TEC/Peltier CPU Chilled Water Cooling

[thequn]

I am not an expert at all, but i have my own sub ambient system and have been water-cooling for over 8 years,

The “work so well it won’t work” was referring to the pelter/TEC stalling out because of the cold added to the hot side from the line assuming you can get a rad large enough to cool off the hot side the cpu will not should not affect the cold water by more than a degree or 2 making the water too cold once hitting the hot side so you may need a second rad to warm up that water back to ambient.
400 watt plet assume 600-800 watt of heat that’s need to be killed and with rad over kill is best.

Also if this works well you can add your cpu, gpus to the line to help warm it back to ambient as well.


(edit) oops my class work was mixed in the post removed it. Lol

 

[threefingeredjack]

Interesting stuff! The only reason I didn't just copy what Ryan has done was that I don't fully understand it, and it seems very large and messy. Chances are, that is because its the only way to do it, building up a cold reservoir.

Now from what Ryan just said, the TEC's hot side needs to be cooled, as if it gets too hot it doesn't work. Thats exactly the opposite of what I understood when I designed this system!

I've looked over google, and can't find anything on TEC stalling besides these threads, which seem to be contradictory... Can someone just clarify this bit, does the TEC stall because the hot side is too hot, and therefore is heating up the cold side (ergo, encouraging highly efficient heat sink on the cold side is the obvious cure) or, does the hot side NEED to get hot in order to function for some reason? Logic dictates its the former.

Thegun, you seem to be suggesting the latter! That the liquid after the CPU will still be cold, to the point that dumping it into the hot chamber will stall the TEC? Are they BOTH the case? And the TEC needs to be operating in some kind of goldilocks zone?

I just played around with the design, to replace the radiator with a heat sink, and have that heat sink sat in a fluid of some kind in a separate chamber. Also in that fluid is a second heat sink in the line from the CPU to the hot chamber, effectively (and seemingly, pointlessly) neutralising the heat/cold.

Probably only relevant if the TEC requires the hot side kept hot or something...I don't even think I know anymore.

What I think I need to understand is;

1. Does the TEC need the hot side kept hot and the cold side kept cold, or, does it want the hot side kept cooler and the cool side kept hotter?

2. How hot/cold do these things actually get? I'm having some thoughts regarding materials for the chambers, and you mentioned you managed to get the heat sink so hot it burned you...

3. If I just include a potentiameter in this circuit with the TEC, will that suffice to control the heat exchange? Ergo, lower the voltage and it'll lower the heat exchange?

4. I already have a liquid cooler system on the CPU... its a cheap corsair one. And sucks. What if I just put the radiator from that into a contained space with the cold side of the TEC (and appropriate heat sink and fan)? The ambient temperature in the box would be pretty nippy... but not so cold as to be dangerous and probably pretty easily controlled... idk.

Thanks for keeping up with my questions here though, you pair are evidently the internet's foremost authority on this below ambient cooling thing, you must get this a lot

By the way, someone over on overclockers has been saying a lot of complicated things about deltaT, and says the radiator won't dump the heat down to ambient. Any ideas why?

EDIT: An image of that heatsink liquid bath thing. Likely pointless.

http://s8.postimg.org/xbtj7duwl/Heatsink_bath.png

 

[mlcaouette]

Your sig link to Coolimus Prime isn't working?

Fixed it, thanks. It must've been broken when they changed the forums a while back.

 

[4Ryan6]

Now from what Ryan just said, the TEC's hot side needs to be cooled, as if it gets too hot it doesn't work. Thats exactly the opposite of what I understood when I designed this system!

I've looked over google, and can't find anything on TEC stalling besides these threads, which seem to be contradictory... Can someone just clarify this bit, does the TEC stall because the hot side is too hot, and therefore is heating up the cold side (ergo, encouraging highly efficient heat sink on the cold side is the obvious cure) or, does the hot side NEED to get hot in order to function for some reason? Logic dictates its the former.

I provided a link in the beginning of this thread so anyone reading could investigate for themselves what a peltier can and cannot do, before you question me because you cannot find something you're looking for on Google, I suggest you read it.

http://www.xtremesystems.org/forums/showthread.php?38367-The-Ultimate-Guide-to-TECs

Stalling out is my explanation of what's happening and you probably won't find any Google references to it by searching for peltier stalling or such, whatever it's terminology is actually Googlable (is that a word? :lol: ), IDK, however, it does happen.

