![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Tubing sizes 3/8" vs 1/2"
- Thread starter xBuddhax
- Start date
Squishi.Dagger
Weaksauce
- Joined
- Apr 19, 2005
- Messages
- 109
1/2 inch provides better flow. makes sense, its bigger. but the 3/8 inch tube is easier to work with. i don't know the difference in performance they provide but with the right pump the 1/2 inch will flow better.
1/2: higher flow, less restriction; needs a high flow pump and water blocks made for high flow
3/8: lower flow, more restrictive; needs a lower flow, higher head pressure pump. works best with blocks that have restrictive interiors and is far easier to route than 1/2 IMHO
truthfully it doesn't matter what you use if you use it properly, if you get a 3/8's dangerden Maze4 cpu block and expect it to work better than it's 1/2 counterpart then you're wrong, but then it was made for 1/2 in tubing and high flow. that's my shpeal, take it or leave it.
3/8: lower flow, more restrictive; needs a lower flow, higher head pressure pump. works best with blocks that have restrictive interiors and is far easier to route than 1/2 IMHO
truthfully it doesn't matter what you use if you use it properly, if you get a 3/8's dangerden Maze4 cpu block and expect it to work better than it's 1/2 counterpart then you're wrong, but then it was made for 1/2 in tubing and high flow. that's my shpeal, take it or leave it.
Erasmus354
[H]F Junkie
- Joined
- Mar 12, 2004
- Messages
- 9,450
cell_491 said:
I agree it can be just as good, or maybe better than a poor 1/2in system, but it will never be better than a proper 1/2in system.
As to 1/2 vs 3/8 , back when people were using low head aquarium pumps, the 1/2 in provided a noticeable performance increase. Nowadays we have these great pumps with high head and high flow, and the performance gain from going to 1/2in is almost next to nothing. Cathar did a good study of this over at procooling i think.
I do not believe that one is better than the other. Looking at it from a pure engineering standpoint, it always boils down to efficency. Products that are engineered with 3/8", 1/2",1/4" in mind are or should be designed to optimize cooling based upon each component in the system. If designed correctly, there will not be a noticable difference. You have to make sure that EACH component is designed specifically to operate with the rest of your components.
Having worked with gas turbines, the principles of flow are something I am quite familiar with, but there is more to it than flow.
Having worked with gas turbines, the principles of flow are something I am quite familiar with, but there is more to it than flow.
Quite simply really.
Waterblocks always increase in performance with more flow.
Larger ID tubing always allows more flow more easily than small ID tubing.
Anyone who argues otherwise is being naive.
The important aspect is whether the size of the difference is worthwhile.
Waterblocks always increase in performance with more flow.
Larger ID tubing always allows more flow more easily than small ID tubing.
Anyone who argues otherwise is being naive.
The important aspect is whether the size of the difference is worthwhile.
Thanks for all the great feedback guys. My original plan was to go with 1/2" system but then I thought well what about 3/8". Reason I was thinking about it was because it seemed easier to route the lines cause it was a smaller tube. Most here seem to agree that the 1/2" will yield better results in a proper setup. The setup I plan on going with is cpu, nb, vga DD blocks, dual 120mm rad or heatercore and dd 12v pump. Thought about a reservoir but I think I'm gonna go with a closed loop system and hook in a T line fill tube.
BellaCroix
[H]ard|Gawd
- Joined
- Mar 25, 2004
- Messages
- 1,054
Cathar said:
You never know, maybe he's being informed with tests he's conducted which we aren't privy to, I've yet to see a practical comparisson of the two which has indicated any signifigant difference between properly set-up loops.
The majority of side-by sides I've seen have always shown each to be within 1 degree of the other.
Even if we throw out that any side-by side comparisson is going to have a margin of error and would require the person conducting the test to perform it several times for each loop then take the averages, 1 degree variant between the two is not a signifigant difference to base an opinion resulting in calling the opposing viewpoint "naive".
Now, if it's that apparent and you have the tests to prove it, please post a link or let me know where I can contact you to provide a post office box for you to forward the analysis of your critique as well as the test logs and results used to arrive there.
Otherwise, please refrain from posting an opinion and stating it as fact, particularly when that results in throwing around insults based on meritless claims.
