to get some opinions on it before I wen't out and spent the cash on it, and being a fan of your thread I thought you might be able to give some advice.

I've been thinking about the buried copper idea(great but alot of work) or the toilet idea(I think I could make it work but pain in the butt) or even making a slush box outta a cold water line and a drain...then I got to thinking-

It's winter time. It doesn't quite freeze where I live(Las Vegas) but it does get into the upper 30's/lower 40's every night. High of mid 60's..meaning the ambient temperature has a great average in the lower 50's or so.

I also have a swimming pool in the back yard, that doesn't freeze. I figure it holds around 10,000 gallons of about 40 degree water. I was thinking of running a supply and return to and from the pool(maybe the deep part?Might take some experimentation) and cycling that through a heat exchanger feeding my cooling loop.

I really can't see the 200-300 watts of heat my computer puts out having much of an effect on that much water. Passing it through a heat exchanger would prevent any contaminates from entering the loop, and of course no heat exchanger is 100 percent efficient so the delta will be a bit higher. Obviously would need something like a pond pump(or maybe 2 in series?) to get enough head.

Sound like a plan or am I wasting my time?(before I waste my money?)

Sounds feasible to me, if you used a high flow round tube radiator as the heat exchanger in the bottom of the pool, the flow restriction would be lessened.

Most of the Watercool brand rads are round tube.

You would need an inline pump to keep all contaminants out of the system, unless I'm misunderstanding your concept?

Whoa, So you want to transfer the cold water from the bottom of the pool to an exchange point closer to the home, is that your intention?

How far is the pool from where you would need the cold water to enter the home?

You would need to insulate the water lines from the pool so daytime sunlight would not effect their cooling.

At first I was imagining just dropping a rad into the deep end of the pool, but I don't think that is your plan, please explain exactly what your concept is so I can get on the same thought page as you?

 

[kermdawg]

Basically your kinda using the pool like a cooling pond, ala a power plant. You would pump the pool water itself up to your computer where a heat exchanger would transfer the heat from your computer loop to the pool water before the pool water would run back down into the pool(preferably as far away from the intake as possible).

I've seen a couple concepts similar to this but the one's I've seen are MUCH more extravagent.

I live in a 2 story house, but my computer is next to the exterior wall adjacent to the pool(sucks in the summer cause the wall faces west, getting the pull brunt of the hot las vegas sun). I figure it's 20 foot at the most to where my computer sits on my desk, plus the 10 feet to the bottom of the pool...gonna need a pump(or two in series) capable of 30 foot of head.

Yes the exterior lines will definately need to be insulated...its a shame I can't run them underground or up the wall in the house(I COULD but my landlord wouldn't like it)

From the edge of the pool to the exterior wall of the house is less than 10 feet. Again though, I gotta get it UP to the second floor.

 

[4Ryan6]

30' of head lift is definitely a serious pump alone, even 2 is running into some bucks for just a wintertime project, are you sure you want to go this route?

Being on the 2nd floor of your home adds some serious challenge and you'd want quality pumps that will last and they aren't cheap, you could lift from the pool bottom to a staging station box containing the 2nd pump to lift from there to your heat exchange box.

Have you considered what you intend to invest in this possible project?

 

[kermdawg]

30' of head lift is definitely a serious pump alone, even 2 is running into some bucks for just a wintertime project, are you sure you want to go this route?

Being on the 2nd floor of your home adds some serious challenge and you'd want quality pumps that will last and they aren't cheap, you could lift from the pool bottom to a staging station box containing the 2nd pump to lift from there to your heat exchange box.

Have you considered what you intend to invest in this possible project?

Ya thats what I'm still considering. Sucks that the pool pumps are on the other side of the yard(probably 50 feet away) But I think I could still route a line around the pool to one of the pumps and tie it in, thus only needing one pump.

I guess it sounds like something I could research further. Ill see if I can figure a cheap way to move water up 30 feet or so...it's windy enough here, maybe a windmill to pump it up to a tank on the roof? lol.

