Watercooling GTX 570 SLI and I7 4ghz Loop Question

[skinnee]

Meant x58 ud7 , Well rev 1.0 still isnt supported and its the best out of the 2 revs = /

My rev2 would disagree with your opinion.

[Johnny87au]

My rev2 would disagree with your opinion.

, still my rev 1.0 > yours

[OC_Nooby]

well I have it flipped 180 so the inlet is pointing towards the front of the case (maybe that's why I didn't have any issues).

[matari]

There are a few issues with your setup. Starting with the fans on the radiator; slow speed fans work better in pull configuration. I would move the fans on the top of the radiator pulling the air outside of the case. Second, (I cannot tell how the rear case fan flow of air is) the rear fan needs to push air into the case so that the radiator gets some fresh air. The third thing that I notice is that your tubing has tight bends. I would try to prevent that if possible, you could have a kink somewhere.

[OC_Nooby]

There are a few issues with your setup. Starting with the fans on the radiator; slow speed fans work better in pull configuration. I would move the fans on the top of the radiator pulling the air outside of the case. Second, (I cannot tell how the rear case fan flow of air is) the rear fan needs to push air into the case so that the radiator gets some fresh air. The third thing that I notice is that your tubing has tight bends. I would try to prevent that if possible, you could have a kink somewhere.

I have the following fan configuration:

4 Fans pushing air from the side panel (120mm ~40cfm per),
1 Fan pushing air from the bottom of the case (120mm ~40cfm),
1 Fan pushing air in from the rear of the case (140mm ~60cfm),
2 Fans pushing air from the front of the case (120mm ~40cfm and 230mm ~110cfm),

Then the 3x typhoons pushing air out. I also added the stock haf 932 fan on top of the case to make a push/pull system without any difference in temps.

The case is a HAF 932.

I tightened the CPU block some more and after running 5 passes of intelburntest I got the following via Real Temp 3.60:

69°c 70°c 65°c 68°c max on each core. In reality it was at 68° with random spikes to 69c/70c which only happened once.

Honestly its either I don't have enough flow or I just have a hotter then normal 920 D0. I'm running HT as well.

Ohh the tubing is not kinked, kinda does look like it from the pics.

[Church]

HT=off probably will drop 5-8 degrees. 70C looks good enough for older Supreme, not HF, waterblock (current leading blocks probably might win extra 3-4 degrees, imho not worth buying new block for that).

[OC_Nooby]

When did they come out with the HF version?

[OC_Nooby]

I was also wondering, how much can an MCP355 push? Can one pump say handle 2x 360 rads, 2 videocard blocks, 1 cpu block and chipset block? I probably wont be using 2x 360 rads but found another 120.1 rad that will help with the video card blocks.

[Johnny87au]

I was also wondering, how much can an MCP355 push? Can one pump say handle 2x 360 rads, 2 videocard blocks, 1 cpu block and chipset block? I probably wont be using 2x 360 rads but found another 120.1 rad that will help with the video card blocks.

Well 355 would be better then a 655 with what you mentioned since it handles more restricive loops, For a loop like that it should be able to run it but i run 2x 355 in serial which works out twice as good And easily enough for any setup i go with, As for the pump they push ~ 120 GPH (454 LPH) with around 20ft of head, Even more with a EK top..

[OC_Nooby]

Alright, I say screw it. Just to watercool the cards would cost me what the card would be worth in 3 months.

I appreciate all the help and support everyone has givin me in this thread/section.

Thanks!

[jadawgis732]

remount your block, and if that's not the solution get a flow meter to ensure the loop is not too restrictive

[morencyam]

How it all flows.

i have always read to have loops run rad>cpu>mb block/gpu/etc>res>pump>rad

maybe your current loop configuration is causing the high temps. the heat from the pump isnt getting dissipated before it enters the blocks

[Holst]

i have always read to have loops run rad>cpu>mb block/gpu/etc>res>pump>rad

maybe your current loop configuration is causing the high temps. the heat from the pump isnt getting dissipated before it enters the blocks

The order of components makes no real difference to temperatures.

[Johnny87au]

The order of components makes no real difference to temperatures.

+1 , just make sure the res is before the pump.. And remember only pump/cpu blocks have inlets/oulets

[OC_Nooby]

Rad to Pump/Res to CPU to Chipset then back to Rad.

I found a 120x1 rad which ill add to the loop. So I'll install it before it goes to the 120x3 and add 2 fans for push/pull. God its going to be fun removing the bubbles from 2 rads in series

Should have results tomorrow.

[Brazen_NL]

Those are some pretty high temps. I only use 1x 120 Black Ice GT Stealth XFlow with a TFC TK-122 and I'm getting 85 max on cores running Linx 25000 for 20 runs. i7 950 is running at 4.23 GHz with HT. Vcore 1.399 under load. Screenshot/photos available if you need to see that.