I agree with that.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
The impact of tubing sizes
- Thread starter Cathar
- Start date
I agree with that.
I appriciate your work and knowledge you bring to the table. I am glad someone with your authority in the watercooling realm has stepped up and told it as it is instead of deceiving the users.
Over time I'm sure we will see people migrate towards the solution that fits their needs rather than trying to fit a square peg in a round hole.
Thats wrong. Those charts are incompatible. Check Swiftechs own comparison between Apogee classic and Apogee GT and youll see that Swiftechs PD for classic Apogee is different than those from SC:
Swiftechs readings for the classic Apogee seem much higher than those from SC. So your GTX PD reading is much higher than it would be if it were adjusted to match the readings at SC. Therefore your 4x figure is off. Interpreting each chart independently may be a better idea to avoid errors. We can compare the relationship between classic Apogee and new Apogee and compare that to SC's relationship between classic Apogee and maze 3.
Lets do a reality check: there is no way GTX is anywhere near 4x as restrictive as Apogee classic. Hence, since Apogee classic performs similarly to maze 3, GTX is nowhere near 4x restrictive as maze 3. For arguments sake, even in the supernatural event that such was the case, its nowhere near your delirious figure of 5-10x. Even then, its ironic for you to choose the more convenient GTX block and omit the equally relevant GT, Fuzion, etc.
Your argument does not hold: even if todays non-restrictive blocks are slightly more restrictive than yesterdays, they are nowhere as restrictive as impingement-based blocks. There is a huge gap between blocks like the Apogee and Fuzion and blocks such as Storm and Nexxxos. Huge. Sure, non-restrictive designs might have a slight increase in pressure drop because of the natural evolution of block design, but such evolution has a minimal secondary effect on PD when compared to the design philosophy of impingement-based blocks. There is a shift in philosophies; at least for now the trend is clearly non-restrictive driven. But you seem to be claiming otherwise which is startling.
I never said nor implied that you were to blame for everything that I see wrong in W/C. I never said nor implied that everything you are is bad. Why are you making stuff up?
I appreciate the fact that you came forward and tried debunking the myth about tubing sizes at the risk of being blamed for your previous work. However, I dont appreciate you picking on, insulting and dismissing people at the slight sign of controversy. The controversy of the history of tubing size and its repercussions is well frakin founded, not trollish as you seem to make it appear.
Hmm, something is not right. I'm arguing with Cathar about Apogee being similarly non-restrictive as maze 3 with Cathar claiming it to be 5-10 less restrictive and you say that maze 3 was worthless in a low flow system? What's going on?
Honestly I'm lost, I thought the 1/2" frenzy was justified mainly by the impigement-based blocks such as g4, not the non-restrictive maze blocks. Even then it's debatable, but why pick on a perfectly fine maze block to justify larger tubing? More flow on a maze design does nothing to performance. It doesn't make sense.
Please
Please TN, we know that you'll find every opportunity to mention AC and champion their cause.
How do you not understand, that higher flow rates is always a good thing. Higher flow rates will always mean better performance. You're attempting to lump smaller tubing sizes with low flow and trying to convince everyone that Cathar's new results on tubing sizes directly correlates with flow rates as well. It does not. Just like Cathar stated, his results compared ONLY tubing sizes while keeping everything else constant. Lower flow rates will always equal less performance.
Now for the off topic issues:
Are we trying to be racist and ignorant now and blame those of different ethnicities?
You are severely misguided if not delusional if you think nurses run the hospitals. All high level management positions are not staffed by nurses, but rather doctor's or those with MBA's. Nurses are in charge of handling the day to day responsibilities. Nurses and doctors work together to ensure the patient receives the quality care they deserve. However, if one had to be removed from the equation out of necessity, the nurse would be the first to go. A nurse can not perform a doctor's duties but a doctor can perform that of a nurse (but granted they rarely want to and see it as a mismanagement of their time and expertise).
I sit on the board for my father at both Long Beach Memorial as well as Orange Coast Memorial hospitals for the majority of the meetings as he usually does not have the time to do so with his busy schedule. I've long worked in health care as I was pushed (often against my desire) towards the field since I come from a family of M.D's. I've worked in almost all areas of health care starting as a receptionist in high school, to an M.A. for a year, and eventually promoted to work on the administrative side for various HMO's (Kaiser) and Hospitals (Hoag, Long Beach Memorial, Orange Coast Memorial) . I've had more than my fair share of run in with nurses.