The link above is some excellent information until you get to the part suggesting not air cooling the peltier, because you obviously can air cool the freaking thing, or I've been running magical fairy godmother cooling for over 8 months!

 

[thequn]

Hi.

To answer your question about the Goldilocks zone, YES 100% Both are true, If The hot sides gets too hot it can and will bleed over to the cold side. If the too much coolness hits the hot side the same will happen though to a lesser degree, Because your rad needs to be able to take the hot and remove it 100% back to normal temps before it is then cooled and chilled down to let’s say 10* it will hit the cpu cool it at max jump to 12* then hit the pelter lowering the heat on that side of the system. which can lower the cooling power. (This will depend upon how you move the heat away from the hot side of the pelter) this is also just bases on you diagram (if change are made the idea will change.) I suggested adding a rad to the cold side after the cpu,gpu to warm the water back up before taping the hot side , in essences it "should" be like ruining two loops, only in one loop, of course this is not necessarily what will happen, because it depends on too many factors , and is my best guess assuming perfect conditions and transferring of heat and cold. Asumeing that the cold side will even get cold having to disperse with

Other thoughts:
Once you have this set up its will not magically work the way it should a lot of refining small items ensuring the rad is dropping those temps to Ambient and then you have to make sure the pelter is cooling well and the Liquid is getting cold, then you have to make sure the coil was long enough, to cool it down to the right temps after the liquid is cold, and so on and so forth, Again Everything is better if overkilled on this project. Too big of a rad on both sides too long of a coil to cool the water.

I have to say that i never seen a system like this and this is why i love it because it can work if done right. Though you are going to have to spend a lot of time a resources testing to and building it to make it perfect.

ask any one whose done any creative cooling solution and most custom watering people there set up is an ongoing never quite perfect system and we always have room for more knowledge more testing is like an addiction. LOLOLOLOL

oh also I would like to point out that no can will be able to tell you how effective your TEC will be cooling ambient temps consistently . It could stall it out just by itself, Ryan is the only person I know to have mixed sub ambient cooling and radiators together.

 

[4Ryan6]

http://s8.postimg.org/xbtj7duwl/Heatsink_bath.png

Addressing this design of yours I see more questions than answers, even if you tried to use radiators to replace the proposed coils, the design has a lot of flaws that need to be addressed.

So in the top cavity is some kind of coolant that the pump is circulating, assuming that is the peltier fixed in the separation wall, and you're planning on coolant being in direct contact with the peltier itself on both the hot and cold sides?

Even with a moisture insulated or (Potted) peltier, that is not a good idea simply because over time of operation, expansion and contraction, will crack even the best insulated peltier allowing fluid to enter and short out the peltier, they are designed to flex off some moisture, but not designed or insulated for direct submersion in fluid.

Lets consider the top chamber assuming it supposedly is cooling the hot side of the peltier some peltiers top temperature range is 125c, which is basically the burn out point of the peltier so if you were to run it safely for the peltier at say 100c, that is the boiling point of water.

So what happens when the pump fails and it eventually will, and the peltier is still energized, have you thought of that?

The cold side of the peltier in direct fluid contact according to the sketch will actually freeze the fluid, have you thought of that?

You don't show any circulation of the fluid in the lower section only flow through a tube coil, you show a circulation fluid flow in the top section but not the bottom?

Maybe you could fill it with some kind of thermal conductive gel with a seriously low freezing point?

How will you keep an air pocket from forming just below the peltier in the lower section?

The Neutralized water/oil bath section, if it worked perfectly would eventually reach a common temperature, which is very similar to the radiator having an equal and opposite effect to the peltiers cold production in my setup, that's why I eliminated the radiator from my setup, it was counter productive to my end goals.

From what I see sketched out in the picture, the top hot/cold transfer would be a nightmare to produce, the lower bath section will neutralize what the top manages to produce, basically it won't work, but feel free to prove me wrong and if you do, I will admit it right here in front of the world.

