Loop tweaking...

[speed bump]

The difference in size between 8mm and 3/8" is about a 1/16" so yeah there isn't much difference between them.

The difference between 3/8" and 1/2" is about twice that though. You end up with a fair amount of difference between them, is it always measurable with a $.05 diode? no its not and part of that depends on the block and pump you use. A storm and an Iwaki RD30 are going to probably show you a couple of degrees between the two. However, you might only see one degree difference on a Apogee and an Ehiem 1048 just becuase of the difference in restrictions and pressures in the loop.

Also GPUs tend to heat up your loop very well, even running a Bip3 with over 300cfm going through it I still was 2 or 3 degrees hotter on my CPU than with just the CPU in the loop. So I have to say if you have a spare loop then use it becuase two GTSes will put out more heat than your cpu anyday. (unless your running dual duals or quads).

My thought on your setup would be if you don't want to buy a buy a new CPU WB then don't, Robotech at syscooling did a test a while ago that showed a Maze 3 was within 1°C of a storm on an IHS loaded CPU.

As far as setup I would go 1/2" all the way and setup it up as two loops like this.

loop one.
Maze 4 CPU
Chipset block (the MCW30 is a good choice I would stay away from the maze 4 though the Acrylic top breaks fairly easily.)
Mag2 LE
single HC or if you feel like upgrading it looks like you could put a double in there.

Loop2
GPUs
Ehiem or if you don't get the mag the strongest pump
Double hc.

 

[GreenJelly]

I understand the math behind it... But take a close look at the gear out-there... Reservoirs with 1/4" tubes... Water blocks that are really restrictive, and then Radiators that also restrict the water (though this is the least guilty party involved).

All I’m saying is that it Depends on your setup. I saw 0 changes in my flow rate when I dropped to a lower diameter tube. This is from physical experimentation.

Theory is great, but it is the beginning of the scientific process, and not the end

 

[torin3]

As far as setup I would go 1/2" all the way and setup it up as two loops like this.

loop one.
Maze 4 CPU
Chipset block (the MCW30 is a good choice I would stay away from the maze 4 though the Acrylic top breaks fairly easily.)

I already have the maze 4, bought it several months back, and with 3/8" fittings. I'll be switching it over to 1/2" ones.

Mag2 LE
single HC or if you feel like upgrading it looks like you could put a double in there.

Well, the pictures may not show it, but there is about a 2" overlap in space of where the larger heater core and the PSU would intersect. Even taking the fans and shroud out and putting them on the outside of the case won't work...still about 1/2" too wide. Believe me, I was ticked off about it when I realized what had happened.

Loop2
GPUs
Ehiem or if you don't get the mag the strongest pump
Double hc.

Thanks. This fits with my gut feelings about the pumps. I know it is more restrictive, but I was thinking about using a 90° just prior to the intake on the Eheim pump as I've got a little bit of a routing issue. If I can turn the pump about another 35° I can mount the Mag II next to it. However, it would push the hose up against the video card and kink it.

Also, I've only got room cut out of the top for 1 fillport. I was thinking of using a Y connector for the T-lines. From what I've seen there isn't a lot of movement of the fluid in the T-line, so I think that will be ok.

I've also got about a quart of MTC-40 that Danger Den threw in as an extra. Is this still considered a good fluid to use in the loop? I've been using Zerex, but I still get cloudy loops.

 

[QuietIce]

I usually have no problems with a 90° bend (though many on here scream about them) but the pump inlet is a very bad place to put one. I can't tell very well but can the pump be mounted on it's "back" (smallest side) with the base up against your drive-cage panel ...?

 

[torin3]

I can't tell very well but can the pump be mounted on it's "back" (smallest side) with the base up against your drive-cage panel ...?

THAT is a distinct possibility!

Thanks!

I may have to angle it slightly to get the holes to line for screws, but if that doesn't work out, I can mount it with one screw, and then use a neoprene pad to rest the 'base' on.

 

[GreenJelly]

Bah... People around here are nuts... 90 degree angles are BAD... they say... 3/8 inch tubes are BAD they say...

Hell, if they could, they would be using 2 inch tube in their systems... well to be fair, If I could I would be using 2 inch tube in my systeml... Just cause its so bad ***

 

[QuietIce]

No way I'd try to run 3/8" tubing out to my rads and back - it would be worse than adding an Apogee to the system. Not that the Apogee is a problem but I'd rather stick with 1/2" tubing and add a REAL Apogee for my wall-mount CPU ...