 

[4Ryan6]

30' of head lift is definitely a serious pump alone, even 2 is running into some bucks for just a wintertime project, are you sure you want to go this route?

Being on the 2nd floor of your home adds some serious challenge and you'd want quality pumps that will last and they aren't cheap, you could lift from the pool bottom to a staging station box containing the 2nd pump to lift from there to your heat exchange box.

Have you considered what you intend to invest in this possible project?

Ya thats what I'm still considering. Sucks that the pool pumps are on the other side of the yard(probably 50 feet away) But I think I could still route a line around the pool to one of the pumps and tie it in, thus only needing one pump.

I guess it sounds like something I could research further. Ill see if I can figure a cheap way to move water up 30 feet or so...it's windy enough here, maybe a windmill to pump it up to a tank on the roof? lol.

A windmill?

If you think pumps are expensive an actual water pumping windmill has got to be expensive?

What happens on a still day, with zero wind blowing?

By the time you invest what it will take to put your idea into actual operation you could have built a TEC cooling setup from scratch that is covered in this thread, for less money.

The peltier electrical cost at 12v dc would probably be much less than 2 120v lift pumps running constantly as well, and the peltiers yield below ambient room temperature, even through the summer when the pool trick becomes useless.

Just something for you to ponder.

 

[4Ryan6]

It has come to my attention that the Swiftech Apogee XTL water block is now a discontinued product as far as Swiftech is concerned, which was the perfect water block to modify for a peltier sized 50mm x 50mm.

Swiftech still produces the Apogee HD and this is the cheapest price I could find on it.

The copper base is exactly the same as the XTL, and the top can be modified for flow the same as the XTL as well, it's just more expensive than the XTL was.

The XSPC Rasa can be used on a 50mm x 50mm as well but would be better if a 40mm x 40mm peltier were used on it, but I used a Rasa on a 50mm x 50mm myself in the beginning and it worked well, Performance-PCs still has the Rasa at a great price.

Performance-PCs also still carries the Thermalright Ultra 120 Extreme heat pipe cooler which is used to cool the hot side of the peltier, and the 110cfm Scythe Slip Stream 120mm fans, at very reasonable prices as well.

I hope all of you are having a wonderful New Year!

The very Best to all of You! Ry

 

[4Ryan6]

Plans for this thread coming in 2014 are posting a detailed parts list with lowest prices available, duplicating this cooling performance solution.

I also have an idea in the brain pan for a new insulated reservoir I am intending to build, it will have much greater cold retention but be the same volume contained, and will have detailed build with pictures to share, of it's entire construction and it's cost.

Since my present peltier setup is now past a year in operation, I will be running some more experimentation testing, along these same successful lines attempting to shrink the size of the cooling setup, and any successful discoveries will be shared.

My goals this year will be more regarding looks and outward appearance, that anyone would be proud to have displayed in their home, I may even house it all in one large case, IDK at this point so we'll see on that one.

But this project is far from over and the best may very well be yet to come!

 

[kermdawg]

So just a little update on my little project.

I managed to find a shallow-well jet pump, with a pressure tank, at harbor freight for 110 bucks. Good for 95 feet of head, and 22 foot of suction-both overkill

I was thinking of having it pump up to some kind of tank, and just let the water slowly drain out, so that fresh, cold water is constantly being feed into it-kinda like a worm barrel/condenser for making alcohol. Maybe just submerge the radiator itself into the tank, negating the need for an expensive heat exchanger(for now, a more permanent setup would definately require a heat exchanger)

I've seen setups for solar heat exchangers that basically put a big coil of pipe into a similar tank setup...maybe the water in the cpu loop could just be circulated through the loop in the tank? Wouldn't need a radiator at all then.

Edit-Found possibly a better pump for this setup, 12 volt dc with 35 foot max head for 40 dollars.