There is so much wrong in that design that you may consider going back to the drawing board, at this point it is just an idea, I'm not attempting to offend you, just save you a lot of time and money.

 

[4Ryan6]

1. Does the TEC need the hot side kept hot and the cold side kept cold, or, does it want the hot side kept cooler and the cool side kept hotter?

The only thing the TEC or Peltier needs is voltage, when you apply voltage the 2 dissimilar metals react to each other with one side getting hot and the opposite side getting cold. It's actually a marvelous invention and discovery, it has no moving parts and as long as you don't crack it with uneven pressure burn it out, with too much voltage or allowing it to overheat, it's MTBF is amazing long, somewhere in excess of 100,000 hours. The peltier does not want anything it will continue to function as long as it is powered, the key is extracting either useable hot or cold from it, and how to go about it.

2. How hot/cold do these things actually get? I'm having some thoughts regarding materials for the chambers, and you mentioned you managed to get the heat sink so hot it burned you...

Peltiers come in various spec ranges regarding voltage requirements, amperage draw, wattage used, hot side maximum temperature, and the Delta Tmax.

These are the exact peltiers I am running.

Specs below are my peltiers specs.

Maximum operating temp: 125 C
Imax: 26 Amps
Qmax: 243.5 Watts
Vmax: 15.4 Volts
Delta Tmax: >68 (C)
Size: 50mm X 50mm X 3.10mm

In my earlier testing when I first encountered the peltier stalling I was using Ice water cooling which is covered in the Exploring Below Ambient Water Cooling thread linked in the first post of this thread. My light bulb in the brain pan came on, I had what I thought was a brilliant idea, since I already had ice cooling available what a perfect way to cool the hot side of the peltier!

Right?

Wrong! It stalled out the peltier and I got nothing useable!

So from that I learned the hard way that for the peltier to produce a useable outcome the hot side has to be allowed to get hot, there's something for your brain to chew on.

Then I was going to use standard radiator water cooling to cool the hot side of the peltier I used a Black Ice GTX 240 shrouded radiator to do it with with a XSPC Bay Res/Pump with the pump inside the Res, the water temperature got so hot flowing through the loop, the radiator burned my hand to touch it, the same with the pump, and today that pump is dead, it died shortly after that experiment.

I then proceeded to the highest quality heat pipe air cooling heat sink testing, which most at OCN tell me won't work but does work and that's what I use today.

3. If I just include a potentiameter in this circuit with the TEC, will that suffice to control the heat exchange? Ergo, lower the voltage and it'll lower the heat exchange?

You can simply control the useable heat exchange by the fan speed on the heat pipe air cooler, that supposedly doesn't work.

4. I already have a liquid cooler system on the CPU... its a cheap corsair one. And sucks. What if I just put the radiator from that into a contained space with the cold side of the TEC (and appropriate heat sink and fan)? The ambient temperature in the box would be pretty nippy... but not so cold as to be dangerous and probably pretty easily controlled... idk.

That sounds like Motos Chiller Box project.

 

[4Ryan6]

By the way, someone over on overclockers has been saying a lot of complicated things about deltaT, and says the radiator won't dump the heat down to ambient.

The Delta Tmax is the best you can hope for between the hot and cold side of the peltier, that sir is the sweet spot I am milking to get the useable cooling results I'm getting.

You can depend on that being a constant if you're supplying the exact spec'd voltage requirement of the peltier, but in my case supplying 3.4v DC less than the spec'd requirement of my peltier changes the other proposed operating parameters of the peltier.

It drops my amperage load down to close to 20amps, it drops my wattage used to 200w, (actual wall tested draw for my setup is 200w per peltier energized), it also changes my Delta Tmax range, which I don't know the actual but is probably dropped to maybe 55c from the spec'd 68c.

That's more than likely why I can use the heat pipe air coolers to cool the hot side because I am not running the maximum voltage to the peltiers, so I get useable cold production from the peltiers cold transfer to the modified flow water blocks by milking the Delta Tmax window.

Now you're not going to find milking the Delta Tmax window on Google, that's my terminology as the cooling I'm running in this thread is new tested and experimented with territory and is apparently Greek to those at OCN! (no offense to the Greeks intended, it's just a commonly used phrase.).