Swimming_Bird
n00b
- Joined
- Jan 25, 2005
- Messages
- 23
i'm sorry but asside from the massive ethos of cathar in this subject you also have common sense working against you. The more water to come in contact with the surface of the block the more heat can be removed. Also due to the design of waterblocks, excess water flow (which will NOT be achived with 1/2" tubing) would not allow water to pass through it without being cooled. Although this is assuming you dont just have a big copper box on top of your die.theseeker said:
EDIT: I should add, while it is true that more water flow increases heat disapaited in highly restrictive systems the change in the cooling ability between 1/2 and 3/8 may not be worth the increased difficulty of setup, with 1/2 you get thiker walls which are harder to bend as well as a much larger OD. If you are in a situation that may require difficult routing of the tubing it wouldnt hurt to look at 3/8 or 7/16. While my current system has 1/2 when i rebuild it with my radbox i plan on using 7/16, especially since i hope to have a water chiller by that point.
BellaCroix, I believe that I said that the real issue was whether or not the difference is significant to the person assessing the importance of the difference.
I said two things:
1) The larger ID tubing is less restrictive than smaller ID tubing.
Now, please tell me how arguing against that point could not be perceived as anything less than naive.
2) Waterblocks always improve in performance with more flow:
Evidence as a result of tests with the accuracy to discover such:
http://www.procooling.com/html/pro_testing.php
http://thermal-management-testing.com/summary.htm
http://www.overclockers.com/articles373/wbsum.asp
All else being equal, excluding positive displacement pumps which raise pressure to sustain a fixed flow rate regardless of (tubing) restriction, larger ID tubing is less restrictive thereby facilitating higher flow rates, thereby increasing waterblock performance. Even with positive displacement pumps, the increased friction of the water through the tubing will raise the water's temperature slightly
As I stated, what is important is whether or not the difference can be coined as significant.
I said two things:
1) The larger ID tubing is less restrictive than smaller ID tubing.
Now, please tell me how arguing against that point could not be perceived as anything less than naive.
2) Waterblocks always improve in performance with more flow:
Evidence as a result of tests with the accuracy to discover such:
http://www.procooling.com/html/pro_testing.php
http://thermal-management-testing.com/summary.htm
http://www.overclockers.com/articles373/wbsum.asp
All else being equal, excluding positive displacement pumps which raise pressure to sustain a fixed flow rate regardless of (tubing) restriction, larger ID tubing is less restrictive thereby facilitating higher flow rates, thereby increasing waterblock performance. Even with positive displacement pumps, the increased friction of the water through the tubing will raise the water's temperature slightly
As I stated, what is important is whether or not the difference can be coined as significant.
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
1/2" will provide better cooling than a 3/8" at any test
Problem is cooling doesn't increase as flow increases. Too much flow can sometimes have negative effects. First of all, the extra energy can be released as heat if not break something else in the setup.
Problem is cooling doesn't increase as flow increases. Too much flow can sometimes have negative effects. First of all, the extra energy can be released as heat if not break something else in the setup.
First of all I post here to have fun and learn from every one's mistakes so that I do not repeat them. I was merely trying to have some fun. Sorry if I ruffled any tail feathers.
I agree and dissagree with certain points here. However, I am not about to conduct numerous tests to prove my point, besides I might be wrong. I have about $4-5,000 deployed in my system and I upgrade frequently. I do it simply for fun.
On the other hand, we have conducted many studies on gas turbines, including air flow, water injection, effects of temp's, special blade coatings and many other tests which would bore the %$#& out of each and every one of you. BTW, gas turbines range in price from $12m-36m, and the testing is also quite expensive. Believe it or not, there are many similarities. Lastly, I totally agree with one point made earlier....chill the water, I am working on that one as we speak.
If I have offended any one, please accept my apologies.
I agree and dissagree with certain points here. However, I am not about to conduct numerous tests to prove my point, besides I might be wrong. I have about $4-5,000 deployed in my system and I upgrade frequently. I do it simply for fun.
On the other hand, we have conducted many studies on gas turbines, including air flow, water injection, effects of temp's, special blade coatings and many other tests which would bore the %$#& out of each and every one of you. BTW, gas turbines range in price from $12m-36m, and the testing is also quite expensive. Believe it or not, there are many similarities. Lastly, I totally agree with one point made earlier....chill the water, I am working on that one as we speak.
If I have offended any one, please accept my apologies.