For the most part, they love their job and enjoy their work. However, there's an increasing percentage of nurses who do not belong in the profession and only remain in employment for a variety of reasons including 1) shortage of nurses 2) strength of the nurse's union 3) lack of staffing for a particular department.
There are those who are there simply because the hospital can't afford to fire them due to the small pool of available nurses. In the event that a suitable replacement can be found, the 800lb gorilla known as the Nurse's union (hello Kaiser) often blocks the layoff for inter-office political reasons (even in situations where fellow nurses believe the offending nurse should be fired). A worrisome number of nurses, especially those who work in the geriatric department should have long been fired for their spiteful attitudes that manifested in their neglect for the well being of their patients. There's a startling amount of daily complaints ranging from cruelty to just simply ignoring a patient's request. It's impossible to find competent, non dysfunctional nurses willing to work in that department since the majority of applicants wish to work either in ICU, ER, or pre/post natal care. In addition, there's a number of rogue nurses who feel as if they have the educational background, the training, and prerogative to diagnose patients and offer second opinions...which they obviously don't. This often serves no reason other to harm the patient into second guessing the physician, affecting the patient's relationship with the physician and their willingness to follow the physician's instructions.
As mentioned earlier, being a nurse is no promise of belonging on top of the almighty altar of professionalism. There are bad apples in every profession (I can tell which guys in my law school are going to be assholes already).
/end off topic
Gabriel Rouchon
n00b
- Joined
- Mar 4, 2003
- Messages
- 12
Hmm, something is not right. I'm arguing with Cathar about Apogee being similarly non-restrictive as maze 3 with Cathar claiming it to be 5-10 less restrictive and you say that maze 3 was worthless in a low flow system? What's going on?
I did state that I was primarily talking about the Maze 4. You're wanting to drag this back to the Maze 3, presumably because my 5-10x statement doesn't quite hold true to that block, and you feel that you've "got me" on that one. Okay, fine. I'm not going to argue it any longer. This is not going to be answered easily without the same testbed data.
Pressure drop graphs are comparable though, unless we believe that pressure-drops at a specific flow rate are variable. They can be, but it depends on the coolant used.
Rather than answer all your points, I'll focus on this paragraph, and maybe we might understand ourselves better.
The "1/2" Frenzy" started around 7 years ago. It was already in ascension by the time I even started to water-cool myself, about 6 years ago now. Half the blocks on the market at that stage were already 1/2", and in the USA at least, were primarily maze designs, or middle-in jet designs, or both.
At the time that I entered the scene, I started looking into half-inch tubing myself, and ran various tests, and suggested that there were some improvements there. I was but a small voice at that time. Bill Adams also ran various tests that suggested, for middle-in JI blocks at least, that 3/8" was better. Various other testers also ran their tests that showed that 1/2" was better. Lots of data about that had huge errors, but some were convinced that 1/2" must be better, and so most just opted for that.
A large portion of the reasons for 1/2" tubing rising even further was due to the Procooling.com tests (see the link I gave above). Poke through the reviews, and look at the flow-performance curvess. Most of the guys were looking to where the curves flattened out, and that occurred mostly at >8LPM flow-rates for popular blocks at the time.
The Cascade & Storm/G4 designs were actually one of the first number of blocks that didn't require staggeringly high flow rates to excel. Heck, even at very low flow rates, they were beating most other blocks of their era that had super-high flow rates. At this time I put forward that 3/8" might be okay, but few listened. I then put forwards 7/16", and only after mathematically AND physically proving that it didn't affect performance at all, did people start to even take 1 step back from 1/2".
Privately I started running tests with 3/8" and 8mm on the Storm, and it worked just fine. For years I've known this, but the greater community is rabid about 1/2".
So, you see, the Storm was never the block that started the 1/2" craze. That fireball had exploded LONG before. The Storm was actually one of a few blocks that those who cared to, would find that it worked equally well as a low-flow block. Heck, I remember stating that many times, and getting shouted down and pretty much laughed at for suggesting that tubing smaller than 1/2" might not be needed.
