WWLR 1/2 tubing

[tok3n]

Will the White Water block use 1/2" tubing or will it be available in 3/8"? tubing size in ID

[ralf_c]

hi tok3n

1/2

[dicki]

i thought it's a 1/2" in and 2 * 3/8" out? anyway it all adds up to 1/2" anyway...

if it was 3/8" in then the 2 outputs would go down to 1/4" which just sounds too small to be practical in computer systems though i know the overall flow will be the same...

[ralf_c]

i stand corrected

[Staudie]

http://www.cooltechnica.com/Merchant2/merchant.mvc?Screen=PROD&Product_Code=DTek-WW-SA&Category_Code=NEWP

Comes with 1/2"OD fittings and 1/2"OD "Y" connector

[dicki]

i wasn't correcting, i was as confused as you mate ;)

[Cathar]

It's all 1/2" OD. Makes for a more consistent use of tubing.

[dicki]

welcome to the boards :)

doesn't 2* 1/2" into 1* 1/2" do some weird things to the flow? like cause a lot of back pressure?

2 *3/8 -> 1* 1/2 " would seem to be a better solution as you have to use the Y anyway... though i'm assuming you've done the maths ;)

dicki

[Cathar]

Originally posted by dicki
welcome to the boards :)

doesn't 2* 1/2" into 1* 1/2" do some weird things to the flow? like cause a lot of back pressure?


Remember it started off as a single 1/2" inlet. Yes, it's merging two 1/2" flows, but each flow is running at half-speed, so it's not a problem.

[Player0]

To answer tok3n's original question, the tubing size IS 3/8" ID, correct Cathar? Saying the that the barbed tube fittings is 1/2" OD is a very confusing way to put that measurement (and incorrect if you ask me). I can buy 5/8" OD tubing that will fit the same as 1/2" OD tubing would, except its thick wall. When you are talking about barbed fittings, the correct measurement to list is ID, not OD.

Splitting water flow from a 3/8" ID line to two 3/8" ID lines *does* create flow problems. You will NEVER get even flow out of the two outlets, and you CANT guarantee that both outlets is getting 50%. One outlet might be 70%, and the other 30%. Will this affect the waterblock's performance? Don't know. But it would be incorrect to claim that both outlets get 50% of the original flow because that is unlikely to be true.

You can only guarantee proper flow through each side of the split by reducing the size of the tubing in the split based on ID tube diameter. You will get much closer to a 50-50 ratio if you used two 1/8" outlets (as well as the equivalent flow area on the inside of the block).

Im not familiar with the waterblock design on the inside. However, given the possibility of unbalanced flow, certain channels in the water block may not be working as well as they could be. You could possibly increase the efficiency of the waterblock by creating the proper flow reduction. But, there are too many other variables coming in to play to say for sure whether or not this actually makes a difference on the temperatures anyway.

[Cathar]

No, the tubing that should be used is 1/2" ID.

The barbs are 1/2" OD.

Naturally since the tubing slips over the outside of the barbs, you would want 1/2" Inner Diameter (ID) tubing to fit over 1/2" Outer Diameter (OD) barbs.

Saying 1/2" OD barbs is not confusing at all as it explicitly tells you what size ID tubing you require to go on it.

The correct measurement to list barb diameter therefore is in OD.

Splitting water flow from a 3/8" ID line to two 3/8" ID lines *does* create flow problems. You will NEVER get even flow out of the two outlets, and you CANT guarantee that both outlets is getting 50%. One outlet might be 70%, and the other 30%. Will this affect the waterblock's performance? Don't know. But it would be incorrect to claim that both outlets get 50% of the original flow because that is unlikely to be true.


Given an equal pressure resistance to take the paths to each exit, the flow split will be equal. This is true for the block.

Im not familiar with the waterblock design on the inside.

Water comes in the middle though a rectangular nozzle, hits the base-plate perpendicularly, and then splits to go to the two exits. CFD analysis, along with measuring the flow rates in the real world, shows that the flow rates to each outlet is equal.

[Staudie]

Water comes in the middle though a rectangular nozzle, hits the base-plate perpendicularly, and then splits to go to the two exits. CFD analysis, along with measuring the flow rates in the real world, shows that the flow rates to each outlet is equal.
Cathar,

That being the case could you run each outlet to a different block (gpu and Chipset) and then recombind?

[Player0]

Hi Cathar,

"" No, the tubing that should be used is 1/2" ID.
The barbs are 1/2" OD. ""

Okay, thanks for clearing that up then. I wanted to be sure that the measurements were being listed correctly. You had mentioned everything was OD, so I assumed you meant the tubing, not the barbs hehe.

"" Saying 1/2" OD barbs is not confusing at all as it explicitly tells you what size ID tubing you require to go on it.
The correct measurement to list barb diameter therefore is in OD. ""

Nope, you are correct, barbs should be listed in OD, tubing by ID. I just misread what you meant and thought you were giving tube sizes in OD.


