Water cooling, radiator question

[Fluence]

Hello, WC experts. I'm about to build a PC, and this is my first foray into liquid cooling, so here are some very simple (for you) questions about the radiator. My #1 case choice is currently the Gigabyte Aurora, so you can see what I'm considering:

Link to the Aurora

1) If I were to use an external, dual 120mm radiator such as the Black Ice Pro II, how would I mount it and fan cool the radiator?

2) Does the radiator benefit from the two 120mm case exhaust fans? The radiator seems to just mount over them, and these fans just blow warm air.

3) On that note, do I need to add two 120mm fans between the case and radiator, blowing through the radiator? Or only on the outer side of the radiator, blowing air away?

Finally, if you can recommend a good radiator to go with this case, please do so. I may just get one of the kits, but I'd prefer to go with the better price/performance of a DIM (do it myself).

Thanks a bunch

 

[Luckyboy1]

Radiators are the critical link that seems to almost always get overlooked in water cooling. If it can take two fans, giv it two fans. they by necessity of cramming it all within the case make it so you have to mount them in the least advantagious manner. that is, they have to be mounted inside the radiator blowing through the radiator instead of pulling air through them which is the best way to handle it.

Get the biggest, fattest radiator that will fit in the space allowed by your inside build and hang as many high quality fans... and yes, skip the dumb lights and get something that moves alot of air at a reasonable amount of noise instead of going for the quietest, or the mostest of air moving per fan. Hopefully the radiator you are choosing is a triple pass unit.

 

[TrevorRC]

Lucky, some people value silence. Hence the use of 3x120 radiators.

I run my SDs @ 5V, can't hear them, and still pull good deltas
[2x120mm.]

Personally, I'd switch to the MCR220.. it's worth the extra 5 dollars. [Over the BIPII]

Mind you, it's an a good 1/3 of an inch thicker. [Maybe 1/4]

If you're spending that much on a case, you might as well opt for something a bit better.. [A la Lian-Li.... a PC-G70 can be had for 170, and is about as close to perfect as you can get for sub-200.]

If you are set on the Aurora though...
Here's how you would go about doing it.

Personally, I'd go for a completely-internal enclosure, at the cost of 5.25" bays [With the G70, there is no 5.25 loss--you can, if you wish, but there are 2 positions for PSUs, making it possible to put it in the back, saving the 5.25s.].... cut two holes in the top over the 5.25s, mount the radiator there [With a Radgrillz.. Google it.]. Stick two 38mm fans in [[You can go with standard 25mm fans, but it'd be better with 38mm... less noise, if you can afford the extra room.]]
Have the fans PUSH air THROUGH the radiator... so, in the case I stated above, it'd be Radgrillz->Fans->Radiator.

PM me if you need anything else.. Hardwarrior is also a good person to talk to.

--Trevor

Edit: Also, I really wouldn't reccomend having it in the back of the case--warm air is being pushed through it rather than cold air. You'll notice a substantial difference in temps.
--T

 

[HardWarrior]

Originally posted by: Fluence
Hello, WC experts. I'm about to build a PC, and this is my first foray into liquid cooling, so here are some very simple (for you) questions about the radiator. My #1 case choice is currently the Gigabyte Aurora, so you can see what I'm considering:

Link to the Aurora

1) If I were to use an external, dual 120mm radiator such as the Black Ice Pro II, how would I mount it and fan cool the radiator?

2) Does the radiator benefit from the two 120mm case exhaust fans? The radiator seems to just mount over them, and these fans just blow warm air.

3) On that note, do I need to add two 120mm fans between the case and radiator, blowing through the radiator? Or only on the outer side of the radiator, blowing air away?

Finally, if you can recommend a good radiator to go with this case, please do so. I may just get one of the kits, but I'd prefer to go with the better price/performance of a DIM (do it myself).

Thanks a bunch

1. You have to be creative if you're going the mod route. The only people who can give you definitive ideas about rad mounting would be those who have the exact same hardware you do. Start by looking at something like this, but be prepared to go further, much further if you want a good mounting.

2. No. You need focused air moving through your rad. They're designed that way.

3. You're talking about a fan mounting approach called "push." For the sake of efficiency, I suggest "pull." Meaning you mount your fans on the rad in such a way to pull air through the rad, thereby maximizing a larger portion of its surface area. Fans in pull also make it much easier to keep your rad clean.

