HESmelaugh

After getting a chance to test the Black Ice GTX 560 quad 140 radiator against some triple-rads I was lucky enough to get my hands on some more radiators. With the TFC Monsta, Watercool Mo-Ra 2 Pro, Black Ice GTS 420 and XSPC RX480 the radiators in this roundup are mostly in about the same weight-class as the GTX 560.


Testing Method

Test-loop
- Laing DDC (MCP355) with EK X-Res
- 2x Inline heaters
- 4x Water-temp sensors
- 6x Air-temp sensors
- GMR flowmeter
- T-Balancer bigNG and Sensorhub for monitoring and logging temperatures amd flowrates

- Fans used: 120 mm Scythe Kama Flex 1900 and 140 mm Sharkoon System Fans

Each test run lasts 35 minutes, 25 minutes are warm up and the data from the last 10 minutes is all averaged out.
I don't control flowrates. The pump is always going at 12V and flow is as high as it can be in the loop. Just like the CPU blocks are tested with different flowrates due to differences in restrictiveness, I'm also testing radiators at "their own" flowrates, depending solely on their restrictiveness.
I do compare radiators of the 120-mm-format with those of 140-mm-format, even though such a comparison can never be 100% fair. The rads are being tested with different fans and (deliberately) different fan-speeds. Just keep in mind that the comparison can only be seen as a rough guide and different fans would (and do) lead to different results.

Now, let's get right to the good stuff...

Contestants



From left to right: Feser / TFC Monsta (huge, expensive, flashy - if you haven't heard of this, you must be new to liquid cooling), Black Ice GTS 420 (polar opposite to the Monsta: While sharing the 3x140 format, the GTS is very thin and has very tight fin-spacing), XSPC RX480 (4x120 rad from XSPC - the RX360 proved to be a great rad for low-speed fans), Watercool Mo-Ra 2 Pro (9x120 radiator, hugely popular in Germany for it's silent cooling capability).


Flow



The TFC Monsta proves to be very unrestrictive. Save for the GTS 420 and the Mo-Ra 2 Pro, I would say that all of these radiators are showing fairly good flowrates. The Mo-Ra 2 is by far the most restrictive of the radiators tested, which isn't much of a surprise, really.


CW, single row of fans, push

NOTE: The TFC Monsta allows for 140 mm fans as well as 120 mm fans to be attached. It is always represented with two sets of data, a "140" or "120" showing which fan size was used.



As you can see, this graph is very interesting, since there are trendlines crossing each other all over the place. Each time two trendlines cross, it means that one radiator is more suitable for higher fan speeds and the other is more suitable for lower fan speeds.
For example, we can see that the Monsta, using 140 mm fans offers better performance than the GTX 560 up until about 800rpm. At higher rpm, the GTX 560 becomes the better choice.
Similarly, we can see the XSPC RX480 out-doing the GTS 420 by a considerable margin at low fan speeds, but above about 900-1000rpm, the GTS 420 is in the lead.
And, comparing the very flat, very densely finned GTS 420 to the very thick, very loosely finned Monsta, we can see how these two radiator designs affect performance: The Feser Monsta offers superior performance up to about 1300rpm, after which the GTS 420 becomes slightly better.
Also worth noting is that the Monsta loses a huge amount of performance when it is equipped with 120 mm fans. With the smaller fans, the Monsta can barely keep up with the much more compact RX480 and is overtaken in performance by the GTS 420 at around 900rpm.
The only trendline that is never crossed is that of the Mo-Ra 2 Pro. This radiator simply offers astounding and superior cooling performance all the way from 400rpm to 1600rpm. At the top end of this range, the GTX 560 is coming close to it in terms of performance, though. It's possible that the GTX 560 will outpace the Mo-Ra at very high rpm. Keep in mind though, that the higher the fan rpm, the closer all values lie together. So even if the GTX 560 does outperform the Mo-Ra at, say, 3000rpm, it will most likely be only by a tiny margin.


CW, two rows of fans, push-pull


*The Mo-Ra 2 Pro cannot be equipped with two rows, or rather "walls" of fans in push-pull like the other radiators (and even if it could, there's a limit to how many identical fans of any type I have and that limit is below 18 ). Just out of curiosity, I added the performance data for the single row of fans on the Mo-Ra in with the data on push-pull for all the other fans. Keep this in mind as you're looking at the results.

