TEC/Peltier CPU Chilled Water Cooling

Page 11 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
Wow that is very cool. No pun intended :) I thought I had a good idea but this is really neat. When I was checking the official site for the Delta fans I came across this thing (Thermo Electric Cooler) Check this thing out!

http://www.deltaww.com/Products/CategoryListT1.aspx?CID=0306&PID=503&hl=en-US&Name=Thermo-Electronic%20Cooling%20Unit

If you could find a way to stick inside the case to lower ambient temps real low to cool the rad and everything. Not sure how much one of these guys would cost though or where to find it other than calling Delta. It's not all that big either.
 


Hey Welcome to the thread! :)

The problem with that assembly when it comes to computer related cooling is the heat it exhausts, to gain any useable cooling the heat has to go somewhere outside of the case.

You'd get about the same results if you ducted a room air conditioner to supply direct cold air into the case.

(Been there done that!, and it didn't work either.) :)

Mainly because of the compressor cycling in the AC unit, did not deliver a constant dependable useable cold, the temperature fluctuated.

You would get a constant cold with what you discovered but I think the end cooling results still relying on air would be seriously disappointing vs what that thing probably costs.

The cooling this thread covers is 100% in operation right now and has been in operation well over 18 months.

It is not the kind of cooling for everyone, that's for sure!

Mostly for those just flat sick of the one size fits all water cooling of today, that is adequate for most that have invested a chunk of dough in traditional radiator water cooling and basically settling for their end results.

When it came to their water cooling investment overclocking results, it really did not bring home the prom queen!

There's nothing worse that investing a serious chunk of money into computer cooling then finding out it does not reach the overclocking goals you had. (That's a serious gut punch!)

(Also been there done that!)

This cooling falls into the unknown category, because it was discovered outside the major brand named water cooling companies and most fear the unknown, that's why I've been running it, to prove it can stand the test of time.

ArthurH is running this cooling as well, but he never comments on it's capabilities.

Once again welcome to the thread and I hope I didn't just run you off.

So My 2 cents! Ryan
 
no no can't run me off. I can appreciate an honest opinion and someone who has hooked an AC to their computer is not someone to be reckoned with 😉 I only posted that because I was looking at that thing before and not knowing any of this and it in essence is the same technology except yours is a Hydra TEC. I can totally see how it is a better solution than cooling ambient. Your CPU-Z tag is just evidence of your success. I think this bridges the gap between Radiators and Liquid Nitrogen. How much and where on earth did you get those TEC plates?
 
These are the actual peltiers I am running, bought them from FrozenCPU.

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

I made my aluminum cold plate adapters out of 1/4" thick x 2" wide, flat aluminum stock I snagged from a local hardware store.

The heat sinks are Thermalright TRUE and the modified flow water blocks are Swiftech Apogee XTL. (Not sure if those XTL blocks are still attainable but if you decide you want to undertake this type of cooling I'll track down a good recommendation for you, just let me know?)

Some of the Swiftech water blocks are great for a 50mm X 50mm peltier as they do not need an adapter plate and get full coverage across the entire peltier face.

You can actually buy a 2" X 2" X 1/4" aluminum cold plate today already machined, as there is a lot more available today than was available when I started my project.

http://customthermoelectric.com/accessories.html#coldplates

The thickness of the cold plate used on the hot side of the peltier can actually be a thermal control factor when it comes to the balance of the heat sinks fan speed vs the peak temperature the peltier is allowed to reach. (That statement will make zero sense to you until you actually have an assembly to start testing with as fan speed on the heat sink can and will affect the peltiers cold side output.)

Meaning it will allow you to manipulate the > 68c Delta Tmax window, which at 12v DC supplied voltage is more like a 50c ~ 55c Delta Tmax window, because the 245watt peltier is designed to run at 15.4v DC.

I chose to power my peltiers from a single 12v rail PC power supply mainly for the sheer convenience of, I already had one on hand! :)

So the peltier is undervolted at 12v DC, meaning the 245w peltier I am using is only actually 200w (Measured from wall current draw) and instead of the peltier drawing it's spec'd 26a at 15.4v DC it is more like a 20a draw.

