New to watercooling, looking for feedback

[iRemainStanding]

I've been admiring watercooling builds for some time now, but never really had enough spare money to do it myself. Soon, I'll be buying parts to make my own finally. I wanted to make sure that there were no flaws in my design/part choices before it's too late.

My components:
Case - Corsair 760T (white)
Motherboard - GA-990FXA-UD3 (Gigabyte)
CPU - AMD FX 8370
GPU - Diamond R9 295x2
RAM - HyperX FURY 2x8G 1866
PSU - Corsair RM Series 1000W
SSD - Mushkin Enhanced Chronos 60G
HDDs - 3x 1TB Western Digital Blue (in RAID 0)

My theme is black and white, and as clean as possible. I want to jump straight into rigid tubing to keep everything as straight as possible.

Parts I'm looking at:
Pump/Res - XSPC D5 Photon 170 combo
CPU Block - EK Supremacy EVO Full Nickel
GPU Block - EK Radeon Radeon R9-295X2 VGA Liquid Cooling Block - Nickel
420 Rad - Black Ice Nemesis 420GTS Ultra Stealth U-Flow
280 Rad - Black Ice Nemesis 280GTS Ultra Stealth U-Flow
T Fitting - (1) - XSPC G1/4 black chrome
Res Top Adapter - (1) - XSPC M20 to G1/4 fillcap adapter - black chrome
Straight Fittings - (15) - Bitspower G1/4 matte black for acrylic tube OD 12MM (BP-MBEML)
Valve fitting - (1) - Bitspower BP-MVV-MBKBK - matte black
Tubing - Bitspower non-champher Crystal Link 12MM OD
Coolant - Mayhems Pastel - Ice White

I've heard that for the R9 390, you can use blocks built for the R9 290X. I was also hoping that there would be no issues keeping the backplate that came pre-installed on there. For RAD fans, I've been looking at the BitFenix Spectre Pro PWM 140mm fans (122.2 CFM, 2.8 mmH2O, 29.2 dBA) to cool the radiators. I drew up a rough sketch of the loop I was planning (not to scale in any way) to get an idea of how to set everything up. From what I've heard, a lot of things can change when actually putting a loop together, so I didn't want to take get to meticulous (Okay, I was feeling a little lazy too ) The different colors have no meaning and only serve to show where two tubes will be side by side.



Slightly above the reservoir (not 100% sure on the placement, but somewhere close) I'd be putting the T fitting leading behind the panel. There I want to put the valve and have a removable piece of tubing to both fill and drain the loop. The way I was planning on doing this was to set the computer upright while filling, making it the highest point and to set it on it's side when draining, making it the lowest point.

 

[Thick8]

I've often read that putting the smaller rad between the CPU and GPU is a good idea.
I use active back plates on my system. Martinlab has a great R9 290 waterblock faceoff you can google to see what the best block/backplate combo is.
If you're going to go with hard lines you should get Monsoon's mandrel kit.

EDIT: Buy some extra tubing as you're going to screw up a few time until you get the hang of it. I found that heating the mandrel (some) with the heat gun prior to bending the tube around it makes for perfect bends.

 

[iRemainStanding]

Thanks for the response

I've heard from a few sources that it doesn't matter where in the loop the radiators are placed as there is only a small difference in temperature at any two points. Is this something you've had problems with in the past? Also I saw the GPU block roundup on XtremeRigs. Is this the faceoff you're talking about? I can't find a different one. I've been looking at other GPU blocks, but I'm a sucker for looks and don't really like the way the active backplate looks, or blocks that don't quite cover the card. I'm thinking about the EK Block now though.

I was planning on finding something round laying around to make the bends, but after looking more into the mandrel kits they seem MUCH better. And I was planning on picking up way more tubing than I need. I know I'll mess up a lot xD

 

[cozumel]

Buy some extra tubing as you're going to screw up a few time until you get the hang of it. I found that heating the mandrel (some) with the heat gun prior to bending the tube around it makes for perfect bends.