Swimming_Bird
n00b
- Joined
- Jan 25, 2005
- Messages
- 23
the only way for that energy to be entered into the setup is through the pump. and the energy repeased into the system differes from pump to pump. and it is the preassure that would break something in the ssytem not the flow. and as far as watercooling goes most stups can handle a LOT more then the pump can possibly dish out. for instance an iwaki MD20rz can put out approximately 10psi whereas my current 1/8" walled tubing can handle well over 30 psi.mohammedtaha said:
BellaCroix
[H]ard|Gawd
- Joined
- Mar 25, 2004
- Messages
- 1,054
Cathar said:
The Procooling link is an amazingly useful tool when deciding which block to use and demonstrates quite clearly the differences that can be expected with a wide spectrum of various flow rates. The problem I have with applying it toward this discussion is that it appears they varried the flow rate at the pump, not by altering the type, ID, composition, etc of the hoses being used in the test rig.
I will conceed that there will be a marked difference between 2 LPM flow through a closed loop and 9 LPM flow through an identical loop (although I was somewhat suprised to note that beyond about 4 LPM most blocks gain little from the added flow).
What was not addressed by any of those reviews was varried flow rates found by comparing 1/2" hose & barbs to 3/8" or 1/4". Yes, I'll agree that the concept of a larger hose will allow higher flow rates, that's elementary.
The third link, from Overclocker's indicates that:
If I allow that to be fact then the only comparissions I will be looking at would be 1 GPM, many of the people on the German manufacturer threads are indicating they are getting flow rates of about 2-3 LPM (.80 GPM) although many reviews indicate 2 LPM is more realistic .Overclockers said:
Using then the chart from ProTesting as reference comparing 1 GPM to .8 GPM and averaging each block, at 1 GPM the average temp was 10.99°C (51.78°F), at .80 GPM the average temp was 11.46°C (52.63°F). That's a difference of .47°C (.85°F).
Just to make the argument fair I'll also compare 2 LPM (.52 GPM). Using that figure the average temperature was 12.56°C (54.60°F) for a difference of 1.57°C (1.97°F).
Okay, I give... you're absolutely correct. There everyone, the debate is over, 1/2" performs better than 3/8".
The differences are not huge, but they are there. They are also even greater when you factor in the effect of higher flow rates on radiator performance:
http://thermal-management-testing.com/ThermoChill.htm
http://www.thermal-management-testing.com/8black ice.gif
http://www.thermal-management-testing.com/10surplus.gif
http://www.thermal-management-testing.com/11be cooling5x10.gif
http://www.thermal-management-testing.com/12big momma.gif
Also, one needs to factor in the degree of resistance to which some tubing size is offering at a particular flow rate. For example, if your system is routed with 3/8" tubing, and flow rate is 0.5GPM because you're using a small weak pump, the resistance of the 3/8" ID tubing (plus 3/8" fittings) is probably accounting for less than 2% of the total system resistance. Swiftching to 1/2" in that case isn't go to do anything significant. Switching to 8mm probably isn't going to change anything much either.
In this case one can easily justify using smaller ID tubing. There really is no need, either actual or performance based, to use larger tubing.
On the flip-side, if you're using a Maze4 waterblock and an Eheim 1250, then you want larger ID tubing otherwise the 3/8" ID tubing will be a significant source of restriction in comparison to the Maze 4 and the radiator.
However, it must be made clear that such a scenario is a function of the pump and the other resistances in the system (waterblocks etc). I think that this is where some people get a little confused about the whole debate.
Small ID tubing can perform the same as large ID tubing IF the flow rates are so low that the small ID tubing is not a significant source of restriction for the pump to work against.
As you noted by the links, some blocks stop gaining in performance in a significant fashion fairly early on, others still gain even after 2gpm.
I am not a hi-flow proponent, but neither am I a low-flow proponent. All tubing sizes have scenarios where they are perfectly suitable for the job at hand. If the flow rates regardless of the tubing size are never going to be high enough to justify some large tubing size, then there is no need to ever use a large tubing size.
If you're using a high restriction waterblock with a pump like an Eheim 1046 or 1048, much like many of these European based kits do, then I too would stick 8mm ID tubing on them. The flow rates aren't high enough to justify using larger tubing because the 8mm ID tubing won't be a significant source of restriction to the setup.
