Two Loops, One Res..Possable?

Associate
Joined
6 Jun 2005
Posts
1,431
Location
Church Fenton
I just started out in the big world of watercooling and Iv got a loops that cools my chipset because its stupidly toasty. 48C just idle.

The kit Iv got for this is:

Danger Den CSP-MAG 12 Volt Pump
Tygon 3603 1/2" ID Clear Tubing
Swiftech MCRES-MICRO™ Reservoir
Black Ice Pro Compact High Peformance Radiator 120mm
Silverprop Cyclone Nexus SX chipset waterblock

That should cool it down a tad, but I have bigger plans to cool this spec:

4400 @ 2.6>
DFI Expert
6800GS @ 500/1200 (cool enough on air cant be clocked anyhigher)
Akasa Eclipse.

This is my plan, note all right angles will be nice curves or y-joints.



My question is Will this work or will the loops eventualy heat up?

If it does I can maybe put in a better clocking CPU :D
 
R B CUSTOMS said:
so let me get this straight... yo havnt bought any of this yet right ?
theres really no need to cool the chipset specialy using a seperate loop !
unless your just doing it for the hell of it i think its a bad idea.

what about your graphics... how are they cooled?


I have the chipset loop already. I cant actualy use my system to its max or it just crashes and burns. Goes to about 60C and crashes thats with the 120mm fan on fun and the zalman (with not pins chopped off)

Maybe I could put the GPU into the CPU loop. its cooled by the vf700-cu at 5v.

CaDLoN said:
:confused:
Correct me if im wrong, but...

It looks like the output of the CPU block & the output of the Ice Pro are both inputting into the lower (what looks like a) radiator on sepaterate connections.

I foresee water out side of the loop
:eek:

Yeh it kinda does look like that. Ill do a flow chart.

Res > (Chipset Loops) Pump CSP > Block > Ice Pro Rad > Joining tube back to res

Res > (CPU/ +GPU) Pump MCP > Block/s > Lower Rad > Joining tube back to res

My idea is that each pump should get its own flow of water, but the question is will the res 1/2" tubes that is feeding two pumps bottle neck.

The CSP doesn have high flow rate it just has pressure so its a matter will the MCP draw to much water through the output of the Res

Also which one of the two tubes on the Res is the input ment to go into, its designed to remove air from the loop but I cant work out which one you have the input hooked into.
 
Last edited:
R B CUSTOMS said:
yikes ! that sound drastic !

as for the res...


the bottom one is taking water OUT where as the top one is squirting it back in... the idea is for it to hit that plastioc tab to stop it making noise / getting air into the system
www.swiftnets.com / products then cloiock ont he res link
it has 3 videos showing it in action ;)

as long as theres enough water int he system it should be fine.
what about just jhaving the 2 loops completly seperate from eachother using the res for one and a Tline res for the other ?

if your planning on going for tyhe CPU loop id take anything over a apogee. apparently swiftech have back tracked on performance with that block tfor the sake of production cost.

Thanks for the link, I was testing out you see and it looked to just put the air back into the system I guess it will eventualy get all the air out.

As in t-line res you mean a fill port? I do have room for two res's so thats ok.

Oops I will be getting the storm not the apogee I just put what was on Ocuk on to the list.
 
R B CUSTOMS said:
ah.
storm... great block ! well one of the best
if yiou fill the res up a bit more it may take some of the problems away but it does take a few hours for air to get out of the loops. you iwill need to run it 'open air' ( so the res aint sealed' for a couple hours to get all the air out.

that pump and the storm will be a wicked combination.
you could use one res but personaly id use two ( well id just use a T line but thats besides the point)

one thign though... on your diagramns... and i know its JUST a diagram but remember that the water goes IN the horiuzontal barb of the swiftech pump and out of the verticle one. you want your res to feed into the horizontal feed.
again i knwo its just a diagram but just incase you had it mixxed up thats all ;)


Thank you for the pump advice I need to keep that inmind anyway I keep forgetting.
 
welshtom said:
I had two pumps and one res.

output of the pumps went to their individual loops.And rejoined at the rad

The input to the pumps was from the same source.

Tom

I was waiting for you to come along :)

So did it work well at all, were they both same pump or different?
 
Back
Top Bottom