Just wanted to second that. I've just built my first loop and bought 10ft of hoses even though the loop only requires probably around 4ft. I've used around 7ft so far and will use another foot today to change the loop slightly for efficiency and to make it look prettier too! Now that I've learnt what I've learnt with this build, I'll know I can get less next time around as I won't make the same mistakes. (Mistakes such as starting with a cut that is too long and then snipping some off and ending with a section that is too short!)

 

[iRemainStanding]

The tubing I'm looking at is only $4 per meter, I was planning on getting around 10 of them. Probably way more than I'll use, but at only $40 I'm okay with that.

 

[Thick8]

I have not personally used more than 1 radiator so my opinion is only parroting other that I've heard and read. Gabe (the owner of Swiftech) has a great article on watercooling system structure.

BTW: I think the active backplates looks badass

 

[GTXJackBauer]

Thanks for the response

I've heard from a few sources that it doesn't matter where in the loop the radiators are placed as there is only a small difference in temperature at any two points.

That is correct. The loop say after 30 mins will reach an equilibrium.

I've been looking at other GPU blocks, but I'm a sucker for looks and don't really like the way the active backplate looks, or blocks that don't quite cover the card. I'm thinking about the EK Block now though.

EK is a great choice. If you do choose EK, make sure to use their Coolingconfigurator that helps find you the right block for your GPU. A few other GPU blocks to look for as well are from Swiftech, Koolance, Heatkiller and Aquacomputer to name a few.

(Mistakes such as starting with a cut that is too long and then snipping some off and ending with a section that is too short!)

After years of watercooling, I can confidently say, been there and done that. It is always a good idea to buy extra for instances like these or you'll be waiting days for your next shipment.

 

[cozumel]

EK is a great choice. If you do choose EK, make sure to use their Coolingconfigurator that helps find you the right block for your GPU. A few other GPU blocks to look for as well are from Swiftech, Koolance, Heatkiller and Aquacomputer to name a few.

@iRemainStanding

In addition to the above, don't forget to check the revision of the board before buying the block. Even though it might be the same model, a different revision can totally change whether a block will fit or not.


And while I'm here, I wasn't sure if you planned on having some sort of drain port. I have personally found it really useful during installation as you may need to drain the loop during install, setup and testing procedures. For example, if it turns out one of your fitting is faulty and leaking. Oh...and maybe buy a couple of extra fittings too. Just in case you have any that are faulty.

 

[GTXJackBauer]

@iRemainStanding

In addition to the above, don't forget to check the revision of the board before buying the block. Even though it might be the same model, a different revision can totally change whether a block will fit or not.


And while I'm here, I wasn't sure if you planned on having some sort of drain port. I have personally found it really useful during installation as you may need to drain the loop during install, setup and testing procedures. For example, if it turns out one of your fitting is faulty and leaking. Oh...and maybe buy a couple of extra fittings too. Just in case you have any that are faulty.

+1

 

[iRemainStanding]

After doing a bit of research, apparently the MSI R9 390 is a pain in the rear and doesn't support full cover waterblocks Thinking about exchanging it for the XFX R9 390, any opinions on this?

EK is a great choice. If you do choose EK, make sure to use their Coolingconfigurator that helps find you the right block for your GPU. A few other GPU blocks to look for as well are from Swiftech, Koolance, Heatkiller and Aquacomputer to name a few.

The Swiftech and Aquacomputer blocks look like they have pretty high VRM temps.
The Koolance block has cooler VRM temps than the above two, and that's another one I may look at as well.
The Heatkiller block looks like it performs pretty well, but doesn't cover the entire card like I would prefer.

The Bitspower block has the highest core temps it looks like (by a few degrees Celsius) but has somewhat decent VRM temps.

And while I'm here, I wasn't sure if you planned on having some sort of drain port. I have personally found it really useful during installation as you may need to drain the loop during install, setup and testing procedures. For example, if it turns out one of your fitting is faulty and leaking. Oh...and maybe buy a couple of extra fittings too. Just in case you have any that are faulty.

