Watercooling: getting pump/radiator far away from PC

[jyi786]

Now that I've gotten a GTX 980 Ti (which does run hot), it has re-sparked my interest in watercooling. There really isn't anything wrong with air cooling, but with the drastic increase in heat output, there are literally two potential goals I have that I think watercooling will get me.

1. Removing the heat and dumping it elsewhere.
2. Removing fan air noise.

For #1, I know that it can only be as cool as ambient allows. My goal with #1 is to use quick disconnect fittings directly outside the PC because I would literally have hoses running from the radiator/pump which will be in a different room altogether. This means that I will be able to store the radiator/pump in a, let's say, server closet or something, which means NO NOISE. Tell me if my idea is wrong or unfeasible.

Doing it this way means I can use a powerful pump without worrying about noise and vibration, larger diameter tubes, and a much larger radiator.

Tell me if any of my thinking is wrong here, and if anyone has pictures of something similar that they can show me as examples, that would be awesome. I scoured the watercooling pic thread, but most everyone has their radiators/pumps localized.

Thanks in advance for any help/suggestions!

 

[Tsumi]

Your idea is feasible and has been done before. Probably the best way to find similar setups is to look for Mo-Ra setups. There have been people that have gone as far as placing the radiator outside a window or on the roof.

 

[jyi786]

Your idea is feasible and has been done before. Probably the best way to find similar setups is to look for Mo-Ra setups. There have been people that have gone as far as placing the radiator outside a window or on the roof.

Thanks for the assist man. That certainly does look intriguing without me having to build everything from scratch.

 

[NoNRG]

Although this setup isn't in a different room, it reminded me of what you're trying to do. Seems like the same setup would work for you but there's no quick disconnects.

http://hardforum.com/showpost.php?p=1041632296&postcount=1340

 

[jyi786]

Although this setup isn't in a different room, it reminded me of what you're trying to do. Seems like the same setup would work for you but there's no quick disconnects.

http://hardforum.com/showpost.php?p=1041632296&postcount=1340

Haha! Nice one!!! I was looking at that post for the past few days because it is indeed kinda close to what I want. Except I'd probably go overboard with the quick disconnects.

 

[Tsumi]

Quick disconnects make everything much more convenient. They are well worth the cost, especially if you can find used ones.

 

[Frode B. Nilsen]

... running from the radiator/pump which will be in a different room altogether. This means that I will be able to store the radiator/pump in a, let's say, server closet or something, which means NO NOISE. Tell me if my idea is wrong or unfeasible. ...

Server closet? Say you output 4-500W of heat, where does that heat go?

Also, using long tubes increase the water volume, requiring a lot more pump power. A pump might not be exactly silent, all depending on your definition of silent.

You also got the issue of condensation, unless you enjoy rotten walls. You need to tackle it.

But sure, it can and have been done, even by Linus on Youtube.

You have yet to define "silent", as that is such a relevant term. But for a really silent in-room-setup, in the sub 10dbA class, you are probably way over your head for a 500W system, posting like you are. That is no trivial task. I have yet to seen it done by anyone, like never, but I am closing in on it now. I speak of cool systems, not cooking ones.

And only to find there is something called "coil whine".

 

[Erasmus354]

Server closet? Say you output 4-500W of heat, where does that heat go?

Also, using long tubes increase the water volume, requiring a lot more pump power. A pump might not be exactly silent, all depending on your definition of silent.

You also got the issue of condensation, unless you enjoy rotten walls. You need to tackle it.

But sure, it can and have been done, even by Linus on Youtube.

You have yet to define "silent", as that is such a relevant term. But for a really silent in-room-setup, in the sub 10dbA class, you are probably way over your head for a 500W system, posting like you are. That is no trivial task. I have yet to seen it done by anyone, like never, but I am closing in on it now. I speak of cool systems, not cooking ones.

And only to find there is something called "coil whine".

You are pretty incorrect on a couple of points.

1) Increase in water volume doesn't make a difference for the power of the pump needed. However the increased length of tubing will increase the restriction, which may need a stronger pump. The overall effect won't be much though, 20 feet of tubing is about the same restriction as adding another waterblock. I would suggest measuring the length of tubing and using any of the freely available calculators to figure out the restriction it will introduce.

