Return to Level1Techs.com

Fan bearing types question

I’ve just seen the video of Jay about the difference in fans and noticed him stating that fluid bearing type fans are not that good when it comes to longevity:

Since I have built a new PC in December and also bought seven SilverStone SST-AB-120R I immediately googled the life expectancy of fluid bearing fans and stumbled over this article by GamersNexus:

Now, I’m a bit confused, because here they applaud the longevity of fluid bearing fans.

Do you know the reason for the difference in both opinions? Or has the market changed that drastically from 2012, when the GN article was written by Steve.

Its fine. I think the best “bearing type” is the magnetic one but it is absurdly expensive for what its supposed to do.

Since the parts dont touch it should be able to operate smoothly indefinitely.


Also I dont watch Jayz… GN is a better source overall even if its dry and frequently boring. Our overlords are also good but they dont spit out content at the pace of LTT or even GN

3 Likes

Woooow, he really said that? I mean I can watch the video, but…
Ok, here’s the thing: I have about 10 fans. I know this ain’t a huge pool but they are all fluid bearing and they are all a decade old.
Maybe he is talking about the overpriced Corsair or whatever fans he has been given as marketing, but I bought mine cheap from the store and fluid bearings are amazing.

2 Likes

Honestly? You don’t even need to educate yourself too much on this.

All fan rotor types can vary from bad to very good; all of them. It’s as usual a matter of implementation; what went with design, prime materials, manfacturing processes.
If you just -had- to pick the best out of all of them?
i) you’d need to keep in mind and never forget that this only applies to the ““PC”” category?
ii) that while exceptional a design, its implementation thus far is… lacking?

The so-called ‘magnetic levitation’ kind.
Sunon proprietary, recently licensed to Corsair and only Corsair. It’s what they use on their ML/maglev fans.
Even so, as you just read, only in theory. The implementation lacks.

  • Sunons aren’t made for ‘noise-to-performance’, nor are they made for ‘performance’. They’re made for ‘performance-to-longevity-to-voltage ratings for regulation meeting server units’.
  • Corsairs, despite being sold at over 30 a pop, are cheap and hasty implementations (ergo as much profit as possible) of this design. And that’s for their 120s. Their 140s are even worse, as they’re just 120s scaled upwards. You never do that for fans. Aerodynamics.

In practice then?
As in all other subjects, forget “best” and stick to individual practices; yours.

For Performance-to-longevity, Noctua IPCs. Disadvantage is noise.
For Performance only, shit as shit can be, i just want “bestest” performance, EK vardars. Disadvantages are noise and longevity.
For Performance-to-noise, the new Noctuas that only come in 120s. Disadvantage is pricing (worth it though) and colouration for some.
For performance-to-noise at 140 until Noctua gets its butt in gear, the Gentle Typhoons (can still be found). Disadvantage is lack of PWM mode for most models.
For low noise above all else, BeQuiet, the expensive ones, Shadow Rock or whatever. Disadvantages are low performance and issues when driven through PWM, too high impedance when daisy-chained. You’d need a good external controller (Aquaero) or they will suffer; even if you’ll never know of it.

And that’s it.
Implementation :slight_smile:
Some of the above are “ancient” for today’s standards, but… implementation.

As always in life, feel free to “research” this deeply.
If my word means anything to you? Can guarantee you that:
i) this is the list you’ll end up with.
ii) knowing by such time that the difference is so, so minimal? Your journey wasn’t worth it ^^

2 Likes

Add Arctic to the picture and you won’t have to spend 30$ for a fan…
Fractal Design ancient Silent Series R2 are insanely quiet even after a decade in my case. And I don’t clean very often cause lazy…

Seriously people are you recommending fans based on the price? Arctic is quarter the price of vardars and similar performance.

1 Like

Arctic never impressed me tbh.
There’s better for any category one might be focused on. And when better amounts to a few measly bucks more, why settle?

Again, just me, not picking on you :slight_smile:

2 Likes

What about when the quality is the same, but you buy either 1 BeQuiet or 5 Arctic?
PS: Sorry, I’m in a bit of a mood today. I will stop, cause I feel I am getting annoying…

2 Likes

Got me a very thick skin, don’t worry, it’s fine!

And it’s not am afraid :slight_smile:
If it were, sure i’d buy 5 for the price of one; but…

The thing with fans is really quality/longevity, goes into noise too.
As stated above, the differences in performance or raw DBs can be minimal up to 1500ish. However, given enough time, you’ll find rattles and whines starting to be audible when none were before. You might hear the distintive sound of the fan powering down before powering back up, even though you run it stable. You will almost assuredly find its sound signature having changed over the first 6 months. Being generous here, many overhyped fans start sounding even shittier even sooner, but you get the idea.

That’s what i pay for; again, personally.
Because some fans, you gotta dismantle the whole PC to replace them and frankly this isn’t a hobby for me.
It’s just that from a certain quality and above, you naturally get other consequent benefits, be they in regard to pure DB or performance.

