Project BSG Scorpion Fleet Shipyards (Lian Li PC-A77B)

[luke997]

So I've swapped 2x450S for 3xDDC 3.25, fired WC up and....

175 l/h

In all fairness, it required a total of 7 45's more than 450s.


That's where I thought I'll try it EXA's comment

I would look to try and reduce restriction instead. Going from serial to paralell gfx again, reducing 45s to a minimum and so on...
I know, 45s doesnt have that much restriction, but they do have some, and with the amount you have it will add up.
http://www.xtremesystems.org/forums/...d.php?t=266003

Im not saying that your loop is restrictive just because of the 45/90s, i dont. If i should take i guess, i believe its the mosfet and nb block thats the biggest sinners.

I just think its un-neccesary to use alot of them 45s when the tubing can do the job just fine

I've dissasembled 2/3 of the loop and removed this merry band of fittings



All replace by straight compressions and hose. Couldn't do any more, my this Primochill is very stiff but the results were worth it:

211 l/h

Considering previous flow with 2 DDC's and even older with 2xD5's - one conclusion - this new EK blocks are much more restrictive than previous wave design.
Nothing wrong with it if the temps are better - will check this later.

Adding more pumps has a really big diminishing returns. Look at Stephen from Swiftechs little (for fun) test. going from 3 to 5 pumps gained nothing.
http://www.swiftech.com/stephens-corner.aspx#tab3
I dont think you will gain much at all going from 2 to 3 pumps in your really restrictive loop.

Here's scaling in my loop with 3 pumps:

1 - 122 l/h
2 - 175 l/h
3 - 211 l/h

3rd pump is worth it in my book, especially if one dies again, I'll still get enough flow to keep rig running for a while

DDC's run @ 4500 RPM, got 55C temperature inside the housing and are much quieter than 2x450S, not to mention lack of any vibrations.