Overclock.net banner
1 - 20 of 38 Posts

FROZENPC

· Registered
Joined
·
50 Posts
Discussion starter · #1 ·
INTRO: (FEEL FREE TO SKIP IF YOU WISH TO GET RIGHT TO THE DEETS. oh god im never saying deets again.)

I was staring at my P.C. the other day (you know, as any normal person does) and i really wanted it to be cooler. I could put a grand into water cooling but wheres the fun in doing what so many have done before?! No no no, I need MORE. Even if water cooling could yield ambient temperatures ( which it cant {dont argue about this...} ) it is just not good enough when you look at Peltiers and N2. Now I am well aware of those who have made what i could best describe as a fridge out of peltiers for their reservoir. They have yielded great results! Some getting the water down to 12C! but then... condensation strikes them down. Ive seen people post threads on lowering the humidity in their system to get that extra 3 degrees out of their peltiers (which is utmost impressive). This brings me back to my long gaze into the LED's of my belove....i mean PC
wink.gif

Water, given the system that it is most always surrounded by in a PC (hot chipsets, processors, drives, the next greatest oven (290x) can only get so cold. If it gets too cold it starts to attract its friends that are just floating around in the air we breath. Probably tired of the hot monotonous work it has to do, it forms a drop right above your gpu, you thought you were smart setting up ur peltier "fridge" until Bloop. the drop hits its mark, and you gotta take a hit right in the checkbook.

BEGIN HERE: Alright lets get serious. Im going to make some propositions, ive currently designed most of this on paper and have found almost everything i need to build the system, and this log will be updated as time goes on (and as i get payed haha )
I HAVE NEVER built a water cooling loop, but i do not need help building one. Its simple fluid dynamics. ive

I do however, understand Fluid and Thermal Dynamics and Statics (which will be a small role in this project)

YOUR CRITICISM IS WELCOME!
YOUR ADVICE IS WELCOMED MORE!
SUGGESTION ARE AWESOME!

PLEASE SEE POST #13 FOR MORE INFORMATION ON TOP OF WHAT IS DETAILED BELOW!
smile.gif


I am going to be completely sealing off a pc case. I have chosen a case that is pretty. Because the looks matter. It is also very simple, this makes it easier to seal. And it has a window because well if im gonna do all this work i damn well better be able to look at it haha.

The case will be under a vacuum of about 100 to 40 microns. As far as the computer is concerned, this will not hurt it. UNTIL it is turned on. With nowhere to put all that heat the processor will fry in seconds!!!

BUT of course the whole point behind this is so I can use water JUST ABOVE freezing. so why not, while we are at it, lower the freezing temperature of water.
A nice little solution of 60% Ethylene Glycol 40% water sounds nice to me SO FAR.

It drops the freezing temperature IN CELSIUS to -52.8 degrees!

THE PLAN IS TO USE A WATER/ ETHYLENE GLYCOL SOLUTION AT -48C
***I UNDER STAND THAT ETHYLENE GLYCOL ISNT AS SAFE AS PROPYLENE GLYCOL BUT ITS 4C MORE EFFECTIVE SO ILL TAKE THE RISK ***


If you are interested heres all the chemical properties one would need to know about the aforementioned mixture. http://www.engineeringtoolbox.com/ethylene-glycol-d_146.html

But how do i get the heat off of the north bridge and the south-bridge, and the hard drives... ok im not putting my hard drives in a vacuum bc you cant even sneeze with those things in ur hand sometimes and i now when to admit that i dont know about something and i have no idea how hard drives are sealed.
But i digress.

I have 2 Options that are good enough for me right now. REMEMBER! SUGGESTIONS ARE WELCOME

1) I take off the stock heat-sink and create my own/ mod the stock to accept the water, but i would prefer it not be too cold as the cpu will get bc i dont wan the board to crack under rapid cooling.

