Got my watercooling setup picked out, have some questions...

I'm finally ready to take the dive into watercooling. I've always kept up with the products and lately done more than usual research. I've finally assembled my list of components, but I have a few questions.

Swiftech Storm
Tygon 1/2ID(5/8OD) Tubing (10ft)
Thermochill 120.2
DangerDen D5 Pump
Swiftech Hydrx

Anyway, my first question is, I have two 6800GT in SLI. Sooner or later, or if I have enough money, right now, I plan to watercool these as well with Acetel MAZE4s. I'm wondering if the 120.2 would be able to handle the heatload. There'd be no pelts, just the cards and right now, a 3200+ Winchester, but I plan to get an X2 soon. I'll be using two TT SFII 120mm, manually adjustable up to ~90CFM each. I usually like a quiet computer but I don't notice such things when I'm gaming.

Second, the Hydrx seemed like the best additive at the moment. I just want something that makes the system as low maintenance as possible. In other words, corrosion inhibitors and algaecide. This claims it handles both, and I've never had a reason not to trust Swiftech... but I'm just wondering if there's something better.

Advice appreciated.

JudgeDredd said:

Second, the Hydrx seemed like the best additive at the moment. I just want something that makes the system as low maintenance as possible. In other words, corrosion inhibitors and algaecide. This claims it handles both, and I've never had a reason not to trust Swiftech... but I'm just wondering if there's something better.

Advice appreciated.

Click to expand...

I don't want to get out of line here and recommend anything for watercooling or stuff I haven't used, considering my first waterblock--storm is a few states away in the hands of FED_X. But I had to ask sneaky what additive he was using to get that rich green color in his case and he said this www.frozencpu.com UV green water dye.....he said it was so concentrated it was dark red in the bottle...6-8 drops to color his entire system........so that's what I ordered. But I ordered on color density alone.

Brand new from FrozenCPU
http://www.frozencpu.com/dye-03.html

I think it would be a good idea to get it if you plan on doing an SLI cooling solution for 2 6800GT's and cooling your CPU.

Also, here is an article that Overclockers did on all the Thermochill Rads that you may also want to check out.

You might want to rethink your tubing. 3/4" OD will be much better if you don't want to deal with kinking. And for 2 6800GT's plus your cpu I would go with a thermochill 120.3 just to be on the safe side, I don't think a 120.2 could handle that much of a heat load.

mysterfix said:

You might want to rethink your tubing. 3/4" OD will be much better if you don't want to deal with kinking. And for 2 6800GT's plus your cpu I would go with a thermochill 120.3 just to be on the safe side, I don't think a 120.2 could handle that much of a heat load.

Click to expand...

I think I might just go with the 7/16ID tubing since it's cheaper as well and should prevent the kinking. If I don't do my video cards right now, I shouldn't have a problem with kinking anyway.

Unfortunatly, I can't fit a 120.3 in my case. Maybe I'll stick with two fans for now, and make a push-pull configuration when I watercool my GTs.

JudgeDredd: the 7/16" id tubing with 3/4" od will not kink and work quite well. Also even though the 120.3 will be quieter than a 120.2 with 4 fans you can indeed use the 120.2 with 4 fans at a small loss in temps. However it is nothing critical.

Even going 4 fans on the rad won't help much. I'm with Sewer on this. I see no need to go for a bigger radiator. You will still acheive far better temps than air even with 3 blocks in your loop.

And Judge, I'll chime in as well and say you need to get tubing with a 1/8" wall for yoru SLI setup (your going to need to run some pretty tight bends)

That would be
1/2" ID - 3/4" OD
7/16" ID - 11/16" OD
3/8" ID - 5/8" OD

For you most common tubing sizes.

Jas: going 4 fans does help since it provides more airflow across the radiator. Radiators by design are very restrictive devices and they do better with more airflow. This could mean another degree or two on a properly mounted setup. So for his setup every little bit helps since he can't fit a larger radiator.

If he were running lower rated fans yes, he would get added static pressure. He's more than likely going to want to run his fans at full speed in an SLI setup which should provide plenty of static pressure.

I have to disagree with the push and pull configurations. They really only benefit when using lower speed fans on restrictive radiators. And adding two more fans just adds more noise and essentially bringing you a similar level of loudness and cooling performance that can be achived with two more powerful fans.