Now that you are thoroughly confused, have a great day! :)
 
# hi guys just picked up the tread here dont know whats been said yet, so have you consider cooling the mobo from the other side also like using a cooling plate or rack cooler! I have had some ideas on chilled water cooling if you removed all your air and moisture out of your case and had a sealed sistem you wouldn't get bottle sweat i think?
 




So what have you actually done or are you in the idea phase?
 


# cool its that you machine are you using a solid .999 fine silver Heatsink plate, Ok so I'm think of adding an instrument cooler to the bottom of my mobo i currently gonna build my own if i can be bothered, but my point is about increasing surface area in pipes on pipes increasing the heat exchange efficiency, so basically improving you radiator pipes and stuff, in industry for exampl you can buy something call a spot cooler, it runs on an air compressor and can maintain/produce temperatures of -15c but generally buy most industrial equipment isn't suitable for home computer use, increase the available surface area allows cooling quicker cooling,

But I'm talking if you using metal pipes, if you could find. A way of adding a heat sink to your pipes and then some way of cooling them, it would reduce your temp, lol i have been looking at Silver and 9ct gold Tubing.

I suppose it you wrapped pure copper wire around your pipes and put a fan blowing on them you liquid will be colder for longer.
 


I run clear flexible vinyl tubing that is also run inside 1/2" wall thickness foam pipe insulation, I run the vinyl because it is cheaper and is all I need to get the job done!

I would not even consider running metal tubing, maybe hard acrylic one day, but I do not want anymore metal in the loop than I absolutely have to have.

I suppose it you wrapped pure copper wire around your pipes and put a fan blowing on them you liquid will be colder for longer.
Click to expand...

Actually without pipe insulation that would warm the water, toward ambient room temperature, basically the same exact effect of putting a radiator in a chilled water loop, it has the complete opposite effect and begins warming the chilled water.



 
So i haven't ever really been able to follow some aspects of this very well. Are you saying the more you cool the hot side of the peltier, the colder the cool side gets or the hotter the hot side wants to be? I'm confused as to how you regulate how hot or cold either side gets, is it just a function of how much power you put into it? Also am I correct in thinking that what you're doing is basically using the peltiers to cool water in a separate box and then routing that to another loop?
 


# yeah i suppose but when I wrote that I was thinking about fan less systems and external radiators, for example running tubes through the PC out the other side into a chiller box, and not using water as such, but running water thought gallery's adding more surface to chill the water faster, that's why I was on about silver and gold tubing expensive but would be great looking.
 


# well i thought if you cool the rear of you entire motherboard with and instrument cooler then you could bring down motherboard temperature thus the system temp! back in the early days most CPUs did not require a heat sink as heat dissipation into the circuit board was sufficient, later they added heat sinks and fans i have seen old system where they only have a Heatsink no fan at all!

So the comments were slightly off topic but about cooling in general, (motherboard temperature!)

Well they might have had heat sinks but didn't need CPU mounted fan systems

 


I sounds like you don't grasp the common workings of the peltier and it's basic capabilities, so for getting a good foundation of information regarding peltiers read this. (Note: Peltier and TEC means the same thing.)

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

The quote below is taken from eweast's guide linked above.

When a low voltage DC current is applied to a thermoelectric module, heat can be moved through the TEC from one side to the other. One side is therefore cooled [cold-side] while the opposite side is simultaneously heated [hot-side]. The temperature differences can be upwards 50C in real world application. TECs help enhance your cooling ability by creating a temperature differential that can be more easily moved out of the system. Water-cooling systems can only cool an object to ambient temperatures (room temp), but they still have excess cooling capacity (provided they have sufficient flow-rate and a capable radiator). TECs allow more of the cooling capacity to be utilized and therefore achieve lesser-than-ambient temperatures.
Click to expand...

The 50c temperature difference he makes reference to is what I manipulate to get the 10c actual water temperature I run, which the lower temperature is stored in an insulated reservoir, the temperature storage begins immediately after the peltiers or TECs are energized.