With double the fans, the tightly finned Black Ice radiators gain a lot of performance. With this setup, the Monsta (six fans) can still hold it's own against the GTX 560 (eight fans) up to about 700rpm and against the GTS 420 (six fans) up to about 900rpm - after that, it's Black Ice territory.
It's also worth noting that the RX480, while being overtaken by the GTS 420 at relatively low rpm, comes pretty close to the Monsta's level of performance at high fan speeds in push-pull.
While it's interesting and amazing respectively to see the GTX 560 (with eight fans) and the GTS 420 (with six fans) surpassing the Mo-Ra's (nine fans) performance at medium to high fan speeds, the comparison is not quite fair, of course. With 18 fans in push-pull, the Mo-Ra would undoubtably be hard to beat.


CW, all radiators tested so far, single and double fans



As you can see, this one is pretty messy due to the sheer amount of data represented.


Data



Details on the individual radiators in the posts below.

HESmelaugh

TFC (The Feser Company) Monsta 360/420



Measurements: 15 x 10.4 x 47.5 cm

Fins per cm: 3.5

The TFC Monsta is the thickest radiator I've ever seen. It has almost twice the thickness of an XSPC RX or Black Ice GTX. It also has very generous fin-spacing.
The Monsta is the only radiator in this roundup that comes with more than the bare essentials: It includes two compression fittings, two stop-fittings (to close off two of the four threads on the radiator), an extensive set of screws that are black, matching the Monsta's color, a small bottle of anti-corrosion fluid as well as two silikcone gaskets, one for installing 140 mm fans, the other for installing 120 mm fans.
The Monsta's compatibility with two different fan sizes is a nice feature, but I can't really see it as a real bonus. Looking at the data, I cannot recommend using 120 mm fans with the Monsta at all. It loses a ton of performance compared to 140 mm fans. Using the larger fans, the Monsta performs very well at low fan speeds. In other words, it's well suited for silent cooling, but not a prime choice for high rpm fans. Highflow-fanatics will also appreciate it's low restrictiveness.

On the downside, there is the minor issue of size (the radiator is too big for internal installation, as are a few others in this roundup) and the greater issue of price. At about 260 USD, this radiator isn't exactly cheap. It's performance alone certainly doesn't warrant the price, but I doubt anyone would have expected as much. With it's carbon fiber side panels and the way it's being marketed, it's clear that you buy the Monsta for it's performance on the one hand, but also for owning something special, something "Extreme", for adding a status symbol to your hardware-shrine.

So, all in all, I think the Monsta concept works, and it's raw cooling performance at low fan speeds is better than I would have expected.



HWLabs Black Ice GT Stealth 420



Measurements: 15.5 x 3 x 46 cm

Fins per cm: 10

As I mentioned above, the main reason I really wanted to include this radiator in the roundup was that it's practically the anti-thesis of the Monsta: The GTS is very thin and has a very dense fin-structure, but shares the same format (3x140) as the Monsta. To me, it's very interesting to see how these two opposites compare to each other.
The GTS 420 is a HWLabs radiator and shares the design and characteristics of all it's peers. That is to say: It has a near perfect, glossy black finish, M4 threads, has the same fin-spacing as all the Black Ice Stealth rads etc.
It also shares a really blatant weakness with all the other HWLabs radiators in the 140 mm format: It's (unnecessarily) slightly too wide to fit into 5.25" drive bays. This means that while it's length will fit into a big-tower case, it's breadth means you'll probably have to get out your Dremel.
Apart from that, this radiator leaves a very good impression and once again shows how much potential radiators for 140 mm fans have. More importantly, among the rads tested here, the GTS 420 is a real bargain with a retail price of just 85 USD. That leaves you with enough change to invest in three extra fans for it, and a GTS 420 sandwiched in between six 140 mm fans probably offers the best cooling-performance-to-size ratio you'll see in a long time. Given that those fans are doing high rpms, at least.



Watercool Mo-Ra 2 Pro



Measurements: 39 x 5.7 x 41.5 cm

Fins per cm: 6

The Mo-Ra 2 Pro is extremely popular, not to say legendary, in Germany and at least parts of Europe. It's known for offering extremely good cooling performance even with fans turned down to inaudiable speeds - perfect for silent cooling. And, as the test-results show, it really is a cooling beast and can hold it's own through a very wide range of rpm.
It's also very well manufactured and beautifully designed. Unfortunately, it has many tiny weaknesses and annoyances that a user has to put up with.
For one thing, fans cannot be attached directly to the Mo-Ra 2 Pro. In order to attach fans, you need to get a separately sold fan bracket thingy (last two pics above). This makes the whole package more expensive than it needs to be. To make things worse, the size of the holes for attaching fans on the bracket is very puzzling: The holes are too small for standard fan screws that usually come with fans, yet no screws are provided with the Mo-Ra or the fan bracket. So you can't use the ones that come with the fans and they don't give you ones you can use, either. Finally, there's the matter of flow. The Mo-Ra 2 Pro does hinder flow more than other radiators, though personally, I think it more than makes up for it with it's raw cooling power.
So, while the Mo-Ra is a great and powerful radiator, it seems that it could be improved with some relatively minor changes.
Good to know, then, that Watercool are working on a new Mo-Ra. That doesn't mean it'll come out any time soon, mind you. They're working on it and that's all that's known so far.