Your choice of rad is going to depend on your mounting to some degree, what sort of fans you want to use (powerful and noisy, less powerful a relatively quiet) and how much you want to spend. Think this through first and there will be less of a chance that you'll end up with something you won't be happy with.

 

[Luckyboy1]

Trevor, actually, I was trying to say his girlfriend lied when she said size doesn't matter... on radiators, it definately does! The bigger the better and girth is where it's at as well. If they've got one that will queeze into his case with 3 fans, so much the better!

 

[Fluence]

Low noise is one of the two reasons I'm going with a watercooling setup. I'll be overclocking, but not so much as to really need water, just that air would be loud. Coolness factor is the 2nd reason, unless the double entendre counts for two.

No gf, Luckyboy, but I'm hoping that ladies will flock to my water cooled case.

Allright, I now understand what I needed to understand. Pulling seems counter-intuitive, but I read that with a shroud it is nearly as effective as pushing. And I like the idea of not having to constantly clean the radiator, which will only further its effectiveness. On that note, is there any way to reduce dust in a pulling configuration?

As for modding, I'm not much of a modder. Meaning, I've never done it, have no tools, and failed woodshop. Ok, I would have failed woodshop, but never took it. So modding is not something I'm going to attempt at the moment, and one reason for going with the Aurora.

I'm pretty set on this case. It has exceptional reviews everywhere, very few complaints, and the closest to my ideal, missing only more intake vents and better positioning of the external USB/Firewire ports. Ironically, I found this while surfing just now:
Nextgen Aurora?

I owned a PC-60USB for a few years, and it was a great, light, sturdy, but unimpressive case. Antec makes good cases, two of which I've owned. I briefly considered the Tai Chi and Titan Robela, and am still lightly considering the Thermaltake Dream. I believe the Aurora is just calling out to me...

Anyway, a few more questions...

1) I don't think I'm going to escape rear mounting of the radiator. The top doesn't seem the best option as my computer is under my desk, and mounting inside seems as though it would increase temperatures significantly more than external mounting. Ok, this isn't really a question.

2) I'm inclined to do something like this Swiftech external setup (link)

The Aurora has two 120mms as rear exhausts. (Link to another view of the case...)

Is this a bad idea? Any better suggestions? Should I reconsider internal mounting? Really reconsider the case?

And can I mount the radiator upside down, or will this restrict water flow? I mean to have the tube connections towards the bottom of the case by the PCI slots instead of up by the PSU where it is mounted, as the Aurora's built-in holes are at the bottom.

3) How much weight does your water system add? I just want to know what to expect... 10lbs? 50lbs? I'll be using 1 GPU and 1 CPU cooler, -maybe- HD coolers in the future, so I figure pump+rad+2blocks+tubes+reservoir = 10lbs empty +water=20lbs full? (counting only the water system components)

It's not a LAN party case, but I live in hurricane land and want it mobile for evacuation season.

Thanks for the advice so far. I'm pulling advice from other threads, so thanks for that, too.

 

[HardWarrior]

Actually, the size of a rad should be determined by what the user is trying to accomplish, space limitations, how many FRU's are to be cooled, etc. There are loads of people with smaller than 120.3 rads who are satisifed with their performance. If the biggest, best blown rad is important to you, then fine. However, just saying "bigger is better" doesn't address all concerns, nor does it encompass all preferences.

 

[aigomorla]

Originally posted by: Luckyboy1
they have to be mounted inside the radiator blowing through the radiator instead of pulling air through them which is the best way to handle it.
.

Proof? from my knowledge it doesnt really matter what you do, and people have stated its better having them suck air as it leads to a lower presure inside.

 

[Fluence]

Here's a good discussion about this question. No 'proof' is provided, except some user tests.

Link @ HardForum

Summing up, pulling with a shroud or pushing are both suggested to work well for cooling.

And if any of you can answer:
Is it ok to mount an external radiator upside down, or will that restrict water flow?

 

[HardWarrior]

Originally posted by: Fluence
Here's a good discussion about this question. No 'proof' is provided, except some user tests.