As for the gains with the old school blocks. See the maths I gave above. The equipment is no longer around, but it demonstrates the reasoning that went behind the justification for 1/2" tubing, and it was real, for that era of blocks at least. Some people tested it, and saw the gains. Some tested, and saw nothing. With ~2C accurate thermometers though, who can tell what the truth was. People relied on the reviews (Procooling), and the maths.
So, you see, I've by gently trying to push towards smaller tubing sizes for a long time now. Please don't confuse people championing 1/2" with the Storm, with my personal views on the matter, which are markedly different. It is this that is angered me the most yesterday about your statements, and it wasn't until I read that paragraph above that I realised that you had no idea that it was never I that had been pushing for 1/2" on the Storm.
With all due respect, I'm afraid that "The impact of flow" may be the next chapter in the series of myth debunking. Actually it should have been the first. Tubing size is just one factor affecting flow. We've seen how such factor is for the most part irrelevant. Let me ask you though, what happens if you add an extra block to your loop? Flow will probably decrease as much if not more than using the "bastarded" 1/4 tubing. But it doesn't necessarily mean that heat absorption will decrease.
I mean lets be down to earth for a minute. On one end we hear that adding an elbow or two to a loop is Armageddon and humanity will perish. But people report no change. People reported no change between 3/8 and 1/2. Something is fishy. The math is telling a tale that doesn't necessarily apply to real life.
Just like Koolance's testing doesn't apply to real life, a lot of the flow nitpicking doesn't apply either.
We need to identify the factors that truly make a difference and exploit those. Everything else needs to be debunked.
No. Local.
Frictional resistance is a factor, yes, but since we can't create from energy we don't have, there is a limit to your statement. Even a Laing DDC2 only pushes without around 3W of hydraulic power. What that means is that no more than 3W of power would get converted to heat in this manner.
Once again, I'm curious as to what actual reviews you've found that support the notion that more flow is harmful. Adding more heat by using a stronger pump can be harmful, yes, but if we keep the same pump and use lower restriction everywhere by using larger fittings and larger tubing, then more flow is never harmful.
I think that perhaps those two issues are being confused here. Also, what sized pump you can get away with is a factor of the radiator. With a radiator with very slow spinning fans, pump heat from a stronger pump is a much greater factor and can make things worse, and it is this that may lead to the observation that more flow is bad, but that is the pump doing that, not the fittings/tubing.
Who is doing the math I ask?
I did the math over at XS, and found that the impact of an elbow would be significantly less than 0.1C per elbow.
Are you sure that math/theory was actually used, or whether or not people were just talking out their rear ends?
Perhaps this is the issue. People are blaming the maths, when the maths never shows that. The real issue is just people making up stuff that they really didn't measure.
The "impact of flow" is what has been explored all along with flow vs performance graphs being conducted by various reviewers. It WAS the first chapter, and it WAS the first thing done.
What has occurred since, at least to my eyes, is an incredible inability to actually use that useful data presented to arrive at the correct conclusions. Over time I've formed the opinion that most will just look at a pretty graph, look at the highest values (or lowest value - whichever the case may be), and decide that's the best thing, no ifs or buts.
The problem never was a lack of information about "the impact of flow". The problem has always been a lack of the required ability to properly dissect the information given to arrive at the correct conclusions.
This is where I try to step in, to show how to correctly analyse the information given, and to demonstrate precisely why some myths are wrong.
Fair enough, thank you for the good will.
This is asking for trouble, and I won't go into that. But yes, I do agree that there is poor interpretation of the theory in general. It's a pest.
Yet in this very thread I hear the claim that the a maze block "needs" and benefits from higher flow which is complete nonsense. There is a huge misconception regarding the conditions in which flow makes any difference, at all. Such is the misconception that people were afraid of using 1 elbow=death (and this was backed up by "math") and it was discussed about 3 weeks ago at XS, that was before your input. I think the topic is well overdue and far from over.
It's no misconception that all blocks benefit from more flow. The Procooling tests, which were run on an actual CPU on a real motherboard running real load, which is why the tests were so popular, demonstrated that all blocks benefitted from more flow. More flow equals a cooler CPU. Reference any well conducted review to discover that.