2)Run copper in the most beautiful way possible to from these components to a sort of peltier array. or straight too the **two main "CPU" blocks**

**this is where a consider able amount of the cooling will happen**

EDITED: THERE USED TO BE A LINK HERE BUT I DECIDED IM GOING TO MAKE THE BLOCKS MYSELF
smile.gif


2 of those in series both with 3 pelitiers on one side of them.

this will then flow into the reservoir which will be very beautified version of this IDEA not actually this really at all just in concept, just wanted to give him cred,

http://www.tomshardware.com/forum/275185-29-exploring-ambient-water-cooling

Now because of the damn laws of physics, this part will have to come out of the vacuum and then enter back in at the end of the series. Everything that youve read is ENTIRELY my idea. (except the refrigerating the RES, i would have probably done this regardless of finding this article in my initial research into finding if anyone has had this idea before, so i cant be for sure that it didnt influence me so im giving him credit
I've scoured the internet for someone who has done this post exploring the idea and all i have found is quite ill informed folks thinking if they make a vacuum in their computer, it will reach 0 kelvin.......... or my personal favorite.....have fans spin around in side so it becomes T<0 kelvin..... i will post pictures of that if i can find the forum again and after i edit out the user names of those involved. haha oh speaking of fans, if you didnt notice, there are no fans in this system, JUST A PUMP IS THE ONLY SOURCE OF NOISE IN THIS SYSTEM. , that vacuum pump will run once every month to year depending on how well i can seal the pc.

IMPORTANT! PLS NOTE! THANK YOU
smile.gif

THIS WILL EVENTUALLY, HOPEFULLY SOON, BECOME THE PROTOTYPE BUILD LOG.

THE ONLY THING I CAN DO A PRIORI IS GIVE THOSE INTERESTED A DETAILED OUTLINE OF THE LOOP AND WHAT THE VACUUM WILL LOOK LIKE. AND THE LOGIC BEHIND IT ALL. THAT IS COMING SOON!

BEING 17 AND A SENIOR WHO IS UNEMPLOYED CURRENTLY MONEY ONLY COMES AT THE END OF THE QUARTER AS I GET PAID TO GET GOOD GRADES. (i dont receive an allowance and never did so that's my form of allowance) SO WITHIN A MONTH I SHOULD HAVE ABOUT HALF OF THIS DONE/ PILED UP IN MY ROOM hehe.

Im shooting for 450 to 550 bucks for just the case all rigged up and ready to be depressurized. NO PC HARDWARE IS INCLUDED IN THAT. but i do have to buy a 2 cycle vacuum pump whos price is included in my estimate so really if one were to do this on the cheap and not the prettiest as i am doing for the prototype you could do it for 300. i just dont have fittings or tubing or a pump haha. gotta put in that initial investment!

AND FINALLY: I will ON THE SAME THREAD, post the final build, given the prototype yields positive results. IM GOING TO BE GOING ALL OUT ON THE FINAL BUILD
im currently thinking of using the 900D because of the monsterous size and simplicity. plus i could hide everything ugly in the bottom bay!
Let me know what you think though
Thanks so much for reading this if you got to the end and i hope to see you in a year when i plan on having the 900 D complete and my wallet emptied!
***if i made a typo i promise you that you can get over it
wink.gif
tongue.gif

COST LOG
2x ALUMINUM L BRACKET: 1 1/2" X 48" $22.00
Looking for my unused H50 cooler 15 calories
plugging in the 8320 stock cooler for the first time: 5 minutes of laughter at just how sad that thing is.
Quick disconects: $5*3
Ball valve: $10
 
Subbed, but a few things;
Proper water cooling gear would be much better than that horrid block you linked to.
You can buy waterblocks specifically designed for chipsets, both universal blocks and ones specific to certain motherboards. http://www.frozencpu.com/cat/l1/g30/Liquid_Cooling.html
For the components you don't want to chill you can create a second, non chilled loop.
 
Discussion starter · #3 ·
Thanks for the advice! I didnt know they made good water blocks that big and so beautiful. i will agree that thing is bad but then again it is just in the prototype phase.
for the final build i will have second loop cooling the board thanks to you! but i wont have fans on it as that is one of the rules of this is there can be no fans. MOAR PELTIERS!! haha

Do you think a HDD block would be the best as it has a large surface area. i may just get radiators at that price and just put em in a fridge like thing.
 