2) Condensation is not an issue. You aren't doing sub ambient cooling so there will not be any condensation. You also aren't introducing moisture into the air because it is a closed system so rotting or mold isn't a concern.

3) Sub 10dbA is ridiculous. Average noise in a suburban household is in the low to mid 20's for being "silent". A bird farting outside your window is probably higher than 10dbA....




OP: This is an eminently do-able goal. Quick disconnects will be useful. I would recommend 1/2" tubing for the external run because that will lessen restriction.

 

[Tsumi]

Everything Erasmus said.

A few additional points:

Having the pump and radiator in a separate room will for all intents and purposes make the computer silent, as long as there is enough soundproofing from the other room.

Also, no one said anything about a server closet. It could be a well ventilated room, or even outside the house. Even the server closet can be made to be well ventilated if necessary.

 

[jyi786]

Server closet? Say you output 4-500W of heat, where does that heat go?

Also, using long tubes increase the water volume, requiring a lot more pump power. A pump might not be exactly silent, all depending on your definition of silent.

You also got the issue of condensation, unless you enjoy rotten walls. You need to tackle it.

But sure, it can and have been done, even by Linus on Youtube.

You have yet to define "silent", as that is such a relevant term. But for a really silent in-room-setup, in the sub 10dbA class, you are probably way over your head for a 500W system, posting like you are. That is no trivial task. I have yet to seen it done by anyone, like never, but I am closing in on it now. I speak of cool systems, not cooking ones.

And only to find there is something called "coil whine".

...

<...snip>

OP: This is an eminently do-able goal. Quick disconnects will be useful. I would recommend 1/2" tubing for the external run because that will lessen restriction.

<...snip>
Having the pump and radiator in a separate room will for all intents and purposes make the computer silent, as long as there is enough soundproofing from the other room.

Also, no one said anything about a server closet. It could be a well ventilated room, or even outside the house. Even the server closet can be made to be well ventilated if necessary.

^^^ What they said.

I'm not a n00b by any means. I'm well versed in the science and physics of doing actual watercooling, so I wouldn't be stupid enough to expect sub ambient temps. Getting close to it is a completely different matter, in a well ventilated separate room. In other words, get the noise AWAY from me. You'll appreciate what I mean aiming for silence if you were a musician where you can detect even the slightest pin drop in a given clip with 16+ tracks.

 

[Zarathustra[H]]

Another alternative is to just stick the entire rig in a different room, and run USB and display extensions to your desk!

Might be easier than running long tubes with water

 

[jyi786]

Another alternative is to just stick the entire rig in a different room, and run USB and display extensions to your desk!

Might be easier than running long tubes with water

I actually toyed with this idea more than you'd know. The problem is that it's not only about silence, although that is a big reason why I'd be doing this. It's also about getting the delta for the component temps closer to ambient temps. For instance, my new GTX 980Ti idles around 46C, but has full load around 83+C. That is a delta of 37 degrees, which is extremely high. I'm trying to focus on getting that much lower, and the silence is an added bonus.

 

[Frode B. Nilsen]

You are pretty incorrect on a couple of points.

1) Increase in water volume doesn't make a difference for the power of the pump needed. However the increased length of tubing will increase the restriction, which may need a stronger pump. The overall effect won't be much though, 20 feet of tubing is about the same restriction as adding another waterblock. I would suggest measuring the length of tubing and using any of the freely available calculators to figure out the restriction it will introduce.

Well, maybe you should try to add those 20feet of tubing, and then prove your point? Its like saying a short freight train is as easy to push, as a long one, and everything is only dependent on the friction of the train. It is sort of why we have the law of inertia.

It has been known for some years now.

You need to get the water going, and that is linearly dependent on the volume, or mass.

Adding two more radiators, even in parallel, nearly starved my pump. The main problem is to get started. And even then, keeping it going, requires a lot more power. Since my rads are doing parallel, restriction is equal to about one rad. It is pretty easy math really.

Water volume really start to matter once you enter the multi liter range for the loop.

2) Condensation is not an issue. You aren't doing sub ambient cooling so there will not be any condensation. You also aren't introducing moisture into the air because it is a closed system so rotting or mold isn't a concern.