Link @ HardForum

Summing up, pulling with a shroud or pushing are both suggested to work well for cooling.

The same sort of "my way is the best way" stuff that makes water forums so difficult to participate in.

And if any of you can answer:
Is it ok to mount an external radiator upside down, or will that restrict water flow?

[/quote]

I doubt it'll make much difference with a dual-pass radiator. With a single-pass you might have problems priming the loop. Using a res would probably help in both cases.

 

[Luckyboy1]

Originally posted by: aigomorla

Originally posted by: Luckyboy1
they have to be mounted inside the radiator blowing through the radiator instead of pulling air through them which is the best way to handle it.
.

Proof? from my knowledge it doesnt really matter what you do, and people have stated its better having them suck air as it leads to a lower presure inside.

Proof? You want an argument, you can either...

1) Marry my Ex

2) Go ask that question of the American Mechanical & Heating Engineers Institute because I'm too tired and old to hash out this very small, but important point Mr. prove it!

 

[HardWarrior]

My AC unit pulls air through the coils. Just because it's an old argument doesn't mean it has been decided. If someone takes it upon themselves to offer advice they should also be prepared to offer some rationale for the opinions. It's only fair.

 

[Luckyboy1]

It's not a rationality or philosophy question. The subject is poorly understood by science. All the Engineers know is that it tests out cooling better pulling the air through rather than trying to push it. The best theory as to why as far as I've seen has to do with the fact that the air is corkscrewing its way forward after the fan. If you've noticed, your radiator fins generally are not arrainged in a corkscrew shape, so there's more resistance and less flow... but it's just a theory as to why. Still, the test results are no theory. Pull the air through the radiator rather than try and push it if possible, but design doesn't always allow you to choose.

 

[HardWarrior]

A rational approach to water-cooling does indeed have its place. Lacking a scientifically proven approach to many aspects of it, what else is there? Simply asking a question about water-cooling can illicit dozens of responses, all offered as fact in many cases. Some of these opinions can even be diametrically opposed to one and other. That?s why it?s important to be able to think on your own, to some degree at least. Also, I've never seen a rad or a core that wasn't a variation on a theme. That's to say that they all pretty much look, and function, in the same manner. This being true, it's not hard to see design as a near constant. If pull works efficiently on one rad, it?ll likely work the same way on another.

 

[liquid51]

my 2 cents?

It doesn't take much time to swap some fans around. Setup a push/push, push/pull, pull/pull, push, and pull. Hell, try a push/push/pull/pull while your at it

Keep the setup that produced the lowest temps.

I think it's safe to say that this is unrefutable.

 

[HardWarrior]

Swapping fans really isn't that easy. Even if it was, why should a person that's new to water-cooling be left with trial and error? If those of us who've been using water for an extended period are good for anything in places like this it should be as a resourse that helps new people not to have to reinvent the wheel.

 

[Fluence]

It is much less the fan configuration changes than the temperature testing, that consumes time. If there were actually a proven best setup, then it should be the one recommended to try.

However, I do agree with you that each system will probably have its own best configuration, liquid51.

As for pushing vs. pulling, I would guess that pulling allows the air to interact with the radiator fins longer than in a pushing configuration, and that is why pushing works as well or better. More heat is exchanged per mm of air, whereas with pushing the cool air is through the radiator before it has a chance to do its job.

 

[HardWarrior]

Originally posted by: Fluence
It is much less the fan configuration changes than the temperature testing, that consumes time. If there were actually a proven best setup, then it should be the one recommended to try.

Really, swapping fans on and off rad isn't as easy as it might seem. Depending on the rads prefered mounting, it could be even more problematic. On the other hand, temp testing is easy if you have the right SW. Something like Everest can be set to record temps into an HTML document at predefined intervals.

 

[aigomorla]

Originally posted by: Fluence
Here's a good discussion about this question. No 'proof' is provided, except some user tests.

Link @ HardForum

Summing up, pulling with a shroud or pushing are both suggested to work well for cooling.

And if any of you can answer:
Is it ok to mount an external radiator upside down, or will that restrict water flow?

well if you remember general physics, the presure of the water going down will compensate for the water going up. Either way your going to get the same restrictions up or down. The only exception would be to mount it sideways. But then you dont have gravity helping the water down the tubes, but at the same time you dont have to worry about the water going up them as well.

