TEC build log start

[toolmaker_03]

http://imgur.com/cd15M1a
ok so I have finished all my testing and decided on a general build for this unit.
this build will be sealed inside of a chest so I will not have physical access to the computer case, I would like some input on the general idea that I am presenting, I use antifreeze so freezing the coolant is not a issue, and seeing as I am increasing the cooling capacity of my water cooling system, I have also chosen new clocks for my hardware, the CPU clock will run at 5Ghz, and the GPU's will be running at 930Mhz, I have already tested theses clocks and found them to be stable for my hardware.

 

[kanewolf]

Seems way too complicated. What benefit do you believe the vacuum environment provides? Condensation avoidance? I would just add sensors and regulate the temps above the dew point.

 

[toolmaker_03]

well I have a little camera in the vacuum chamber to show that nothing is freezing up. With no hardware running on the water loop but 1 TEC running, the water gets to a chilling -35C, with only the CPU running the water will be right around 0C. I have tested this with the CPU and one TEC only so far, but with a 5Ghz clock on the CPU it held at 38C under load.
the idea for this build was to be able to clock my hardware further, while still keeping the same temps that I am getting now, I want all the hardware to run between 25C and 50C under load.

 

[kanewolf]

Wouldn't replacing the vacuum with nitrogen give you the same effect? No moisture to condense. But much easier to work with.

 

[toolmaker_03]

don't know I have never worked with nitrogen, but I have worked with vacuum chambers, as I build satellites and test them for use.
so as a result I knew that this approach would work, the hard part was finding a chest large enough, I ended up with the cube chest for $47, I did have to drill a hole in it, and then I had to seal the hole up, after I ran all the wires and soft copper tubing through it, but still a rather cheap way of solving the issue of ice forming all over my water blocks.

 

[toolmaker_03]

to give a idea of how close I am to completion, I still need two power supplies, I am using the powermax 12V 45A power supply units, one for each TEC, I have one now, I still need 2 X 360mm radiators, I have 4 now, and I still need some fitting for the water loops, I have most of them already.

these things below will be possible upgrades for this build.

I would like to use a new motherboard the asus rampage 4 black edition for this build
I would like to get a 3960X CPU for the final build
I would like to get motherboard blocks for the build

so I am about 6 months, to a year from completion, depending on how many of the upgrades I decide to get for this build.

the new rampage 4 black edition motherboard is almost a must have for this build.

 

[computers_are_freaky]

Ok, I am really interested in this build log, and I've seen you on ryans TEC sticky, but didn't he say he tried watercoolng the hot side of the TEC and it resulted in stalling? You will probably need to rig up some sort of software to regulate the waterflow when you turn on the peltiers to make sure the temps don't start getting ridiculous. Anyway, really cool idea, and I also want to say this:
IT IS SO AWESOME THAT YOU WORK ON SATELITES!!!! I'm still a kid, but wow.

Also, won't this use up a lot of power? The main reason I can't do this at home is because my dad said no way I was going to be using 200W just for cooling. My personal idea revolves around using a solar panel and some sort of massive battery. Anyway, awesome build and I'm really interested to see how it proceeds.
Good luck.

 

[toolmaker_03]

http://imgur.com/w0oxpEZ
ok so this is more close to what it will look like, I have 3 X 400mm tube reservoirs, so I think that will be enough of a cold bank, the cold side loop will also have a 250mm reservoir on the outside of the vacuum chamber.

the hot side loop is just a longer version of what I have on my system now, instead of two radiators it is six radiators.

 

[toolmaker_03]

ok so weight wise building a radiator box out of wood for all six radiators is too heavy to move around, so I am going to mount the radiators to the top of the chest.
Igloo 60-Quart Ice Cube Roller Cooler
the case inside
http://www.newegg.com/Product/Product.aspx?Item=N82E16811108464
I had to heavily modify the case to get it to fit inside of the chest, I striped the case down and then I cut the bottom off, then made the case 3" shorter from top to bottom all the way around, reattached the bottom to the case and it just fits, lots of fun.
I do not recommend this kind of build for anyone, as it is time consuming, lots of tools are needed, and expensive, but if your hobby is water cooling, and looking to take that hobby to the next level in performance, I would say this is it, TEC's are lots of fun to play with and provide hours of tinker time.