My system is looped like this; Insulated Reservoir, > Pump, > CPU, > TEC Assembly 1, > TEC Assembly 2, > Insulated Reservoir. The Insulated Reservoir, Pump, and both TEC Assemblies are housed in an AZZA 1000 case, the insulated water lines are the only water cooling elements entering into my main computer going direct to the CPU water block.

The water in the insulated reservoir after the machine has been shut down for some time actually warms to ambient room temperature because the insulated reservoir cannot hold the cold indefinitely, without the TECs creating the cold temperatures, this has turned out to be a benefit as any cold boot problems that were experienced when I was using ice in a cooler to run below ambient temperatures are completely gone.

The difference between the peltier chilled water cooling and traditional radiator water cooling is when the radiator cooling is powered the water temperature rises above ambient room temperature and eventually reaches a point of temperature equilibrium which will always be above ambient, that water temperature above ambient is your deltaT.

With radiator water cooling the only control over deltaT is more radiator cooling field to drop the deltaT as close to ambient room temperature as possible, but it will still be above ambient.

When the peltier chilled water cooling is powered the water temperature starts dropping below ambient and begins chilling the water mass in the insulated reservoir, which when the TEC assembly is fine tuned and hot and cold balanced, the chilled water system also reaches a point of equilibrium, but it is a negative deltaT and usually around 15c below ambient room temperature.

Regarding the bold underlined above:

It took a lot of trial and error to get the cold output from my TEC assemblies pumping out useable cold to store in the reservoir, however if this cooling setup is duplicated that this thread covers, you will get the exact same results I am getting, it has already been duplicated by another forum member, not an exact duplication in his component layout but the same component duplication, and he gets almost identical cooling results.

Does that help your understanding?

Edit: Your sig download speed results make me sick! :)
They stomp my results in the dirt as I live out of the city in the country and I have a satellite internet connection and my ping is over 1,000, and I am not joking!
You have some killer download speed! :pfff: :)



 
 


I was really more questioning how you regulate the temperature of the cold side of the peltier.
Click to expand...

By controlling the temperature of the hot side.

Like does it have a constant draw regardless of how many volts you plug into it or do you have to adjust that manually?
Click to expand...

Yes it has a constant voltage, wattage, and amperage, draw unless you add some kind of a voltage regulator.

I didn't really follow what you were saying when you mentioned in the original post about using water to cool the hot side and burning your hand on the pump.
Click to expand...

That was the radiator was too hot to hold your hand on, and the pump was an inside the reservoir model and burned out.

That gave me the impression that the more you cool the hot side the hotter it will get and by comparison the cold side will get cooler.
Click to expand...

How could the hot side get hotter the more you cool it?

Wouldn't the hot side get cooler the more you cooled it?

****************************************************************

The peltier itself is a very simple device that can actually do three actions depending on what you use it for, you can cool with it, you can heat with it, or you can get electricity from it.

When power is applied to the peltier one side gets hot and the opposite side gets cold, if you do not control the hot side heat and allow it to overheat past it's specs, the peltier will simply get too hot and burn out. Your Bad!

So somehow you have to control the peltiers hot side heat so it does not get to spec burn out, the hot side can be controlled numerous ways, many told me it flat out could not be done with an air cooler, but they were wrong!

The link below is the peltiers I am running.

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

These are the 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

The maximum peltier operating temperature is 125 C that is basically the peltiers burn out point if it is allowed to reach that temperature, so you probably think that maybe 100 C would be a safe peltier operating range well away from it's burn out point, Right?

Well there's only one major problem with that thinking and that is 100 C is waters boiling point, so if you plan to cool the hot side of the peltier with water cooling, you will have to operate the peltier below 100 C, because you do not want 100 C even possibly building in your water cooling loop, because the system could build steam and explode as none of the components are designed for that kind of temperatures, pump, tubing, etc.

Is the veil of understanding beginning to lift for you, I would like to know before I spend the rest of the morning typing and you still don't understand?

So please help me, help your understanding?

Edit: I mentioned you could get electricity from the peltier, and if you artificially supply the hot and cold to the peltier, the electrical leads will read a voltage output.

