Originally Posted by IT Diva
With normal flow rates, this is true . . . . .
But if the flow rate is compromised for some reason, the flow rate thru the rads may be low enough that it's become laminar.
The reason why rads use a flat tube design is to ensure turbulent flow at relatively low flow rates, but that doesn't mean that the flow rate can't drop so low that it becomes laminar.
If that's the case, all those big rads aren't going to mean squat
There are a couple reasons I've seen over the years that cause really low flow rates where they should have been Much higher.
I've had fittings with very small bores where you'd expect at least twice the bore size, and some fittings have at least a 2mm longer thread length than others.
With the long thread lengths, they can go too far thru what they screw into and just barely not hit what's at the other side of the thread. (outlet of a CPU block for example)
All the pumps in the world aren't going to move watermelons thru a hummingbird's butt
Steezy really needs to see what the flow meter tells him at this point to really know what direction to move in to get the temps better managed.
Not the best pics, but you see what I mean about unusually long threads, and strangely reduced bore diameters versus normal on the left.
The stock Bp anti cyclone res fitting is another issue, especially with multi pump setups, it's amazingly restrictive when compared to their hi-flow version.
Originally Posted by B NEGATIVE
Surely when he filled and bled the loop,he would of noticed a low flow problem?
After all,it is a dual gpu loop,low flow should be patently obvious to my mind,especially during the fill as it would be so very difficult.
Steezy,how easy was the loop to fill and bleed?
The fitting thing is bang on,The old push locks i used to use had 10mm of thread on them,many of hours of filing was required......
Originally Posted by Costas
True but something doesn't quite add up with his build as it seems he is utilising fairly standard components and it does not seem all that complex - he is currently running 2 D5's and considering a 3rd...!
We need a tad more detail pics/info on his config as we are guessing somewhat as to what the exact cause is. Maybe he has an oddball fitting etc which is really restricting the flow...?
Flowmeter to verify then process of elimination I guess if that is the cause.
Originally Posted by TheCautiousOne
I just am still having a hard time with understanding the problem of the water temps.. Really just trying to comprehend the frustration. If steezy wouldn't have a water temp meter, what would his outlook be then? The purpose of this is to cool the GPU's better so they do not run in the mid 50C range correct?
Has a remount of the gpu blocks been advised? Is the CPU and GPU loop seperate? Running 2 x D5 pumps isn't a bad thing and not unheard of but possibly unnecessary. I run a single DDC for 2 x 980 EK waterblocks that are attached to a 480mm and 240mm rad.
I might be speaking crazy, but just thinking of other solutions. I feel if we are this deep in the build, then yes, get the flow rate locked in (What is it, is it stable yet, what is flow increase with 3 x D5 pumps)
What is the FPI of the rads you are using and what is the Fan speed at which you are running them at? Are you dissipating heat as efficiently as one possibly can. What is the ambient temperature in the room?
Of the Rads that are used in the loop for the gpus, what are set as intake and exhaust? I see some have mentioned an exhaust route for heat in the lower compartment of the SMA8, do you have even one fan in that compartment? Or is it only exhausted by the motion of the fans used on the outside of the rads?
I will first add the pump, making it a giant single loop using three pumps total. I installed the flow meter, but I still need to put all the tubing in. When I get it all installed, I will check the flow. The flow should be a lot better.
At first I really thought the issue was all the heat in the bottom compartment, and I ask geggeg about it a few weeks ago. I even tried to move the 360 rad out from the bottom compartment and laid it on the side of the case where it wasn't directly in the compartment. Temps were the same. This is why I don't think it is directly related to the heat down there (but what do I know right lol).
BNEG, when I first had two pumps in the loop (CPU including 480mm), my XSPC pump wasn't running a low speed because it didn't work with the corsair link software. So like my Swiftech pump was running 4800, while my XSPC one was only running 2000).
But since it's an extremely large loop (going from from compartment to middle compartment, to too compartment), there are slots of ups and downs and 90's. I'm trying to get rid of some 90's.!but overall, the full wasn't too bad. I will fill it sometime this or next week and let you know.
As of now, I am only using 1 pump (1 pump for CPU and one pump for GPU's). But when I was running two pumps in a single loop, like I said before, there weren't running right.
TCO. Let me simplify my reasons. People have less rad space than me. People also run dual Titans. People run high voltages. People get 5c delta.. What the heck would explain a 15-17c delta?