GomerPyle42

What size tubing would you guys recommend and why?

__________________

TheKraken

1/2'' ID tubing for high flow/low resistance loops. I find that it's best for my system because it dissipates the heat well, but you've just unlocked a whole theory of high pressure vs low resistance in loops.

The argument for high pressure (3/8'' ID) vaguely is: higher pressure of the coolant will keep the coolant moving (bringing cooler water faster) and it breaks away that "film" of non-moving coolant that can form on the waterblock. There's a lot to it, and I'm sure you'll run into the right guys for this topic. Good luck.

__________________
CPU-Z Validation

Click to view my rig's showcase! Watercooling pics UP!!!
2nd Rig: P4 630 @ 3.68GHz, Abit Fatal1ty, GeIL DDR2 667 2x 512 MB, Asus Geforce N6200TC, Aspire 520W, Aspire X-plorer Case, WD 160GB.
Laptop: Dell Inspiron 9100: Pentium 4 3.2GHz HT, 100GB ATA, 1GB (2x512) Samsung DDR 200, WUXGA 15.4'' Widescreen.

pauldovi

1/2 inch interior diamater and 3/4 inch outside diameter is the best and most common size.

__________________

Thumper

The trouble with fast moving water is that it doesn't spend enough time in the radiator (or heater core) to discharge all the heat that it is carrying. This leads to higher temps as the still warm water cycles back to the water block and picks up more heat but not as much if it was cooler.

This similar principle is at work in cars (which is where I draw my knowledge from), which is why when the thermostat in the car sticks open, you can expereince higher temps (although not as high if it were stuck closed! LOL)

DeathONator

There is really no advantage whatsoever with the 3/8" ID tubing in most watercooled setups. Yes, pressure is important, but more so is your GPM. If you really want pressure then you can run 1/2" tubing everywhere, but reduce the tubing right before the waterblock (or at the actual waterblock w/ a nozzle). Then you have pressure where you need it, but don't restrict flow elsewhere.


Also, in regards to the idea that "faster flow rates means less time in the heatercore." That is only true from a limited point of view. To show you what I mean, let's look at an example. Let's compare a system with a high flow rate to a system with a low flow rate. Let's say that it takes the water 20 seconds to make one complete pass through the sytem for the low-flow system, and that it takes the water 10 seconds to make one complete pass through the high-flow system. We will also say that the time the water spends in the heatercore is 1/3 the time of one complete pass through the entire system. So here's what we got:

Low-flow system:
With the low-flow system, the water spends 6.667 seconds in the heatercore per pass. The water would make 3 passes through the entire system per minute.

High-flow system:
With this sytem, the water spends 3.333 seconds in the heatercore per pass, but the water makes 6 passes per minute.

So:
(Total time in heatercore per minute) = (Time in heatercore per pass) * (number of passes in a minute)

Low-flow: 6.667*3 = 20 seconds
High-flow: 3.333*6 = 20 seconds

So yes, the water spends less time getting rid of heat in the heatercore per pass, but it also spends less time collecting heat in the waterblock. And overall, the water spends an equal amount of time in the heatercore for both systems.

__________________
Fold for the cure & team 37726!!
H2O cooling: Chevette heatercore, 172mm 258CFM fan, custom waterblock (made by me), DangerDen GPU block, and Ehiem 1250. Load temp 1.90vcore 45C.