No? So your in house temp is always equal to the outside temp, for any building? I sort of lost you on that one. It is the main challenge for all these heat pumps being sold these days.

If you start to pump water into a system, that is far below the ambient of the room, be my guest. Good luck on not generating condensation for such a set up.

Doing it to a neighboring room, is less of a problem, as the temp delta is low.

There is this basic law of how much water air can hold at a specific temp and pressure. It should apply in the living room, or when water cooling.

... 3) Sub 10dbA is ridiculous. Average noise in a suburban household is in the low to mid 20's for being "silent". A bird farting outside your window is probably higher than 10dbA....

You got some really strange birds where you live. I do hear my PC, but I cannot recon ever hearing a bird farting.

Seriously, silent and quiet, is a moving target. Some people describe a 1000rpm rad fan as silent and inaudible, while I have to go below 350rpm with the same fan, as to not hear it.

... OP: This is an eminently do-able goal. Quick disconnects will be useful. I would recommend 1/2" tubing for the external run because that will lessen restriction. ...

Yep. And it will add volume. The restriction difference is low, but once you get into 4-5 liter of water, you get like 4 times the volume, thus mass. You need a trade off between the two.

Frost, or sub zero temps, would be a something to consider as well.

 

[Erasmus354]

Well, maybe you should try to add those 20feet of tubing, and then prove your point? Its like saying a short freight train is as easy to push, as a long one, and everything is only dependent on the friction of the train. It is sort of why we have the law of inertia.

Prove my point? Ok.

20 feet of 1/2" ID tubing is roughly 0.6 psi of pressure drop at 1.5gpm with water at 20C. Compared to a CPU block which ranges between 1 and 3 psi of pressure drop at 1.5gpm, you can see fairly easily that 20 feet of tubing is a negligible pressure drop (and hence low restriction).

It has been known for some years now.

You need to get the water going, and that is linearly dependent on the volume, or mass.

Adding two more radiators, even in parallel, nearly starved my pump. The main problem is to get started. And even then, keeping it going, requires a lot more power. Since my rads are doing parallel, restriction is equal to about one rad. It is pretty easy math really.

Water volume really start to matter once you enter the multi liter range for the loop.

No it isn't dependent on volume of water at all. It is dependent on the overall restriction of the loop, or the pressure drop vs. head pressure of the pump if you prefer to think of it that way (they are directly related). Want to know how to figure out what your flow rate in your loop will be? Plot your pressure drop curve (flow on X axis vs. pressure drop on Y axis) of your loop with the pumping pressure curve together. Where the two lines intersect is what your flow rate will be. Notice how the only two things on these charts are flow and pressure. Volume doesn't matter.

Want a really easy science experiment you can do yourself? Try sucking up liquid through a tiny coffee stirrer straw. It is really hard to do, but there is such little volume in that tiny straw it should be easy according to you! Repeat the same with a really large diameter straw and it is really easy, but according to you it should be impossible because the large straw holds so much liquid! It is because the volume doesn't matter, the restriction does! A small straw has a huge amount of restriction and therefore takes a much higher pressure drop when compared to the large straw.

No? So your in house temp is always equal to the outside temp, for any building? I sort of lost you on that one. It is the main challenge for all these heat pumps being sold these days.

If you start to pump water into a system, that is far below the ambient of the room, be my guest. Good luck on not generating condensation for such a set up.

Doing it to a neighboring room, is less of a problem, as the temp delta is low.

There is this basic law of how much water air can hold at a specific temp and pressure. It should apply in the living room, or when water cooling.

I don't think you understand the setup that is being proposed. Water is not being pumped into any system. There is no cooling below ambient occurring. He is simply wanting to move the pump / radiators outside of the PC case and locate them in a different room. The water is still completely enclosed inside the cooling loop. Water is never coming in contact with the air, it is still being cooled at room temperature and no sub ambient cooling is occurring. Therefore it is physically impossible for condensation to form as a result of this cooling loop.

The advantage of doing this is silence. The main noise producing components (the pump(s) and fans) are located in a different room entirely. No noise producing components in your room means a silent PC

 

[Tsumi]

I love it when people try to sound smart and superior and end up making a complete fool of themselves.

