Overclock.net banner

1 - 8 of 8 Posts

·
Registered
Joined
·
92 Posts
Discussion Starter #1
Heres my upgraded layout idea



So i was going to get a waterblock for my card, and a slim 480 radiator for the top of my case. I was hoping to run the loop like so, res->pump->gpu->rad->cpu->res->rad->res.

My plan was to have the cpu go into the top of my res, force the water to move threw the res, and be drawn out the bottom. The idea is to cause move water movement in my res, while also avoiding tubing going infront of my res. Currently before adding anything my res doesnt have any tubing infront of it, i like the look, and hope to keep it that way.

My main concern is, will the water flow well? The bottom would be used as 2 outlets and one inlet. one port going out the side into the front rad, the coming out of the radiator and into the bottom of the res, and then coming out threw the bottom again into the pump.

I also plan on running a monsoon ccfl tube in my bottom rear port, so all 3 of my in/out ports will be coming from the same channel as seen in the following picture


Is that too much water flow for that small area? will i achieve the water movement in my reservoir im looking for?

I do plan on purchasing a multiport top with 5 ports, and a internal tube, that way water going into the top of the res wont splash, because ive read that can cause air bubbles

Is there a better way for me to set this up? im open to idea's. I know it would be easier to simply go from one radiator to the other and avoid going back to the reservoir again, but i really want movement in my res. Kinda bothers how i have it set up now, because its only cycling the water in the bottom of the res
 

·
Registered
Joined
·
32 Posts
I might be missing something but I see you've thought about a drain system but what about filling?

Using the top of the res as an inlet is no issue but I'm not sure how you're going to be able to fill the loop easy?
Turn the case upside down and use the drain port? Because in that case, you'll have to close the drain port, turn the case over and cycle the pump, before turing the case over again and adding extra coolant.
 

·
Registered
Joined
·
92 Posts
Discussion Starter #3
So i was going to purchase an ek multiport top like this


And i was going to get an internal tube to avoid splashing, like this


I was thinking since the top will have a internal tube and multiple ports, i could use one of port on the top to use as a fill port.

redid my mock up so its easier to read
 

·
Registered
Joined
·
32 Posts
In that case it should work just fine!
smile.gif


Put a ball valve on top that you can open for filling and then just close when you're done.



So this screwed on the top of the res, perhaps a 90° fitting on top and then a soft tube fitting. Just attach the barbe when filling?
You can also do it without but if you're going to maintain the loop, I'm personally more a fan of a fixed fill system but that's personal preference.

Good luck!
thumb.gif
 

·
Registered
Joined
·
1,166 Posts
I can't see how you would have any flow at all through the front rad like that. The front rad having its inlet and outlet on base of the res will make it an unpressurised branch of the loop with zero flow.

Its a bit like connecting two ports on the same side of a gpu block to the loop instead of offset, There is no reason for the fluid flow to not just completely bypass the block. You are essentially doing the same thing with the front rad because the two base ports of the res have zero pressure differential to induce any flow.

What I think would look better and actually work would be to flip the front rad so the ports are on the top and have the loop go Top Rad - CPU - Front Rad - Res (top inlet) - Pump - rest of the loop is fine
That will give the front rad flow and create lots of movement in the Res like you want.

The way the loop is set in your pics with just the CPU should be providing plenty of flow and movement through the res. Even with both the Res inlet and outlet on the bottom ports they are separated enough that fluid swirls around quite a bit. It may just not look like its moving a lot because its quite deceptive unless you have extreme flow rates and vortexing going on.
 

·
Registered
Joined
·
92 Posts
Discussion Starter #7
for real? i did some reading before doing my initial build, and people were saying its fine, just have to tilt to case to ensure all the air is purged out.
 

·
Registered
Joined
·
1,166 Posts
Quote:
Originally Posted by Gobstoppable View Post

for real? i did some reading before doing my initial build, and people were saying its fine, just have to tilt to case to ensure all the air is purged out.
Sure tilting the case around is good for bleeding air but that won't help with the loop.

Anyone that actually said you could go Res - Front Rad - back to res - pump and have it work was either not reading / looking closely enough or just doesn't understand how loops work well enough.

Water has to be forced to flow where you want it to go. With the front rad inlet and outlet both connected to the base of the res (or even the top and the base) there is nothing to force flow through the rad when it has a completely free path through the res. Bypassing the rad completely.

You won't find any other loops with parts set up like that for this reason.

To put it another way, fluid flow always finds the path of least resistance, and further, in the presence of two paths the flow will be split such that the flow through each path causes the same pressure drop or resistance. If one path is twice as restrictive it will have 50% lower flow.
In your proposed loop the fluid arrives at the base (or top) of the res from the CPU block and has two paths. One path is through the res body to the outlet. The second path is through the front rad and to the res outlet. The path through the front rad and some short tubing sections offers quite low restriction ... but ... the path through the body of the res offers effectively zero restriction. Its just no contest. The flow will pass through the res body to the outlet with no resistance to cause any parallel flow through the front rad.
 
1 - 8 of 8 Posts
Top