So I discovered that I could control the useable cold output from the peltier by controlling the fan speed on the heat pipe air cooler, with useable meaning producing more cold output than the heat output CPU was creating, countering the CPUs heat with more than needed to do so, which the cold overage is stored in the insulated reservoir, which drops the water temperature in the insulated reservoir.

So in my case and setup I am using the Delta Tmax window to my advantage.

Not everything regarding using peltiers is cut and dry, the math only works with solid input, variables change everything, that's where experimentation comes into play, and I am not through experimenting by a long shot.

My forward progress of course will be building from what I already know for a 100% solid fact works, to attempt to improve upon it!

Which I have already done some improvements by going to the same water blocks and heat pipe air coolers, and when I get the time, which is something I'm extremely short of these days, I'll update this thread.

I am presently not overclocked and with only one peltier energized, my water temperature is 10c below my ambient room temperature vs about 7.5c below ambient when overclocked to 4.5ghz. (not sure why I threw that in other than looking at my present water temperature readout in the reservoir.).

 

[threefingeredjack]

Hi Ryan, thegun,

Been reading that info thread you linked me to regarding the peltiers, excellent stuff. Going to go build a spreadsheet to do all the maths for me, should make this a bit easier, I'll send you a copy if you like, might come in useful with your continued experiments. PM me an email address and I'll send it through when I'm done.

Meanwhile, thanks for all the advice! Trying to get as close to expert opinion as possible (since nobody here seems to want to give themselves the title even though they deserve it!) in order to save myself some money was exactly why I was asking you guys. Been thinking on this thing more and more, once I get an idea into my head I don't drop it too quickly!

Regarding the air bubbles etc in the cold chamber... the drawing was literally just as simple as I can make on this (friend's) laptop, I'm away from my PC (the monster I want to chill) at the minute so no CAD I'm afraid. Plus, not everyone is familiar with engineering diagrams so I figured on mspaint doing the job. Anyhoo, yeah, the diagram is all hypothetical, I have a few ideas to ensure no air gaps in the lower chamber, and also to increase heat transfer in the upper chamber using a vortex induced by the flow, as well as a cowling system and heat pipe system to take away heat from the radiator and get it out the room.

I wasn't aware of the TEC cracking over time though, makes sense I just figured the type of insulation they'd use on them would take into account the high/low temperatures involved...you know, in its function? If you're saying that isn't the case then this design wants some additional thinking on... I'm thinking of putting a small heat sink on the hot side anyway before it enters the fluid, so maybe I'll build the chamber around this heat sink.

Regarding the pump stopping working, yeah I figured on all kinds of failsafes assuming something goes wrong with any component it would mean the TEC just boils over or the chambers freeze or the processor could be damaged. Wouldn't be too hard to build a simple transistor based safety into the circuit, killing the TEC, pump or PC if draw across components changed.

So the bottom line is, the hot chamber wants to be kept as close to its deltaT sweet spot temp as possible? This 55 degrees? In which case, I think I can redesign this, and I also think I know what needs to be automated and what can be controlled... just maintaining a constant voltage on the TEC seems inefficient, but you're probably right in it being the only way, however, if I can build this to automatically increase the pump speed/fan speed to ensure the hot side water is always 55 degrees, surely I can then manually control the TEC voltage to determine what I want the processor temperature to be? Maybe. TO THE MATHS!

I'll keep you posted... cheers guys!

 

[4Ryan6]

So the bottom line is, the hot chamber wants to be kept as close to its deltaT sweet spot temp as possible? This 55 degrees? In which case, I think I can redesign this, and I also think I know what needs to be automated and what can be controlled... just maintaining a constant voltage on the TEC seems inefficient, but you're probably right in it being the only way, however, if I can build this to automatically increase the pump speed/fan speed to ensure the hot side water is always 55 degrees, surely I can then manually control the TEC voltage to determine what I want the processor temperature to be? Maybe. TO THE MATHS!

No! That is not what the 55c meant when I was referencing my possible Tmax.

If you look at my peltier specs the max heat is 125c, that is the point if reached of peltier failure or burn out, so you cannot actually run the peltier at 125c, but if you could with a 68c Tmax your cold side assumed temperature would be 125c - 68cTmax = 57c cold side temperature.