 

[Sleepy_Steve]

Also... Is there any way to mount the PSU on a spacer of some sort to have it sticking out the back of the case just enough to get a second 2x120 HC in place??

Im not at all sure how feasable or practical that would be... I was just letting the ideas do thier thing.

 

[speed bump]

Bah... People around here are nuts... 90 degree angles are BAD... they say... 3/8 inch tubes are BAD they say...

Hell, if they could, they would be using 2 inch tube in their systems... well to be fair, If I could I would be using 2 inch tube in my systeml... Just cause its so bad ***

Its actually the science of fluid mechanics that so those things.

I'm playing around with the pressure-drop.com calculator currently and I think I might be able to give you some numbers to back this up once I have figured out exactly what I am looking at in at and how the results are to be interpreted. I think I have a way to optimize tubing size for you pump though.

 

[torin3]

Well the GTS derlin blocks have been ordered from EK. Now I just have to wait for international shipping to get them to me. Plus I pulled the trigger on the other GTS card. Spendy day today.

 

[torin3]

Also... Is there any way to mount the PSU on a spacer of some sort to have it sticking out the back of the case just enough to get a second 2x120 HC in place??

Im not at all sure how feasable or practical that would be... I was just letting the ideas do thier thing.

Not really possible. The support the PSU gets is from a sheet of acrylic inside the case under it. Plus the mounting holes aren't mirror imaged, so that won't work either. And without support, it is going to pull on the side of the case pretty hard from leverage and gravity.

 

[GreenJelly]

Its actually the science of fluid mechanics that so those things.

I'm playing around with the pressure-drop.com calculator currently and I think I might be able to give you some numbers to back this up once I have figured out exactly what I am looking at in at and how the results are to be interpreted. I think I have a way to optimize tubing size for you pump though.

Your confusing Scientific Theory for Scientific Experimentation. Theory is great, but it must be backed up by sound Experimentation for it to be true.

Now their are ALLOT of variables in a cooling system like the ones we use, and those variables are not easily accounted for. They definitely cant be proven by a simple calculator.

With that said, I have done the Experimentation, and I have found that their are larger restrictions on my system that are so restrictive that it doesnt change the flow when I drop my tube size from 1/2 to 8mm.

The Storm is VERY restrictive. It comes with both 1/2 and 3/8" ID barbs. I found that at my pressure their is no difference between the two and their performance. Again this is backed up by experimentation. My VGA cooler really is unrestrictive, and does change flow slightly when I reduce it to 3/8"... but already I got a restriction that is greater then that of the VGA block preventing water from moving through any faster, because the Storm is restricting the flow.

With that said, I run 1/2 Too my Storm, and 1/2 from my resivor. Its easier to do this then to convert the pump to another ID. Inside my resivor is a 8mm OD plastic tube. So no matter what I do, it will be restricted to this flow.

With that said I have spent hours AT Frozen CPU, looking at almost every part they have, and I always see major restriction points that equate to 1/4" ID.

So How can you pump the water contained in a 1/2" tube through a component that only has a 1/4" ID restriction? Can you compress water, were as the rest of us are left stuck with the rules of nature? If you can, please tell me, cause I would REALLY like to use this to push more water through the system.

The truth is that I have ALLOT of pressure, and I have VERY fast moving water, and that after all is said and done, I get great cooling, with very little need to change anything.

Now if you dont want to physically test out your equipment, then use a 1/2" or 1" or 2" tube, cause it may be beneficial for you... If you want to get down and actually test, and take numbers with your system then you can find a tube size that is smaller yet doesnt effect performance.

When I run my tests, I have a jar to empty the water in, and I have a jar to suck water from. The test has to start the pump and the whole system has to have water in it, then you can start with a timer. The amount at start, with the amount after 60 seconds can be calculated as your overall flow.

With my system, I had to do allot of things to accomplish this. I needed to put on a splash cap of my outlet water container or it would spray out so violently that it would make a mess everywhere. BTW, I use 8mm tube to and from my Radiator. I am unable to use anything else do to physical restrictions of the construction and placement of my radiator. Yet I dont have ANY heat issues. In addition when I pulled my radiator out and put on some 1/2" barbs on it, I found that I had the same actual amount of flow that I had with the 8mm Compression fittings.

There is no point in quoting or calculating the difference of restriction and flow, when the end result is that through the scientific process (which ends with Experimentation) you receive results that are contrary to your theory's.

Mike