Extreme WaterCooling Setup, Tips/Ideas needed
 

Hello everyone. I've got a HUGE project that I want to undertake while I'm off work for the next 2 weeks. It involves lots of 1/2" ID hose (50+ feet), Swiftech Storm Water block, MASSIVE radiator and an old (but still functional) air conditioner.
A little bit of background...
I've been a computer junkie for many years, and a watercooling junkie for about a year. I also have a VapoChill XE2, so I am not new to the world of super-cooling experiments. However, this will, by far, be the biggest, craziest thing I have ever tried.
I need some tips or ideas on what the best components for this experiment would be, or if it is even POSSIBLE.
My computer is sitting on a table, which is about 4.5 feet high. The air conditioner outlet is about 5.5 feet high and 12 feet away from the computer. The way I would like to do this is to run tubing across the ceiling, using approximately a 45 degree angle to and from the ceiling, which would leave approximately 5-7 feet running level at the ceiling height. My ceiling is about 7 feet high.
I'm planning on using some sort of triple-fan radiator and just attaching it to the outlet of the air conditioner, with possibly a fan sucking the air from the air conditioner through the radiator (if necessary).
I have figured 2 Hydor L35 pumps would work, but I'm not 100% certain about that, so I really need some input there. Will that be enough to push the water ALL that way, including up both inclines (to/from the air conditioner).
Would I need to find a way to mount the reservoir HIGHER than the tubing to alleviate air in the system? If so, I think I can drop the tubing down a few inches and affix the reservoir to the ceiling, although I am not sure where yet.
This is what I have figured so far, please let me know if this sounds reasonable:
1- Swiftech Storm Waterblock, 1/2" ID
50 feet of Tygon, anti-kink tubing, 1/2" ID
2- Hydor L35 AC pumps @ 450 GPH
1 (or 2) Danger Den Black Ice Extreme III Radiator(s).

Since I live in Missouri, a ground loop isn't really feasable. This is more of a fun project than anything. If I needed more EXTREME cooling, I'd go with another VapoChill.
Also, would I have to worry about condensation with this particular setup? I really don't have any idea how cold the water will get, I'd have to get some temp measurements to figure that out.
Any ideas would be GREATLY appreciated and thanks in advance for your replies.
-baron_iv

wow, nobody? It's kinda slow around here this weekend.
Or maybe this has been done a million times and nobody cares anymore.
:)
Have a nice weekend everyone.

Sorry, laptop is out of action so I have no approximator to let you know the results.
However, to date, once you put a G4 in a system then the tubing restriction is almost nil.
Not sure of 50ft though as I've never done that much tubing.
Keep in mind that if the ducted rad can get the coolant temp sub-ambient then the tubing will be working against you as the coolant will heat up again.
Yes, bubbles will tend to collect in the highest part of the system, but I'd put the reservoir before the pump and, if I saw bubbles, just take down the tubing to get it out.

If your res isn't higher than the rest of the system, the high point will collect microbubbles over time. Make sure you can do something with those bubbles.

It's been done before. But that doesn't mean it isn't a good idea! LHGit is right though. Insulate your tubing. It'll bleed off all that sub-ambient temp before getting to your computer and you'll be left with a really complicated method to get ambient temp water to your CPU. You can get insulation for tubing at home improvement stores for plumbing or HVAC.

All in all, your implementation plan sounds quite practical. You might want to place your pump right next to the rad by the AC though. Get as much heat out of the water before you send it back to the computer.

Most importantly, remember to have fun and take pics! After all, if you just wanted lower temps, you could buy a Kryotech system. :D

The position of the res doesn't matter. It'll be no more or less effective if its at the top of the system or the bottom. Eventually any good res will drain out all air, and initially any very low flow velocity spots with no easy way for air to escape will gather air bubbles. But in the end if the res works, it'll get rid of your air.

Although putting it near the pump will certainly making priming a lot easier.

Yeah, the main reason I'm DOING this is for fun. I already have a VapoChill XE2 that cools my main system down to sub-zero temps. I'm just looking for something to burn up a little bit of time. Plus, I love experimentation. A couple of weeks ago, I went to walmart and bought 40 feet of dryer hose. I ran it from the air conditioner in the kitchen, to a computer I conveniently set up in the dining room. Needless to say, the little lady wasn't very happy about this at all, but it did cool my CPU (via watercooling) from 40C at 100% usage to 28C at 100% usage. So it was certainly a success, at least a cooling success...not a domestic success in any stretch of the imagination. In fact, I'm pretty sure if I had left it up one more second, I'd have been single again...which WOULD be condusive to more experimentation...
No, that wouldn't work...I'd have wires, hoses, cables, computer parts, etc strung out all over the house so that nobody could walk through if someone wasn't here to keep me in check. I'd also probably forget to eat and starve to death. A body can't survive on Mountain Dew and pop tarts.
Anyway, thanks for the ideas guys. If you think of anything else, please post. I'm going to go ahead and buy everything I need to see if I can get this working.
By the way, someone told me that if I got the radiator to cool the water down sub-ambient, that the tubing would be working against me...I'm not quite sure I understand the meaning of that... If someone could clear that up, I'd be greatly appreciative.
Have a great weekend!

