High-air-flow exhaust Radiator?

[unimatrix0]

I've been trying to design a silent water-cooled PC with as few fans as possible for a loooong while and through the quest, I've stumbled into a radiator problem.

It's become obvious to me, that the best water-cooled system is one where the radiator heat exists the case as efficient as possible. It's no good if the heat stays inside. And, in a typical ATX air-flow design, I've concluded that the radiator must be mounted right below the PSU, where the system exhaust fan is at, and that the hot air must get out of the case as efficiently as possible.

Now, the only way for the air heat to exit as efficiently as possible is for the radiator to have good air flow at its opposite side. But most thin radiators don’t seem to allow for much air flow.

I'm mainly looking at a radiator such as thermaltake's bigwater radiator but for some reason I feel this radiator doesn't have good airflow. To make things worser, thermaltake allows for exterior mounting which makes it look as if there is no air flow... because if there was then the heat would be drawn inside the case? I'm inclined to think it's a giant design flaw from Thermaltake.

So my question is, does anyone know a good thin radiator that allows for good air flow exhaust? A possible design I’ve found so far is the “finsâ€

[andyb]

Why are you interested in watercooling.???

Watercooling CAN be quiet, but its a lot more expensive, complex and dificult to achieve.

Cant you cool what you want with air, there are some excelent coolers on the market, that can cool as well as some water cooled setups with far less noise.


Andy

[unimatrix0]

Hey! thanks for the reply.

What I can tell you is that, yea, water cooling is more expensive and more complex, but I really believe it's mileages ahead of air cooling. Let me share with you what I've found. With water cooling I can use one 120mm fans to cool the CPU and GPU and efficiently redirect their heat outside the case (given you’ve got a good design).

With air cooling you typically have many smaller noisier fans, whereas with water cooling you’d use one or two big fans. And I've found that a 120mm fan is mileages quieter than any smaller fan. Also, heat transportation seems better in water tubes than air. I’m thinking of adding a water block to my HDD too. This way I could also possibly seal the front noise of my case.

I'm personally so atracted by the possibilities and wonders of water cooling, that I've scratched air cooling out of everything, except my PSU, for now

[frankgehry]

You should go over to the watercooling forum at
www.xtremesystems.org and read their stickies. They are somewhat silence oriented and have good product recommendations.

[HammerSandwich]

It's no good if the heat stays inside.

Having the radiator exhaust into the case will be okay if your CPU & GPU (and maybe HD) are WCed, and the PSU has an intake duct.

But most thin radiators don’t seem to allow for much air flow.

Thin beats thick with all else equal.

To make things worser, thermaltake allows for exterior mounting which makes it look as if there is no air flow... because if there was then the heat would be drawn inside the case?

I don't understand this. What keeps the fan from pulling air through the radiator - i.e. away from the case - in the picture you linked?

So my question is, does anyone know a good thin radiator that allows for good air flow exhaust?

Check into Thermochill's PA series, which was designed for low airflow resistance.

[unimatrix0]

Having the radiator exhaust into the case will be okay if your CPU & GPU (and maybe HD) are WCed, and the PSU has an intake duct.

It'd be ok but not ideal and far less from best! What I was suggesting though, is that, ideally if one can take heat outside directly, it'd be much better than otherwise.

To make things worser, thermaltake allows for exterior mounting which makes it look as if there is no air flow... because if there was then the heat would be drawn inside the case?

I don't understand this. What keeps the fan from pulling air through the radiator - i.e. away from the case - in the picture you linked?

Yea, I guess you can do that, you can flip the fan and make it exhaust outwards rather than inwards. By default thermaltake has it so that it blows inwards rather than outwards. Hence the perplexity..

Check into Thermochill's PA series, which was designed for low airflow resistance.

Thanks for that reference, it looks very nice, only thing killing me is the price tag :’(

[zds]

Yea, I guess you can do that, you can flip the fan and make it exhaust outwards rather than inwards. By default thermaltake has it so that it blows inwards rather than outwards. Hence the perplexity..

For what I know, almost all the radiators have been designed so that you can easily attach fans to either side and either way. There's just no standard place in ATX chassis for radiator, so you need to pick the right installation case by case. And with axial fans, pulling works better if you have to put the fan close to the radiator. Check the Chylld's guide to watercooling on this forum for more details.

[ZeroR3D]

What case do you have? Have you considered externally mounting the rad? A thin rad with high flow design is inherently noisy.

[unimatrix0]

What case do you have? Have you considered externally mounting the rad? A thin rad with high flow design is inherently noisy.

Hey, I'm using a CoolerMaster Centurion case which gave me some trouble with the radiator, but I was finally able to mount it right below the PSU. I decided to cut out the fan grill from the back to eliminate any air resistance and inherently noise. I've scratched out any possibility for the radiator to be outside; I thought I'd have better chance to sound-dampen if it was inside. And also the fact that the one 120mm fan would act as a system fan and radiator fan at the same time, and I liked the idea

About thin radiator noise, to my amazement and contrary to Chylld’s guide, I found that blowing air into the radiator is much quieter than sucking air through it. Way quieter. Airflow is also better. Through some experimentation it’s not hard to conclude that the less obstruction at a fan’s intake the more airflow. When sucking air, the fan’s intake side is “obstructedâ€

[ZeroR3D]

Traditionally, suck fans were usually mounted with a shroud and away from the rad itself. This might account for some nose reduction in addition to the difference in air turbulence area.