 

[4Ryan6]

So just a little update on my little project.

I managed to find a shallow-well jet pump, with a pressure tank, at harbor freight for 110 bucks. Good for 95 feet of head, and 22 foot of suction-both overkill

I was thinking of having it pump up to some kind of tank, and just let the water slowly drain out, so that fresh, cold water is constantly being feed into it-kinda like a worm barrel/condenser for making alcohol. Maybe just submerge the radiator itself into the tank, negating the need for an expensive heat exchanger(for now, a more permanent setup would definately require a heat exchanger)

I've seen setups for solar heat exchangers that basically put a big coil of pipe into a similar tank setup...maybe the water in the cpu loop could just be circulated through the loop in the tank? Wouldn't need a radiator at all then.

Edit-Found possibly a better pump for this setup, 12 volt dc with 35 foot max head for 40 dollars.

Sounds like you are really getting serious about this but I do have one question and that is what happens with this setup mid summer when the pool water is around 85f, if you have high overclocking intentions you may not have the cooling you really would need?

In the winter this idea rocks, what backup provisions are you considering that will allow this to be a year round cooling solution?

Don't take me wrong I am not trying to discourage you.

That said, some of our best ideas as human beings come to us in our planning and exploring ideas, we may stumble on discoveries in our quests we never even considered before, and that was from following a idea path that led to our discoveries.

For myself I had already bought half the parts to build a bong type evaporative cooler, and was on my way to the store to buy the rest when I remembered an alternative cooling solution a friend of mine had experimented with, using a styrofoam cooler and ice and water, which for me led to the exploring below ambient water cooling thread.

Discarding the bong cooler idea, and returning the parts I had bought back to the store, since then it's been one step after another of tried ideas, some worked and some didn't, until I discovered this present peltier chilled water cooling, so any ideas you undertake will lead to some great discoveries I'm sure!

 

[kermdawg]

Ryan-Thanks for the encouragement! You have no idea what that means to an aspiring tinkerer/wanna be inventor like myself. Everyone else just seems to think I'm crazy! ESPECIALLY the wife! Instead of the "Why?" type of person, I'm really more of the "Why the hell not??" kind of guy. I guess it's just wanting something that I built myself, is unique, and I can take pride in.

As for summer time, since I've been wondering about that myself for the 8 months out of the year its not 40 degrees but closer to 140 degrees outside-I was thinking about adapting something like this. It's ancient technology used in the middle east for thousands of years to cool their homes. Basically a primitive swamp cooler.

Also, the realization has come to me that perhaps a much better(and cheaper) source of cold water is from the house plumbing itself. It's buried about 3 foot deep( I just had to dig up the sewer main to fix it, and low and behold I stumbled across the water main as well) and should remain between 40-60 degrees year round..not quite as cold in the winter, but much cooler than the pool water in the summer. Plus, no pump needed to bring water up the side of the house.

Same concept would apply(I'm a plumber so no problem with the craftwork involved), would just need to figure out the drain...either running a drain outside and down into the pool or back into the bathroom sink drain(they make a Y fitting just for condensate drains...this would be perfect)

Plus side is, anyone with running water and a bit of plumbing experience could potentially install a system like this as well.

 

[06yfz450ridr]

Ryan-Thanks for the encouragement! You have no idea what that means to an aspiring tinkerer/wanna be inventor like myself. Everyone else just seems to think I'm crazy! ESPECIALLY the wife! Instead of the "Why?" type of person, I'm really more of the "Why the hell not??" kind of guy. I guess it's just wanting something that I built myself, is unique, and I can take pride in.

As for summer time, since I've been wondering about that myself for the 8 months out of the year its not 40 degrees but closer to 140 degrees outside-I was thinking about adapting something like this. It's ancient technology used in the middle east for thousands of years to cool their homes. Basically a primitive swamp cooler.

