- Jan 21, 2019
- 2
- 0
- 1,510
Hi!
I'll start with specs:
XSPC Raystorm Pro CPU block
EK-FC1080 GTX TF6 GPU block
Alphacool D5 Pump running at 3000 RPM
HW Labs GTX 480 + HW Labs GTS 420 Radiators.
So, after a lot of testing, I noticed that GPU reacts very well on increased flow, while the CPU doesn't care much. In the case of GPU, it's a 5C difference between the lowest speed (~1800) and the highest (~4700).
Currently, my loop is in series: Reservoir > Pump > CPU > GPU > Rads > Reservoir.
What I want to do is: Reservoir > Pump > Y-splitter > Parallel to CPU and GPU > Y-splitter > Merging CPU and GPU outlets into one tube > Rads > reservoir.
Here is the picture of how I want to make it, and also how it's routed now: View: https://imgur.com/a/qir7uUQ
Why I expect it to be beneficial - Both blocks receiving the cold water of the same temperature which theoretically should be transferred to better cooling. -1C -2C is a win-win with my OC.
Why I expect this idea to possibly be flawed - drop of the pressure in the place where the flow is splitting; Univen destructiveness of the blocks.
I know from reviews that XSPC Raystorm is a pretty low restriction block (0.67 PSI).
I found only one pressure drop data for EK-FC blocks which shows that my GPU block has a 0.79 PSI drop which results in a 16.4% difference between blocks.
Assuming that the pump speed will remain the same, (3000 RPM is the highest silent point this pump has), what is going to happen with a flow/pressure when it hits the Y-splitter on the way to blocks?
Does 16% difference is so dramatic? Will Y-splitter on the other end (that merging block's outlets that go to the rads) will equalize the flow?
I would greatly appreciate any info and I especially looking to here some input from rubix_1011.
Thanks!
I'll start with specs:
XSPC Raystorm Pro CPU block
EK-FC1080 GTX TF6 GPU block
Alphacool D5 Pump running at 3000 RPM
HW Labs GTX 480 + HW Labs GTS 420 Radiators.
So, after a lot of testing, I noticed that GPU reacts very well on increased flow, while the CPU doesn't care much. In the case of GPU, it's a 5C difference between the lowest speed (~1800) and the highest (~4700).
Currently, my loop is in series: Reservoir > Pump > CPU > GPU > Rads > Reservoir.
What I want to do is: Reservoir > Pump > Y-splitter > Parallel to CPU and GPU > Y-splitter > Merging CPU and GPU outlets into one tube > Rads > reservoir.
Here is the picture of how I want to make it, and also how it's routed now: View: https://imgur.com/a/qir7uUQ
Why I expect it to be beneficial - Both blocks receiving the cold water of the same temperature which theoretically should be transferred to better cooling. -1C -2C is a win-win with my OC.
Why I expect this idea to possibly be flawed - drop of the pressure in the place where the flow is splitting; Univen destructiveness of the blocks.
I know from reviews that XSPC Raystorm is a pretty low restriction block (0.67 PSI).
I found only one pressure drop data for EK-FC blocks which shows that my GPU block has a 0.79 PSI drop which results in a 16.4% difference between blocks.
Assuming that the pump speed will remain the same, (3000 RPM is the highest silent point this pump has), what is going to happen with a flow/pressure when it hits the Y-splitter on the way to blocks?
Does 16% difference is so dramatic? Will Y-splitter on the other end (that merging block's outlets that go to the rads) will equalize the flow?
I would greatly appreciate any info and I especially looking to here some input from rubix_1011.
Thanks!
Facebook
Twitter
Instagram