I don't think that's what you're getting at though. You're asking whether or not it makes any real difference on the CPU, aside from merely the temperatures? Is that where you're coming from? Like a flow vs over-clock stability graph?
Flow vs overclock, it was established with the Maze blocks that more flow meant slightly better overclocks. Couple that together with the 3DMark craze, and phase-change at the time barely being touched by more than a handful of individuals, water-cooling was where it was at for the "benchmark races". For a significant number of individuals, an extra 10-20MHz was a big deal. It might mean the difference between 20th and 5th on the benchmarking charts for example. This is where the intense pursuit of the maximum possible cooling obtained through the highest possible flow facilitating the highest possible overclock was born. I remember the days well, and yes, more flow did equal more overclock, and did equal better results. That's not a misconception. It was demonstrated many times.
With IHS capped CPU's, it's less significant. A 1C difference may not mean anything, and if you're referring to the state of the cooling world in reference to IHS CPU's, then yes, 1 or 2C differences may mean nothing.
It was never stated that maze blocks NEED high flow to work, although that might be how some interpreted it, but they did indeed benefit from high flow rates to achieve the best possible over-clocking results, and possibly eke out that elusive extra 10MHz. To understand the extremity of people's statements (and in some ways the silliness of them) you have to understand where they are coming from.
Heck, for a while there I ran my entire water-cooling system off a $5 pump the local hardware store. It was 1.5"x1.5"x1" in size. Tiny thing. Put out around 3lpm peak, and had about 50cm of pressure head. I barely noticed any difference. CPU overclocked 30MHz lower, temps were 3C higher overall, but that was about it.
The way some people carry on about "You MUST match so-and-so block with a HUGELY powerful pump" is just ludicrous. The results from the reviews were showing that if anything, you could use an extremely weak pump and be perfectly fine. I pointed this out MANY times, but got ignored. I personally put a fair amount of blame onto some of the statements by Joe Citarella from overclockers.com for that. He'd often review a block, would find that it was somewhat restrictive, and even though he was able to demonstrate that at 1LPM flow rates that the block was better than some other low-resistance block, would then turn around and say that the block NEEDED a strong pump. I emailed him a number of times about those statements, but to no avail.
So you see guys, I know where you're coming from. Yes, you don't NEED high-flow to have a good performing system. You don't NEED a strong pump to get good cooling performance. You don't NEED big-bore tubing to get a good overclock. All the reviews that are out there have demonstrated that many times right in front of many people, but most have overlooked it. Even after it was pointed out to them, they overlooked it.
Having said that, if you want the absolute best possible performance, that last 0.5-1.0C, that last 10MHz at any expense, then yes, big-bore tubing, a strong pump, and high-flows are your best bet to achieve that. What's gotten lost over time is the sense of perspective. There's a vast difference between what really happens, vs what some people think happens, when you use a weak pump, small tubing, and low-flows.
My personal perspective is that I'm happy to focus on getting to within 0.5C of the best possible achievable performance in the most practical, quiet, efficient manner.
I hope that this clears up some confusion on where I stand on the matter. If you want me to agree that "Low flow is better than high flow, performance/temperature wise", I won't be saying that. It's not true. If you want me to agree that "Very few people will actually notice any real difference between low-flow and high-flow", then yes, I'll agree to that 'cos I know it's true.
The way some people carry on about "You MUST match so-and-so block with a HUGELY powerful pump" is just ludicrous. The results from the reviews were showing that if anything, you could use an extremely weak pump and be perfectly fine. I pointed this out MANY times, but got ignored. I personally put a fair amount of blame onto some of the statements by Joe Citarella from overclockers.com for that. He'd often review a block, would find that it was somewhat restrictive, and even though he was able to demonstrate that at 1LPM flow rates that the block was better than some other low-resistance block, would then turn around and say that the block NEEDED a strong pump. I emailed him a number of times about those statements, but to no avail.
So you see guys, I know where you're coming from. Yes, you don't NEED high-flow to have a good performing system. You don't NEED a strong pump to get good cooling performance. You don't NEED big-bore tubing to get a good overclock. All the reviews that are out there have demonstrated that many times right in front of many people, but most have overlooked it. Even after it was pointed out to them, they overlooked it.