If I supply the full 15.4v to my peltier the 68cTmax does not change so if I wanted to run my actual raw peltier temperature on the cold side at 0c, theoretically I would need a hot side temperature of 68c or 154.4f, that 68c range is the useable window I was referring to not a proposed temperature for your top side hot projection.

Since I am supplying only 12v to my peltier and estimating my new Delta Tmax because the voltage is below spec., my Delta Tmax window is smaller than 68c so I estimated it being approximately 55c, meaning if I want to run a 10c water temperature, I have to use the lower end of the window to give me CPU heat countering water cold side temperature lower than needed to counter the heat from the CPU.

So actually I'm running a hot side temperature around 130f or 55c which yields close to 0c at the peltier itself.

That probably just confused you but is actually what's happening with my setup, I hope that clears the 55c misconception?

The Delta Tmax under perfect spec'd voltage used, is the difference between what the the hot side is allowed to reach and how cold the cold side actually gets, of the raw peltier.

The reason I said raw peltier, is other things affect the peltier, the thickness of the copper on the cold side pick up has a direct relation to whether the cold is transferred fast or slow and whether it will cause the peltier to stall out.

The reason the modified flow water blocks work is the copper base is only 3mm thick or just shy of 1/8th of an inch, with the constant water flowing through the block it transfers the cold from the raw peltier through the copper to the water fast.

Which effectively keeps the cold side from getting too cold and overriding the heat of the hot side, which causes the peltier to stall, in my setup the peltier does not stall, it keeps constantly outputting cold.

And that sir, is why it works!

 

[thequn]

I went "shopping" for a Pelter Yesterday I checked 32 stores - 6 computer shops 7 Radio-shacks 2 fry's 2 best buys 1 toys are us 3 home depots 2 lows 1 harbor freight 5 indie shops 3 music stores and the La conversion center. No one even knew what a TEC, pleter, Thermoelectric cooler was, closest response was a 3.5g bottle of Arctic sliver 5... LOLOLOLOL Thank god for the internet and shipping companys cant imgain how people use to get things they wanted 20 years ago...

All in all spent 68 dollars in gas drove about 300 miles and honestly it was fun and I had a good time.

 

[4Ryan6]

I went "shopping" for a Pelter Yesterday I checked 32 stores - 6 computer shops 7 Radio-shacks 2 fry's 2 best buys 1 toys are us 3 home depots 2 lows 1 harbor freight 5 indie shops 3 music stores and the La conversion center. No one even knew what a TEC, pleter, Thermoelectric cooler was, closest response was a 3.5g bottle of Arctic sliver 5... LOLOLOLOL Thank god for the internet and shipping companys cant imgain how people use to get things they wanted 20 years ago...

All in all spent 68 dollars in gas drove about 300 miles and honestly it was fun and I had a good time.

Is this a joke? :lol:

I would have loved to have seen the faces at Best Buy when you asked for a peltier! ROFLMAO :lol:

Most times now when the Geek Squad see me walk in they go on break! :lol:

I had a poor little Best Buy newbie come stumbling over himself to assist me one day, took about 30 seconds of conversation for him to realize he didn't know squat!

Two minutes later he was ready to run from me and was wondering why he had ever walked up to me! :lol:

Best Buy, the Idiot Squad! ROFLMAO :lol:

You could have bought a peltier and paid for shipping in the gasoline you used on that adventure, of course you've now learned that the hard way.

http://www.shop.customthermoelectri...14A2EF1DC40DAFC60B1335B961C477.m1plqscsfapp03

FrozenCPU is very limited to what they carry they only have 2 models the 245watt linked below that I am using and a 80watt.

http://www.frozencpu.com/products/2408/exp-01/245W_Potted_Peltier.html

 

[thequn]

LOL,,, yes. I knew going out before hand that I was not likely to find one. I was out shopping for other things and wanted to make a day of it. but i thought it was funny how every place I went to every one 100% knew what I was looking for....... only for them to come back with something ridiculous....

 

[threefingeredjack]

Firstly, lol yeah I was in town the other day and asked in a couple of PC shops... I got some blank looks regarding TEC plates.

Secondly, to Ryan, no that makes sense now, I get the deltaT thing and how keeping the hotside within the window is the important part of the TEC magic trick.