Yes, they collect air bubbles just as good anywhere in the loop.
However putting a res between the pump's outlet and the first highly restrictive peice in the loop will definately make a difference. Tested at OCAU for bad results - it adds restriction and also will be more prone to linking. Res's belong just before pumps to reduce the NPSH, with as thick and short a tube as you can get away with.

Next, to "working against you".
What I meant was that if the coolant is colder than ambient, then as it travels along the long pipes the ambient temp air around those pipes will be warming the coolant up. So, insulation or thick walled tubing is better than mega-thin-walled tubing, and insulation around the pipes is better still. That is, if you care....

Thanks for the quick answers, I really appreciate it. I hadn't really put much thought into insulating the tubing, but that makes complete sense. My mind is usually running in a thousand different directions, so gaining heat through the tubing walls probably would have COMPLETELY slipped by me.
Since the little lady isn't home tonight, I hooked up my dryer hose to the air conditioner. It's dropped my temperatures an additional 10C from the plain ol' watercooling setup. The dryer hose is ran from the air conditioner output to the fan on the front of the radiator. It's actually got my temps equal to ambient at the moment and it hasn't risen more than 7C even at full load. I saw it dip below ambient when I allowed the computer to sit and do nothing, so I will probably have to watch out for condensation. Yay, I just checked again and it's BELOW ambient temp!! My temp for the room is 22C and the CPU temp is 19C! Wow, this is cool! I expected temps just above room temperature!
I am really looking forward to sticking a radiator to the front of the A/C unit. I think if I can keep the water from freezing, it will cool me down well below ambient.
Again, thanks for all your help guys. I'll be sure to post pics and keep track of all the items I use, as well as an approximate build and time cost. I'm pretty sure I'm nowhere near the first to do this, but I'd betcha there is someone out there who really wants to do it, but thinks it would be too difficult (like me a year ago).
Have a great remainder of your weekend.
-Scott

Well, a badly positioned res along with a rad that is the highpoint can lead to the air in the rad. However, you are correct for most sane situations.

I've already figured it all out, the radiator is going to be a couple of feet down from the ceiling and the tubing is going to run along a wall-board about 4' high (that way I can hide it). I'm going to build a hide-away box for the reservoir that is higher than the tubing/radiator, so I won't have that problem anymore.
All of my hardware should be here either tomorrow or thursday. I usually like getting my stuff sent to me 2nd day, but there was just SO much stuff that it would have cost like $70, so I figured regular ground was good enough this time. I think UPS makes plenty of money without me donating an extra $50 for 1 day faster delivery.
Anyway, hopefully I'll have time to get most of this set up this weekend. I also ordered some blocks of copper and aluminum to make my own waterblock. I have some theories that could make a pretty spectacular block. I'm going to have to give 'em a test first though, before I release it to the world. :dome:

Sounds like you are having fun, all right. Great!

If you are interested in the waterblocks, you should be following the thread about Cathar's G7 block. Interesting stuff there if you have a head for thermodynamics and cooling.

Thank you, I'll be sure to check it out.
My block is going to use the quick heat-releasing properties of Aluminum to get the heat out of a copper bottom...
well, if it works anyway. hehe

Sigh, here we go again.

Have people never heard of thermal barriers?

Baron iv: try going to a junkyard and buying the largest automotive rad you can. If your going to do this in an extreme fashion, you might as well do it right. Try to find something made out of Cu (good luck, I'm yet to see one besides a HC). Maybe just one Iwaki instead of two hydors? An RDL will pack more than enough punch for what your doing.

I have actually looked at a large car radiator, but I found a transmission cooler that would be much more suited for my needs. It's got 2 individual cooling areas instead of using the entire radiator for one loop. That actually works out pretty well because I have 2 different computers that need to be watercooled. I haven't decided if it would be a good idea to run them both on the same loop, that's going to add a LOT of heat. This requires a lot more thought than my vapochill!!
I doubt that I will be trying the block in aluminum AND copper after 2 days of reading on thermal properties of various substances. My "idea" was just another example of me over-analyzing relatively simple things. I think I have the OPPOSITE of ADD...maybe Attention Surplus Disorder. Once I focus on something, it stays in my mind and I think about it constantly until I think TOO much and make it far more complicated than it really is. :shrug:
Now if I can only figure out how to drill holes in copper that are SMALLER than .25mm. :confused:

Please start a new thread with evidence that aluminium is better than copper at cooling. Evidence is not supposition based on a link to the properties. Evidence can be a fully simulated model.