Thermochill 120.3 or PA160?
 

Currently I have a setup as follows

Cathar Original LRWW #92
DD ZChip
Aquatube
Aquadrive
Eheim 1250
HTF2 Single 120mm Rad

I'm looking to replace the radiator with one I can mount on the side of the case, should I go for a 120.3 with slow moving fans or a pa160 with a slow moving fan?

There could be possibilites of using 3x 120mm slow moving fans that produce the same noise as a single silent on the pa160.

Thoughts are most welcome.

Thanks

...This really sort of depends on what you NEED.

It would really help if we knew what you are cooling and what sort of temperatures you need to maintain. A PA160 has shown itself to be an excellent low noise radiator. One fan will be more quiet than (3) as long as its the same type of fan and the RPMs are within 10%. That's my opinion anyways. Do you intend to mount the rad externally regardless?

Noise per noise the PA160 is the rough performance equivalent of a 120.2, but the 160.1 is more compact and will fit in the lower-front of cases that do not have drive bay racks there in the way.

A 120.3 is going to outperform a PA160.1 with 3 x turned down fans in comparison to a PA160.1 with a higher-speed single fan. It will beat the 160.1 both in terms of raw performance, and in terms of performance offered per noise level.

Again, the drawback is the size of the 120.3, being over twice the length of the 160.1, so it's not something you're gonna fit anywhere comfortably.

It's the old adage:

Low Noise
Low Space
High Performance

Pick any two.

The PA160.1 is designed to tread a middle ground between all 3 by targetting "Low-Moderate Noise", "Low-Moderate Space" and "Moderate-High Performance" (in comparison with existing commercial radiators).

If your desire is to favor high performance and low noise more strongly and the cost of more space means little to your situation, then the 120.3 would be the better choice.

Well said.

Well, there is one way to "cheat" that, Cathar, which is to spend lots of money. Peltiers can offer a great boost to thermal performance while keeping space consumption low. If you can fit the high current power supply in the solution, you can use a much smaller radiator and/or pump. Also, adding peltiers does reduce the system reliability and the resilience to failure for the system. Any peltier failure basically means replacing the CPU, motherboard and maybe nearby components unless you have some darn good watchdog electronics.

But, it's all tradeoffs. There is no "perfect" design, as Cathar correctly states.

Edited to add more components that a peltier can destroy in a failure mode

lol for some reason that strikes me as funny.

But speaking of throwing money at the 2of3 issue, how loud are compression systems nowadays. Perhaps some would disagree with me, but I don't consider them on the large side of things to too great extent. (except, maybe if you're used to SFF systems). I would think you could shoehorn a compression system in the larger, wider midtowers out now, with careful planning.

DrCR

Are you talking about phase-change?

I hope he is. If he's talking about an automotive compressor (read: turbo/super charger) than he's out of his tree.

If space is not an issue whatsover, i.e. you have the radiator outsideina garage what rad would be possible?

Whatever floats yer boat.

A nice copper based radiator from a high volume capacity car engine would be good. Build up a radiator box to stick it in, and install four or so 15cm fans onto the box for air-flow. Get 400+ CFM through it, and it could handle 250W of heat load for just a 1C rise in water temps.

Yes

As for a 'remote' rad box, I would consider routing it underground first. I think it was someone here a while ago that did this, though perhaps somewhere else. I'm on a 2nd story myself, or rather my computers are, and I've ruled it far too much of a hassle to have such a setup.


My current (or rather, that last thing I came up with before budget gave me a '06 at the earliest deadline) is this radbox layout to sit under the computer. Kind of like a "stand" if you will. I have a fully tower right now, and so since I would be using a midtower with this setup, the overall dimensions would be the same.

Above view, non-proportional:

http://img.photobucket.com/albums/v235/drcr1/Rev4b.png

Airflow Intake: Front, Sides. Extake: Rear
Fans: Dual Sunon A1123/A2123 HST 120x38mm AC

Best thing I've came up with at the time in an attempt to attain 'all three'.


Anyway, how loud are phase-change system nowadays i.e. how loud are the best noise/power compressors nowadays? I thought about building my own at one point (link), but I don't really have access to the equipment needed.