I've had a play with the spreadsheet to map the possible temperature range and the amount of wattage-dispersal required by a cooling radiator or heatsink in order to keep the hot side within range. I still can't see a reason why the second design i posted wouldn't work, it would effectively enable you to run a hot and cold loop and just use the cold loop for cooling and the hot loop for heating...something else. Coffee say.

The only issue I'm having is working out how the heat is transferred and in what quantity from the CPU to the cold line, and what impact changing the pump speed would have. I'd like to balance the system if I can, ensuring neither the TEC nor the pump are doing more work that needed. Rather than just run the fans at different speeds to compensate for over-abundant heat production from the TEC.

SO! I'm just trying to get my head around this whole Q=CMT something or other equation and how that relates to fluid flowing through a waterblock... I'll get there. PM me both your emails and I'll send it through when I'm done. Might help others build such a system as it details the amount of wattage radiators need to disperse in order to stay on top of the TEC heating + processor. On the plus side, the hotter the radiator is, the faster the heat disperses from it, so adding the CPU heat to the TEC heat actually makes the radiators more efficient, thus the pump/fans can be run slower... which is pretty ace. I just want to balance the whole damn thing somehow...

 

[4Ryan6]

@ threefingeredjack

Unfortunately no matter how much planning, or ideas, or designs, the only way to know if your design is a viable cooling solution is to build it and test it.

 

[4Ryan6]

Note to anyone considering duplicating my setup:

The Swiftech Apogee XTL is the best block for flow modifying to use with a 50mm x 50mm size peltier, I'm recommending it over the XSPC Rasa, the Rasa will still work, but not as well as the Apogee XTL.

Due to the XTLs larger copper base plate the entire 50mm x 50mm peltier fits in solid contact fully covering the contacting face, whereas the Rasa base plate is exactly the same size as the peltier, requiring filling the screw indents with thermal compound.

The Apogee XTL is exactly the same design of the original Apogee XT, but the top is machined from black delrin, which is much easier to cut and modify, and is also cheaper, the copper base is wider allowing further coverage and cold pickup of a 50mm x 50mm peltier.

I must start this next comment with apologies as my camera is malfunctioning so I have no pictures to show of the modifications I did to the XTL to increase its flow, but it works very well.

I additionally had a HDD failure which cost me every picture I had taken that had not been backed up, we tend to take things for granted since things seem to be running so smoothly, but that failure beast is still out there, make sure you backup regularly what you do not want to loose!

I will be taking new pictures of the setup as it sits now and replacing the pictures in the beginning of this thread once I either discover what's wrong with the camera or use another to take them.

PerformancePCs is one of the only places you can still acquire the Thermalright TRUE without the fans, which is great if you already have the fans you need.

They also have an excellent price on the XSPC Rasa, if that is the block you prefer.

 

[4Ryan6]

I modded my Rasa Water Block much more than I had remembered!

I remember starting with the wagon wheel design, but had the block begin to freeze up, not to the point of splitting it open, but to the point of freezing up enough to choke off the water flow, so I obviously modded it again.

This is the end result and what I've been running.

My apologies for the pictures!

My digital camera is giving me a fit presently, I took about 20 pictures just to get these 2 and I'm not sure why it's happening?

This camera has been fantastic up to this point I fear I may have to replace it.

Major changes from the original the first step protrusion was completely removed.

 

[kauedg]

I received my TECs two days ago. One day ago I fried a lousy PSU while testing. Good, one extra fan, extra wires and connectors! Got -22C tho... Waiting for the appropriate PSU (12V ~ 20A) to arrive now.

 

[thequn]

Wow Ryan,
lol
its smooth did you sand it down a bit? or did the water flow do that? also how far did you lap the copper plate?

 

[4Ryan6]

Wow Ryan,
lol
its smooth did you sand it down a bit? or did the water flow do that? also how far did you lap the copper plate?

Used the small Dremel polishing wheel to smooth it over, takes light touch so you don't melt the acetal.

Copper base plate was not lapped, filled the screw head indents with thermal compound as the Rasa copper base plate is exactly 50mm x 50mm.

If you use a Swiftech XTL water block it's large enough so there are no screw hole indents to fill, using a 50mm x 50mm peltier.