Also, the realization has come to me that perhaps a much better(and cheaper) source of cold water is from the house plumbing itself. It's buried about 3 foot deep( I just had to dig up the sewer main to fix it, and low and behold I stumbled across the water main as well) and should remain between 40-60 degrees year round..not quite as cold in the winter, but much cooler than the pool water in the summer. Plus, no pump needed to bring water up the side of the house.

Same concept would apply(I'm a plumber so no problem with the craftwork involved), would just need to figure out the drain...either running a drain outside and down into the pool or back into the bathroom sink drain(they make a Y fitting just for condensate drains...this would be perfect)

Plus side is, anyone with running water and a bit of plumbing experience could potentially install a system like this as well.

maybe I should have my father( whose also a plumber) run a geo thermal system for my loop haha. free energy basically but idk how well it would work though haha

i may need to venture back into this if I can get my over clock higher

 

[kermdawg]

"maybe I should have my father( whose also a plumber) run a geo thermal system for my loop haha. free energy basically but idk how well it would work though haha

i may need to venture back into this if I can get my over clock higher "

Geothermal systems are WAY beyond plumbing bro. From what I've looked into, true geothermal systems involving drilling a loop several dozen, if not hundreds of feet down into the earth. Not very cost effective at any rate.

From the relatively minute wattage that a processor(or a CPU and a few GPU's) put out, I believe burying your loop 4-5 feet would be pretty sufficient. Remember, the earth is the best insulator known to man, basically because of its mass-it takes ALOT of heat to cause a temperature change in that much mass. Couple that with appropriate insulation and you might be able to make something work...most of your expense would be labor, and you can get that for free....right?!?

You ever dig a hole in the ground?I mean a DEEP one? (I have, lots of grease traps, sewer/water mains...some of these things are 10 feet deep or more). When your down there on a hot day (100degrees plus) the dirt around you is pretty darn cool to the touch.

And lastly, I'm not sure the temperature delta this is capable of yet. Just because we have a good source of cold water doesn't mean your loop will get to that temperature..depends on the efficiency of the water blocks and the heat exchanger. That'll be the true test to determine weather or not this is worth anyone else's time (and money)

I'm thinking the "weak link" so to speak is going to be the CPU water block. Also, since I'm running Ivy Bridge, I'm thinking about doing a couple test runs, first stock, second lapped IHS, third lapped IHS/delidded.

 

[4Ryan6]

Ryan-Thanks for the encouragement! You have no idea what that means to an aspiring tinkerer/wanna be inventor like myself. Everyone else just seems to think I'm crazy! ESPECIALLY the wife! Instead of the "Why?" type of person, I'm really more of the "Why the hell not??" kind of guy. I guess it's just wanting something that I built myself, is unique, and I can take pride in.

As for summer time, since I've been wondering about that myself for the 8 months out of the year its not 40 degrees but closer to 140 degrees outside-I was thinking about adapting something like this. It's ancient technology used in the middle east for thousands of years to cool their homes. Basically a primitive swamp cooler.

Also, the realization has come to me that perhaps a much better(and cheaper) source of cold water is from the house plumbing itself. It's buried about 3 foot deep( I just had to dig up the sewer main to fix it, and low and behold I stumbled across the water main as well) and should remain between 40-60 degrees year round..not quite as cold in the winter, but much cooler than the pool water in the summer. Plus, no pump needed to bring water up the side of the house.

Same concept would apply(I'm a plumber so no problem with the craftwork involved), would just need to figure out the drain...either running a drain outside and down into the pool or back into the bathroom sink drain(they make a Y fitting just for condensate drains...this would be perfect)

Plus side is, anyone with running water and a bit of plumbing experience could potentially install a system like this as well.

ArthurH ran tap water cooling a long time, he ran his output down the drain, but seeing as how you have a pool that always suffers from sun evaporation you could drain to the pool with zero guilt.

I never ran that cooling because I just couldn't see wasting the water and running it down the drain, but you constantly have to add fresh water to a pool, it's a fact of pool life, so it really wouldn't be wasted directly down the drain and you would get cooling use of it before it got to the pool.

