How To Setup Watercool Loop?

[JoeOnePack]

I am trying to figure out how to run my loop and incorporate the motherboard ports into the equation. Here is something I drew up. Any suggestions?

 

[Tsumi]

From GPU to GPU, just go straight up. There is no designated in/out for GPU blocks.

Otherwise, it looks fine to me.

 

[JoeOnePack]

So, the water will distribute evenly in the blocks?

 

[Tsumi]

Yes. There are channels within the blocks to direct flow, but the design of the block is to be bi-directional.

 

[DogChainX]

Looks good. Just bleed it at full-speed if the pump is variable (after you do leak testing), otherwise it'll take forever to get the "glug glug" sound out.

 

[JoeOnePack]

Looks good. Just bleed it at full-speed if the pump is variable (after you do leak testing), otherwise it'll take forever to get the "glug glug" sound out.

Yessir. Pump is variable. What setting should I run the pump on for leak testing?

 

[Tsumi]

Full speed/max settings.

 

[Crosshairs]

Tsumi, correct me if I'm wrong....I think it would be a lot more efficient with the front radiator inlet/outlet at the top rather than the bottom....with the inlet/outlet at the bottom you are sure to have air trapped at the top not to mention making the pump work harder to pump the water up rather than letting gravity help with the flow.

also, dont you want the reservoir bleed as the highest point in the system?

just my .02

 

[JoeOnePack]

Tsumi, correct me if I'm wrong....I think it would be a lot more efficient with the front radiator inlet/outlet at the top rather than the bottom....with the inlet/outlet at the bottom you are sure to have air trapped at the top not to mention making the pump work harder to pump the water up rather than letting gravity help with the flow.

also, dont you want the reservoir bleed as the highest point in the system?

just my .02

Well I won't rule out switching the front radiator. I just drew up a loop to represent the front radiator facing up. What do you think?

 

[Mr. K6]

Well I won't rule out switching the front radiator. I just drew up a loop to represent the front radiator facing up. What do you think?

Better, although I would get rid of the res altogether: in that configuration any small air bubble that i collects will be immediately sucked back into the system (defeats the purpose). In general I think reservoirs are superfluous and dangerous - they really only help with filling and bleeding (which you do once) and the rest of the time they're a big risk for leaks/damage. That said, if you do plan to use the res, outlet goes parallel to or below the inlet, not above.

Btw, what case is that? Looks like you have a lot of room.

 

[JoeOnePack]

Better, although I would get rid of the res altogether: in that configuration any small air bubble that i collects will be immediately sucked back into the system (defeats the purpose). In general I think reservoirs are superfluous and dangerous - they really only help with filling and bleeding (which you do once) and the rest of the time they're a big risk for leaks/damage. That said, if you do plan to use the res, outlet goes parallel to or below the inlet, not above.

Btw, what case is that? Looks like you have a lot of room.

TBH I probably won`t get rid of the reservoir. If the res ever leaks, at least the loop looked good. haha

The case is a Corsair C70. I have more room than I need. I almost wish I went with a smaller case.

 

[DogChainX]

Yessir. Pump is variable. What setting should I run the pump on for leak testing?

Not to contradict Tsumi, but I'd start of at the lowest setting initially. If no "slow" leaks, then crank it up and retest for leaks that occur under the highest setting. Especially since your motherboard isn't inverted (where inverted is cpu socket is at the bottom, yours is at the top) and have maximum risk of damage if top radiator and/or cpu connections leak...maximum damage meaning you fry your graphics cards too since they're right under the CPU.

I did leak testing once where two of the barbs on an external radiator I was using were old school kind, and didn't seal properly with the hose clamps. Both leaked a little under low speed, but leaked a lot during full speed. I just like the extra step just in case.

 

[BababooeyHTJ]

As long as you have the res over the pump nothing else matters. Just do whatever is easier.

Most importantly be sure to flush the rads and I would cut in a drain line. I find them very handy.