Slightly above the reservoir (not 100% sure on the placement, but somewhere close) I'd be putting the T fitting leading behind the panel. There I want to put the valve and have a removable piece of tubing to both fill and drain the loop. The way I was planning on doing this was to set the computer upright while filling, making it the highest point and to set it on it's side when draining, making it the lowest point.

This was my plan, I'm not 100% sure how well it would work though. Thanks for the heads up on fittings, That wouldn't hurt

 

[cozumel]

Slightly above the reservoir (not 100% sure on the placement, but somewhere close) I'd be putting the T fitting leading behind the panel. There I want to put the valve and have a removable piece of tubing to both fill and drain the loop. The way I was planning on doing this was to set the computer upright while filling, making it the highest point and to set it on it's side when draining, making it the lowest point.

I missed that paragraph first time around as it was hidden underneath your loop diagram

From one watercooling noob to another, I quite like that idea. For what its worth, it works when picturing it inside my head.

 

[tachi1247]

Total noob to this as well, but putting the radiator between the cpu and gpu seems like it would lower the temp of the gpu just because the water would be cooler at that point.

It seems like simple thermo dynamics, there is a fixed amount of cooling being added and a fixed amount of heat so after awhile the temp in the loop will stabilize when measured at the same point, but the temps measured at different points in the loop once stabilized will still be different.

The only thing I can think of that might change that I'd the extra length of the loop creating more pressure loss, but I would think that would be negligible considering the total difference in length would probably only be a foot.

 

[GTXJackBauer]

After doing a bit of research, apparently the MSI R9 390 is a pain in the rear and doesn't support full cover waterblocks Thinking about exchanging it for the XFX R9 390, any opinions on this?

You might have to call in the other companies to see if the MSI is compatible with theirs before you decide to trade for an XFX. The only thing I could say about XFX is as far as I thought, I thought they were under, meaning they were gone and I've heard their CS department barely exists. No phone communications, nothing. I'd try to stick with MSI or another reputed company with great CS.

If you're not doing extreme OCing, it shouldn't really matter which GPU block you're going with. Just go with the one that looks best but I should say in some tests, the Bitspower showed to be one of the most restrictive blocks in the market. Search for some GPU WB roundups. My personal GPU WB preference would be EK, Koolance, Heatkiller or Swiftech. Those are my top 4 I'd go with 90%+ of the time.

Total noob to this as well, but putting the radiator between the cpu and gpu seems like it would lower the temp of the gpu just because the water would be cooler at that point.

Noob part, yes, the other part, no. I kid, I kid.

Here, let me put it to you guys this way. In some instances at the beginning, the liquid temp will be pretty much the same as the ambient temp. Once the system starts heating up, sure there might be some cooler and warmer spots in the loop as its flowing around. About 30 minutes plus, it will reach a "neutralized" or for better terms, equilibrium. Loop order will not matter. Loop order only matters when having a reservoir prior to a pump with a natural flow to it. If you're looking to bring the Delta-Temps down, having constant efficient flow with enough heat surface and premium radiator fans would be part of the equation. There's a whole assortment of things that can be done.

For those really interested in H20, please browse our intro to water cooling stickies and martinsliquidlab. Lots of great information.

 

[cozumel]

After doing a bit of research, apparently the MSI R9 390 is a pain in the rear and doesn't support full cover waterblocks Thinking about exchanging it for the XFX R9 390, any opinions on this?

...but I should say in some tests, the Bitspower showed to be one of the most restrictive blocks in the market. Search for some GPU WB roundups. My personal GPU WB preference would be EK, Koolance, Heatkiller or Swiftech. Those are my top 4 I'd go with 90%+ of the time.

@iRemainStanding

I kinda did what GTX is saying (I think). I did some research on the current blocks available and then decided to go with the company that offered low restriction with full cover blocks and backplates. I then went through their database to see what cards were compatible with their blocks, trying to find a listed card that was using binned chips.

Another method I considered was just selecting cards that used reference boards, as that would mean they would be compatible with just about anything. Some of the block manufacturers have lists of current cards that have been verified as having reference boards which is quite helpful.