Touch an Arctic fan blade. Then touch a Noctua IPC. Real simple stuff, forget mechanics, hydraulics or aerodynamics :slight_smile:

P.S. Forgot to mention this, though by now you probably won’t see it. The kinder the environment (heat, static, fields, UVs, waves), the more moot the above, should go without saying. Had the privilege or misfortune to test in harsher than average conditions and just took my lessons home, literally :slight_smile:

2 Likes

I did, it almost cut my finger off…
The arctic fans are of the quality of Noctua and BeQuiet. Even their old series is super good.
I have used a bunch of cheap crap fans before and since I started using Arctic I haven’t bought another fan brand. I don’t know. I don’t believe a fan of 25 euros can give me so much more, to skip my 5 euros fan that I am absolutely happy with.

I don’t know, some times I just sound like an ass, even when I don’t want to. I am not trying to, it just happens sometimes.
I am a bit enthusiastic, because some people have dismissed Arctic on the basis of “they are cheap, so how good can they be” and “if they were good, the tech media would talk about it”, not understanding the tech media sits on their asses and wait for review samples of everything. Arctic started sending review samples here and there and what do you know, turns out their stuff is pretty good, but nobody knew about it.

2 Likes

Was enthusiastic myself, so i get you just fine. But it’s late down here and i’m old :slight_smile:

In short and for everyone’s benefit:

  • forget “reviews”. The only way to really test a fan is with a special chamber, running some very expensive equipment attached to it and that’s assuming you know how to use them.
  • ‘a’ fan may work for you, in which case awesome and honestly. But the more you broaden the scope, the more the curve will align to my earlier statement. Whether one opts to go for safer or smarter… individual.
  • sound works in funny ways. One Aenra fan has 10DB, three identical Aenra fans running at identical speeds will have 16. NOT 10. If the psycho_666 fan has 17DBs but pulls as much as the three combined, which with do i benefit from?
  • To choose fans with numbers rather than what Aenra says, you need know the following:
    i) the airflow your cooler requires. Doesn’t matter the type or kind, all require a certain volume.
    ii) the resistance your filters and grills add.
    iii) the static pressure of the fan in question so as to calculate how much airflow it’s actually bringing in.
    iv) that companies test in optimal/specific setups and that their numbers don’t apply to real-life scenarios. A good rule of thumb?
    iv)-a) For Noctua, deduct 10% from anything they advertise
    iv)-b) For select other good companies, deduct 15%
    iv)-c) for everyone else, at least 20%

Or, you spend a few bucks more and you don’t have to do any of that.

Didn’t you say the benefits are minimal?
Yes i did, between ‘x’ and ‘y’ exceptional fan. Specifically. Not generally.

O.K., that’s it for tonight, lol

2 Likes

I agree with that.

I will disagree with that.
I don’t buy a fan to put it in an expensive equipment and what not. I buy it to put in on a case and cool some hardware. So I would expect a fan testing at least in the “media” to be focused on putting a fan on a damn case and cooling some damn PCs. No, even GN are buying expensive hardware. Their CPU coolers reviews doesn’t include CPUs in any ways. It’s absolutely worthless…

True, and that’s why I don’t like discount options. I discount stuff like Cooler Master and Thermaltake fans, cause they aren’t anywhere near any decent level of fans. Even something like Enermax TB fans, that had the fan controller integrated in the fan itself have uses.

We are essentially arguing about 1 degree celsius and 2dBa for 20$ or so, so yeah…
Night :slight_smile:

1 Like

There are only three main bearing types: sleeve, ball and maglev. All others are a variation of sleeve which sometimes affects longevity.

Now as for fans. I have used many over the years and have a couple of boxes full of all sorts of PC fans. I have sleeve bearing fans that are more than a decade old and still work like new. That thing that I found kills fans the most is dust, so clean them out every once in a while.

I stopped spending more than $20 for a single fan a long time ago because it’s simply not worth it. Personally I have fallen in love with the Phanteks PH-F120MPs (original version) since they are very quiet even at full speed and cool really well in the 1000-1500 rpm range compared to others. I threw out my ML120s in favor of those since I found them quieter and better performers at similar rpm ranges compared to the Corsairs (and the ML120s are very good fans). The 140mm versions (PH-F140MP) are not that great but still serviceable. Both versions can be mounted horizontally with no issues.

Thermalright has some good fans and they are mostly sleeve. Scythe fans are also good. Noctua fans I would never consider unless I get them with a cooler (and I like Noctua coolers); I have no problem with their brown color scheme, it’s just that the fans are too expensive for what they offer in terms of performance and even noise metrics.

1 Like

Yeah, I also prefer GN and L1, but Jay has some experience in building computers as well, especially when using a custom loop. Furthermore, he also analysed the New World frying 3090s which GN did not do due to time and resource constraints.