 

[toolmaker_03]

http://imgur.com/a/CA6Iy
this is completely covered in two types of clay
http://www.amazon.com/Malleable-Polymer-Modelling-Earser-Plasticine/dp/B015MEW9ES/ref=sr_1_27?ie=UTF8&qid=1450507000&sr=8-27&keywords=eraser+putty
this is about a 1/8 inch thick around the entire cold side blocks, and fittings, and this kind of clay is covering that
http://www.amazon.com/FEITONG-TM-Malleable-Modelling-Plasticine/dp/B012CLIL0A/ref=pd_sim_sbs_21_8?ie=UTF8&dpID=41Nna9QotpL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=1S4KVNKV4ATQ8PHS0FDW
this stuff is about 2" thick around the entire cold side assembly, yea its ugly, but functional, the cold side blocks and piping does not end up as a block of ice this way.

 

[toolmaker_03]

to insolate the reservoir, I will be covering it in foam insolation to start, and maybe some clay on top of that if it starts to freeze up around it.
they make real nice fittings today for making the conversion from G1/4 to soft copper tubing. the copper tubing will also be covered in clay.
http://www.performance-pcs.com/new-koolance-nozzle-single-cu-pipe-for-od-13mm-1-2.html

 

[toolmaker_03]

power consumption :lol:

for lots of giggles yea, its high, I will have a total of 5 power supplies for this build, a 1250W PS for the motherboard and video cards, a 600W PS for the 4 water pumps, all the fans, fan controllers, temp sensors, and flow meters. 3 X 12V 45A powermax PS for the TEC's
so its a lot of power consumption.

 

[computers_are_freaky]

Oh my pc!!! That is A LOT of power
Why are you worried about rad box mobility, it's not like you are going to be taking this pc to LANs (unless you own a flatbed and a forklift truck). XD
I was thinking about that clay on the tubing. If the tubing moves ever so slightly, then there will be a gap, won't there?
Are you going to insulate the mobo? I saw the vacuum pump and think it is cool (get it?) but what if something leaks or some moisture gets in, couldn't that pose a problem?
This is really awesome, please carry on updating this thread too, it's something that I doubt anyone has been crazy/devoted enough to do before and I can't wait to see how it goes.
Good luck.

 

[toolmaker_03]

thanks but really all I am doing is a simplified version of a industrial vacuum chamber for testing electronics, the difference is that there chambers can get hot as well they also cost a lot more.

if I was only using the clay for sealing the chamber then yes a leak would be very possible but I used silicon sealant the good stuff and flex seal spray on rubber to seal the vacuum chamber.

also once I find out how thick I need the clay to prevent condensation or freezing on the cold side blocks and tubing I will wrap it plastic wrap and coat that with flex seal to keep it all together as a single unit.

and next I will build a submarine out of a old truck, using nothing but flex seal :lol:

 

[toolmaker_03]

well let me explain a couple of the issues I ran into while testing this setup for use.
cold coolant is a good thing for hot components, but rubber seals get hard when there cold, and as a result can get little cracks in them causing a leak, I did 4 things to resolve this issue, first I replaced all the rubber O-rings with ones that are rated for -65C, two I used gas line Teflon tape on all the connections, three I used pipe thread sealant on all the connections, and four after the connection is tight, I cleaned it off with a Q-tip and some alcohol, so that I could apply gasket sealant rated for -65C at the seam of the connection.
overkill, maybe, but I don't want it to leak.
freezing of the blocks is not really a issue for the coolant inside the system, but when the system is shut down, all that ice turns to water and creates a puddle under the blocks, in testing I put a pie pan under the blocks to collect the water, but for the rig I am trying the clay out as a way of preventing the ice from forming.

 

[computers_are_freaky]

What sort of fluid temps are you getting? I realize you probably haven't plugged any hardware in, but have you reached the thermal limit of your cooling system yet (sorry for my had terminology, by this I mean: what is the coldest temp you have gotten in your loop, where it can not get colder? Again, sorry, but in my defence, all my classes are in italian ). Anyway, since your in the testing phase and you clearly have no qualms about power, why not have a dehumidifier going in your work room, that way you won't get as much ice buildup in the testing phase.
This is awesome, looking forward to the projects completion.

