The Corsair RS140 Max and RS120 Max
With the RS Max fans, Corsair is aiming really high. Across the most popular 120 and 140mm formats, it has incorporated technically what should define perhaps the most efficient fans of their kind. Especially when it comes to the 140mm variant. It too is 30mm thick, which combined with the curved LCP blades makes it a theoretical favorite to win many “first places”.
Based on the design features, we can say that the new Corsair RS Max fans are definitely not going to be only making up the numbers. We can support this statement with our knowledge of the behavior of certain elements on other, competing fans. And Corsair combines those very well in this case. We’ll go through them all in turn in this article.
The first thing that characterizes the Corsair RS Max is its 30mm thick profile, which is 5 millimeters thicker than the vast majority of other fans. On the one hand, this will admittedly worsen compatibility, but when it comes to “performance”, it will be higher. And the more restrictive the environment, the more this aspect might matter. The greater thickness mainly causes higher static pressure, which makes the drop in airflow through an obstacle lower. We know this from testing the Phanteks T30 (also a 30mm fan), which the smaller, 120mm Corsair fan (RS120 Max) also bears a striking resemblance to – seven LCP blades of very similar geometry.
The biggest difference (compared to the T30), in terms of airflow, is that the Corsair fan reaches “only” 2000 rpm. One thing for sure, then, is that compared to the Phanteks T30 (with 3000 rpm), the RS120 Max will achieve a lower maximum airflow (Corsair claims 122,33 m3/h).
But at lower speeds, with comparable airflow, the Corsair may actually be quieter (than the T30). It may or may not. It depends on several things, which will only become apparent in tests. The elite Phanteks fan still has some room for improvement in this regard and there is a lot of tuning to be done, especially at higher speeds, to achieve lower noise levels. The question of course is what is caused by the basic rotor shape (the RS120 Max has a similar one) and what by the other elements.
Among competing models, the 140mm RS140 Max variant (with a max. of 1600rpm, or 176.70m3/h) is likely to stand out even more (than the RS120 Max). It will face significantly less competition (there are fewer high-end 140mm fans than 120mm ones). Like the BeQuiet! Silent Wings (Pro) 4, it has a small number of blades (7) for 140mm fan standards, with relatively large gaps between them. However, this does not automatically mean that top-notch results will not be achieved. They actually are so with the 140mm BeQuiet! Silent Wings Pro 4, which achieves greater airflow than the 9-blade Thermaltake Toughfan 14 Pro in non-restrictive environments.
On obstacles, however, the less dense impeller design of the BeQuiet! fan (similar to that of the Corsair RS140 Max) is already starting to show, and it lags behind the Thermaltake fan due to the lower static pressure. It’s the static pressure that Corsair increases with the RS140 Max with the thicker profile (than the competitive Silent Wings Pro 4 has). Whether such a measure will be enough to catch up to the Toughfan 14 Pro even on radiators, on which the Thermaltake fan is extremely efficient, we will hopefully soon find out (in tests). Liquid crystal polymer (LCP), by the way, is also used on the Corsair RS140 Max impeller.
Common to both variants of the Corsair RS Max fans are “magnetic” bearings promising extremely low friction, and noise at this level is supposed to be very low.
The price for a single RS120 Max fan in Corsair’s European e-shop is 44 EUR and for a triple-pack you’ll currently pay around 119 EUR.
In the case of the RS140 Max, it’s priced at about 55 EUR apiece or 94 EUR for a two-pack. 140mm fans are clearly sold more in pairs (for example, for 280mm liquid cooler radiators), while 120mm ones are sold in threes. This will probably be because of the wide popularity of 360mm radiators, but also of cases where just three fans will fit behind the front panel or under the ceiling.
English translation and edit by Jozef Dudáš
RS140 Max seems to have strangely high airflow for its RPM (176.7 m3h @ 1600RPM). Based on this spec, it beats the similarly shaped Silent Wings Pro 4 @ 2215 RPM (151.13 m3h), and also the San Ace 140x38mm fan (https://publish.sanyodenki.com/San_Ace_E/book/#target/page_no=184), which achieves only 139.3 m3h at the same RPM. Granted, it’s a no obstacles environment, but the data still looks strange to me. The data for the 120 mm variant looks reasonable though.