What about natural oil to seal the air away? Don't know (and can't look right now) for the freezing temperature of it but i think it's really low.

Subbed as I'm interested.
biggrin.gif
 
About the HDD, is replacing them for SSD's not an option? I know the income is limited in your case but they're more resistant overall. (I'm assuming they could handle the vacuum I may be wrong so don't quote me on this, just throwin an idea).
Also how do you plan to replace a part if it breaks?
Other than that, subbed since I want to see the outcome
biggrin.gif
 
Discussion starter · #6 ·
I have considered an SSD, but to me, as i am one who likes having large storage amounts, SSD's just haven't reached a point where a 1tb is viable and worth the cost considering the limited times it can write data. Its getting better though! And the final build may very well contain ONLY SSD's.
As far as what if something breaks?
Even a failed result is a result, given i dont break something due to my own ignorance or neglect. But I never purchase stuff without having more money left over in case i need to buy more stuff.
A sort of emergency pc fund if you will

@NVNW
I know very little about Natural oils, so thanks for giving me something to do tonight!
but i would imagine that unless they are VERY thick viscosity, basically solid, as there will be a 760 mmHg difference in pressure (14.7 PSI), that it would fail. But if it is easy to clean up then i will definitely try it before i permanently seal the system.

Im trying to find a good copper block, literally just a rectangle (preferably 120mmx60mm {LxW}) that just has an in port on one end and an out port adjacent to it.
Given that i have access to a 3D printer and can solder quite well (i could be better but this gives me a chance to improve my "skills") i may just make this block myself.
 
Large storage capacity can easily be solved by using an external solution. Figure out what you will use day to day and need instant access to and go with that in SSD form. I have a 3tb external drive that is configured to be on all the time and not go to sleep so it is no different than having the drive inside my case. Not to mention the transfer speed with moving data from it is quicker than some of my internals.
 
Discussion starter · #8 ·
GIVEN THE LOW TEMPS I WANT, I CANNOT HAVE 0 FANS, THE PELTIERS WILL NEED PROPER COOLING. This being said, i will be running them at the quietest possible setting that yields the results i want.

As for the SSD inside and the external HDD.
Im aware of this. I planed on running the HDD on a section inside the PC that will not be under vacuum (isolated from the rest of the vacuum)
Also i have a HDD from an old laptop that i will put in the vacuum to see if it will effect it and i will post that up when the time comes.
 
Discussion starter · #9 ·
LOOK WHAT I FOUND!


HAHA my dad is an HVAC tech for the city and has been for years. The dude can fix anything!
well i completely forgot that he installed our home ac and his friends in the summer when **** gets hot as hell.
I was looking around to see if he had a micron gauge, which he didnt, BUT he did have that pump. Its overkill, ITS FRICKEN LAB GRADE!
6.2 CFM and its a 2 stage. guarantees AT MOST 1000 microns In very large systems. For the PC i would expect about 25 to 100 microns
HAHA so now that i have the most expensive thing out of the way hopefully i can finish a little quicker.

I will be posting pictures of plans and schematics.
I will also outline some parts i would like to build; Feedback would be much appreciated on those.

I DONT WANT TO TRIPLE POST SO IM EDITING THIS ONE
I have changed my mind on the case. I have ALWAYS wanted the coolermaster HAF XB but the hard drive cages and all that bs wasnt removable. BUT the XB EVO does have that. The bottom can be completely cleared out. and with the height it has it can basically become 2 computers on-top of each other and it is just awsome. also, now i can just make a box out of thick acrylic and just seal that and place it inside. ill make it look all pretty of course. i may move sealing the case itself to the 900D, but honestly the more i think about the mods the more better off i am just getting an aluminum box. thats what the 900d will become. idk but what i do know is that the ideas are flowin!
 
You'll have to do more than just cool the CPU, GPU and the like with the system in a vaccum.

For example, the VRMs/MOSFETS also need cooling along with everything else that's part of the system itself, normally those components will pass off their heat to atmosphere without a problem but in a vaccum, there is no atmosphere to pass it along to.