XSPC RX480



Measurements: 12.4 x 5.8 x 51.6 cm

Fins per cm: 4

At first glance, it might seem that the XSPC RX480 is pretty much out of it's league in this comparison. And that's true as long as you're only looking at the data itself. The XSPC RX480 has one very considerable advantage over all the other candidates in this roundup: It can be installed inside a case with relative ease. A good big-tower case will accomodate this rad and you won't have to entirely rip out the case's guts and rebuild everything around a humongous radiator in order to make it work. In fact, this is probably the only radiator in this roundup that will comfortably sit in the bottom section of a TJ07 without you having to use more than a screwdriver to install it. So, for anyone who likes their rads internal and doesn't think of a huge MountainMods box when he hears the word "Case", this radiator is one to consider.
Looking at it's cooling performance, it's clear that it is very strong with low-speed fans. With silent fans, this radiator will match the performance of a GTX 560, beat the GTS 420 by a good margin and outperform a Monsta equipped with 120 mm fans. At higher speeds, it is outpaced by most of the competition, though. I'm not intending to understate that.

Also worth mentioning: If the radiator I got is any indication, the problems the first batch of RX's had with the finish are a thing of the past. The finish on my XSPC RX480 is flawless.


Thanks for reading. I hope you enjoyed this review and comments are, as always, welcome.
Check out the original article in German on www.DeXgo.com

Cheers,
Shane

ira-k

Thanks....Its going to take a while to digest all that, very well done!

I just wonder how the thicker higher fin per inch rads would do with some 38mm thick higher pressure fans on them? 25mm low static pressure fans are kind of anemic on higher fin per inch rads, at least from what little I've run them on them they seem to be.

Thanks again for all the work, I'll put this one with the rest.

__________________
...***Beware all ye Men born of Woman, beware the Hemogoblins***...
**E8600 @ 5751Mhz ***OCN Water Cooling Club And Picture Gallery ***150 Opty @ 3242Mhz **

kimosabi

Excellent test. Great work!

__________________

OCN Forza 3 Club

nafljhy

WOW! what a test! thank you so much HES! much appreciated!

__________________
SALE! - Custom Brass Mounts - Custom Stinger V8

I Fold, Do You? - Multi-GPU2 Guide - Brass Bottom Boys GPU2 Team

Projects: Serenity - Quantum
Completed: Valkyrie - Techstation - BrassIe - FaHRack

shemer77

thanks for this, im planning to buy a gtx 560 soon, glad i know its good!

__________________
Don't worry about the world ending today; it's already tomorrow in New Zealand.

Life is too short to be bound by the moral, ethical and legal constraints imposed on us by modern day society. - 003
GO BIG OR GO HOME
"Wanting to be someone else is a waste of the person you are" (Kurt Cobain)
"Time you enjoy wasting, is not wasted time" - John Lennon.

HESmelaugh

There's probably more to be explored in high rpm, high pressure setups. Not really my thing though, as you may have gathered.

I tested an RX360 vs a GTX 360 with Ultra Kaze 3000s once but my results didn't show a significant advantage for the GTX due to the thicker fans. Ultra Kaze's might not be the right fans for the job, though...

Thanks for the comments, everyone.

Mason92

the monsta really is huge. It could cool my car. amazing!

OC_RoB

Great Job HESmelaugh! I appreciate your hard work! The quality of your data is awesome!

__________________
#################################

Nautilus 500 Mod (EASY)

What's a WaterKeg?

HESmelaugh

Thanks.

While I try my best to get good data, I do want to remind everyone that the quality of my data might not be all that awesome. I'm still looking for ways to improve precision.
The reason I write this is simply that graphs with lots of curves make anything look very sciency. Keep in mind that there still is a margin of error involved here. I'm confident that the error is quite small and that the data is relevant (i.e. the error is a lot smaller than the differences between the rads tested), but it could be better... As you know, it could always be a bit better.