I apologise for coming across a little strong before, but blanket statements like "small ID tubing is the same as large ID tubing" as were being made in this thread are overly simplistic.
http://thermal-management-testing.com/ThermoChill.htm
http://www.thermal-management-testing.com/8black ice.gif
http://www.thermal-management-testing.com/10surplus.gif
http://www.thermal-management-testing.com/11be cooling5x10.gif
http://www.thermal-management-testing.com/12big momma.gif
Also, one needs to factor in the degree of resistance to which some tubing size is offering at a particular flow rate. For example, if your system is routed with 3/8" tubing, and flow rate is 0.5GPM because you're using a small weak pump, the resistance of the 3/8" ID tubing (plus 3/8" fittings) is probably accounting for less than 2% of the total system resistance. Swiftching to 1/2" in that case isn't go to do anything significant. Switching to 8mm probably isn't going to change anything much either.
In this case one can easily justify using smaller ID tubing. There really is no need, either actual or performance based, to use larger tubing.
On the flip-side, if you're using a Maze4 waterblock and an Eheim 1250, then you want larger ID tubing otherwise the 3/8" ID tubing will be a significant source of restriction in comparison to the Maze 4 and the radiator.
However, it must be made clear that such a scenario is a function of the pump and the other resistances in the system (waterblocks etc). I think that this is where some people get a little confused about the whole debate.
Small ID tubing can perform the same as large ID tubing IF the flow rates are so low that the small ID tubing is not a significant source of restriction for the pump to work against.
As you noted by the links, some blocks stop gaining in performance in a significant fashion fairly early on, others still gain even after 2gpm.
I am not a hi-flow proponent, but neither am I a low-flow proponent. All tubing sizes have scenarios where they are perfectly suitable for the job at hand. If the flow rates regardless of the tubing size are never going to be high enough to justify some large tubing size, then there is no need to ever use a large tubing size.
If you're using a high restriction waterblock with a pump like an Eheim 1046 or 1048, much like many of these European based kits do, then I too would stick 8mm ID tubing on them. The flow rates aren't high enough to justify using larger tubing because the 8mm ID tubing won't be a significant source of restriction to the setup.
I apologise for coming across a little strong before, but blanket statements like "small ID tubing is the same as large ID tubing" as were being made in this thread are overly simplistic.
Nate Finch
Limp Gawd
- Joined
- May 5, 2005
- Messages
- 298
Screw math and the scientific method... 1/2" ID tubing just looks 33% more badass.
-Nate
-Nate
BellaCroix
[H]ard|Gawd
- Joined
- Mar 25, 2004
- Messages
- 1,054
Nate Finch said:
Ah, thank god... back to the subjective!
Nate Finch
Limp Gawd
- Joined
- May 5, 2005
- Messages
- 298
BellaCroix said:
Glad to help get things back on track.Common sense does seem to indicate that higher flow = more heat removed. An internal area of .44 vs .79 square inches seems like a huge difference. I have no idea if that'll translate to a noticible difference in temperature, but if you actually get nearly double the flow, you should have significantly better cooling.
And now the completely air-cooled guy is going to shut up
-Nate
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
Swimming_Bird said:
The problem is not the tubing .. as stated earlier, if flow is low then it might be equal in performance to a smaller ID tube.
A tube may be able to handle 30psi but the waterblock may not. Heat which is the way the energy will be released in the system can be caused due to friction not because the pump is hot.
The pump also may cause a huge difference in cooling. A low flow pump may not provide any performance gains for 1/2" over a 3/8" with a stronger pump.
In general as I stated before, a 1/2" ID tubed system will provide better cooling than a lower ID tubed system if all other factors remain the same.
Nate Finch said:Screw math and the scientific method... 1/2" ID tubing just looks 33% more badass.
-Nate
Dude, your so right, why sit around and worry about all the small downsides and rant on the forums about how your smaller tubeing is mor e effective bla bla bla . Just be a man and use 1/2" from the get go and worry not! I guss some jsut defend smaller tuebing because they got sucked into buying cheapo blocks..
theseeker said:
HAHAHA, you have no Idea who Cathar is and what he dose do you?
zer0signal667
2[H]4U
- Joined
- Jan 11, 2002
- Messages
- 3,847
Hehehe, welcome back Cathar. You've been gone for too long, apparently
I'm glad to see people are acting intelligently in a thread based on such a controversial topic. Thanks, and keep it up!
I'm glad to see people are acting intelligently in a thread based on such a controversial topic. Thanks, and keep it up!
