So this is what happens when Attila takes a shot at water cooling, I absolutely love it 100% awesome .
You are going with a Al + Copper loop? I assume you will be using 50/50 mix then? Personally, I think you should have gone with a full Al loop. Your temps will be a bit worse, but you won't have problems with mixed metals at the blocks. However, you will still have problems with the rads.
Beautiful block- it's a fantastic idea to have it go behind the board! One of the best works I've ever seen! Rep.
you can get your hands on an all aluminum rad, not really designed originally for PC use but with some slight mods which Im sure he can handle you can use it for such. I used a Outlaw race kart aluminum rad once for a build project
If you were closer I have a Koolance 4x120 rad with shroud that's all AL. In fact, for shipping it's yours if you want it. We get home tomorrow, I'll find it a box.
Not wishing to state the obvious, but... the mixed metal debate isn't ending Attila. Surrender the cooling system to one or the other may be the only viable option.
Thanks for the replies and thanks for the concern at the haphazard way I've approached this. And Kayin, thanks a lot for the offer of the rad, that's very generous of you. You know I've never really been a fan of water cooling but I do like a challenge, and this is certainly a challenge. I really wanted to see if I could build a w/c system, just once, and build it my way. When ever I start thinking of something, a picture forms in my mind, a picture of how it will all look and work, of the materials I will use, colours, finishes even how I can make the parts. This picture is very hard to shake. The blocks have to be anodised alu. They just have to, that's all there is to it, I need to make this work. The only reason I've become attached to the copper bottoms is from the experience of 'working' the metal. This stuff is very different from the alu I'm so accustomed to. Sure it's more difficult to work but it's what I noticed when cutting and filing that won me over (as far as a material for cooling). When I cut or file or drill alu, the work piece and the tool become hot, but when working with copper, the piece got hot but the tool stayed cool. It appeared to me that the copper was like a heat 'magnet', sucking the heat away from the tool, I imagined that sitting on top of a processor, the copper would 'suck' the heat out and pass it on quickly to the flowing liquid to be removed. But if I want to keep the alu blocks then the copper will have to go. Not only that but the brass radiator will have to go as well. In this scenario the only small problems will be the nickel plated fittings. Now nickel is even further away from alu than copper on the galvanic scale, but because the ratio of alu to nickel will be so great (alu blocks, water channels, radiator), that with some finely tuned anti-corrosives, the system should have a long working life. Oh, and I wont be using 'anti freeze' or anything with ethylene glycol. The other problem may be the efficiency of the bottoms. But these will be easy to remake and fit if they need to be (because of really bad cooling) down the track. Now if (dreaming again) I had a mill right here, then I would simply make the entire block from copper and mill a chunky alu 'sleeve' to cover it. But too much time and effort has been expended making what I have and I don't have the will to do it all again. Now if you'll excuse me, I need to PM Kayin.
I think everyone is over thinking this whole galvanic corrosion deal. Inside a computer cooling system is a very controlled environment. ie. no external unpredictable scenario's to contend with. if you look at the" Anodic Index" of Aluminium and Copper, and work out the galvanic difference it's only 0.55 Volts, which is almost with the allowable tolerances (0.5Volts diff). with a reduced conductivity agent in the coolant this would be no problem at all. And he's going to Anodize the Blocks!! as long as it's both inside and out, I see no problems with the setup that Attila has. if all the aluminium is anodized then brass, nickel and copper are all that's left in the mix and they have very close galvanic index's, with only a 0.1volt spread. No problem as I see it. PS. Great Work with the blocks by the way, they look really good.
The only problem with that is that anodizing is tough to get to stick on complex surfaces. If you want to use copper bases, get them electroless nickel plated. You shouldn't have any issues then.
Actually, it's not an over blown issue. Some people have run mixed metal loops pretty fine for months with out many issues. However, there have been quite a few people that this: Source for the pics: http://forums.koolance.com/viewtopic.php?f=5&t=207 http://www.xtremesystems.org/forums/showthread.php?t=203601 Koolance was notorious for this problem. This is why I STILL don't buy Koolance stuff. As you can see, it's not really a minor issue. Go through XS threads and you will find plenty of corrosion stories + pics,etc. edit: How the hell do I re-size images so they appear at 800*600 in the Forums? I'd rather not do this manually if it is possible....
Those blocks used a sacrificial layer of zinc in between the aluminum and an outer coating of nickel. With bare copper, the zinc was a sacrificial anode, and actually caused the aluminum to corrode faster. The better choice would have been in that case a delrin top and a nickel plated base. Nickel is pretty noble, so it's tough to provoke a reaction from it. While the GTX case was a severe one, I've actually seen far worse than that. It all depends on loop chemistry, specific processes used in manufacturing and if you held your mouth right while you put it together. Also, that center pic is from the rusting steel Koolance fittings that were pulled from the market. Koolance cleaned up its act, enough that I'd rather recommend them over EK right now.
I don't care how long people have been running loops with mixed metals. At one point or another, physics will win. And when it wins, it will kick your ass. Anodizing isn't always thorough, things get worm down and chipped, etc. Mixed metal + water = bad. On a fun note, you're like a damn factory. Don't you ever take a break?
Even factories need a break now and then - this guy is much more persistent. He must be an alien. Or an unstoppable hun.
I've seen those pictures before, and worse! I'm not concerned about the corrosion likely to occur in the blocks or water channels, The feature sizes here are very large and the material very thick. My main worry was inside the radiator where the channels are small and the material very thin. Corrosion in the system could lead to blockages here and after time, even perforation of the thin walls here. I still need to read up on which metal in a mixed metal system will suffer the most deterioration. If anyone knows this please reply before I ditch my rad. Anyway, life goes on. I had about an hour spare today, so a start was made on the GPU block. This time I will do this a bit differently by drilling a bunch of holes. This will make routing the channels a lot easier. It got a little messy. And ready for routing tomorrow, if I have time. I should be able to remove a fair bit of the bulk of this towards to front (top of the case) so I need to think of a design that can make this part look a little better.