I think he states that dust can get caught in the fan more and also that the fluid can dissipate.

@Aenra Thanks, you a probably right that shoddy implementation can mess up the best design.

GN just recently purchased gear for fan testing. :slight_smile:

1 Like

The only way to test fans is to put them into a few workstations in an office and see how many years under less than favorable-conditions they last.

/thread
:wink:


I had 2 fans fail. One was a BeQuiet Silent Wings 2 (worked as an exhaust fan sitting in the stream from the CPU cooler) and the other was an Arctic I had in some project of mine that got killed when the regulator controlling it blew up.

3 Likes

to be honest i found that fan orientation and spindle type to be 2 of the biggest deciders on where to put fans.
fdb should only be mounted horizontally. mounted vertically the spindle sinks throught he fluid and rests on the metal of the housing when the fan is turned off. as a result you get wearing on the spindle as the fan spins up and down during power cycles.
same issue with maglev except when there’s no current the spindle sinks in the housing.

sleeve bearings best mounted vertically
twin bearing, best mounted horizontally inverted. (gpu fans typically)

all my results while anecdotal are garnered over time, with trial and error. :slight_smile:

1 Like

Good evening;
Am not arguing; on the contrary, am putting some effort into helping you grasp that; each post addressed to you begins with an intro so as to showcase that :slight_smile:

On the meat of the matter, i will repeat that it’s not just* a couple of degrees that dictate my preferences, but mostly the qualitative factor and what it entails, empirically, in the long run. Long run.
Will also add as i neglected to last night, that warranty is a factor here as well.
Have had fans showing issues well before their warranty expired.
You try getting one replaced with zero hassle, close to no time wasted and zero shipping expenses on you.
Really, go ahead. See how many companies offer you this.

(you as in figuratively, we’re not arguing i’ll humbly remind)

  • when the heat exchange is done by air (as is the case in both air and “liquid” cooling), a few degrees of Celsius difference is all one can aim for; if not just one. Context. And within said context, yes, 2-3 degrees lower temps at about the same DBs is exceptionally indicative in terms of quality/performance. Indicative. Whether any individual’s interested in 2-3 degrees, that’s subjective and on a per-individual basis.

And since i saw the price factor mentioned yet again in one of the following posts…
In life you get what you pay for. Or you get what you deserve, based on what you worked for. Period.
I don’t split hairs for a measly bucks, i do for higher amounts, sure, but not for a few measly bucks.

if one has time to waste so as to save a few bucks, by all means. Long as we do not invalidate the facts that one choses to ignore (because they can and good for them) or bypass/sidestep/risk.

A long run to get back into the topic and my original reply, which is “bearing/rotor type don’t matter”.
Only implementation does.

One user here says sleeves can potentially affect longevity. Seen another user often mention ball bearings and their noise levels.
Whereas there are fans out there today with sleeve bearing that will run fine for over a decade and are known for it, or ball bearing fans renowned for their noise-to-performance ratio.
Implementation.

1 Like

OK, I stopped reading mid post…
I am tired of this so I will just say it like that - I am not arguing about quality. I am not saying go buy Aerocool fans or whatever.
I will stop posting, cause I thing we are not understanding each other…

That’s where he should have remained IMO. All the “reviews” he’s doing in the last what… 2? 3? years bring nothing new to the table and I never felt like he knows his stuff. It’s the same garble that every outlet does just in less detail.
I enjoyed watching his watercooling stuff but he just kinda gave up on that.


Anyway though to he topic…
“Fluid bearing” is also a very broad description since every manufacturer kinda does their own thing and call it differently. In the past when I was researching fans for a bit before a purchase I tried looking up the few different names various manufacturers use and all it turned up was the respective manufacturer’s website. There’s not really any one “standard” fluid bearing.
What I’m getting at is that IMO you can’t judge by the type of bearing, but only specific manufacturer’s implementations.

Essentially I would just look at how it performs (noise vs. temperature mostly) and go from there. Because realistically how important is longevity in a fan really? Unless it’s a complete crapshoot it will probably last you 2 or 3 PCs minimum, and personally that’s not worth my time doing research.
Another issue with that is also that you can only go by historic data from previous models from the same manufacturer, but even that doesn’t mean that the new model is gonna last just as long. And it’s just hard to impossible to find historical failure rates of fans.

1 Like

I will always watch jayz over GN. I have the capability of obtaining tech data myself if I wanted to listen to a dry long haired dude all day he would be in my sub list. hell no

This is accurate and it has to be expensive. Its not your typical magnet. It has to deal with the heat of the fan hub which demagnetizes normal materials

2 Likes

It seems like I have started a more heated discussion here than I ever expected, I wasn’t even sure if I would get an answer but all these various opinions are appreciated :slightly_smiling_face:

So, you assume that fans last anyway between 6-9 years and thus it does not matter to begin with?