Yes, you’re thinking correctly. Considering only 1600 rpm and the “well-known” geometry, the airflow of the RS140 Max seems to be really a bit high. Nevertheless, I have no doubt that at comparable noise it will never achieve a lower airflow than the 140 mm BeQuiet Silent Wings Pro 4. If only for the LCP blades (instead of PBT) and the possibly eliminated critical vibrations. Of course, it also depends on the thickness with which Corsair works. We will discuss this in detail (hopefully soon). The results of the RS140 Max will also be a very useful reference for the evaluation of the Noctua NF-A14x25(r).
With the San Ace, a relatively lower airflow per unit of speed will also be achieved because of the oversized impeller hub, which has a larger than standard proportion of the total cross-sectional area. This is the reason why the EK-Meltemi 120ER has never particularly excelled.
Quasarzone has recently released their review on the RS140 Max: https://quasarzone.com/bbs/qc_qsz/views/1741327
Without obstacles, they tested 68.39 CFM (116.19 m3h) @ 1582 RPM (max speed), quite a bit lower than the Silent Wings Pro 4 (77.52 CFM/131.7 m3h @ 2200 RPM). Granted, their setup introduces some resistance as seen in the lower airflow measured on the SWP4 vs. yours, but I think it’s enough to show that the official figures from Corsair are overestimated for some reason.
Not by HWBusters testing, the found the RS140MAX CFM at max speed to be only ~3.58% overestimated:
https://hwbusters.com/cooling/corsair-rs140-max-fan-review/9/
I think I have read that Corsair uses the Longwin machine internally too, so them reaching similar results make sense.
Honestly, I’m starting to doubt the accuracy of the Longwin machine, crazy as it may sound. I have started to notice some strange results. Perhaps I should analyze the results more…
What I do understand currently is that the Longwin LW-9266 operates on a very different principle vs HWcooling/Quasarzone and many others.
The later measures airflow based on velocity by using anemometers, and laminarize airflow with a wind tunnel (plus airflow straightener if measuring in “push”).
The former measures the pressures in two chambers connected by an opening, and calculate airflow by the pressure difference and hole opening size. The airflow is laminarized by a perforated sheet which “stops” the air. The extra resistance is compensated by an exhaust fan on the other side (which also allow simulation of different impedence).
On paper, both approaches seem valid, but there must be things I am overlooking that can (sometimes) cause large differences in measured values.
It’s more about precision here than accuracy, Longwin has much greater precision, and lower margin of error for all the metrics involved compared to custom built solutions.
Quasarzone Tunnel is not very advanced, it’s only a single step away from the tunnels Corsair and Phanteks used to send to reviewers (that VSG from Techpowerup uses for instance). I’m looking forward to seeing the numbers from Expreview, and HWcooling. At first glance the HWcooling setup seemed closest to Longwin results, with Expreview close, If I’m not misremembering
But yes Idk what the default Longwin Resistance setting is, or if HWBusters/Cybenetics input custom Resistance setting numbers.
The Quasarzone tunnel is much better than the TPU/Corsair one. Actual flow straightener, sealed interface between anemometer and wind tunnel, much longer tunnel.
I don’t think the HWcooling tunnel lacks precision or accuracy either. Ľubomír clearly knows what he’s doing and a lot of thought are put into the design to ensure consistent and accurate results. The main limitation here is the lack of ability to do PQ curves, which is hardly relevant to PC applications anyway.
51972 has the right idea. They are self-proclaimed “newbies to fans” and they’re smart to avoid doing in depth airflow tests by themselves, instead relying on subjective experience by multiple people (though the way they did it, not normalizing to dBA at all, is a bit weird), and on temperature tests which is infinitely better for someone who don’t thoroughly understand fans. The things actually done by (external) professionals in that video are max airflow and max static pressure tests (13:55, two models from Longwin and one model from AMCA) and vibration tests (10:00), though the criteria for the later isn’t that clear. The former is interesting in that it is a rare direct comparison of results between different professional equipment.