Having said that, if you want the absolute best possible performance, that last 0.5-1.0C, that last 10MHz at any expense, then yes, big-bore tubing, a strong pump, and high-flows are your best bet to achieve that. What's gotten lost over time is the sense of perspective. There's a vast difference between what really happens, vs what some people think happens, when you use a weak pump, small tubing, and low-flows.
My personal perspective is that I'm happy to focus on getting to within 0.5C of the best possible achievable performance in the most practical, quiet, efficient manner.
I hope that this clears up some confusion on where I stand on the matter. If you want me to agree that "Low flow is better than high flow, performance/temperature wise", I won't be saying that. It's not true. If you want me to agree that "Very few people will actually notice any real difference between low-flow and high-flow", then yes, I'll agree to that 'cos I know it's true.
Major bummer. Like I said before ask to speak to a higher up. Nurses are the worst patients because we know what's up and have no problem complaining to get things taken care of.
More drivel...
I wrote about 2 pages of response to your nonsense and then decided I really didn't want to trash this thread with something off topic. Any thoughts I would have had about working at either LBM or OCM died with those words you espoused. Add to that many local doctors have told me that nurses are highly screwed by LBM and are expected to be perfect 110% of the time. That's really bad when local doctors tell nurses not to work at a particular hospital.
However, if you do want to discuss this further we can start a thread in GenMay.
More drivel...
I wrote about 2 pages of response to your nonsense and then decided I really didn't want to trash this thread with something off topic. Any thoughts I would have had about working at either LBM or OCM died with those words you espoused. Add to that many local doctors have told me that nurses are highly screwed by LBM and are expected to be perfect 110% of the time. That's really bad when local doctors tell nurses not to work at a particular hospital.
However, if you do want to discuss this further we can start a thread in GenMay.
Please. Which is why LBM and OCM are the fastest growing hospitals within the region. Specialists have flocked from Cedar Sinai and Hoag to LBM and OCM. The nurses are paid better there than any of the nearest competitors within a 30 mile radius. Hoag has officially lodged a complaint that we're stripping their better doctors/nurses/administrators. The hospitals have been so successful that new wings have funded (Children's Hospital in LBM) and full fledged research wings are in the works.
And yes, the nurses at LBM/OCM are expected to have a close to perfect record. LBM/OCM pay the best and expect the best out of those employed. Those who don't perform up standard are let go. LBM/OCM's name speaks for itself when its one of the few hospitals not plagued with rampant staffing shortages which lead towards harmful conditions for patients. If you research the state's auditing, you'll find that OCM/LBM rates in the top 10% of all California hospitals. Don't hate on the hospitals that most likely denied your application
Anyhow, back OT: You can get back to starting shit with Cathar.
"Low flow is the best flow" round 2:
I'll make it easy on you.
Given the right pump/radiator/block(s), tubing size impacts performance about .7c when moving from 1/2" tubing to 1/4" tubing. Pretty much debunks the myth that bigger is soooo much better.
I’m glad to see that kind of mindset, and yes we need to recuperate that sense of perspective.
I’d like to add something to the analysis. I’m convinced that flow is one of the least important factors in cooling performance (at least with today’s pumps) and I’ve seen it a zillion times over and over in the real world. Arriving at an objective explanation of why, it seems to me that there is a flow range by which the cooling remains practically constant. Judging by most thermal resistance vs flow graphs on W/C blocks, if we look at the exponential curve, the biggest hit happens well below the 1gpm mark (example: the GTX). Anything above, say, 1.5gpm to 2.5gpm will practically remain negligible. Since most systems fall within the later range, most people never find a significant difference in temperatures.
So the biggest thing to look for when it comes to flow is to be above the ~1gpm mark.
ikellensbro
[H]ard|Gawd
- Joined
- Feb 14, 2004
- Messages
- 1,410
Yes, get 1-1.5gpm, and you're good to go. Depending on the block, go below 1gpm and you start seeing exponentially increasing deltas. Below 0.5gpm (1.9lpm/114lph), temperatures start to go up much higher, which is why I dislike very low flow rates. Personally, I don't advocate 90* mitre elbows when a 45* sweep (or no elbow at all) will do just fine, because it's an unnecessary performance loss. And the smaller the tubing size, the easier it is to route tubing anyway. I like to get the most for my money.