 

[4Ryan6]

I received my TECs two days ago. One day ago I fried a lousy PSU while testing. Good, one extra fan, extra wires and connectors! Got -22C tho... Waiting for the appropriate PSU (12V ~ 20A) to arrive now.

What are the peltier specs and size?

 

[threefingeredjack]

So!

Been a while, I've been busy with many things so haven't dedicated all that much time to this TEC concept.

I've been trying to get my head around heat transfer rates and its interaction with the actual and rectified temperature of the CPU. In theory, my system should work just fine...as long as I only wanted about 25 degrees on the CPU... Hardly seems worth it. I'm struggling to work out how in your (Ryan's) system you managed to get these low temps.

In my calculations it works out that the water block on the CPU, running a pump at 5 L/sec, pumping 5 degree water across a copper plate manages a heat transfer rate of about 200w.

My issue: I have no idea what this means. No, rather, I understand this to mean it can transfer heat away at a rate of 0.105330775 degrees per second. So it basically drops the temp by 1 degree every ten secs. Great right? Provided no additional heat is being added by the CPU, then yeah. However, it probably will be, so, is the main problem that I need to work out the Watts of the CPU? If so Wikipedia says mine does 95W.

I don't think there is a way to calculate all this in 2 dimensions. I need some sort of bloody CFD system or something. I think I can start to see why you've got 2 TEC in your system.

Might build it all when I've got some cash together. I'll keep the forum posted if I do, and if there is anyone who wants to take a look at this damn spreadsheet and let me know where I'm going wrong I would be willing to throw a parade for you.

 

[4Ryan6]

So!

Been a while, I've been busy with many things so haven't dedicated all that much time to this TEC concept.

I've been trying to get my head around heat transfer rates and its interaction with the actual and rectified temperature of the CPU. In theory, my system should work just fine...as long as I only wanted about 25 degrees on the CPU... Hardly seems worth it. I'm struggling to work out how in your (Ryan's) system you managed to get these low temps.

In my calculations it works out that the water block on the CPU, running a pump at 5 L/sec, pumping 5 degree water across a copper plate manages a heat transfer rate of about 200w.

My issue: I have no idea what this means. No, rather, I understand this to mean it can transfer heat away at a rate of 0.105330775 degrees per second. So it basically drops the temp by 1 degree every ten secs. Great right? Provided no additional heat is being added by the CPU, then yeah. However, it probably will be, so, is the main problem that I need to work out the Watts of the CPU? If so Wikipedia says mine does 95W.

I don't think there is a way to calculate all this in 2 dimensions. I need some sort of bloody CFD system or something. I think I can start to see why you've got 2 TEC in your system.

Might build it all when I've got some cash together. I'll keep the forum posted if I do, and if there is anyone who wants to take a look at this damn spreadsheet and let me know where I'm going wrong I would be willing to throw a parade for you.

You have enough information in your calculating I'm not sure why you can't get your head around it.

copper plate manages a heat transfer rate of about 200w

Watts of the CPU? If so Wikipedia says mine does 95w

So 95w of CPU produced heat is subtracted from the 200w which = 105w of cooling produced over the CPUs produced 95w of heat, that 105w cooling overage is what is stored in the insulated reservoir.

That 95w CPU heat load increases the higher you overclock the CPU when you increase the CPUs operating voltage, when the 95w increases, it takes more away from the 200w constant.

Even in the CPUs highest overclock max heat range it still does not counter the 200w produced cooling it may knock the overage down to maybe absolutely worst case possibly 90w, but that overage still builds on the cold in the reservoir.

However, Here's what most seem to miss, the theoretical 105w cooling overage going to the insulated reservoir is not only countering the CPUs heat output, but at the same time is constantly dropping the temperature of the water in the reservoir.

As that reservoir water temperature drops, it improves the overall cooling for even better performance.

I think I can start to see why you've got 2 TEC in your system.

The 2nd TEC is mainly for high end gaming it allows a constant gaming temperature around 10c water temperature, and also for running stress tests to stabilize higher overclocks.

 

[4Ryan6]

There are many various applications of peltier use!