The one problem ArthurH encountered was the effect of his municipal treated water and the effect it had inside his CPU water block, it was actually corroding the copper inside the water block, now that may not even happen with your supply line as the treatment process may be totally different than what they use where he lives.

It's not something to be overly concerned with as you could count the water block as expendable and just keep and eye on it for any type negative water effects, I recommend the XSPC Raystorm water block as it has a copper base plate and an acetal top, the top itself is pretty much resistant to just about anything, and it is an excellent cooling performer water block.

Maybe ArthurH will stop by and give you some tips as he has the most actual experience running tap water cooling here.

 

[4Ryan6]

maybe I should have my father( whose also a plumber) run a geo thermal system for my loop haha. free energy basically but idk how well it would work though haha

i may need to venture back into this if I can get my over clock higher

It's a lot more involved than you may realize!

That's something I have also considered but I was going to have to go down at least 12' where I live in South Carolina and after working so hard to get a nice yard covered in grass I just couldn't see digging it up to cool my computer.

My wife already thinks I'm totally nuts for the cooling experiments I've already run, so some of them I don't even let her know about, but digging up my back yard first off will not be done by hand, it would take a backhoe, and it would be kinda hard to hide it from her.

"What the hell are you doing?" I'm digging a hole dear, "Obviously!" what's it for?"

Trying to explain the next part would probably get me committed! :lol:

It would have been nice if I had thought about ground thermal cooling when we first moved out where we are now, as there was a backhoe on site that was used for the septic tank installation, I can run one and could have dug it out myself, but back then, my computer was the last thing on my mind.

There are many options available today to get you cooling lower than any radiators can reach and still be above the condensation point, but sadly it's all in Do It Yourself territory.

 

[oroboros86]

Hi everyone, I've been reading over the past week over this thread, and registered specifically for this thread.
I have thought about how to do a watercooling TEC setup now, I'm looking to do a below ambient setup like yours (4ryan6) and would like to have some input from you guys. While this is all theoritical so far I'd like to put this experiment into fruition sometime in the future when I have all the parts and numbers set.

So far

This is what I'm looking to do and would like your input on how viable it is
wb = waterblock


fx9590------------Formula Z-------------r90x
CPU----------------Mobo---------------- GPU
WB==>==>==>=WB==>==>==>==WB==>pump/res
TEC 245w
WB==>==2x140-rad==>==1x120-rad==>==2x120mm rad==>pump/res
HS(heatsink for the extra cooling and backup)

The Waterblock will go on the CPU which will have the TEC and another waterblock ontop of that, along with the heatsink.



I'm having a hard time doing the math, but my few questions are
1-is there enough cooling to for the TEC hotside?
2-will the TEC provide enough to cool the CPU/mobo/GPU?
3-would i need another TEC to keep it at a steady 10C~15C?

 

[4Ryan6]

1) As long as it is a high quality heat pipe air cooler to cool the hot side of the TEC, I tried cooling the hot side with a rad and it did not work and some at overclock.net said I was using too small a radiator to do it, as I was using a 240GTX Black Ice rad, it's much cheaper just to put a heat pipe air cooler on it with a 110cfm cooling fan to cool the hot side.

Heat Pipe Air Cooler

Scythe 120mm 110cfm Cooling Fan

2) No absolutely not!

Cool the CPU with the TEC assembly using an insulated reservoir to store the cold it produces.

Cool the GPU with it's own separate loop using a Traditional Radiator/Pump/Reservoir/Water Block, assembly.

Most don't even worry about water cooling the Mobo but if you do include it into the GPU loop, not the CPU loop.

3) It depends on how far you intend overclocking the CPU, if your overclock target is as high as 5ghz then you will need two peltiers.

For more information of what did and did not work that I experimented with there's a link at the beginning of this thread to the Exploring Below Ambient Cooling thread and about 2/3rds through that thread the peltier experiments begin to show up, check it out as well.