 

[Tsumi]

Tsumi, correct me if I'm wrong....I think it would be a lot more efficient with the front radiator inlet/outlet at the top rather than the bottom....with the inlet/outlet at the bottom you are sure to have air trapped at the top not to mention making the pump work harder to pump the water up rather than letting gravity help with the flow.

also, dont you want the reservoir bleed as the highest point in the system?

just my .02

You're going to end up with air pockets either way. You'll either have the air pocket on the inlet that needs to be pushed down to the bottom of the radiator and then back up again, or the air pocket at the top that needs to be pushed down to the outlet.

Also, once the radiator is full of water, the water coming down the radiator sucks on the water coming up the radiator, creating the siphon effect, and causing no additional stress on the pump. Gravity only helps with the initial passthrough where there is still air in the system, but once the system is full of water, height has absolutely no meaning. It is also for this reason that where the reservoir is doesn't matter.

 

[Crosshairs]

You're going to end up with air pockets either way. You'll either have the air pocket on the inlet that needs to be pushed down to the bottom of the radiator and then back up again, or the air pocket at the top that needs to be pushed down to the outlet.

Water is heavier than air, you will never be able to "push"air down ,as it will always be at the highest point in the system, which is why putting the inlet/outlet at the bottom is just an invitation for a trapped air pocket which will reduce the efficiency of that radiator an amount equal to the surface area of the air pocket.

This is the reason that closed loop cooling systems in which the radiator is not the highest point are always equipped with a bleeder that is located at the highest point.

 

[Tsumi]

Water is heavier than air, you will never be able to "push"air down ,as it will always be at the highest point in the system, which is why putting the inlet/outlet at the bottom is just an invitation for a trapped air pocket which will reduce the efficiency of that radiator an amount equal to the surface area of the air pocket.

This is the reason that closed loop cooling systems in which the radiator is not the highest point are always equipped with a bleeder that is located at the highest point.

With a sufficiently high flow rate, yes you can push air down. And as I said before, no matter what orientation you have, you will have an air pocket that needs to get pushed out.

Additionally, how does it reduce efficiency? The pocket will be trapped at the highest point, the plenums of the radiator. Any air that does make it into the channels will be rapidly pushed down, the force of the water in those small channels do not allow air to escape back up.

 

[JoeOnePack]

Whichever way I decide to run my front rad (decision will be made in next couple days), how do you guys think I should run a drain?

I was thinking of running a Koolance QD on the lower GPU, but I don't think you can actually drain from these fittings. It's a shame because it would have been a perfect setup.

 

[BababooeyHTJ]

I just cut in a tee to the bottom of the loop.

 

[Mr. K6]

Whichever way I decide to run my front rad (decision will be made in next couple days), how do you guys think I should run a drain?

I was thinking of running a Koolance QD on the lower GPU, but I don't think you can actually drain from these fittings. It's a shame because it would have been a perfect setup.

I wouldn't waste my time putting in a drain. When are you going to drain your loop? Why?

 

[Reality]

Tsumi, correct me if I'm wrong....I think it would be a lot more efficient with the front radiator inlet/outlet at the top rather than the bottom....with the inlet/outlet at the bottom you are sure to have air trapped at the top not to mention making the pump work harder to pump the water up rather than letting gravity help with the flow.

also, dont you want the reservoir bleed as the highest point in the system?

just my .02

you're correct, having the ports at the top will allow the radiator to more easily fill up with water versus having the ports at the bottom

 

[JoeOnePack]

Alright well, I am now leaning towards facing the front radiator up.

I am also leaning to not put a drain. I don't think I will be draining it any time soon. Worst case I can just unscrew one of the compression fittings on the lower GPU or pump inlet with the case titles sideways.

Here is a new drawing someone did up for me. I changed a couple lines in picture #2. Which one do you prefer?

OR

 

[sazuzaki]

either one of your drawn up configuration would work, I like the 2nd one better, it would look more clean.

 

[Mr. K6]

I'd recommend strongly considering dumping the MOSFET blocks as 1) they don't do anything and 2) they severely restrict flow. With two full cover GPU blocks and a CPU block there's already a lot of resistance in that loop, adding the MOSFET block might really make your temps suffer. This is especially true if you're trying to pump up to the CPU block through the two fullcover blocks (which may already give sub-optimal performance).