I see elements in the Longwin LW-9266 that can cause significant differences, but I don’t want to comment and speculate on them. It would be irresponsible. A detailed and clear analysis could be made if we had the device available for “review”. I know fairly accurately what would be examined, but, well, we would have to have that Longwin. Anyway, yes, its evaluation metrics are fundamentally different and many people consider them to be a benchmark. Personally, I wouldn’t bet my money on them. One thing is that it’s a “standard” and another is what measurement technology is “closer to reality”.
Accurate airflow measurement is based on the assumption that the pressure is uniform within each of the chambers, which can only be true if the perforated sheet does its job perfectly and lets air evenly pass through it. If the later isn’t always true, different fan exhaust patterns can skew the results in different directions. I’ll stop my speculation here 🙂
Hope you can check out these machines in person one day.
The discussions on Kaze Flex vs Slim inspired me to look into the geometry of Silent Wings Pro 4 (140) and RS140 Max more closely.
As it turns out, I have fallen into the same trap by comparing the blade geometry only based on the front profile. When viewed at an angle, SWP4 140 has an angle of attack of about 20° at the base, while the RS140 Max is at about 25°.
They still don’t seem *that* different, but it might be enough to explain the higher-than-expected airflow spec of the RS140 Max (and also its relatively poor performances against obstacles).
Also, as a suggestion to improve the fan reviews on HWcooling, taking several additional photos of the reviewed fan at different angles might be something worth doing.
–“…When viewed at an angle, SWP4 140 has an angle of attack of about 20° at the base, while the RS140 Max is at about 25°…”
And were from do you know the angles? Do you have the fans in real? Or do you have some pretrained neural network…If so, I would be very thankful if you can share it. 😉
Or do you have manually solved the angles from the pictures? …I am too lazy for that 😛
Sorry, if I bother you with so much questions, but you needn’t to answer them, I am just interested.
Estimated from these photos using a digital protractor:
https://img2.quasarzone.com/editor/2024/05/21/2dddc1242c7d2d0049c7704547b1296e.jpg
https://img2.quasarzone.com/editor/2022/10/25/7d9d38e065c8cc3d8eec049b0203b6ed.jpg
On second thought, I was wrong as I didn’t attempt to correct for perspective at all, which means the values must be off by a lot… so please ignore those values. In fact, the differences seem to be larger based on the mental 3D model inside my head. Well, I guess that’s why I would like to see images taken at several identical angles. It would make comparisons of blade designs easier.
Even so, we can deduce something from the frontal view alone: if the projection of the blades are the same, one has a deeper blades, the one with the deeper blades must have a higher angle of attack. Compare: https://www.bequiet.com/volumes/pim/fans/silentwings/silentwingspro4/xr/140mm/001.jpg. The frontal views do show a striking resemblance, and the Quasarzone pictures do show that the RS140 Max has deeper blades.
–“…which means the values must be off by a lot…”
Well, I made some crappy calculations in these two cases (sin, cos, Pythagoras), and I got something less than 53° in the case of the RX140 Max and something less than 46° in the case of the BQ SW 140 Pro (note, the angle at the rotor HUB, the angle at the tips is milder).
–“…Well, I guess that’s why I would like to see images taken at several identical angles. It would make comparisons of blade designs easier…”
Probably, it will be sufficient to take picture from 45° angle, from the identical distance and with the identical focal settings and centered on the middle of the rotor HUB.
And the root of one blade should be set at 12 o’clock …if the blade is perpendicular to the viewer the angle is 45°, if the viewer sees the inner part of the blade, the angle is steeper, and when the viewer sees the outer surface, it is milder. 😉
Have you ever done something from photogrammetry? I haven’t done anything from this sphere, yet. Who knows, maybe I give it a shot (but raising small children is quite energy and time demanding, especially in my age 🙁 )
Your estimations of the accuracy of photogrammetry for complex shapes appear slightly exaggerated. The difference in thickness of individual parts is too high, making the process too time-consuming for good results. Buying and measuring a real fan would be a better option from a cost perspective.