"" Given an equal pressure resistance to take the paths to each exit, the flow split will be equal. This is true for the block. ""

That is true, if there is equal pressure resistance. However, there isn't equal pressure on each 'side' for three reasons. First is imperfection in the channels, barbs, friction differences, tube swelling, etc. This has a minimal impact. Second, you can only vouch for what goes on in the block, but how people connect the two exit barbs back together is left in the air. Unless they do a perfect job of making the tubing lenghts (and heights) equal, there will be a difference. The third and biggest factor is gravity. You are trying to force some of the water up, and some of the water down. And this can become extremely troublesome depening on the blocks orientation (as some motherboards rotate the socket 90 degrees). True, in a closed water system, water will tend to stick together and try to balance out pressures, but its been in my troubled experience in splitting water loops that I could NEVER get the pressure to equalize through the split. If the block refocused the split back in to one output barb, youd have more control over it. Or, if you used smaller channels to force a balanced situation...im not sure it would work better or worse, I only mean to say that it could be an improvement for your next versions :)

Maybe the flow rate isn't too off though. I think it could vary depending on who's building the system, GPH, etc. But since I havent even seen the waterblock, Im only making observations, not comments about the block. It looks really good (especially the new waterjet you have out) and is highly rated, thats for sure. I'd love to get my hands on the blocks for a review, thats for sure too. I always remain skeptical of new waterblock designs since I've never really seen that much performance difference in any waterblocks really. All the top line ones seem to work within a few degrees of efficiency from each other. Theres only so much good design can do when you have limited space over the core, and your using the same metals, TIM for heat transfer. But Im glad to see people like you finally making smart waterblock designs, rather than just carving a channel in a block and sellign it as the next great thing.

The jetted block looks damn intriguing tho. I can see how it could work better. It looks like a damn nuclear reactor ;)

[Cathar]

Originally posted by Player0
The jetted block looks damn intriguing tho. I can see how it could work better. It looks like a damn nuclear reactor ;)

You mean this one?

http://www.employees.org/~slf/c2/c22.jpg

http://www.employees.org/~slf/c2/c25.jpg

http://www.employees.org/~slf/c2/c210.jpg

[dicki]

yeah i know a few new blocks using that idea, very nice idea in my opinion all these advances can only be good for us users :)

[dicki]

opps just adding a few more comments...

re: splitting the flow.... in all my experiments i've never been able to balence systems that slpit the flow... built a water block with 2 outputs which was a dismal failier because i simply couldn't ensure the same amount of water was flowing out of each side. this means that one side gets starved while the other is fine.... within the water block this may not be a big deal but then going to NB and GFX you'd end up with one getting significantly hotter than the other so i still recommend series use.

re: the showerhead attachment ;) doesn't that create serious back pressure? would a very powerfull pump increase its effectiveness by really jetting in the water?

dicki

[Cathar]

Originally posted by dicki
yeah i know a few new blocks using that idea, very nice idea in my opinion all these advances can only be good for us users :)

You do? Which ones? Links?

re: the showerhead attachment doesn't that create serious back pressure? would a very powerfull pump increase its effectiveness by really jetting in the water?


It creates some amount of back-pressure, but not excessively so.

All waterblocks benefit from a stronger pump/higher flow rates.

[dicki]

sorry i'm under NDA ;)

do they? how come the swiftech Quiet power system performs so well with a (relativly) weak pump would it be better with a real monster in there or does the system need to be balenced... i'm personally more for balence... admittedly you will generally see a performance boost if you just shove more water through a system but i'm sure more is not allways better

[Player0]

Yep, thats the one, so what do I need to do to get a review sample for the next roundup? :)

[Cathar]

Originally posted by Player0
Yep, thats the one, so what do I need to do to get a review sample for the next roundup? :)

Got a link to your last roundup so I can check it out?

sorry i'm under NDA

Seems to be the usual story. I develop something, show it, and suddenly everyone's under NDA about bringing out something similar. Sorry, I have just gotten a little cynical over time.

"Balance" with respect to flow only applied when talking about excessively hot pumps that dump more heat into the water than they help through pushing more flow through the block. The Swiftech Q-power is a nice little kit, but we're talking a totally different league with respect to waterblock performance here. Kits are more balanced to a budget and to ease of use, rather than top-end cooling performance.

[Player0]

Here are some of my waterblock reviews on this site:

http://www.liquidninjas.com/reviews.php?op=showcontent&id=17

http://www.liquidninjas.com/reviews.php?op=showcontent&id=29

http://www.liquidninjas.com/reviews.php?op=showcontent&id=8

The next round up will be much larger. I just bought a whole new AMD system to create a watercooling test bed.

As for the NDA block, I will be reviewing that one in this round up as well, since I have a pretty good rapport with it's developers. So we'll see how things perform :)

[dicki]

yes and i have a prototype and hopefully the real thing shortly :)

i know what you mean about being cynical cather :( everyone is allways after that slight edge no matter what but i know this block's been in development for ages and i was asked to keep it quiet so until i hear otherwise i'm afraid i do have to keep quiet.

fair point about the QP vs the LRWW they are in different water cooling leagues, sadly i don't have a powerfull enough pump or any flow meters or i'd run a few tests on the water blocks i have...

[Cathar]

I guess we'll just have to see then eh?

Be sure to let us all know when the final thing gets released.

[dicki]

i certainly will i've asked the manufacturer for details that we can post... we'll just have to see

[Cathar]

Originally posted by dicki
i certainly will i've asked the manufacturer for details that we can post... we'll just have to see

Actually they've since gotten in touch with me after seeing this thread here. Very nice people. I like their design and I applaud them for creating a unique interpretation on the problem of effective water-cooling. It wasn't quite as similar as was being made out earlier in this thread.

[dicki]

yup they are a good team arn't they :)

the prototype i have is very early (damn thing wasn't even screwed together!) so i've not seen the newest version in the flesh yet, it will be nice to see exactly what been done