For the Record: This system has been in operation now for 20 months.







 
I guess where I got lost was you made it sound in your original post as if the water wasn't cooling the hot side effectively enough because it was getting hot enough to burn your pump which seems strange to me because you're doing ok with just a fan. Water should be more efficient as far as moving heat to the air I would think. If the peltier is putting out a consistent amount of heat shouldn't the water cooling cool it better than just a fan? Other wise I just have to accept that it's PFM (pure f'ing magic) and believe a fan cools better than water does which I know can't be true.

Like I get what you're saying about how the water is going to get really hot especially over time, but is that with just a single radiator of a specific size with a fan at a specific speed or did you try different radiators and fans/speeds? It seems like at some point the water should reach a point of equilibrium where it is dissipating as much heat as the peltier is putting in. Also with the fan what is the normal operating temperature of the peltier? I totally understand what you're saying if you have fan that is keeping the peltier at like 110 or something so that it's far enough from the max temperature because yea that makes way more sense than having water that is so hot it breaks components in the loop.

But other than all that yea the rest makes sense now, but just for clarity the cold side will always produce the same amount of cooling and hot side will always produce the same amount of heat right?
 


I don't respond well when this type of "PFM (pure f'ing magic)", attitude comes across, because obviously you do not realize there are some air coolers that do out perform some water coolers, and magic has nothing to do with it!

http://www.tomshardware.com/forum/id-2196038/air-cooling-water-cooling-things.html


 
No disrespect intended, I'm a submariner and as you might expect we have a lot of complex systems that all work together. There are things that much smarter men than me designed and when it comes to the nuts and bolts of it I just have to assume it's PFM because I just can't understand it without digging into it and actually putting my hands on it. So that's just a regular term in my field of work is all. Plus my demeanor can sometimes come across the wrong way so, I apologize if that's the case.

I understand that some CLC coolers are junk and something like a EVO 212 will certainly beat those but I meant a custom water cooling loop. Have you ever tried that or was the CLC cooler you used the only thing you tried as far as water cooling? I'm willing to experiment on my own, but so far you're the only expert that's been there and done that with this. I just want to know how far you've gone with attempting water cooling for your tec or if you just tried one CLC and decided it was easier to just use a fan cooler.
 


OK I understand where you're coming from with the comment, I was Navy as well but on a surface ship an old Destroyer the Richard E. Kraus DD849 long since retired from service.

Cooling the hot side with water the very first attempt I made was attempting to cool it with chilled ice water.

All that was covered about 2/3rds the way through to the end of the "Exploring Below Ambient Water Cooling Thread", which has a link to it at the beginning of this thread.

Since I already had a 54qt cooler filled with ice and water I thought it would be the perfect thing to cool the hot side of the peltier because my thinking at the time was keeping the hot side cooler was going to give me much lower cold side.

That was before I learned that for the peltiers Hot/Cold reaction to properly take place the hot side had to be allowed to get hot, the chilled water had the hot side of the peltier too cold and it caused the peltier to stall out.

So cooling the hot side of the peltier with chilled ice water quickly was a discarded idea, that seemed super logical to me to be the perfect way to cool the hot side, but was too aggressive and did not work.

What I am referring to was in the very beginning when I first got my peltier in hand and began experimenting with it, I do not recall if I even wrote about that discovery in the other thread?

Before going any further you need to understand that if a peltiers hot side gets too hot it will affect the cold side being able to get cold enough and the peltier will stall and literally produce nothing useable, in the opposite direction if the cold side gets too cold it will lower the hot side temperature and also stall the peltier.

That's the finicky part of peltier cooling you have to find the sweet spot, where the hot and cold are in a balance and for this type of chilled water cooling the cold side is constantly producing cold, with zero stalling.

Once I discovered that an air cooler could actually do this, it was just a no brainer to continue down that road of testing leading to the results I have today.

Some of this has already been covered in the beginning of this thread.

Logically it would seem the best solution would be to keep the hot side as cold as possible, but that's not the case, you have to allow the hot side to get hot for the cold side to get cold, but since we're only talking a approximately 1/8th inch thick plating, that's a very close distance to play the hot/cold balancing act.