Yes, a high pressure pump like any DDC or a D5 can get past most restriction and still deliver high flow rates to get in the nice 1gpm + flow rate. Lesser pumps (especially 12v'd 1046/1048s that cost more than either aforementioned option, with a larger footprint) produce less flow in highly restrictive situations, and start to fall several degrees behind. You don't need a moderate to high pressure pump for a somewhat restrictive block to work - you just get the most out of it by doing so.
Yep, this has been precisely my thinking for a while now.
So long as we're above 4LPM (1.1gpm), the differences get smaller, fast. Past 6LPM, pretty much all gains are so negligible as to be not worth the bother (IMO). Couple that with the pump heat dump from the more powerful pumps to push those flow rates, and if one runs the physical tests (I have), we can even see temperatures almost completely plateau between 6-10LPM, before taking a downwards turn at >10LPM flow rates as pump heat becomes a barrier to further improvements (for most normal setups).
I ran a significant number of real-world tests on CPU's to prove this first, not just mathematically, years back with a number of pumps and feeding different voltages into them. It was the practical foundation real-world testing work that led to the mathematical theory in this thread.
You can go as low as 3LPM (0.8gpm) and be okay. Temperatures will climb a bit (1C or so over 4LPM), but it shouldn't affect overclocking in any dramatic fashion. Might get 10-20MHz lower overclocks than at 4LPM, or it might even overclock the same. Every CPU/GPU is different though, so it's something that you need to test. Get much below 3LPM and things get warmer, quickly.
My recommendation always has been that an Eheim 1048 is about the lowest powered pump we'd want to use. The frequency modded 1046's sold by some German companies will match or exceed a 1048 though, and so happily fall into the same category as being pumps of suitable power.
Now, by suitable power, and to enforce a measure of context here, I specifically mean being suitably powered such that we're at the point where a more powerful pump will yield very minimal, if any, benefits to the overall system aside from granting marginally lower temperatures, and certainly not much chance of any additional tangible benefit, such as being able to overclock any higher. Might get an extra 10-20MHz, if lucky. We do need to understand though that this what some people are after, and so it becomes very justified to them.
Still, even having said that, if I cast a thought back to my little Jebo pump which is a quarter as powerful as a stock Eheim 1046, and only pushed 1.25LPM (0.33gpm) flow rates in my test system, and things still weren't that bad in comparison to 10.5LPM (2.8gpm).
It is this "happy medium" of 4-6LPM that when coupled with a well powered pump, that allows us to get away with using 8mm or 3/8" tubing, and still not see any appreciable impact on temperatures as a result of reduced flow from the smaller tubing/fittings. So long as flow remains above 4LPM, we shouldn't see any real system impact in comparison to a higher flow-rate. 1/2" tubing is really only justified if we're chasing flow rates higher than 6LPM, but I've conducted plenty of real-world tests to show that flow rates above 6LPM yield next to no benefit.
I think this thread should be a stickie or the pertinent info added to one of the FAQs. It took me 3 days (I am horrible at searching ) before I was able to find a thread like this one that contains good, solid data regarding the difference in 3/8 ID vs 1/2 ID.
) before I was able to find a thread like this one that contains good, solid data regarding the difference in 3/8 ID vs 1/2 ID.To bump a 2 year old thread....
To confirm, the results of OP is that tubing size doesn't matter at all. Any review you read with comparisons of air CPU coolers almost always say +/- 2C is within the margin of error and thus is no difference. Looked like from skimming the OP post there was less than 1 C
Im currently running Primochill 1/2" ID (3/4" OD) and it is a pain in the ass to work with in many situations. Admittedly this is my first ever custom loop so I dont know whats normal and whats not, but I found it hard to work with.
I want to redo the system next time I take it apart with 3/8" ID (1/2" OD) and see if it looks better and is easier to work with (as long as the temp diff is +/- 2 C). I could also add ram banks (possibly) with smaller more flexible tubing. No way that I could do that in my current case with the 1/2".
Here is it is now with the 1/2" tubing:
To confirm, the results of OP is that tubing size doesn't matter at all. Any review you read with comparisons of air CPU coolers almost always say +/- 2C is within the margin of error and thus is no difference. Looked like from skimming the OP post there was less than 1 C
Im currently running Primochill 1/2" ID (3/4" OD) and it is a pain in the ass to work with in many situations. Admittedly this is my first ever custom loop so I dont know whats normal and whats not, but I found it hard to work with.