More or less every and anything that generates heat will need to be at least considered - This will make it a real challenge to cool it properly all around. Since in this config there would be no air at all cooling these components the risk of heat buildup to them is a real threat not to be underestimated.
 
some constructive criticism:

1. If you've never been trained to handle hazardous chemicals, do NOT use hazardous chemicals. you may understand that it is more dangerous, but do you actually understand what that means in terms of how to handle the chemical and what a potential exposure would mean to your health and quality of life? Your comments appear very flippant (no offense intended, just stating how they appear). You say that you want to use ethylene glycol because it means an extra 4C, but you do not state why that is significant. The ONLY reason to use a more dangerous chemical is if your project requires it. So please prove to yourself that it is REQUIRED! That means calculating the expected deltaTs of your water loop for different conditions. Then, you need to figure out how those deltaTs will correlate to something useful. Since you are risking your health and quality of life, those reasons should be pretty important.

2. Its not clear to me how you are achieving the very low temps that would necessitate using chemicals to depress the freezing point. It looks like you only plan to use peltiers and those only have a max deltaT of around 50C with zero load. You need to calculate your expected heat output, and then figure out how many peltiers you can afford to deal with that heat output at a deltaT that you desire.

3. Before you buy a case, have you calculated if it can withstand the forces of the vacuum? My guess is that most commercial cases can not. It is likely that they will deform. If they deform, then you will most likely destroy your vacuum seal. Please do the calculations first before purchasing anything!

4. Making an airtight seal around different types of components like water cooling tubes and peripheral cables is extremely difficult. Making a vacuum seal is even harder! As soon as it experiences a pressure differential it will want to deform! Furthermore, you want to design it so that you reduce the number of seals you need to make as much as possible.

5. I'm still unclear on the purpose of the vacuum. Is it just to prevent the possibility of condensation? If that is the only purpose, than it is admirable that you are trying a new way to achieve that (I genuinely mean that). However, preventing condensation is a solved problem. Insulating a motherboard is very effective. Chillboxes are also extremely effective. Trying new things is great because you learn a lot, but it can be extremely expensive to be a trailblazer!

6. A vacuum is an extremely effective insulator. If you want to add that as a component to your build, I think it would be best suited as a way to insulate your chilled reservoir.

7. To sum up, I believe your idea as presented so far will not work as you expect. I also think that you are going to spend a lot of time and money and will not be able to get the performance that you would be able to achieve with methods that are already available. I highly suggest doing extensive research on both chillboxes as well as all the peltier threads you can find. This thread has details on chillboxes: http://www.overclock.net/t/1533164/the-24-7-sub-zero-liquid-chillbox-club. Additionally, this is the only thread I've ever seen that built something that could withstand a slight pressure differential: http://www.overclock.net/t/1346823/subambient-full-submersion-phase-change-cooled-pc, however, it was designed to resist much less pressure than the full vacuum as you describe.
 
Discussion starter · #13 ·
????
1. Ethelene Glycol is not "hazardous" and its not even recommended to use a gas mask when handling small amounts. (That being said i WILL be wearing a gas mask just in case) Please look it up as it is quite harmless, someone of my weight (130 LBS) would need to consume 300Ml to die, thats ROUGHLY a third of a liter... Id have to drink it on purpose to wind up dead via consumption or exposure. Contact with the skin isnt dangerous as well. Its alternative is also very crapy as far as thermal conductivity is concerned There for i need to use Ethelene Glycol I know how to properly dispose of it (pay someone) im not just gonna dump it down the drain or in my backyard. Also if it was so 'hazardous" it wouldnt be Unregulated. its diamond numbers are 1-1-0 Blank, lava lamps are more dangerous. LOL

2. While everything about the max delta T being about 50C is correct... It is MY OWN personal opinion that the idea that this means that the use of chemicals to lower the temperature in not justified, is COMPLETELY WRONG.
My room is 24.4C right now. Last I Checked, and i could be wrong, 24.4-50 is -25.6C HENCE I MUST USE ETHELENE GLYCOL
And since i will be using more than one peltier, it should sustain a - temp easily. Also i plan on using N2 to cool the water (its inefficient i understand, but i have my reasons haha)Possibly at some point, hence my want for a -48 degree operation temp. i did not state this as i planned on doing a new thread for that, if i decided to do it.