 

[oroboros86]

Back to more reading, I'm on page 4 of the Below Ambiet Cooling thread. I've been reading more on TEC's in general as well.

"The Waterblock will go on the CPU which will have the TEC and another waterblock ontop of that, along with the heatsink."

i was looking to making a block that consisted of the waterblock/TEC/Heatsink/waterblock all-in-one for the CPU,

What my goals are for this build would be:

1)low noise
i'll add more rad/low rpm fans if I have to

2)low CPU/GPU temps

3)compact~ (fit all into a mid tower)

I'm having some troubles wrapping my head around some of the stuff such as the peltier temperature difference.

here's an example.

cpu temp = 60C

peltier temp difference of 66Cs~

Theoretically cpu would be -6 and the hotside be 126C without all the factors of resistance and what not.
That's how I'm thinking it works, but doesn't seem to be correct.

For the cpu to remain at 15C, what would the temperatures be on the hotside?

I'm also looking at making a voltage regulator to control the peltier temps.
Running 2xTEC's regulated should be enough to cool both CPU/GPU water loop.
Reservoirs would be small on both hotside and cold side to be able fit inside the case.

 

[4Ryan6]

Back to more reading, I'm on page 4 of the Below Ambiet Cooling thread. I've been reading more on TEC's in general as well.

"The Waterblock will go on the CPU which will have the TEC and another waterblock ontop of that, along with the heatsink."

i was looking to making a block that consisted of the waterblock/TEC/Heatsink/waterblock all-in-one for the CPU,

What my goals are for this build would be:

1)low noise
i'll add more rad/low rpm fans if I have to

2)low CPU/GPU temps

3)compact~ (fit all into a mid tower)

I'm having some troubles wrapping my head around some of the stuff such as the peltier temperature difference.

here's an example.

cpu temp = 60C

peltier temp difference of 66Cs~

Theoretically cpu would be -6 and the hotside be 126C without all the factors of resistance and what not.
That's how I'm thinking it works, but doesn't seem to be correct.

For the cpu to remain at 15C, what would the temperatures be on the hotside?

I'm also looking at making a voltage regulator to control the peltier temps.
Running 2xTEC's regulated should be enough to cool both CPU/GPU water loop.
Reservoirs would be small on both hotside and cold side to be able fit inside the case.

You're going to discover that the theoretical and pure logic regarding a peltier does not always work because of the peltiers finickyness of only constantly producing a useable cold temperature if the peltier HOT~COLD stays in the zone.

Meaning if the peltier HOT side gets too hot the peltier stalls, if the peltier COLD side gets too cold the peltier stalls, and that's a bad thing!

The peltier itself is only about an eighth of an inch thick so too much HOT side heat can warm the COLD side temperature so the cold output is not useable and it throws the peltier into a stall and does not produce useable cold, the exact same happens in the opposite direction if the COLD side gets too cold it affects the HOT side, so you have to find a balance so neither side is overridden, then you can get useable cold side output you can store in an insulated reservoir to use to cool the CPU.

There is no getting around that information I just shared with you, so it has to be factored into you setup, which means you will not get an entire peltier cooling setup inside one case, unless you insulate the motherboard and go for a direct mount, and this thread is not about that kind of setup.

I am running a two case setup, one case houses all the computer and only the cooling tubes enter the case, the other houses all the cooling, it is jumped to start up when the computer starts.

Trying to water cool the 245w 26a , peltier HOT side with a water block did not work for me, because all you have to cool it is ambient room air, the radiator got so hot I couldn't lay my hand on it, and that begins releasing heat into the case, get the picture?

The heat actually caused premature pump failure, so if you are determined to go that route Good Luck To You!

Remember this!

What this peltier thread covers works 100% proven and in operation now for 1 year and 2 months, and is also now duplicated by another forum member ArthurH and is working for him with almost identical results as mine.