You may not be able to make this trade-off:
When slowing down an axial fan that's moving air through a resistance there comes a point where the fan "stalls" in the sense that it isn't producing enough pressure to move much air through the resistance. Each of the three fans is going to "see" more resistance because it's moving air through a smaller aperture (and resistance is proportional to the square of the velocity).
There's a formula for the increase in dB when adding equal-energy sound sources (which I don't remember, but I think it's add 3dB to double, add 2dB more for a third - maybe someone here can refresh me on this but for now I'm guessing three fans are about 5dB louder than a single one). For comparison 10 dB is double the sound level (or half, depending on which way you're comparing).
So... you'd need to reduce the speed on all three fans so that each was about 5dB quieter than the single one on the PA160 in order to get the same noise level - which I'm not sure is possible without stalling the fans.

dB is a log10() kind of equation*. So, doubling the power (not the apparent noise, the actual acoustic power) is about 3 (10*log10(2) = 3.0103). Tripling the power is about 4.77, and quadrupling the power is about 6.02.

So, add about 3 for the second fan, and add another 1.7 for the third fan. Four the fourth fan, just add another 1.3dB.

But, that's the power measured by the ability to do work. For apparent sound level, an additional 10dB is about twice as loud to the human ear.


* The dB is a logarithmic unit expressing the RATIO of two powers. It is defined as:
dB = 10 * log10(P2/P1).
When measuring pure amplitude, you need to modify things a bit, since the proportion is based upon the square of the two amplitudes. log(x^2) = 2*log(x). So,
dB = 20 * log10(A2/A1)

Thanks for all the advice, I think i'll go for the 120.3 as I will be mounting the radiator outside the case regardless. I currently have a 120mm radiator mounted where a PA160 would go but there is really a lack of space and having it external would be more convenient plus being away from the temps inside the case.

On the side, why the choice to go with 3/8" I/O on the PA160?

I'm not dissing the 120.3, its a great rad, but if your mounting external wouldn't you be better off with a larger copper based car rad like the one cath suggested? Surely you can acquire for = or < the price of a 120.3. Make your own shroud, mount the sucker on the side of your case, stick four low speed 12cm fans on it so they are sucking air away from the case and you'll be good to go.

...Then again a 120.3 would look cleaner....

Something like this?
My RadBox

The 120.3 would be a cleaner install and having the inlet/outlet on the same sid of the rad is ideal for access into the case.

How about PA160.1 times 2?

L

I am having a similar problem that I want to decide myself. My main aim is for low noise and performance second. But I want to keep it all self contained within the case. The case I am planning to use is a Chieftec Mesh Full Tower case. I have attached some pics as to what I plan on doing. What I would like to know is, will the performance with the HE 120.3 be any better than the PA160 with the config shown in my pics (excuse my l33t pics :p) I am planning on only using a single 12cm fan in a pull config with either placement. I want to keep it quiet.

I plan on cooling an AMD 3800+ X2 with this and a single Nvidia 7800 GTX. Probably the northbridge as well depending on what nForce 4 motherboard I get, as the NB's position on the DFI Lanparty isn't ideal.

From your sketches, the 120.3 most likely would be better with the ambient intake. Duct the 160's intake in the same way, and the differences will be very small. What fan are you planning to use?

I don't know about very small, the 120.3 has a much greater surface area than the pa160. Since you have the room you might as well use the 120.3. Too bad there isn't a 160.2, that would probably be best for you.

HammerSandwich, I was planning on using a low speed Papst that I have from my last water rig. I was thinking similar to what Razor6 said. With the same amount of airflow in both examples, I'd think that with more surface area with the 120.3 it would be better performing.

BillA's Thermochill article and Cathar's PA160 testing show that radiator efficiency is very high. We're talking about something like 95%, which means that getting significantly better performance requires more airflow, regardless of radiator size. And that leads to the PA160's design focus: as little airflow resistance as possible in a reasonably compact size. The 120.3's larger frontal area might allow a bit more airflow (though I'm not convinced this is so) but probably not enough to deliver even 1C.

Cathar?

As a user of a 120.2 & a PA160.1, I much prefer the PA160!