 

[JoeOnePack]

I'd recommend strongly considering dumping the MOSFET blocks as 1) they don't do anything and 2) they severely restrict flow. With two full cover GPU blocks and a CPU block there's already a lot of resistance in that loop, adding the MOSFET block might really make your temps suffer. This is especially true if you're trying to pump up to the CPU block through the two fullcover blocks (which may already give sub-optimal performance).

Ahh man... Why tell me this now.. lol

If I don't dump the MOSFET and decided to arrange the loop more efficiently to combat the heat from the GPU's, would this be more appropriate?

Can it be that restrictive? I can't find the dimensions of the copper pipe being used, but I am already using 3/8 tubing.

I'm going to have to draw up something without using the MOFSET, but looks like its going to be a bit harder.

 

[sazuzaki]

Ahh man... Why tell me this now.. lol

If I don't dump the MOSFET and decided to arrange the loop more efficiently to combat the heat from the GPU's, would this be more appropriate?

Can it be that restrictive? I can't find the dimensions of the copper pipe being used, but I am already using 3/8 tubing.

I'm going to have to draw up something without using the MOFSET, but looks like its going to be a bit harder.

TBH, it doesn't matter what block is first to go in the loop, the temps will be the same. I am speaking by experience, had my CPU first on my loop, then tried my GPU then CPU, even tried putting the radiators between the two blocks, no temp difference at all.


And I would suggest using hardline tubing, Hellfire toyz went in their build from traditional tube to acrylic tubing from primochill and they saw their flow rate doubled. probably because there are less pressure from the tubing.




And another thing, I believe that MOSFET/Chipset block from ASUS is aluminium based not copper, and I believe mixing aluminum and copper blocks can be bad due to corrosion?

 

[JoeOnePack]

TBH, it doesn't matter what block is first to go in the loop, the temps will be the same. I am speaking by experience, had my CPU first on my loop, then tried my GPU then CPU, even tried putting the radiators between the two blocks, no temp difference at all.


And I would suggest using hardline tubing, Hellfire toyz went in their build from traditional tube to acrylic tubing from primochill and they saw their flow rate doubled. probably because there are less pressure from the tubing.




And another thing, I believe that MOSFET/Chipset block from ASUS is aluminium based not copper, and I believe mixing aluminum and copper blocks can be bad due to corrosion?

Thank you for the input. I believe that none of the aluminum actually touches the water.

If you had to choose between the 3 loops, which one would you pick?

 

[jbean7457]

One thing I've learned watercooling over the years is that order isn't as important as some people make it out to be. Water will find and equilibrium temperature and it won't matter. I'm in the process of adding a second 240 rad and will have a similar setup to you, I'm interested to see how your temps work out.

Right now with a 3750k @ 4.4 and a 290 with full cover block I top out around 56c on water playing bf4 for hours on end. This is with my 1450 rpm fans @ 100% on the rad, I'm hoping adding a second rad I can lower fan speeds and stay in the 40's.

 

[Mr. K6]

Ahh man... Why tell me this now.. lol

If I don't dump the MOSFET and decided to arrange the loop more efficiently to combat the heat from the GPU's, would this be more appropriate?

Loop order doesn't matter until your flow rate is very low, and at that point you should be fixing your flow rate. Pretty much the only thing worth considering is pumping high to start out so that gravity helps you with the rest of your loop.

Can it be that restrictive? I can't find the dimensions of the copper pipe being used, but I am already using 3/8 tubing.

I'm going to have to draw up something without using the MOFSET, but looks like its going to be a bit harder.

I have yet to see a MOSFET block that isn't extremely restrictive. They just aren't a good idea for a performance loop. The 3/8" tubing isn't terrible as most parts in the loop are already more restrictive (1/4" can be, but that's another story).

And I would suggest using hardline tubing, Hellfire toyz went in their build from traditional tube to acrylic tubing from primochill and they saw their flow rate doubled. probably because there are less pressure from the tubing.

They must have screwed up the traditional tubing installation, laminar flow is laminar flow.