I found that Yate Loons sound the same either way.

[unimatrix0]

Traditionally, suck fans were usually mounted with a shroud and away from the rad itself. This might account for some nose reduction in addition to the difference in air turbulence area.

I found that Yate Loons sound the same either way.

Thanks man, yea I read about shrouds after I posted. A shroud makes a lot of sense, I'd like to try that to see what happens. Now.. what's catching my attention is that almost all of the commercial water-cooling rads/kits don't have shrouds! Either they don't make much difference, or they are unjustly neglected? Beats me.

[ZeroR3D]

Shrouds are usually used with heatercore rads I think due to the fin design and smaller size. I recall some of the earlier commercial 80mm rad/fan assemblies being designed with shrouds. There may be diminishing returns when shrouding 120mm fans on a commercial rad's lower fin-density design. Performance differences vary by user experience it seems.

You could always just hold the fan in place where it would be if the shroud was there to see if that affects noise output

[unimatrix0]

You could always just hold the fan in place where it would be if the shroud was there to see if that affects noise output

Wouldn't work... there'd be an open space where air would escape. Suppose I try sucking between fan and rad, it'd suck from the sides rather than from the rad so I wouldn't be able to judge fairly about noise levels. What could work is having some card-board and taping the fan and radiator and then try noise differences. I will try that and let you know.

Now the main issue I have is a Maxtor HD which is driving me nuts, it makes the tipical subtle HD hissing, it makes noise when it seeks, and worst of all: it ticks!

[zds]

There may be diminishing returns when shrouding 120mm fans on a commercial rad's lower fin-density design.

Isn't the point of the shroud to remove the "dead spot" at axial fans hub? However thin/thick your fins are, still the fan has the hub and thus a dead spot.

[unimatrix0]

I tried the shroud. Airflow was a bit better in suction. But noise levels were still quieter with blowing. At this point I'm inclined to think radiator design might make a difference. But this merits more research.

About commercial rads not having shrouds, it seems they have "mini" shrouds which explains why they work. But manufacturers like Thermochill also advertise shrouds to enhance performace. So it seems it's open to testing and subject to each user's configuration. In my case, blowing air to the radiator was much quieter than sucking. I'd like to hear from others...

[zds]

In my case, blowing air to the radiator was much quieter than sucking. I'd like to hear from others...

Interesting. Did you measure the temps? I mean, if pulling is more effective in terms of cooling per fan power, you could run the fan lower and end up having same temps than pushing -> might counter the noise advantage.

But this is just speculation.. we'd need some real tests :-]. It seems it would pay off to measure with your particular system before fixing everything.

[ZeroR3D]

There may be diminishing returns when shrouding 120mm fans on a commercial rad's lower fin-density design.

Isn't the point of the shroud to remove the "dead spot" at axial fans hub? However thin/thick your fins are, still the fan has the hub and thus a dead spot.

Note I said "diminishing" returns, not "no returns at all"

With 80mm and 120mm having roughly the same size hubs, the increase in blade length on the 120mm means it's able to blow/suck over a proportionately larger area and somewhat negate that dead spot. At least that's my theory.

With regards to fin-density, picture an area 80mmx80mm with high fin-density (traditional heatercore) and another area 120mmx120mm with half the fin-density. If we pretend that the same amount of heat is uniformly distributed across both areas, then superimpose a similar-sized circular dead spot in the middle of them, you can see that more heat is trapped in the spot on the 80mmx80mm area. So it becomes less necessary to shroud the 120mm system.

Of course shrouds are available on 120mm rads (I think Thermochill has a bunch available) but it seems to me that most people seem to run most 120mm sized rads fine without a shroud.

[unimatrix0]

In my case, blowing air to the radiator was much quieter than sucking. I'd like to hear from others...

Interesting. Did you measure the temps? I mean, if pulling is more effective in terms of cooling per fan power, you could run the fan lower and end up having same temps than pushing -> might counter the noise advantage.

But this is just speculation.. we'd need some real tests :-]. It seems it would pay off to measure with your particular system before fixing everything.

I already tried pulling, it was my first configuration, with fan at ~1200 RPM. With pushing, keeping same fan RPM, I'm getting much better results, about 2-3 C degrees difference. Like I said earlier, airflow was better when pushing which explains the 2-3 degree diference.

Going back to the thread topic, I was looking into Swiftech MCR120 which is below $30 and is designed for low noise. I refuse to get a ThermoChill PA120.1 is way too expensive, and I don't see where the magic is at. Someone can compare these two? I'm almost already going with the Swiftech.

[HammerSandwich]

...high fin-density (traditional heatercore)...

Don't be too quick to generalize. Many heater cores have 11-14FPI, comparable to the Thermochill PA radiators, while the HE series and Black Ice rads have about twice that. But airflow is also affected by core depth, fin thickness (i.e. the metal's gauge), fin shape, and probably a million other variables. The main problem we have is that there is so little real information available to us.

After far too much reading and a couple years experience with WCing (though not very many different components), I've concluded that reasonable differences in radiator design don't change much for our very low airflow builds. Most SPCR-approved systems are more limited in both noise and cooling by airflow than any other factor; probably than ALL other factors combined. Don't use something obviously disadvantageous (e.g. a 2" thick core with 24FPI) and go for the biggest frontal area you can fit.