Hope this helps with your decision making process

 

[iRemainStanding]

You might have to call in the other companies to see if the MSI is compatible with theirs before you decide to trade for an XFX. The only thing I could say about XFX is as far as I thought, I thought they were under, meaning they were gone and I've heard their CS department barely exists. No phone communications, nothing. I'd try to stick with MSI or another reputed company with great CS.

I bought my card from Micro Center last week, so it's still within the 30 day return period. I was looking at the XFX due to having the best compatibility with the EK waterblocks. As long as it isn't the new version, that is... With Micro Center's 2 year warranty, I shouldn't have to deal with XFX customer support at all

If you're not doing extreme OCing, it shouldn't really matter which GPU block you're going with. Just go with the one that looks best but I should say in some tests, the Bitspower showed to be one of the most restrictive blocks in the market. Search for some GPU WB roundups. My personal GPU WB preference would be EK, Koolance, Heatkiller or Swiftech. Those are my top 4 I'd go with 90%+ of the time.

I want to push it as far as I can go with decent temperatures. I'm using eyefinity to use three monitors at 1080, so any increase in performance is welcomed.

 

[Otter]

Total noob to this as well, but putting the radiator between the cpu and gpu seems like it would lower the temp of the gpu just because the water would be cooler at that point.

It seems like simple thermo dynamics, there is a fixed amount of cooling being added and a fixed amount of heat so after awhile the temp in the loop will stabilize when measured at the same point, but the temps measured at different points in the loop once stabilized will still be different.

You are correct about the physics, but if you do the math, you'll see why people say it doesn't matter. The water will be coolest just after the radiator, but not by much. With a good flow rate, there is so much water flowing past that it doesn't heat up very much before it gets to the radiator, and so won't cool very much in the radiator. There will be a measurable difference, but it probably won't be much more than 1C, if that. You might get 2-3C on a heavily OCed system with two high-end GPUs and everything on the same loop, but machines like that tend to have killer pumps and/or multiple loops. For a typical system, just avoid tight bends that might kink the tubing and make sure your pump immediately follows your res or t-line.

The only thing I can think of that might change that I'd the extra length of the loop creating more pressure loss, but I would think that would be negligible considering the total difference in length would probably only be a foot.

Right again, with modern pumps and blocks, the length of tubing makes little difference.

 

[tachi1247]

You are correct about the physics, but if you do the math, you'll see why people say it doesn't matter. The water will be coolest just after the radiator, but not by much. With a good flow rate, there is so much water flowing past that it doesn't heat up very much before it gets to the radiator, and so won't cool very much in the radiator. There will be a measurable difference, but it probably won't be much more than 1C, if that. You might get 2-3C on a heavily OCed system with two high-end GPUs and everything on the same loop, but machines like that tend to have killer pumps and/or multiple loops. For a typical system, just avoid tight bends that might kink the tubing and make sure your pump immediately follows your res or t-line.


Right again, with modern pumps and blocks, the length of tubing makes little difference.

Appreciate the feedback. What you're saying makes sense, but wouldn't the radiator provide the same temperature delta regardless of the input water temp (for the same fan speeds)? It seems difficult to believe that the radiator only drops the temperature 1 or 2C, if that was the case couldn't you just eliminate it all together and use a larger reservoir?

 

[iRemainStanding]

[QUOTE ] Myth: The order of components has a significant impact on temps
(eg, the radiator must be before the CPU).

Reality:*The order of components makes a difference of less than 0.5ºC
in most watercooling systems. The physics:

pump—->radiator—->CPU—->pump
pump—->CPU—->radiator—->pump

There is only one difference, and that is the position of the pump
in the loop, be it before or after the CPU.

Assuming the pump dumps about 50 watts of heat into the water and the flow
rate is 1 gallon/minute (gpm – very reasonable assumptions):

Water has a thermal capacity of 4186J/Kg-C at 22ºC and a density
of about 1g/mL

With a flow rate of 1 gpm, that’s ~3.75 liters/minute (lpm).

3.75 lpm / 60 seconds= 0.0625 liters or kilograms through the
waterblocks per second.