@Frode: in a completely closed system with air bleeded out and/or a reservoir kept filled, it is physically impossible for you to starve your pump. You were doing something wrong, and your logic is entirely flawed.

As for your freight train example, again your knowledge is flawed. To accelerate the freight train (i.e. turning on pump), it requires more force to accelerate a heavier train (inertia, i.e. getting the flow rate up to speed). However, given the same constant force, once the acceleration phase is over, the only thing that determines the train's speed is friction. Mass only matters for the rate of acceleration, and if applicable, the friction force. Speed is entirely dependent on friction and force applied, and does not depend on mass if the friction factor does not depend on mass. For a train, rolling friction does depend on mass, but for a fluid, friction depends on the restriction of what it is moving through.

 

[Omegas]

I did something similar when mining bitcoins. I put a 480 rad outside and put copper tubing through the wall similar to what AC systems use. Had 1/2" tubing feeding between the wall and rads/computer with Koolance VL4 QD's to add with adjustments. I used a 110AC to 12V DC converter outside to run four 120 fans under my porch and covered the system in pantyhose to use as a removable and cleanable filter.

This was run with a dual MCP355 (probably over kill) cross over system, , where one reservoir fed both pumps. One pump pushed to the 480 and one pump ran through 3 MCW82's in parallel and a CPU block. There were fans in the Case also so never had a problem with moisture. When it got to cold out side (the house heater was kicking on) I would bring the rad in. The system ran flawlessly 24hrs a day for many months while it was profitable.

 

[jyi786]

I did something similar when mining bitcoins. I put a 480 rad outside and put copper tubing through the wall similar to what AC systems use. I used a 110AC to 12V DC converter outside to run four 120 fans under my porch and covered the system in pantyhose to use as a removable and cleanable filter.

This was run with a dual MCP355 (probably over kill) cross over system, , where one reservoir fed both pumps. One pump pushed to the 480 and one pump ran through 3 MCW82's in parallel and a CPU block. There were fans in the Case also so never had a problem with moisture. When it got to cold out side (the house heater was kicking on) I would bring the rad in. The system ran flawlessly 24hrs a day for months while it was profitable.

Thanks for the input bud.

Now, I just have to see if it's justified going with custom watercooling vs. an AIO. I know there are people firmly on both sides of the camps.

So for a pump, how about an Iwaki and some clear Tygon tubing from Home Cheapo? Will that work well?

 

[Tsumi]

It should run fine. I wouldn't use the word AIO nowadays, as it can mean both a CLC or a preassembled custom loop like the Swiftech.

A CLC will provide better bang for buck, be cheaper, but also highly inflexible. A custom loop (preassembled or not) will have the best raw performance potential and most flexibility.

 

[MaDSpartus]

I have a external MORA 3 (9 x 140mm) setup with two computers in series, each with dual R9 290, about 1400 watts full load for both systems. The GPUs never break 50C even with full load extended operation. I have PEX in the wall to take the water to the radiator / pumps in another room. Used koolance QD3 on the cases, on the tubes in the wall, on the radiator, comes apart into like 5 pieces for transport or maintenance when needed.

Basically it sounds like what you are after, and its great!
You definitely end up spending a good chunk of money on the setup though for frivolous things like the quick disconnects, nearly $200 on those alone. I hope they are a very long term purchase though. I'd be very happy if they last 10 years, and I think they will (its been 2 so far).

 

[Erasmus354]

Thanks for the input bud.

Now, I just have to see if it's justified going with custom watercooling vs. an AIO. I know there are people firmly on both sides of the camps.

So for a pump, how about an Iwaki and some clear Tygon tubing from Home Cheapo? Will that work well?

Iwaki's are great pumps, potentially overkill though. The tubing from Home Depot is not Tygon, it is just crappy vinyl tubing which is very stiff and hard to work with. I would not recommend it. If you want to get cheap tubing I would recommend Masterkleer tubing from McMaster.

 

[jyi786]

I have a external MORA 3 (9 x 140mm) setup with two computers in series, each with dual R9 290, about 1400 watts full load for both systems. The GPUs never break 50C even with full load extended operation. I have PEX in the wall to take the water to the radiator / pumps in another room. Used koolance QD3 on the cases, on the tubes in the wall, on the radiator, comes apart into like 5 pieces for transport or maintenance when needed.

