No announcement yet.

Need opinion on TEC water Chiller??

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Need opinion on TEC water Chiller??

    My idea is to build a 4 stage TEC cooled water chiller to use in my PC. I do not care how much power it draws and how inefficient it is the object is to cool the water to whatever temperature I want or can get it to hopefully at least 5 to 10 degrees C. This will be controlled so that I only use however many stages I need depending on the water temps. Something like above 5 degrees C stage one will be one, stage 2 will come on at 10 degrees C, and stage 3 will come on at 15 degrees C and stage 4 at 20 degrees C. This is just an example and will need to work out the temp settings depending on dew point and level of insulation I use. </SPAN>

    What I plan on doing is like the Attached Drawing.</SPAN>
    Each block will have 1 40x40 TEC like the below.</SPAN>

    Imax (Amps) 15.5/ Qmax (Watts) 225/ Vmax (Volts) 24.0/ DTmax (°C) 64°C /Tmax (C) 125 C</SPAN>

    So at 12 volts this will be about 8 to 9 amp draw I think. This will require testing to verify.</SPAN>

    I will be using one of the below water blocks on each side of the TEC</SPAN> The water blocks will be purchased from Custom Thermoelectric at http://www.shop.customthermoelectric...4.qscstrfrnt02

    All copper water block Liquid exchanger to rapidly remove heat from the hot side of peltier TECs or TEGs. Most efficient method of removing heat from a thermoelectric TEC or TEG. </SPAN></SPAN>
    TECs and TEGs can be mounted to both sides thereby allowing heat removal from two 40x40mm TECs at the same time. </SPAN></SPAN>

    All copper construction allows for parts and components to be soldered to water block for maximum thermal conduction. </SPAN></SPAN>

    Measures: 1.62 x 1.62 x 0.55 inches </SPAN></SPAN>

    Inlet and Outlet on the same side. </SPAN></SPAN>

    Usable flat area of 1.62 x 1.62 inches (41 x 41 mm) per side </SPAN></SPAN>

    Inlet & outlet tapped 1/16" NPT: Comes with 2 pieces Nylon barbed tube fittings for 3/16" ID tubing. </SPAN></SPAN>

    Will accommodate per side; </SPAN></SPAN>

    Single TECs or TEGs up to 40 x 40mm </SPAN></SPAN>
    2, 3, or 4 TECs up to 20 x 20mm </SPAN></SPAN>
    CNC Machined from a solid billet of copper. Brazed together to a 100 psi rated watertight seal. </SPAN></SPAN>

    Minimum flow rate: 0.1 gpm. Maximum flow rate: 2.5 gpm</SPAN></SPAN>


    The drawing I have attached shows one block of TEC coolers each one a stage. The first one is first stage; the 2nd one is the 2nd stage and so on. </SPAN>

    Cold side with will do the below</SPAN>
    I am going to take a ½ inch tubing from the computer water blocks and go to a 5 way Koolance splitter that has 4 3/16 lines coming out of it going to TEC water blocks like on the drawing going to another 5 way Koolance splitter then coming out with 5 to 10 degree C water in a ½ inch tube to the computer water blocks. </SPAN>

    Hot side will do the below.</SPAN>
    ½ tube from the radiators and go to a Koolance 5 way splitter that has 4 ½ lines coming out of it each going to a Koolance 5 way splitter, from this splitter will have 4 3/16 lines coming out of it to the hot side water blocks. This way I do not series the water through the hot side water blocks and making it hotter and hotter. It should only absorb 100 to 125 watts total and this should be no problem.</SPAN>

    I work as a maintenance supervisor on a off shore drilling rig. And I have seen this in resal life like this. Have cooling water coming out of 4 engines joining a common pipe and the temperature of the water will be the average of the water coming out of each engine. I have seen this as we have temp sensors where the water leaves the engine and also where it goes into the radiator so only having to cool the average water temp of each engine. Just need a good flow rate and enough radiators to cool this final temp. You add the temps together to get final temp, I have seen and used this on several rigs just like this. We use big engines to turn generators to power our rig. So I do a lot of engineering and use fluid and thermal dynamics to cool a lot of items with water so this should work like I am showing. I will be doing experiments and testing outside of the PC using I will have 2 6” pieces of pipe with Tee on each end and in the center so that I can put a combination of immersion heaters in this to heat the water and simulate the computer components I want to cool. I will have a 100 watt, 250 watt, 375 watt, 500 watt and 1000 watt so I will be able to use different combinations to be able to heat the water like it is coming from the computer water blocks and these will be a TEC cooled cpu block, 2 GTX 680’s and mother board blocks. I will have 2 140x420 radiators, 2 140x140 radiators and 1 120x360 radiator in series to cool the water coming from the hot side water blocks and I think this is going to be over kill.</SPAN>

    I need feedback and people to tell me if I am doing anything wrong if they think this will work or anything like this?????</SPAN></SPAN>
    Click image for larger version

Name:	TEC Water Chiller.jpg
Views:	1
Size:	118.0 KB
ID:	756121
    Last edited by seross69; 12-15-2012, 07:08 AM.

  • #2
    Re: Need opinion on TEC water Chiller??

    I'm a bit confused as to what part is what, as some of the lines you have depicted above go no where, and are fed by nothing. it probably makes sense in your head.

    The big thing to watch for is all of the blocks attached to the ins and outs of the splitter have about the same resistance to water flow. the single out must have less resistance to water flow than any single one of the multiple ins. other wise you will have serious water flow issues and in some cases back feed.

    having a single path makes for a really simple setup and has the same amount of heat to shed in the end. (however the last component will obviously have warmer liquid than the first)

    splitting up the path to multiple paths does have the benefits of having cooler water on the in feed of each block, however has the downside of pressure and flow (pump needs to be bigger to flow the same amount of water through each wb as compared to a single path) and the difficulty of keeping the pressure and flow equal to each waterblock.

    TEC is extremely inefficient as you know, but it's primary benefit is lost here (that being no moving parts or water flow) you have appeared to have designed it with both of these in play, which pretty much makes the solution the worst possible. Not ragging on you, it's a very fine thought, and usually one that all of us whom have gone down this path thought ourselves. But in reality you just appear to be making really cool/fancy heat sinks, but still dissipating the heat the "old fashioned way" via water/pump/fans/radiator.
    You would probably do better by finding better heatsink materials and thermal couple solutions to feed into your 4-5 way water chiller.

    Comment


    • #3
      Re: Need opinion on TEC water Chiller??

      Originally posted by synack View Post
      I'm a bit confused as to what part is what, as some of the lines you have depicted above go no where, and are fed by nothing. it probably makes sense in your head.

      The big thing to watch for is all of the blocks attached to the ins and outs of the splitter have about the same resistance to water flow. the single out must have less resistance to water flow than any single one of the multiple ins. other wise you will have serious water flow issues and in some cases back feed.

      having a single path makes for a really simple setup and has the same amount of heat to shed in the end. (however the last component will obviously have warmer liquid than the first)

      splitting up the path to multiple paths does have the benefits of having cooler water on the in feed of each block, however has the downside of pressure and flow (pump needs to be bigger to flow the same amount of water through each wb as compared to a single path) and the difficulty of keeping the pressure and flow equal to each waterblock.

      TEC is extremely inefficient as you know, but it's primary benefit is lost here (that being no moving parts or water flow) you have appeared to have designed it with both of these in play, which pretty much makes the solution the worst possible. Not ragging on you, it's a very fine thought, and usually one that all of us whom have gone down this path thought ourselves. But in reality you just appear to be making really cool/fancy heat sinks, but still dissipating the heat the "old fashioned way" via water/pump/fans/radiator.
      You would probably do better by finding better heatsink materials and thermal couple solutions to feed into your 4-5 way water chiller.
      Thank you for your reply and you are right it is very ineffeciant and is just being done for myself. I also do not think heat sinks can remove the amount of heat that i will be removing with this water. follow my trials and see it might actually be a good idea but time will tell. i know i want to be able to keep my CPU at dew point and this is the best way.

      Comment


      • #4
        Re: Need opinion on TEC water Chiller??

        Originally posted by seross69 View Post
        Thank you for your reply and you are right it is very ineffeciant and is just being done for myself. I also do not think heat sinks can remove the amount of heat that i will be removing with this water. follow my trials and see it might actually be a good idea but time will tell. i know i want to be able to keep my CPU at dew point and this is the best way.
        save your money and effort and chill your water. It's much more efficient than TEC and much easier/cheaper to obtain. you can easily keep the water at 5c. if you feel brave and experimental there are a load of "DIY" refrigeration units using a compressor and "building supply store" copper plumbing. Just check out youtube for some great ideas. You can still more or less replicate what you are trying to do above there but instead use a good size heat sink, good thermal coupling and damn cold water. lol

        Heck as a maintenance supervisor i bet you have a ton of this stuff available to you already.

        I'm not trying to steer you away, but the reason for TEC not being popular is not for lack of trying. It's a very special use case solution, and it's my opinion, that this isn't it.

        Comment


        • #5
          Re: Need opinion on TEC water Chiller??

          Actually the more I think about it synack, I know you can not efficently remove this much heat load with heat sinks.. if you think so it makes me question what you actually know about TEC's????

          Comment


          • #6
            Re: Need opinion on TEC water Chiller??

            this idea is to build something that can be contained in a super tower computer case that will remove this much heat and keep my components at a couple of degrees above dew point in my home. you are right i could build a refrigeration chiller and cool the water easier with this but I want this small and quiet. the compressor units make noise. i know fans on radiators make noise also but not as much. this is really for me and i believe it can be done.
            hope i did not insult you with last post i made as now i read it and i see it is a little strong so sorry about this..

            be sure that if my experiements do not work and i am not able to accomplish in real life what seems to work on paper i will be the first to say i was wrong or maybe someone can read what i was doing and see where i went wrong and make this work. i do not know. but i do know if i do not try i can not succeed.

            i also know very well how much use engineering figures are that work on paper in the real world. i spend a lot of my time repairing, re-engineering and making equipment work that engineers keep telling me will work. Most of them have no idea of real world conditons and how to translate this to figures on paper but it kills me when they argue with me when i prove them wrong. i know i have not proved anything yet and it is all therory so far but i will try and see. really i do like constuctive Criticism and feel this is how both people learn is from this. have a good day.
            Last edited by seross69; 12-17-2012, 08:59 PM.

            Comment


            • #7
              Re: Need opinion on TEC water Chiller??

              I read somewhere that TEC's are about 10% as efficient as refrigeration. The way you are talking about how much heat you have, i'm thinking you won't be able draw off that heat either with TEC's. or you would need to use an obscene amount of power. And you still need to get the heat off of the TEC's!

              I'm using the term "heatsink" in a general way. I'm not referring to just the metal block with fins. A water block is a heatsink as well. The hunks of brick type stuff on the shuttle is also a heatsink. lol. A heat sink to me is any thermally conductive device that transfers/distributes heat from one place to another.

              And if you think heatsinks/waterblocks can't absorb that much heat... My system when running fully loaded and overclocked pulls about 600 watts at the plug. my water temps (with a single 3x140mm rad and 3 medium speed fans) don't raise by more than 3c over a span of about an hour. My waterblocks are cold to the touch and I can actually feel the room temperature rise over that hour lol...

              Comment


              • #8
                Re: Need opinion on TEC water Chiller??

                Originally posted by synack View Post
                I read somewhere that TEC's are about 10% as efficient as refrigeration. The way you are talking about how much heat you have, i'm thinking you won't be able draw off that heat either with TEC's. or you would need to use an obscene amount of power. And you still need to get the heat off of the TEC's!

                I'm using the term "heatsink" in a general way. I'm not referring to just the metal block with fins. A water block is a heatsink as well. The hunks of brick type stuff on the shuttle is also a heatsink. lol. A heat sink to me is any thermally conductive device that transfers/distributes heat from one place to another.

                And if you think heatsinks/waterblocks can't absorb that much heat... My system when running fully loaded and overclocked pulls about 600 watts at the plug. my water temps (with a single 3x140mm rad and 3 medium speed fans) don't raise by more than 3c over a span of about an hour. My waterblocks are cold to the touch and I can actually feel the room temperature rise over that hour lol...
                Yes I plan on using a obscene abount of power!!!! As far as room temps I have a A/C for summer and this will help heat home in the winter.

                I will be using water blocks on the hot side of the TEC this is the water the rads will cool. the cold side of the tec will have water blocks and this will cool the water to cool my system. i will have 2ac 3x140 rads, 1ec 3x120, 1ea 4x120 and 1ea 3x180.

                i will be able to remove more that 1500 watts of heat from the hot side of the TEC. I will be able to keep water to the hot side at room temp easily. so i will be able to cool the water i am going to use to cool my system to dew point or actually sub-zero C if i want to and depends on level of insulation i will be using. hope you understand better now.

                I have also changed my case to a Magnum STH10.

                Comment


                • #9
                  Re: Need opinion on TEC water Chiller??

                  Originally posted by seross69 View Post
                  this idea is to build something that can be contained in a super tower computer case that will remove this much heat and keep my components at a couple of degrees above dew point in my home. you are right i could build a refrigeration chiller and cool the water easier with this but I want this small and quiet. the compressor units make noise. i know fans on radiators make noise also but not as much. this is really for me and i believe it can be done.
                  hope i did not insult you with last post i made as now i read it and i see it is a little strong so sorry about this..

                  be sure that if my experiements do not work and i am not able to accomplish in real life what seems to work on paper i will be the first to say i was wrong or maybe someone can read what i was doing and see where i went wrong and make this work. i do not know. but i do know if i do not try i can not succeed.

                  i also know very well how much use engineering figures are that work on paper in the real world. i spend a lot of my time repairing, re-engineering and making equipment work that engineers keep telling me will work. Most of them have no idea of real world conditons and how to translate this to figures on paper but it kills me when they argue with me when i prove them wrong. i know i have not proved anything yet and it is all therory so far but i will try and see. really i do like constuctive Criticism and feel this is how both people learn is from this. have a good day.
                  Not insulted at all.
                  Have you considered the noise from the water pump too? they tend to be the noisiest component.
                  You would really have something there if it worked as you hope. I applaud you for trying and tinkering. But you did ask for an opinion... :)

                  I myself looked into TECs at one time for cooling. but after combing over tons of data presented by others who have tried it, I decided it was not a good solution for me.
                  The cost and inefficiencies were just way too high for what you get in my opinion.

                  Lol I don't have the numbers anymore.. but I think I even factored in geothermal as a cheaper more efficient solution.

                  i'm curious how you plan to shed off the heat from the hot side of TEC. with only 32c of Delta Tmax to play with you need to keep the hot side between 37-42c.

                  Comment


                  • #10
                    Re: Need opinion on TEC water Chiller??

                    Haha, never mind on that last question lol

                    Let us know how you make out. but i'll go on record as saying that has got to be the most inefficient heat pump i've ever seen hahaha (any insult is not intended)
                    I hope you can prove me wrong, and have something that is somewhat efficient and quiet. i'll gladly eat that crow.

                    Comment


                    • #11
                      Re: Need opinion on TEC water Chiller??

                      it will not be efficent at all but will be quiet!!! i know i will use an obscene amount of power for what i get but if it is small enough and dose what I want a lot of people will not are about power usage if it is dependable, easily controlled small and can give then a cold PC to over clock. they cann it custom PC and modding for a reason!!!!!!!!!!!!!!!!!!!! lol

                      I just hope i am not the one to eat the crow. I also know it will cost a lot to do this. but also know most people like us in these forums do not care about that. it is all about who has the prettiest, fastest, coolest and most rad PC in the world!!! LOL

                      A friend tells me everyone one on these forums have a serious case of ***** envy!!!! not trying to be vulgar or insult anyone but he is right in a way because of what i said before about having, the prettiest, fastest, coolest most rad PC... lol

                      Comment

                      Working...
                      X