 

[toolmaker_03]

fittings, yea I used a lot of them for this build, the good thing is that my box of spare parts is almost empty now, the bad part is I still need about $500 in fittings to complete the build, they kill me with the cost on fittings.
when I am done I will do a cost out for this build, I can tell you now, its high, too high. by the time I am done with this rig it will be about three years since I started saving up for, and buying the parts for this build.

 

[toolmaker_03]

the TEC's I am using have a 70C delta T what that means is that it can lower temps 70C from the ambient temp, and it will get close to that with cooling only the water, and no components running on the loop, the water got to -35C, that is really cold.

with the CPU on, putting heat into the water loop, one TEC held the water at 0C at load, and the CPU was at 38C at load, while at a 5Ghz clock. that is as far as I took my testing when I decided to go ahead and build the completed model for cooling my whole system.

 

[computers_are_freaky]

And you're adding another 2 TECs to those numbers right? Wow. It means that you can get down to like, ~-25°C at iddle, depending on your ambient of course (I'm assuming you sit around 25°C). What sort of coolant are you using, I read anti-freeze, but what brand exactly? How's the clay comming, because I bet that will look hilarious when you're done (did you get the rainbow like you linked, or just one colour?). Great build, this is really interesting. Looking forward to seeing the benchmarks and finished build. Are you going to do anything for cosmetics, like a custom paint job or anything?

--Update--
Just had another look at your plan, why can't you run your peltiers at full wattage? You have amps to spair. I figure (please bear in mind I'm a noob at electronics and you build satelites, so I know you have a perfectly good reason, or my maths is wrong), but couldn't you run your peltiers at their max wattage and still have a large margin for safety. And I remember you recommended giving a 25% safety margin, but you have 55% margin. Is it to be closer to the highest efficiency point of the PSUs (I know they reach their efficiency rating at 80% load, but I don't know how much it fluctuates based on load). Have I just done my maths wrong?

 

[toolmaker_03]

no, you are right, PS for the TEC's are over powered, I do not like to run my PS at max load capacity, is it overkill? maybe.
the TEC's are 16V and at 16V and 26A they reach there max ability, at 12V there at a little better than 3/4 of there max at 19A. also it is my hope, that the TEC will not burn out quickly at this voltage setting.

 

[computers_are_freaky]

Ok, thanks for clearing that up. What are noise levels like? I mean, you have 18 fans for the rads, a vacuum pump and a fan on each PSU. As well as a few pumps.
Another though:
I realise you are using full cover waterblocks on all components, but very little heat is transfered in a vacuum (I'm not sure about this, please correct me if I'm wrong) so won't thing like RAM, maybe some parts of the mobo that aren't covered by a waterblock, and storage devices start overheating? In a normal case, it is no prob because even bad airflow will keep them cool, but you won't have any airflow (that would kind of defeat the purpose of a vacuum pump, come to think of it ), so won't those little things start to overheat?

 

[toolmaker_03]

good question, the chamber can hold a static vacuum for 36 hours, so I only need to run the vacuum pump for 3 min once a day to prevent condensation inside the chest.

also because the water will be around 0C, running into the vacuum chamber, the water blocks inside are around 5C, so the ambient temp in the chest is around 15C, with leaves all the hardware staying rather cool.

 

[toolmaker_03]

the chest(vacuum chamber) and pump.
http://imgur.com/a/j51mM
had to throw the inwin case away, made too many changes to it, and the case became weak, I have another case, its a HP case, and it fits in the chest, so this time I will be a little more careful about how many modifications I make to the case. :lol: I will try anyway.

 

[computers_are_freaky]

Ok, for some reason I was picturing the chest more like the double chest in minecraft
Are you going to rig up some sort of outer shell, to keep everything together or just have it lighing around?

 

[toolmaker_03]

there is a HP computer case striped out completely with no sides or front on it to make it small enough to fit in the chest, the motherboard and video cards fit inside of the HP case, because the case has been completely gutted of all its insides, the PC case is screwed to the floor of the chest to keep it from moving around in the chest, there are also 4 X 250mm reservoirs in that chest, attached to the outer wall, with 4 X 120mm fans attached to two of the reservoirs, with two fans on each reservoir.
this has been a difficult build, I am on my second chest, and second PC case. :lol:
lots of modifications are needed for this kind of build, at least the way I do them, but I have the tools and the time to waste, so I don't mind the little issues I run into from time to time, but issues also mean cost and extra time to the project.