From direct CPU mounting, to chilling air in a cooler, they can be used to cool, and even be used to heat something that you absolutely do not want to freeze like for instance, a control valve, if you artificially supply the hot and cold to the peltier you get electricity from the leads in return.

The simple key to using the peltiers capabilities is getting what you specifically want from it, and how to go about it.

That is what this thread is all about, one method of acquiring a useable cold output from the peltier, that can be harnessed and stored and used to cool a CPU below ambient room temperature, without condensation worries, if you choose to run the setup above the moisture forming water temperature range.

If you choose to run below the moisture forming water temperature range, you will have to run an anti-freeze coolant, and moisture insulate the motherboard, in and around the CPU socket.

My choice is running above the moisture forming, with no worries, but this cooling can and will go below zero, if you do not control what it is allowed to do.

I have been blasted up one side and down the other, and grilled by some, that have gone to great detail to explain to me why what I am presenting in this thread will not work.

Yet it is now crossing the threshold of 9 months in operation, so for those determined to explain to me in detail, either in PMs, or at other websites, put 9 months in operation in your thought processes and explain to me what you do not understand about it!

And while you all wallow in your doubts, and you know who you are, I'm successfully running this cooling that obviously does work very well!

I am not a practicing liar!

If I tell you this works, then by God you can count on it as 100% solid fact!

If I am supplying false information, that is a ban-able offense at THGF, and as a retired moderator, I know the rules of this forum better than most of the active moderators of today do.

So if any THGF administration wants to come to my home and see this cooling in operation, and verify these results my door is wide open to you!

This is a ground breaking cooling solution, and it is important IMO, that it was discovered and successfully put into operation by a THGF forum member.

Ryan

 

[threefingeredjack]

There are many various applications of peltier use!

From direct CPU mounting, to chilling air in a cooler, they can be used to cool, and even be used to heat something that you absolutely do not want to freeze like for instance, a control valve, if you artificially supply the hot and cold to the peltier you get electricity from the leads in return.

The simple key to using the peltiers capabilities is getting what you specifically want from it, and how to go about it.

That is what this thread is all about, one method of acquiring a useable cold output from the peltier, that can be harnessed and stored and used to cool a CPU below ambient room temperature, without condensation worries, if you choose to run the setup above the moisture forming water temperature range.

If you choose to run below the moisture forming water temperature range, you will have to run an anti-freeze coolant, and moisture insulate the motherboard, in and around the CPU socket.

My choice is running above the moisture forming, with no worries, but this cooling can and will go below zero, if you do not control what it is allowed to do.

I have been blasted up one side and down the other, and grilled by some, that have gone to great detail to explain to me why what I am presenting in this thread will not work.

Yet it is now crossing the threshold of 9 months in operation, so for those determined to explain to me in detail, either in PMs, or at other websites, put 9 months in operation in your thought processes and explain to me what you do not understand about it!

And while you all wallow in your doubts, and you know who you are, I'm successfully running this cooling that obviously does work very well!

I am not a practicing liar!

If I tell you this works, then by God you can count on it as 100% solid fact!

If I am supplying false information, that is a ban-able offense at THGF, and as a retired moderator, I know the rules of this forum better than most of the active moderators of today do.

So if any THGF administration wants to come to my home and see this cooling in operation, and verify these results my door is wide open to you!

This is a ground breaking cooling solution, and it is important IMO, that it was discovered and successfully put into operation by a THGF forum member.

Ryan

Like a senior management official.

So I'm liking the potential. I'll begin building asap and get back to you. My personal thoughts regarding sub-condensation temps are that I might submerge the MOBO in mineral oil. I've seen it done, looks cool as hell. Again, something which can be done relatively cheaply and would have the overall added bonus of being able to cool the surrounding fluid bringing the RAM and GPU temps down also. (Or, be more of an issue with the added temps of each) That 95w goes up to 105 overclocked, and was stated as a max wattage rather than continuous output. So I think I'll be ok... think I'm going to have to find a peltier of similar spec to yours or bigger though if I go mineral oil.

Cheers for all your advice! You are a pioneer in the field! If my system all goes wrong, I'll come back for a technical diagram and maybe we can put together a product of some sort. I've seen a lot of interest in these peltier systems, considered holy grails of CPU cooling.

Cudos!