Some of what needs to be shared is the peltiers what I call weirdness, I originally thought the cooler I kept the hot side the cooler the cold side would be, however the hot side has to get hot for the cold side to get cold, it is a perfect example of, "For every action there is an equal and opposite reaction.".

There was quite a bit of testing and fan changing to get the peltier to run at it's optimum cooling side capability, at one point I actually had the hot side too cool and it was not performing as expected, it's been a learning experience that's for sure.

Simple discovery the hot side can override the cold sides effectiveness but news flash the opposite can also happen, the cold side getting too cold, can override the hot sides effectiveness,

So you're after a balancing act of hot and cold to get what you're after, the true key to this things performance is give it the heat range it operates best within or sweetspot, and it will give you the results you're looking for. It needs to be in the mid range of it's capabilities, so it's hot enough to deliver the cooling performance, but far enough away from it's limitations to keep from burning it out.
Click to expand...

 
Oh I see now, so it does have a natural sort of equilibrium that it needs to try and maintain so it's not really about cooling the hot side as much as possible, you just need to cool it enough to prevent it from burning out. That all makes perfect sense now. I think it's time for me to start piecing some things together and see what I can come up with. Is it a particularly bad idea to make a small test loop with only the tec water block and a reservoir in it? And do you have any thumb rules as to how much wattage a single 245w tec should be able to handle?
 


Specification wise numbers change regarding the 15.4 Volt DC, 245w TEC, when it is only supplied with 12v DC, the voltage acquired from a single 12v rail PC style power supply.

Imax: 26 Amps changes to approximately 20 Amp.
Qmax: 243.5 Watts changes to 200 Watts actual wall outlet power.
Vmax: 15.4 Volts supplied with 12 Volts from a PC power supply
Delta Tmax: >68 (C) changes to approximately 50 (C)



 
Hello Davil, I have duplicated Ryans cooling system on my PC and get exactly the same results he says it can do. So a for what it is worth it does work and very well! Been up and running for about 6 or more months now with no problems what so ever. Best home built cooling system for the money as far as I'm concerned.

Good luck to you.
 
Ok so I'm playing with some 6 gauge bare copper wire while I wait for my tec to show up and I was thinking about different ways to use it for cooling. I was considering a different idea as far as how to use the peltier. Instead of cooling the water directly using the peltier via a cold plate, what if I do something that's more like the way heat pipes work on coolers. Except of course it would be more like a cold pipe. I feel like if I coil up a bunch of this wire and run a lead out to the cold side of the peltier it could be possible a more elegant solution drawing to come..

Alright here is a quick 3D mockup i put together to illustrate my idea.
www.davidsirritation.com/Holdplateoff.gif

Pretty much the idea would be to place the peltier on top of the case then route the coil through a tube opening on the back of the case and then coil it on the inside of the reservoir to transmit the heat. I did some experimenting with a similar setup using a glass of ice water in place of the peltier and it did in fact transfer heat along the wire but requires some insulation to extend how far the cold will go along the wire.
 
The heat pipe thermal principle works because of fluid inside the heat pipes using a wicking action to constantly transfer the hot/cold temperature exchange.

The hot collected at the base plate wicks toward the end of the pipes where the cooling fins are located, the liquid is cooled and drops back to the base plate to start the hot collection action over, it is a constant transfer action.

I don't think your bare wire idea will yield useable results, but many told me my ideas that are 100% operational wouldn't work either so good luck and best to you, and let us know what happens?
 
Yea I may have to re-watch some bill nye videos since I'm a bit rusty on the physics of heat. But I believe hot and cold aren't really plus or minus energy, just varying degrees of it. So the wire should just want to be at the ambient temperature regardless of hot or cold. It should behave the same way it does with heat, which is not as efficient. The heat goes along the copper in a cooler and radiates as it goes which dissipates it quicker. I want to actually contain the cold on the copper and channel it so that means I need some kind of insulator to wrap around the copper which will prevent condensation and channel the cold. Getting some kind of styrofoam or polystyrene wrap is my next step for testing.
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom