one last noob question for my first custom loop

[Kalm_Traveler]

I have all the parts (I think) ready to put together my first custom loop:

 

cpu/vrm/chipset monoblock, video card blocks, two 480mm radiators, pump/reservoir combo, compression fittings for each inlet/outlet, and a 3 meter roll of soft tubing (gotta start somewhere).

 

I have seen many threads about how it doesn't matter what order you have the components, but they were always talking about a single radiator setup. With 2 radiators is this still true? If not, is there a preferred way to split them up?

Both will be in the top of the case (Thermaltake Core X9) so I was going to just run them in series if that's fine.

 

Thanks guys!

[RollinLower]

series would definitely be fine, tough to have an extra balanced loop you could do radiator 1 -> cpu monoblock -> radiator 2 -> VGA's.

that way the coolant gets chilled before entering the next hot components.

 

it might not look as good as just a serial setup tough. and the difference is likely minimal

[NoxiousOdor]

Just be sure to put your reservoir before the pump in the loop order. Other than that the order shouldn't matter at all 

[thelordofwarr]

I doesnt matter how you order them (mostly).  But I'm searching a video on that for you

[NoxiousOdor]

Just now, RollinLower said:

series would definitely be fine, tough to have an extra balanced loop you could do radiator 1 -> cpu monoblock -> radiator 2 -> VGA's.

that way the coolant gets chilled before entering the next hot components.

 

it might not look as good as just a serial setup tough. and the difference is likely minimal

It doesnt matter where the rads are, all of the fluid maintains the same temp regardless of where it is located

[RollinLower]

Just now, WhisperingKnickers said:

It doesnt matter where the rads are, all of the fluid maintains the same temp regardless of where it is located

not quite. if you feel the tube after a radiator the liquid feels quite a bit cooler. 

i mean, it needs to be for watercooling to be even effective.

[thelordofwarr]

There you go  

(just to say I personally trust in what Jay says you don't have to)

[NoxiousOdor]

2 minutes ago, RollinLower said:

not quite. if you feel the tube after a radiator the liquid feels quite a bit cooler. 

i mean, it needs to be for watercooling to be even effective.

Actually yes, that is how fluids work, after the system has been running all of the fluid will equalize in temperature. There may be very marginal differences here and there but generally it will all be the same, that is they way the physics work

[Kalm_Traveler]

Just now, WhisperingKnickers said:

Actually yes, that is how fluids work, after the system has been running all of the fluid will equalize in temperature. There may be very marginal differences here and there but generally it will all be the same, that is they way the physics work

not to argue but that doesn't make logical sense - if the fluid temperature was the same before the radiator and immediately after the radiator, it would mean the radiator was not facilitating the dissipation of heat. 

[NoxiousOdor]

5 minutes ago, Kalm_Traveler1 said:

not to argue but that doesn't make logical sense - if the fluid temperature was the same before the radiator and immediately after the radiator, it would mean the radiator was not facilitating the dissipation of heat. 

Hey man then do what seems right to you. If it was just a question of what seems logical then you could have figured this out on your own and you shouldn't have asked the question

[Kalm_Traveler]

2 minutes ago, WhisperingKnickers said:

Hey man then do what seems right to you. If it was just a question of what seems logical then you could have figured this out on your own and you shouldn't have asked the question

well I like to defer to smarter people when I have no experience with something  I can see that a system-wide average temperature would become present but the radiators have to help remove heat from the water or water cooling wouldn't be a viable way to maintain acceptable component temperatures.

 

Anyway I do appreciateall  the input and will either run the radiators in series, or maybe zig zag like @RollinLower suggested - I am much more concerned with function than form in this case.

[tikker]

1 hour ago, WhisperingKnickers said:

Actually yes, that is how fluids work, after the system has been running all of the fluid will equalize in temperature. There may be very marginal differences here and there but generally it will all be the same, that is they way the physics work

 

1 hour ago, Kalm_Traveler1 said:

not to argue but that doesn't make logical sense - if the fluid temperature was the same before the radiator and immediately after the radiator, it would mean the radiator was not facilitating the dissipation of heat. 

The fact that the temperature doesn't keep climbing indefinitely already tells you that heat is being dissipated.

Starting fresh there may be a noticable difference, as everything is nice and cool, but eventually the loop will equalize to a certain temperature, since the heat transfer and dissipation is not 100% effective. That is, starting at room temperature, the water coming from components will be hotter than room temperature (obviously). The water coming out of the radiator will not be room temperature again, however, but slightly hotter. So the loop will slowly heat up until an equilibrium temperature is reached. What temperature this is depends on the heat output of your components and the cooling capacity/efficiency of your loop.

 

On-topic: I'd go for what is aesthetically most pleasing as the difference will likely be small.

[Kalm_Traveler]

37 minutes ago, tikker said:

 

The fact that the temperature doesn't keep climbing indefinitely already tells you that heat is being dissipated.

Starting fresh there may be a noticable difference, as everything is nice and cool, but eventually the loop will equalize to a certain temperature, since the heat transfer and dissipation is not 100% effective. That is, starting at room temperature, the water coming from components will be hotter than room temperature (obviously). The water coming out of the radiator will not be room temperature again, however, but slightly hotter. So the loop will slowly heat up until an equilibrium temperature is reached. What temperature this is depends on the heat output of your components and the cooling capacity/efficiency of your loop.

 

On-topic: I'd go for what is aesthetically most pleasing as the difference will likely be small.

I understand the first part, but it would seem that with the radiators dissipating heat, the water temperature would always be slightly hotter going into the radiator than coming out of it. If the entire loop ended up being exactly the same temperature, wouldn't that mean that no heat was being removed, so the temperature would rise indefinitely?

 

IE, let's say you have 50C water going into a single radiator that is very effective and is able to remove enough heat energy that the water is down 25C by the time it exits the radiator, until that cooled water picks up more heat energy from the next component it reaches, it should remain a lower temperature than the water that has just passed through a GPU block, CPU block, etc and not yet passed through the radiator, right?

[tikker]

48 minutes ago, Kalm_Traveler1 said:

I understand the first part, but it would seem that with the radiators dissipating heat, the water temperature would always be slightly hotter going into the radiator than coming out of it. If the entire loop ended up being exactly the same temperature, wouldn't that mean that no heat was being removed, so the temperature would rise indefinitely?

 

IE, let's say you have 50C water going into a single radiator that is very effective and is able to remove enough heat energy that the water is down 25C by the time it exits the radiator, until that cooled water picks up more heat energy from the next component it reaches, it should remain a lower temperature than the water that has just passed through a GPU block, CPU block, etc and not yet passed through the radiator, right?

Yes true, the water going into the radiator will be warmer than what is coming out. It will be a small difference though practically due to the flow rate, so everything will be approximately the same temperature.

When the loop equalizes it means that the radiator is dissipating as much heat as the components are producing, hence there is no net temperature increase as a whole.

And as the water heats up less heat will be transferred from the components to the water and more heat will be transferred from the water to the radiator/outside air. This will eventually balance.

[JR88]

It get's even weirder a stronger Pump or try a couple with Ts/ splitters and the Parallel stuff loop. That's more Pump Head/headroom volume. you're 480mm is probably overkill anyways. the 360mm has plenty of cooling capacity in theory.

 

But yeah loop order isn't real significant but there some differences and makes more since at least to go out of the Rad into the blocks First there has to be a cooler potential there.

 

No one wants to mess with some crazy Loop though and will probably just run series.

 

[For Science!]

2 hours ago, Kalm_Traveler1 said:

IE, let's say you have 50C water going into a single radiator that is very effective and is able to remove enough heat energy that the water is down 25C by the time it exits the radiator, until that cooled water picks up more heat energy from the next component it reaches, it should remain a lower temperature than the water that has just passed through a GPU block, CPU block, etc and not yet passed through the radiator, right?

 

You have to consider pump speed, what you outline above would be true if the water spent a very long time in the radiator. However your average watercooled system operates at a flowrate of at least 1 gpm, or 4.4 L/min (with a D5 or DDC). Given that a standard water cooled PC loop is about 1~2 L in volume, it is easy to visualize that any segment of water will go around the whole system about 2~4 times per minute. And therefore the time that any segment of water stays in contact with the same area (say, radiator, for example) is so short that the whole system temperature equilibrates as a whole.

 

In practice, the water in a water cooled PC heats up as a whole, and cools down as a whole for this reason. This is why component order does not matter and that the one radiator is not hotter than another, or the input is hotter than the output, etc. Unless you use something super viscous, or a very very slow pump speed, the difference is negligible.

 

This is visualized fairly well in this time lapse of load + water. You will see that the water color uniformly turns warmer (Except for the color bleed of the VRMs, but that is explained).

 

 

[Mick Naughty]

Would be dumb to put a rad after each component. Impractical as well unless you have a tiny loop. Would also be pointless to create bridges for gpus. Water isn't getting hot enough to be a problem. Its not a car.

 

Only thing that matters is pump speed based of serial or parallel and that the pump is being feed directly or indirectly from the res.

[Kalm_Traveler]

4 hours ago, JR88 said:

It get's even weirder a stronger Pump or try a couple with Ts/ splitters and the Parallel stuff loop. That's more Pump Head/headroom volume. you're 480mm is probably overkill anyways. the 360mm has plenty of cooling capacity in theory.

 

But yeah loop order isn't real significant but there some differences and makes more since at least to go out of the Rad into the blocks First there has to be a cooler potential there.

 

No one wants to mess with some crazy Loop though and will probably just run series.

 

gotcha, I just did EK's configurator and it recommended the dual 480's so that's what I went with. The 6900k gets pretty toasty when pushed to it's limit as I have seen with the old 360 AIO, and I imagine two shunt-modded Titan V's will be putting out some decent heat as well. Unsure about the chipset and VRMs though...

[andrewmp6]

With dual rads that big you could run the fans at a slower rpm and still get the cooling you need.And if you want tempered glass for that coe x9 http://www.thermaltake.com/Chassis/Accessories_/_/C_00003114/Core_X9_Tempered_Glass_Upgrade_Kit/design.htm

[Kalm_Traveler]

16 hours ago, andrewmp6 said:

With dual rads that big you could run the fans at a slower rpm and still get the cooling you need.And if you want tempered glass for that coe x9 http://www.thermaltake.com/Chassis/Accessories_/_/C_00003114/Core_X9_Tempered_Glass_Upgrade_Kit/design.htm

huge thanks!! I had no idea that existed and wished they would make one... ordered!

[andrewmp6]

10 hours ago, Kalm_Traveler1 said:

huge thanks!! I had no idea that existed and wished they would make one... ordered!

No problem i like that case and found the link in another forum about the core x snow line up.And with the side panels being the same you can run 2 tempered glass sides or just one whatever you like.

[Kalm_Traveler]

On 3/22/2018 at 11:46 PM, andrewmp6 said:

No problem i like that case and found the link in another forum about the core x snow line up.And with the side panels being the same you can run 2 tempered glass sides or just one whatever you like.

So I got one first to see how they look, seems like it doesn't fit correctly though... Look at the gap between the glass and case frame