 

[pkme2]

LL has cases like the PC-60 for example, that have 2 80mm to draw air over the hdds for cooling, and 2 80mm for exhaust.

I've replaced the top with Yate Loon 120mm and the remaining 80mm with YL (28dBA) My CPUs are 44C at load.

Maybe going back to basics would help all these overheating threads.
I don't consider myself as an 'expert', but I try my best when it comes to offering help.
Hope this helps some......

 

[Fluence]

You may have posted to the wrong thread, pkme2. I'm looking to set up watercooling for the first time and asking advice on radiator placement and setup. Though, my current configuration could certainly use some work, if I weren't upgrading.

Aigomria, thanks. I don't remember general physics, but that verifies Hardwarrior's assessment so I'll go with it. Gigabyte's Aurora case has two holes towards the bottom, so having the radiator live upside down will make it much easier to set up. (If I do end up with the Aurora...)

And Hardwarrior, I mean the time it takes to thoroughly test each configuration. Boot-up, running various programs to ensure a good CPU load...
Obviously, I've never changed a fan on a radiator, so I'll take your word that it's a pain.

 

[HardWarrior]

Originally posted by: Fluence
And Hardwarrior, I mean the time it takes to thoroughly test each configuration. Boot-up, running various programs to ensure a good CPU load...

Indeed.

Obviously, I've never changed a fan on a radiator, so I'll take your word that it's a pain.

Just to be clear, a lot of the inherent difficulty comes from radiators and their mounting hardware not being nearly as user friendly as they should be. I mean, designed for computer rads have been around for a while now, yet they have the potential of being very labor intensive to deal with. When I bought my latest rad it was one of the top appliances on the market and compartively expensive, but I actually had to go out and buy screws for it because none were included. Of the 5 rads I own each has its own set of problems.

To add to the suck-factor, I can't remove or change fans without pulling my current rads plumbing. So yeah, I made sure that I had a damn good idea what would work best before final assembly.

 

[aigomorla]

Originally posted by: HardWarrior

Originally posted by: Fluence
And Hardwarrior, I mean the time it takes to thoroughly test each configuration. Boot-up, running various programs to ensure a good CPU load...

Indeed.

Obviously, I've never changed a fan on a radiator, so I'll take your word that it's a pain.

Just to be clear, a lot of the inherent difficulty comes from radiators and their mounting hardware not being nearly as user friendly as they should be. I mean, designed for computer rads have been around for a while now, yet they have the potential of being very labor intensive to deal with. When I bought my latest rad it was one of the top appliances on the market and compartively expensive, but I actually had to go out and buy screws for it because none were included. Of the 5 rads I own each has its own set of problems.

To add to the suck-factor, I can't remove or change fans without pulling my current rads plumbing. So yeah, I made sure that I had a damn good idea what would work best before final assembly.

LOL i would have to agree with him on this. To change the fans on a mounted RAD is a serious PITA. Why couldnt they just make a universal clip that would allow us to just clip the stupid fan on the rad??? this question the world may never know!!!

But my rad is mounted externally sucking air from the window behind it. For me to change my fans' i would have to first drain the system, then remove the rad from the mount, then remove the tubing, then lay the rad down and unscrew the fans. This would most likely take half the day, not to mention i would have to reprime/bleed/leak test the entire loop all over again just for changing the fans or the direction of the fans. Seriously the time isnt worth a extra 1-2C. :\

The last time i did anything with my loop was when i dropped in my 7900GT's. Even then i made sure i had both b4 applying the waterblock and that they both were working. Bleeding and leak testing is such a PITA, but the extra care in prevention makes up for it times infinity.

 

[Fluence]

Ah... I was definitely not considering the time to drain the entire system before fan replacement. I'll just go with a pull configuration, it seems to be the best overall.

 

[leegroves86]

I have a 3 X 120mm radiator than has 2 Scythe 120mm fans pulling air thru it.... It works find and I idle at high 20, low 30's. Max cpu is 36 so far.
My setup consist of:
D5 pump
Black ice III radiator
DD TDX cpu block
Dual 5 1/4 reservior
1/2 inch ID tubing