4186 * 0.0625 = 261.625 W/C

So that’s 1ºC warmer for every 261 watts; but only 50 watts of heat are present, so:

50 / 261.625 = 0.19ºC

Ergo there is a 0.19ºC difference in water temperature between the
inlet and outlet of the pump. This does not mean the water is only
0.19ºC warmer than air – that is an entirely different calculation.

And that’s with 50 watts. If you’re running a smaller pump, such as
the D4, you’re looking at about 15 watts.

So, do what allows for the simplest tubing runs – tubing
length/kinking will have a greater impact on temps[/QUOTE]


http://www.overclockers.com/watercooling-myths-exposed/

Science for anyone who cares

The heat is trying to reach an equilibrium with everything attached. The waterblocks have a higher temperature, and transfer heat to the cooler liquid. Without anywhere to put this heat, the temperature will keep rising until the blocks and liquid are the same. Then it will just raise higher and higher until you fry your parts. Add a radiator into the mix, and it transfers heat to this as well. The radiators are being cooled by the air surrounding it, dissipating heat from the loop.

 

[GTXJackBauer]

The heat is trying to reach an equilibrium with everything attached. The waterblocks have a higher temperature, and transfer heat to the cooler liquid. Without anywhere to put this heat, the temperature will keep rising until the blocks and liquid are the same. Then it will just raise higher and higher until you fry your parts. Add a radiator into the mix, and it transfers heat to this as well. The radiators are being cooled by the air surrounding it, dissipating heat from the loop.

My previous posts must have been missed I guess.

Nonetheless, welcome to the basics of water cooling folks. Still lots more to learn.

 

[Otter]

Appreciate the feedback. What you're saying makes sense, but wouldn't the radiator provide the same temperature delta regardless of the input water temp (for the same fan speeds)?

Yes. In fact, even the fan speed doesn't matter. That 1 or 2 C is the difference in temperature between the inlet and outlet of the radiator. That doesn't say anything about the amount of power (heat/time) that is being dissipated or the difference in temperature between the coolant and the air entering the rad.

Turn the fans up and the system will run cooler, but you'll still have the same differential between inlet and outlet, because that is determined only by the flow rate and how fast heat is dumped into the system. If you increase the speed of the pump, though, the the difference between the inlet and outlet will be smaller.

It seems difficult to believe that the radiator only drops the temperature 1 or 2C, if that was the case couldn't you just eliminate it all together and use a larger reservoir?

We could dissipate any amount of power with a small temperature differential between inlet and outlet if the flow rate was high enough. Consider the following examples.

With 100 watt load and a 1 GPM flow rate, we'd have a 0.38 C differential across the radiator. Using the math iRemainStanding presented, that is how much warmer 100 watts would make one gallon in in one minute.

200 W at the same flow rate would raise that to 0.76 C.

Increase the flow rate to 2 GPM, the we're back down to .38 C.

If we needed to cool 800 W and our flow rate was only 1 GPM, we'd have 3 C between the inlet and outlet of the radiator.

But the pumps in an insanely powerful system like that would probably push at least 1.5 GPM (because whoever built it obviously believes that if some is good, more is better), so the difference across the radiator(s) would be only 2 C.

See how it works? Measuring 1 C across the radiator doesn't tell us anything about how fast heat is coming out it.

If we took the radiator out of any of those systems, most of the heat added to the loop would remain in the coolant. Let's assume the reservoir and tubing hold in our waternotcooling system hold one gallon of water. 100W would raise that .38 C per minute. Ignoring losses from the tubing and res, the coolant would reach be 38 C over ambient in 100 minutes. Obviously we need a radiator.

So what happens if we use a radiator half as large? The differential across the radiator inlet and outlet is still only .38 C for 100 W and 1 GPM, because we still have the same amount of power and the same flow rate. But the difference between the coolant temp and the ambient air temperature will double. Double the radiator, and the difference between coolant and air is cut in half. This is why radiators like the Phobya Xtreme SUPERNOVA 1260 and the Watercool MoMO-RA3 9x140 exist.

Is it starting to make sense?