3. Im quite positive i have stated this in a previous post but the case is no longer going to be sealed, i am just going to make a box out of Lexan and place that over the motherboard tray. BUT Please inform me how an aluminum case is not strong enough to handle a vacuum. a vacuum on a system results in 14 psi on the outside. over 21*22 inches (average Computer height), thats 6468 pounds of force total over the entirety of a side panel. This will not, IN ANY WAY deform the case... i shall even refer to this source so you can see for yourself
smile.gif
http://www.americanmachinetools.com/tensile_strength.htm
http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA6061t6

4. thanks for the advice
smile.gif
i am well versed with sealing HIGH pressure systems, and i would assume that these are harder to seal as the psi is much higher. But i shall find out.

5. I find the vacuum a more sure-fire way of preventing condensation. One cant really mess it up as long as they have a vacuum, they have no condensation. also, i am one who likes the look of his work, and insulated motherboards arent exactly pretty HAHA. and the same goes for chillboxes. also, the cost of doing this is actually only about 300 for me and thats because i have 0 water cooling equipment. something that some have have 100s if not 1000s of dollars worth in their home. in which case they would only need some lexan and 50 dollar to 200 dollar pump, depending on what level of quality they want.

6. A reservoir in the vacuum WAS initially my first idea, but then i decided i would cool down the reservoir in a box of peltiers, ( containing at most 4 peltiers) and have the box insulated. i will locate the res close to the vacuum and i will be making sure to insulate the tubing as well.

If i came off as an ass at all im sorry. i didnt mean to be and you can blame my own ignorance in how to deal with people ( when i say deal with i dont mean as in like you are some sort of obstacle or something just getting in my way and pissing me off)
BUT i do appreciate your advice and concern for my own safety
YOU HAVE INSPIRED ME TO KEEP A TOTAL COST LOG NOW!
i will be posting updates to the thread post with costs.
 
Peltiers need adjustable power supplies to tune the voltage.

Use a single output PSU (i.e. Meanwell).
 
Discussion starter · #17 ·
Quote:
Originally Posted by Madog View Post

Add individual wattage together. Purchase PSU with output greater then total wattage.
haha well duh. but what ports can i connect them to besides molex plugs.
what wires basically do i need to use from the cpu and pci connectors
 
I am no expert but if you are working with a vacuum environment, and using blocks on the main components, isn't the only way the PCB of the Mobo can exhaust heat as IR radiation? They aren't designed for that and is the chilled loop enough to cool that as well as the components directly connected by metallic connections?

P.S.- I am curious, not critical.
 
Discussion starter · #20 ·
Quote:
Originally Posted by dave12 View Post

I am no expert but if you are working with a vacuum environment, and using blocks on the main components, isn't the only way the PCB of the Mobo can exhaust heat as IR radiation? They aren't designed for that and is the chilled loop enough to cool that as well as the components directly connected by metallic connections?

P.S.- I am curious, not critical.
Great question, i do plan on cooling a graphics card, and its ram of course, cpu and the chipset as well as the ram. i may also ( as this may be all it needs) lay the tubing against the motherboard. but im going to refrain from doing that as much as possible. im hoping cooling the chipset with such cold water will result in the cooling of other components around it.

ALSO, COULD I NOT JUST USE A CAR BATTERY? I have a couple just laying around
obviously i wouldnt use it in the final product but for prelim testing while i convince myself i should spend 250 bucks on a powersupply.
imma just get some cheap ass chinese **** and do this.
http://www.wikihow.com/Convert-a-Computer-ATX-Power-Supply-to-a-Lab-Power-Supply
found the supply ill mod its 20 bucks an actually has good reviews.
http://www.superbiiz.com/detail.php?p=PS-I500&c=fr&pid=1123077f64e9912b012b1b58de2e29c73dfde602bf94ecbed89ecfbc29366ff5&gclid=CMDnq77ZvcMCFQiDfgodED8AOA
its quite simple. i just dont like screwing around with capacitors. but ill just wear rubber gloves haha. i dont know TOO much about electrical systems yet
 
1 - 20 of 38 Posts