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
Truth be told ... cooling performance has less to do with the flow of the water but the cooling of the water itself.
Even if you have a 1/2", 3/8" or a 1/16" tube ... if the rad or heatercore aint cooling right .. your system is gonna be hot regardless ..
Even if you have a 1/2", 3/8" or a 1/16" tube ... if the rad or heatercore aint cooling right .. your system is gonna be hot regardless ..
BellaCroix
[H]ard|Gawd
- Joined
- Mar 25, 2004
- Messages
- 1,054
mohammedtaha said:
Agreed 100%! The (marginal) gains you will see using 1/2" over another will mean nothing if your system is inproperly set-up or if the components cannot make use of the additional waterflow.
Look at the link Cathar posted from procooling and compare an Innovatek waterblock to a DangerDen waterblock. Innovatek is designed to run with very low water (their website claims you can actually, safely opererate with .3 LPM flow rates). As water flow increases there is very small improvement. Compare that to a DangerDen block optimised for high-flow systems and it makes huge improvements as flow increases.
The secret is to research each of your components and build a balanced system where all components utilize and make use of whatever flow rate (or tubing size) you are actually using.
diredesire
Supreme [H]ardness
- Joined
- Feb 18, 2003
- Messages
- 4,516
...
Don't argue with cathar, it makes me die a little inside...
I must be clicking on all the wrong threads today though, because there are a ton of people just trying to get a word in on things they really know little to nothing about... no offense to the OP/anyone else in the thread... i'm just a little frustrated right now.
Don't argue with cathar, it makes me die a little inside...
I must be clicking on all the wrong threads today though, because there are a ton of people just trying to get a word in on things they really know little to nothing about... no offense to the OP/anyone else in the thread... i'm just a little frustrated right now.
Well now that everyone has had some time to espouse their religious views I am ready to say my piece. Cathar is correct. You are all incorrect. The important thing as he or she pointed out is whether the performance gains from using 1/2" tubing or bigger is worth all the trouble and what you find to be the perfect balance in all the variables. There is no right or wrong here, just what makes sense to you. Bigger tubing and more flow = better cooling. Now whether that translates into anything measurable in what you are doing with the hardware after cooling it is the key question. Which is why I use Aqua Computer and Koolance gear, because it does what I want it to do without a lot of hassle.
zer0signal667
2[H]4U
- Joined
- Jan 11, 2002
- Messages
- 3,847
mohammedtaha said:
The efficiency of the radiator is much enhanced by having the water in motion
Hence, performance has EVERYthing to do with flow.mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
zer0signal667 said:The efficiency of the radiator is much enhanced by having the water in motion
I think what everyone failes to realize is this statement:
It doesn't matter the diameter of the tubing as long as you achieve the desired overclock.
I have had systems that consist of 8mm, 3/8", and 1/2" based configurations and have found the temperatures vary anywhere from 1 to 3c. The best cooling solution didn't always equate to the best overclocking solution, so that should be first and foremost. If I were to rank them they would be as follows:
1. Cathar G4 with 3/8" tubing
2. WW (Cathar design) with 1/2" tubing
3. AquaComputer CuplexXT with 8mm tubing
4. Danger Den RBX with 1/2" tubing
5. Dtek TC-4 REV2 with 1/2" tubing
6. AquaComputer EVO with 3/8" tubing
7. Danger Den Maze4 with 1/2" tubing
8. $piral with 1/2" tubing
So as you can see the largest tubing didn't always equate to the best overclock, in my systems.
It doesn't matter the diameter of the tubing as long as you achieve the desired overclock.
I have had systems that consist of 8mm, 3/8", and 1/2" based configurations and have found the temperatures vary anywhere from 1 to 3c. The best cooling solution didn't always equate to the best overclocking solution, so that should be first and foremost. If I were to rank them they would be as follows:
1. Cathar G4 with 3/8" tubing
2. WW (Cathar design) with 1/2" tubing
3. AquaComputer CuplexXT with 8mm tubing
4. Danger Den RBX with 1/2" tubing
5. Dtek TC-4 REV2 with 1/2" tubing
6. AquaComputer EVO with 3/8" tubing
7. Danger Den Maze4 with 1/2" tubing
8. $piral with 1/2" tubing
So as you can see the largest tubing didn't always equate to the best overclock, in my systems.