Quick question on loop length...

S

schoenda

Gawd
Joined
Apr 8, 2003
Messages
860
So due to an unfortunate bit of architecture the only place I can put my new system is in a closet basically that has been converted to an office nook...it has very bad ventilation and gets quite warm with a large dead space at the ceiling where hot air is trapped. I know there are more than one ways to skin a cat but what I would like to do is to switch my radiator to external and move it about 3-4 ft. away. I am using a swiftech mcp 655 pump and the loop will cool a 950 i7 at 4ghz and 2 5870's at stock. The loop is 1/2 inch id.

My question is, with this setup how many feet can I extent the tubing...a 7 ft round trip total distance would allow me to put the radiator in a much cooler area.

I am not sure how the issue of adding the extra distance/resistance weighs against the alternative of running the system in an area with such poor airflow. I would appreciate your insights.
 
(IMO) The additional length will little or no impact on that pump. Since it is a closed loop and the "head" will remain zero, meaning the weight of the water entering the pump is the same as the weight of the water the pump is trying to move, only the added resistance of the water against the sides of the extra length of tubing will come into play. That is very small and that is a powerful pump that can handle it with no issue at all. I would go as far as to say a 20ft loop of 1/2 id would not pose an issue.

It might be a concern if you had a ton of stuff, mainly a very restrictive jet type block plus mosfet blocks, north and south blocks and 3 or 4 big rads and a couple of vid card blocks but even then I would only be concerned with performance because of low flow rates due to restriction than the issue of the pump not working or early failure.

Plenty of people have done what you are suggesting with no issues at all. 7-8 feet of extra 1/2 tubing is not a problem with that pump.
 
there is another solution......

if you have an open attic space above it may be possable to mount a "small" exuast fan or link into air ducts to suck away hot air or install a small air vent to deliver house air.
 
BillParrish said:
(IMO) The additional length will little or no impact on that pump. Since it is a closed loop and the "head" will remain zero, meaning the weight of the water entering the pump is the same as the weight of the water the pump is trying to move,
Click to expand...

Bill, I hope you don't mind a few n00b-ish questions a week after you did this post.

I'm not sure I follow this explanation as to why the "head" is zero. Wouldn't that depend on the vertical distance between the reservoir and the pump? And also the amount of water in the reservoir?

I'm assuming that the reservoir is lower than some other components in the the loop, like the radiator or CPU block.


only the added resistance of the water against the sides of the extra length of tubing will come into play. That is very small and that is a powerful pump that can handle it with no issue at all. I would go as far as to say a 20ft loop of 1/2 id would not pose an issue.
Click to expand...

Somewhere (skinnee.com ???) I read that one foot of tubing creates a pressure drop of 1" (yes, inch) of head. But I'm not sure how to relate that fact to anything else in the loop.

It might be a concern if you had a ton of stuff, mainly a very restrictive jet type block
Click to expand...

What is a "jet type" block? What isn't? If you think of leading CPU blocks from say Swiftech or EK, what are they?

Thanks.

x509
 
The 1' tube length to 1" head rule would only work for a specific tubing size/material/finish/flow rate. The nature of fluid flow is that the faster the fluid is flowing, the larger the pressure drop across a particular component or fitting. 20 extra feet of 1/2" diameter tubing would have an effect, but most likely negligible for WC flow rates. How much gage pressure is common on the pump output or does anyone measure this? Is it typically less than 1 bar(15 psi)?

Keeping all components on the same elevation will eliminate particular zones being too high or low-pressure. for example if you had a radiator 15 feet above the rest of the loop and then a hose falling from the rad into the res, you might see enough gravity feed for air bubbles from the res to start creeping up towards the radiator.
 
I'm not sure I follow this explanation as to why the "head" is zero. Wouldn't that depend on the vertical distance between the reservoir and the pump? And also the amount of water in the reservoir?