I want to redo the system next time I take it apart with 3/8" ID (1/2" OD) and see if it looks better and is easier to work with (as long as the temp diff is +/- 2 C). I could also add ram banks (possibly) with smaller more flexible tubing. No way that I could do that in my current case with the 1/2".
Here is it is now with the 1/2" tubing:
Looks fantastic!To bump a 2 year old thread....
To confirm, the results of OP is that tubing size doesn't matter at all. Any review you read with comparisons of air CPU coolers almost always say +/- 2C is within the margin of error and thus is no difference. Looked like from skimming the OP post there was less than 1 C
Im currently running Primochill 1/2" ID (3/4" OD) and it is a pain in the ass to work with in many situations. Admittedly this is my first ever custom loop so I dont know whats normal and whats not, but I found it hard to work with.
I want to redo the system next time I take it apart with 3/8" ID (1/2" OD) and see if it looks better and is easier to work with (as long as the temp diff is +/- 2 C). I could also add ram banks (possibly) with smaller more flexible tubing. No way that I could do that in my current case with the 1/2".
Here is it is now with the 1/2" tubing:
DWolvin
2[H]4U
- Joined
- Nov 13, 2002
- Messages
- 3,536
Nice looking rig, but I have a fairly similar routing with 3/8-5/8 inch tube, and I'll guess you are going to be happy when you try the small stuff- I have XSPC FLX Tubing (solid black), and it doesn't seem to pinch at all, even with 90 degree bends with about a 2 inch radius (not used that way, just tested). I'll try to post a pic later...
Pardon the pic, but you can see how easy it is to route the tube- 3/8-5/8 is the black, the clear is EK3/8id and needs the spring looking stuff to keep it from pinching in tight turns (and partly looks cool).
Pardon the pic, but you can see how easy it is to route the tube- 3/8-5/8 is the black, the clear is EK3/8id and needs the spring looking stuff to keep it from pinching in tight turns (and partly looks cool).
Last edited:
Tsumi
[H]F Junkie
- Joined
- Mar 18, 2010
- Messages
- 13,817
3/8" ID 5/8" OD is the way to go if you want easy to work with tubing. The thicker wall prevents kinks, 3/8" ID 1/2" OD will kink easier and not allow as tight of turns.
Tsumi
[H]F Junkie
- Joined
- Mar 18, 2010
- Messages
- 13,817
If you leave the test long enough, it will reach an equilibrium point where volume of water does not matter.
![Zarathustra[H]](https://cdn.hardforum.com/data/avatars/m/9/9298.jpg?1715876891){kind=avatar}
Zarathustra[H]
Extremely [H]
- Joined
- Oct 29, 2000
- Messages
- 39,216
Wish I had seen this before ordering my tubing and fittings
Hopefully 3/8" ID 1/2" OD will work OK for me.All that being said, I don't understand why tubing size would have much of an impact on cooling performance at all. The flow of the system should be determined by its most restrictive point, and surely that is in one of the CPU/GPU blocks, right?
I would have thought that as long as your tubing is less restrictive than your most restrictive block, it should have no impact at all.
Wish I had seen this before ordering my tubing and fittings
All that being said, I don't understand why tubing size would have much of an impact on cooling performance at all. The flow of the system should be determined by its most restrictive point, and surely that is in one of the CPU/GPU blocks, right?
I would have thought that as long as your tubing is less restrictive than your most restrictive block, it should have no impact at all.
Fluid moving in the tube creates hydraulic friction. In loops this small, it's a non-point though.
Tsumi
[H]F Junkie
- Joined
- Mar 18, 2010
- Messages
- 13,817
Wish I had seen this before ordering my tubing and fittings
All that being said, I don't understand why tubing size would have much of an impact on cooling performance at all. The flow of the system should be determined by its most restrictive point, and surely that is in one of the CPU/GPU blocks, right?
I would have thought that as long as your tubing is less restrictive than your most restrictive block, it should have no impact at all.
Restriction is cumulative, not based on a single component. In general for watercooling systems, 3/8" ID tubing and larger have negligible restriction, while 1/4" ID tubing has been shown to impact flow rates.