Pump placement disagreement
 

Hi, I'll shortly be putting together my W/C setup, and I've got a Swiftech MCP600, 2 heatercores, a MCW-6002 and a Maze 4 GPU.

I was going to Link it up so the system was MCP600 -> Both rads -> MCW-6002 -> Maze 4 -> Back to pump.

But he told me that that'd mean the water would be going slower to through the block as it's not connected to the pump, and that the radiators wouldn't perform as well as if it was RAD-> MCW-6002 -> RAD -> MAZE 4, as the water would be warmer when it reached the Maze 4 in my setup.

To me the remark about the slower flow makes no sense, as water must go the same speed through the whole setup as water can't be compressed, just like electrical current is exactly the same everywhere in a series circuit.

I also thought that his remark about seperating the rads wouldn't make much effect as the water barely changes in temperature at all in just one pass?

So basically, is he right? And even if he is/isn't, would you advise I put things in a certain order anyway?

Many thanks,

Cougem

Your original sounds fine to me. Having the two rads in parallel reduces the amount they impede flow--they'd each get less flow, but rads are more sensitive to air flow than water flow. Beyond that, just like you said, the flow doesn't depend on order, so hook stuff up however is easiest for you.

Yeh I guess the parallel idea is good, but what you'd gain from extra flow you'd lose from heat dissipation performance wouldn't you, as the water is moving slower and would pass only pass through 1 of the rads before rejoining the system.

The temperature drop across a rad is on the order of tenths of a degree. You'll gain more by the decreased restriction and increased water flow through the blocks.

Even with an MCP600? I thought since the MCP600 had so little pressure loss that the series setup would be lovely.

Anyone else got any opinions on it?

I'd try and get the pump right before the cpu water block--and just like Groth said, the water really isn't all that warmer (like, you wouldn't be able to tell which was warmer if you could somehow touch the water) after it goes through the block. A lot of people seem to believe that cool water goes into the waterblock and then comes out hot, but this is not really true.

Bruce

Why try and get the pump right before the CPU block though, doesn't water have to be the same speed all around the circuit?

.... That's a really good point. I had this notion in my mind, and I swear it made sense--but now, I don't guess it would make a difference. My bad, thanks for pointing that out :)

Bruce

if i'm not mistaken, the best placement would be the one that suits your setup the most, reduce the number of bends, kinks, and make it as clean as possible..

seconded... putting the rads in parallel would gain you very little extra flow (if any once you add in 2 Ys) in a run with a block as restrictive as the 6002...

when i reversed the flow in my rig, i saw no temperature drop. i'd do as above and use the most convenient and least restrictive flow routing.

Speed, well, the same *flow* but not the same *velocity*. Any change in cross section area in the loop will give a proportional change in velocity. Second thought: *Not* proportional, since change in cross section area will give a change in total back pressure, wich also will give a change in velocity.
(sh*t, I´m confusing myself again)
Well, it wouldn´t make much difference where in the loop You put the pump. The velocity in the water inlet in the blocks would be pretty much the same.
Otherwise I go with Groth, hook it up the easiest way.
regards
Mikael S.