Basically it sounds like what you are after, and its great!
You definitely end up spending a good chunk of money on the setup though for frivolous things like the quick disconnects, nearly $200 on those alone. I hope they are a very long term purchase though. I'd be very happy if they last 10 years, and I think they will (its been 2 so far).

Where did you get your MORA 3? I can't seem to find it anywhere for purchase (or the site that it is on is in German).

Iwaki's are great pumps, potentially overkill though. The tubing from Home Depot is not Tygon, it is just crappy vinyl tubing which is very stiff and hard to work with. I would not recommend it. If you want to get cheap tubing I would recommend Masterkleer tubing from McMaster.

I was reading that maybe 2 Laing D5s in series are better and quieter. Maybe something like this?

http://www.xs-pc.com/water-pumps/twin-d5-dual-bay-reservoirpump-combo

 

[Omegas]

I was feeding my MCP's with XSPC's dual pump dual slot res, worked really well. D5s for volume, MCP's for head. Either can be quiet.

 

[MaDSpartus]

Where did you get your MORA 3? I can't seem to find it anywhere for purchase (or the site that it is on is in German).



I was reading that maybe 2 Laing D5s in series are better and quieter. Maybe something like this?

http://www.xs-pc.com/water-pumps/twin-d5-dual-bay-reservoirpump-combo

I got it here:

http://www.performance-pcs.com/radiators/shopby/radiator-size--1260-9x140mm/

 

[Frode B. Nilsen]

Prove my point? Ok.

20 feet of 1/2" ID tubing is roughly 0.6 psi of pressure drop at 1.5gpm with water at 20C. Compared to a CPU block which ranges between 1 and 3 psi of pressure drop at 1.5gpm, you can see fairly easily that 20 feet of tubing is a negligible pressure drop (and hence low restriction).

The pump still produce the same power output, as in pushing force, and you add mass. The result is a given, unless you can falter the basic laws of physics.

That would win you a noble prize, but this argument will not. It is all about pressure, and that is it.

No it isn't dependent on volume of water at all. It is dependent on the overall restriction of the loop, or the pressure drop vs. head pressure of the pump if you prefer to think of it that way (they are directly related). Want to know how to figure out what your flow rate in your loop will be? Plot your pressure drop curve (flow on X axis vs. pressure drop on Y axis) of your loop with the pumping pressure curve together. Where the two lines intersect is what your flow rate will be. Notice how the only two things on these charts are flow and pressure. Volume doesn't matter.

That is because the chart is a simplification, that is fairly valid given some assumptions. It is not negating the laws of physics because of it.

Want a really easy science experiment you can do yourself? Try sucking up liquid through a tiny coffee stirrer straw. It is really hard to do, but there is such little volume in that tiny straw it should be easy according to you! Repeat the same with a really large diameter straw and it is really easy, but according to you it should be impossible because the large straw holds so much liquid! It is because the volume doesn't matter, the restriction does! A small straw has a huge amount of restriction and therefore takes a much higher pressure drop when compared to the large straw.

Wrong again. This is your claim. I never said that. The sample given, is often used to discuss particular forces in physics, but naturally, not the law of inertia. For bloody obvious reasons.

It also seems to elude you completely that the amount of water in a cpu block is small. Adding one does not add much mass, thus not affecting the result much.

Also, that there is something that is called restriction, that can be described by math, simply does not kill the law of inertia. You are more than welcome to throw a hundred valid example of restriction math, it simply will not change a single thing.

I don't think you understand the setup that is being proposed. Water is not being pumped into any system. There is no cooling below ambient occurring. He is simply wanting to move the pump / radiators outside of the PC case and locate them in a different room. The water is still completely enclosed inside the cooling loop. Water is never coming in contact with the air, it is still being cooled at room temperature and no sub ambient cooling is occurring. Therefore it is physically impossible for condensation to form as a result of this cooling loop.

The advantage of doing this is silence. The main noise producing components (the pump(s) and fans) are located in a different room entirely. No noise producing components in your room means a silent PC

No. This is a common misunderstanding. This is all about heat transfer, not noise. Noise is easy to contain in the same room as the computer, even for the super picky like me.