You are undoubtedly right, these things should also be considered and focused on when micro-comparing fans with similar impeller design. That is, if you want to be more confident about the reasons why some fans behave the way they do in comparison to each other…
We’ll see if we can incorporate these details into reviews… I don’t want to make any promises. But theoretically I could have room for these things, since the graphics cards will be tested by a new colleague. But then again, I’m taking PC case tests instead and some other stuff too, so the time saving (and the related more time for more detailed study of fan design details) will be relative, very relative. All things have to be set up to be sustainable in the long term. 🙂
BiliBili Channel 51972 also has a Longwin machine so I’m looking forward to a 3rd set up numbers to compare with (maybe they set different resistance values):
https://space.bilibili.com/64391344
Their testing is really cool, they do some tests I haven’t seen done anywhere else, and they tear apart every fan, to double check the type of bearings
https://www.bilibili.com/video/BV1Xk4y1X78K
For instance Longwin results from Phanteks T30-120 are for
1. CFM Hwbusters – 105.17, Bilibili 51972 ~104, Hwbusters
2. Static pressure HWbusters – 6.37, Bilibili 51972 ~7.32
That’s strange. The small difference in airflow could probably be attributed to different manufacturing tolerances, but as for the almost 15% difference in static pressure… plus, at comparable speeds… it occurs to me that it’s probably possible that some of these static pressure situations don’t measure at zero airflow? I don’t know myself, you have it studied better for sure.
I had a few ideas what it might, but recently, 2 things caught my eye as I was going through pdf presentation, on how to operate Longwin machines, one there’s 4 different nozzle types that can be used in front or behind a fan (page 14 of the presentation), and second, there’s different chamber shapes for different Longwin models, and no standardized perforated plates. They leave the number of plates to use up to the user, and how much open area they have, their porosity as well (with only a suggestion of 50 – 60% open area of square mesh with round wire (starts on page 20)
Now the presentation does mention a specific performance specification to be met, and first I thought maybe that’s in the AMCA 210 standard documentation, but AlwaysCounterclockwise on the discord where I mentioned, doesn’t believe it’s anywhere there, so definitely seems like something with a lot of user choice
http://www.longwin.com/download/presentation/AMCA-210-07-WT-Introduction-OP-App-20131024.pdf
http://www.longwin.com/english/presentation.html
It’s the second link with the red NEW letters
there is a few formulas for both the flow measuring mode and duct measuring mode, so maybe no matter what combination of plates and nozzles are used they need to produce a specific performance value outlined in more detail in the manual.
Ironically (as there’s a bigger difference in Static pressure numbers between the bilibili channel and HWbusters xD) the presentation literally starts with:
Flow measuring mode is low cost and large uncertainty
Pressure Measuring Mode is higher cost to compare with Flow mode, but higher accuracy and Repeatability
On the other hand, those bilibili numbers are not exact, They have so many different numbers for Longwin tests, and they all vary by a little bit (nowhere near the static pressure difference though)
They essentially visited three labs, each equipped with a different fan tester model (different brands even).
Here’s a screenshot of said results for easier viewing: https://imgur.com/ojd9Weo
The three models are “莞造AMCA-210”, “Longwin LW-9015-250”, and “Longwin LW-9015S-300”.
Can mounting pressure differences lead to such large deviations?
I find that extremely unlikely. There may be some difference, but not that significant. The mounting pressure of the fan, of course, has an influence on the performance characteristics of the fan, but really minimal. For each fan it can be a bit different. It depends on how the size of the gap between the fan frame and the structure to which the fan is attached changes due to the mounting pressure. Sometimes there may not be measurable differences at this level even with low mounting pressure (typically fans without anti-vibration pads in the corners), other times there may be. In cases when the anti-vibration pads offset the fan frame more significantly. That is why the NF-A12x25 fan has a full-circuit silicone gasket to suppress vibrations while maintaining maximum fan airflow/pressure.