A common (and incorrect in my opinion) aproach is to mount a push pull configuration without shrouds. This does virtually nothing. You still have dead spots, and any added pressure from the second fans is being under utilized. If you shroud a radiator, and use fans matched for that radiator, there is no need to go push and pull.

This was a similar argument that Cathar and BillA were adressing when the orignal design concept of the pa160 was being talked about over at pro cooling. Efficient and proper use of a radiator is all about matching the fans.

Wow, I appreciate all the advice, this helps a bunch.

I had originally planned on sticking with my SFII 120mm fans in just a pull configuration. Since then I changed my mind and want to move back to some good old Panaflos, most likely the 120mm Hydro Wave Ultra High in a pull configuration. I figure with the thicker blades and more maximum airflow I should get really good temps for the processor and cards when I'm gaming with the fans on full, and really good full load temps on the processor with the fans turned down and the video cards on idle.

I found the 1/2-3/4 Tygon at DD, so I'll probably be getting that. I might as well just get the GPU blocks while I'm doing all this.

JudgeDredd: the 7/16" id and 5/8" od is either to work with in reality and has minimal flow restriction. So I recommend that

Jas: Push and pull gives 1-2C across the board assuming same shroud and fans pH over at procooling did a good test on this which I have uploaded below if you look at what the CPU Diode reads there is a roughly 1C difference so it may be worth it or it may not be depends on how much you like your ears .

Ok, with the ultra panaflos (what I use), you are getting some serious static pressure. Some quick math, (using some guestimations here, don't take this as a 100% correct for your loop) But you shoudl be looking at somethign like .03 C/W for your rad. Going with a push pull gets you like .028 CW (I'm using some of BillA's calculations for the thermochill rad here, asuming average flow rates). VERY negligable inscrease in performance with push pull.

Higher flow rates will raise the over all performance of radiator (I'm assuming your flow rate will be higher than what I was going off of), and the delta between just a pull and a push pull will shrink and be even more neglidgable.

EDIT:Good find on the chart Sewer. Which rad was that for?
If you are just cooling your CPU, chances are you will see VERY little increase in performance with push pull. Going SLI you are dumping more heat into the system, and with that, you would probably see the 1-2C performance gain.

EDIT # 2: Looking at that chart, there is the same Delta on the Water temp going into the block for a push pull VS just pull, unless im reading this wrong.
My next question is, what is the CFM for the various fans shroud combos? He's obviously gettign more airflow using the push pull (exaust Vs intake air temps have a very low Delta, which seems to imply more air is moving)

Jas: BIXII if I'm not mistaken. To your calculations I'm doubing your numbers since you have no way of getting the pressure with two fans. If you simply added them thats probably not very accurate and again you have no way of getting your own pressure since that can vary from fan to fan depending on number of blades, size and shape of blades and everything. So look at the image I posted which shows a 1C gain on just a cpu.

Sorry for the off topic

Actually I used Billa's numbers using comparable fans (hence why I said not 100% accurate). What's important to note here, is we have no way of determing what teh coolant to air temp delta is, which is signifigant. I again used Billa's generalization of 5C. Now if he has something like a 10C difference, then those numbers go out the window.

This is all great information. I've read long enough to know that the BIX are a lot more air restrictive than most radiators, like the BIP or the 120.x series, so if push-pull affects so little on that radiator, I doubt it would do much good on a thinner radiator.

So I've got most of everything sorted it. My main problem was still deciding on the fluid additives, and it seems most people are using 10% antifreeze and some sort of iodine/algaecide. It got me wondering, does anyone know for sure if there's any aluminum in the components I picked out? I know the Storm is copper, but it doesn't have an aluminum top does it? I don't know about the radiator but I would guess at the very least copper tubing with aluminum fins, or just all copper.

That being said, if everything is copper I might just go with a 5% antifreeze and the recommended ~10 drops/quart of iodine. This makes me feel better than just using Hydrx as I've not heard much of anything about it.

-JD

I use all copper and 95% distilled water and 5% anti-freeze and I have had no problems. The Storm has a delrin (not a metal) top. Also push and pull gives about 1-2C on almost every setup and maybe a little more if you dump more heat in however it will not be much.