As for the water not being in contact with air, you miss the point completely. The water might be contained, but the heat is not. This is getting utterly boring.

Also, just go ahead and pump a 1000W into a small neighboring room. Unless you are doing something insane, the pressure, moist, and temperature of that room, will not be the same as in the room the heat is pumped from. Again for obvious reasons. Ventilating such amount of heat from small rooms, is challenging. A complexity you do not seem to master the slightest, as you claim a room being pumped with a 1000W will have the exact same ambient as a room not being fed a 1000W. That makes no sense at all. None.

This is getting utterly boring, as to the level of knowledge and understanding in here.

 

[Tsumi]

The pump still produce the same power output, as in pushing force, and you add mass. The result is a given, unless you can falter the basic laws of physics.

That would win you a noble prize, but this argument will not. It is all about pressure, and that is it.



That is because the chart is a simplification, that is fairly valid given some assumptions. It is not negating the laws of physics because of it.



Wrong again. This is your claim. I never said that. The sample given, is often used to discuss particular forces in physics, but naturally, not the law of inertia. For bloody obvious reasons.

It also seems to elude you completely that the amount of water in a cpu block is small. Adding one does not add much mass, thus not affecting the result much.

Also, that there is something that is called restriction, that can be described by math, simply does not kill the law of inertia. You are more than welcome to throw a hundred valid example of restriction math, it simply will not change a single thing.



No. This is a common misunderstanding. This is all about heat transfer, not noise. Noise is easy to contain in the same room as the computer, even for the super picky like me.

As for the water not being in contact with air, you miss the point completely. The water might be contained, but the heat is not. This is getting utterly boring.

Also, just go ahead and pump a 1000W into a small neighboring room. Unless you are doing something insane, the pressure, moist, and temperature of that room, will not be the same as in the room the heat is pumped from. Again for obvious reasons. Ventilating such amount of heat from small rooms, is challenging. A complexity you do not seem to master the slightest, as you claim a room being pumped with a 1000W will have the exact same ambient as a room not being fed a 1000W. That makes no sense at all. None.

This is getting utterly boring, as to the level of knowledge and understanding in here.

Read my post. You are the one showing a complete lack of knowledge and understanding. In case you can't find it, I'll put it here:

I love it when people try to sound smart and superior and end up making a complete fool of themselves.

@Frode: in a completely closed system with air bleeded out and/or a reservoir kept filled, it is physically impossible for you to starve your pump. You were doing something wrong, and your logic is entirely flawed.

As for your freight train example, again your knowledge is flawed. To accelerate the freight train (i.e. turning on pump), it requires more force to accelerate a heavier train (inertia, i.e. getting the flow rate up to speed). However, given the same constant force, once the acceleration phase is over, the only thing that determines the train's speed is friction. Mass only matters for the rate of acceleration, and if applicable, the friction force. Speed is entirely dependent on friction and force applied, and does not depend on mass if the friction factor does not depend on mass. For a train, rolling friction does depend on mass, but for a fluid, friction depends on the restriction of what it is moving through.

To further elaborate in case you don't get it:

Given the same amount of air friction and disregarding rolling friction, I can push any weight train to the same speed. The only thing weight does in this case is affect how long it takes me to get it up to that speed. Similarly, water in tubes work the same way. It doesn't matter how much water there is, what matters is the friction between water and the tube wall. In other words, the pressure drop due to friction. Mass is entirely irrelevant in these equations.

You completely misunderstand the law of inertia. The 2nd law of motion only applies to an object that is accelerating. Hence, F=ma. What happens when the object is finished accelerating (the water in the loop has achieved maximum flow rates)? Inertia no longer applies until the force keeping it in motion is removed (pump turned off).

You also assume the 2nd room is sealed. It does not have to be, simply leave a window open and it will ventilate the heat. A box fan set in the window of the room would easily vent out all 1000 watts without a significant rise in ambient temperatures. There are so many easy ways this can be handled, it's almost as if you're arguing this point just to be argumentative.

Learn some physics before you make a fool out of yourself again.

 

[Frode B. Nilsen]

Yeah.

https://en.wikipedia.org/wiki/Friction Hopefully, a quick look at that, should refresh your physics, and you should be able to correct yourself.

I currently also use a system with 4*480 rads, connected in both series and parallel. Once you start playing with a setup like that, the influence of mass will become apparent. If you want any flow that is. The "accepted" arguments of resistance simply fail on that. There sort of is no flow using the same pump power, and resistance is about the same. I simply need more pump power.

Sadly, running a rig that proved the current consensus on restriction to be wrong, and thus knowing it to be, I clearly am a "fool" and "superior". Stupid network effects at play. The irony is, that the "superior" argument probably is correct in this case, and you just lost your window of opportunity to learn that "superior" knowledge. That probably happens a lot in here.

Bye.

 

[Erasmus354]

:/ I just don't know what else to do, I tried to explain the science to him. I tried to give him a nice practical example that explains it simpler, and he just wouldn't accept any of it.... Frode if you would like I can give you some reference material for you to read over that will explain the physics behind closed loop water systems so you can understand it better. I realize it can seem counter intuitive sometimes, lots of people share some of your common misconceptions about the way things work.

Back to regularly scheduled programming. Are you still moving forward with this ajm? I would love to see some progress pics!

 

[jyi786]

Back to regularly scheduled programming. Are you still moving forward with this ajm? I would love to see some progress pics!

I am leaning closer and closer to doing this, yes. But the problem that I'm running into is that I am utterly and completely lost with the next step:

COMPONENTS.

Namely, fittings and pumps are killing me. There are literally 5642948032032490180 options out there and I am so lost.

Help.

 

[Frode B. Nilsen]

... Frode if you would like I can give you some reference material for you to read over that will explain the physics behind closed loop water systems so you can understand it better. I realize it can seem counter intuitive sometimes, lots of people share some of your common misconceptions about the way things work. ...

Great. Do so.

I have argued my case, and is done doing so by now.

But if you have great papers, please do share. Just make sure to share great papers, and not crap. In particular, assumptions needs to be stated and make sense, and references need to be solid, not just a bunch of postulates.

I am out of here now. This sub forum is not challenging enough, so I just bore in here. I will not read anything in here, as I will be gone. Other than checking back on the links of yours.

And no. I will not comment on them. I got better stuff to do. Even if they turn out a complete waste of time.

Bye now.

 

[Tsumi]

Yeah.

https://en.wikipedia.org/wiki/Friction Hopefully, a quick look at that, should refresh your physics, and you should be able to correct yourself.

I currently also use a system with 4*480 rads, connected in both series and parallel. Once you start playing with a setup like that, the influence of mass will become apparent. If you want any flow that is. The "accepted" arguments of resistance simply fail on that. There sort of is no flow using the same pump power, and resistance is about the same. I simply need more pump power.

Sadly, running a rig that proved the current consensus on restriction to be wrong, and thus knowing it to be, I clearly am a "fool" and "superior". Stupid network effects at play. The irony is, that the "superior" argument probably is correct in this case, and you just lost your window of opportunity to learn that "superior" knowledge. That probably happens a lot in here.

Bye.

Inertia. Based on F=ma under all circumstances. In no part of that equation is friction involved.

Darcy-Weisbach equation. The equation for calculating head loss in a pipe due to friction. The variables involved do not include mass.

Please try to educate yourself. If not, your "superiority" and bad science won't be missed.

 

[Wizzard]

Why not just have the whole system in another room / closet, and run your interface cables to the intended location?

 

[jyi786]

Why not just have the whole system in another room / closet, and run your interface cables to the intended location?

If you read earlier in this thread, I noted that I didn't want to do that because it doesn't solve the issue of actually letting components run cooler. My GPU runs about 84C full load. I want to cut that in half. Simply moving the system to another room won't help me achieve that goal.

 

[Dayaks]

I have my 9x140mm radiator/res/655 pump in my basement that pumps to my computer about 10' higher in elevation. I have no problems at all.

My basement is cooler, especially in the winter, so I have a cheap process controller that modulates the temp of the fluid to whatever I set it to. I give off around 800 watts.

Why not just have the whole system in another room / closet, and run your interface cables to the intended location?

I've also done this. Cables are cheap on newegg. No issues... just don't run your sound wires by your power wires.