I'm assuming that the reservoir is lower than some other components in the the loop, like the radiator or CPU block.
Click to expand...

In a closed loop the weight of the water in the loop above the intake is the same as the weight of the water above the output. In a loop with components above and below the pump and a res thrown in it is not intuitive. The best example I can give is that what happens when your res is at the bottom of your case and you open the top ? Nothing. Water does not drain out as one would expect. Its freaky. But the air and water pressure on both the inlet and outlet side is the same so the water goes nowhere. Along the same lines the pump sees the same "pressure" or "water weight" or "resistance" or whatever you want to call it at both sides so the hydraulic head is zero or very close to it. Sorry I am EE, need a good ME to explain better. Or Google "closed loop head pressure" might show up with a better explanation. Not even sure my explanation is technically correct, I know the head is zero, as to exactly why the above is how I understand/explain it to myself. I am likely incorrect in some of the details but the end is the same.

Somewhere (skinnee.com ???) I read that one foot of tubing creates a pressure drop of 1" (yes, inch) of head. But I'm not sure how to relate that fact to anything else in the loop.
Click to expand...

One can calculate that fairly closely (have to make assumption as to the "smoothness" of the tubing inner walls), but I dont care to take the time. Bottom line is that the MCP pump used in this instance is very powerful and I know from experience that even 20ft of 1/2 ID would not even make it work hard and the resistance would be insignificant compared to the other restrictions/impediments to flow in the loop. So not saying that is incorrect, just saying it does not matter in the big picture.

Jet block
Click to expand...
Back in the single core P3 and P4 days there were jet impingement blocks that used a very restrictive inner jet insert to achieve very high velocity over the cpu in expense of flow rate. Some block still use those "nozzles" although I do not think that they are that restrictive anymore. It was mentioned as something to think about is all. I am not sure it even still applies. I probably should not have even mentioned it.

This is an interesting read, it is more about tubing sizes but the impact of increased lengths can be inferred easily.
http://www.hardforum.com/showthread.php?t=1200495
 
Last edited:
BillParrish said:
In a closed loop the weight of the water in the loop above the intake is the same as the weight of the water above the output. In a loop with components above and below the pump and a res thrown in it is not intuitive. The best example I can give is that what happens when your res is at the bottom of your case and you open the top ? Nothing. Water does not drain out as one would expect. Its freaky. But the air and water pressure on both the inlet and outlet side is the same so the water goes nowhere. Along the same lines the pump sees the same "pressure" or "water weight" or "resistance" or whatever you want to call it at both sides so the hydraulic head is zero or very close to it. Sorry I am EE, need a good ME to explain better.
Click to expand...

Bill,

Thanks. I'm neither an ME or EE., and besides, college was a long time ago. :D

So, here is what I have in mind. Corsair 800D case, but you could probably use an HAF 932 and the issues would be the same.

I have the pump at the bottom of the case, the radiator at the very top of the case. The motherboard is a P58 chipset, and the loop has only a CPU block and a reservoir, plus the radiator of course.

The reservoir is off to the side of the motherboard and is one of those cylinder-type models. Now, once my loop is full, I could run the system with the reservoir 1/4 full, 1/2 full, or 3/4 full, or even "almost full." Depending on the reservoir, there will be a different total weight of water on the intake side of the pump. But I haven't changed anything else.

So, what does this "thought experiment" prove? That I'm confused as all get-out?:confused:

x509
 
As Promised, here is the finished system and the reason I started this thread in the first place.....thanks for the advice...everything running great! 120x4 Fesser with 8 Yate Loons, grill from PerformancePC and some balsa wood painted black for the fan/rad case...sticker that came with my Lycosa keyboard...heh... Koolance low water loss couplers to easily remove the external part of the loop for easier portability...

The PC is usually further back in the closet and the radiator further out but for the picture I squeezed them together a bit...

IMG_2878.jpg
 
Last edited:
You must log in or register to reply here.
Back
Top