We made this video about how ball bearings (and grooved bearings) are made! th-cam.com/video/HUtEzjrF-To/w-d-xo.html Watch our video about beat frequency noises, also featuring Jakob! th-cam.com/video/F7ia_FZcthQ/w-d-xo.html Watch our "Forbidden Interview" with Jakob of Noctua last year: th-cam.com/video/82LZkglNiQ0/w-d-xo.html Watch our video (also featuring Jakob) about Noctua's pressure scan results and new Noctua NH-D15 G2: th-cam.com/video/nDDxYlkp-_A/w-d-xo.html Find all of our engineering interviews here: th-cam.com/video/nDDxYlkp-_A/w-d-xo.html&list=PLsuVSmND84Qsv6Q_9GERaAKQ_FsQkOQ7H
it shows that the people in charge actually know about what's going on instead on relying on fancy presentation. This is why Intel's Battlemage has me excited despite their upcoming specs.
Agreed. Imagine a world where sales reps actually understand their products and the engineering behind it. 😮 So sick of suits, pr sales men, models and company executives that don't actually know $#!+ about their products. Well, the models are okay. The point is we need more dudes like this representing their brands.
this guy can sell me Noctua products better than any marketing even without telling me to buy something :D I hope we will see him again! Thank you I like this conversations a lot
Marketers aren't (all) stupid, and they know this, so they will send our real engineers to market to consumers like you (and me; I got got by the Nvidia cooling engineer into desiring a 4090)
@@Ladioz What i bather from this video is that every bearing has it's use case. Brushless may be loud but would that even matter in an industrial setting where the noise floor is much higher anyway? Also things to consider are fine particles in the air, temperature, whether the fan is placed near a radiator or is it in "free air" etc.
An interesting distinction between "fluid' and "hydro", is that gases are also fluids. Air is a fluid. Which can lead the discussion to "air bearings". Hydro on the other hand is specifically liquids. I used to work on Navy gyro compasses and stabilisers back in the 80's and some of those units used what they were then referring to as "air bearings". They were literally "lubricated" with nothing but air, in a similar manner to how an old school spinning hard disk head flies over a disk platter.
@@fatavocado7291 "Hydro" is the Greek root for "water", it's where "hydrogen" gets its name from, not the other way around. But, for instance, "hydraulics" also uses the root "hydro" and it refers to anything using liquids to transfer motion, including oils. So, in engineering, there's a long tradition of using "hydro" to refer to liquids. In physics, "fluid" refers to both gases and liquids.
i mean engineering is basically applied sciences, so it better be applied to life. It's one of the key functions of engineering. (Science is there to learn and understand how something works and what it is, engineering is solving problems by utilizing what science has discovered)
Honestly, companies need to get more of their nerds in front of a camera. The PR-laden C-suite eggheads never give up any info id actually find interesting, but this fella here, id have a beer with.
The problem is finding the nerds who can and want to do it, you do keep them away for their actual job. The eggheads have nothing better and do so do it with pleasure.
you guys are demonizing "c suite" suits but yall dont realize a silent majority of the c suites well know their strengths and empower their nerds. you all are middle management at best. simply don't know.
One of the very old "choice" of bearing type (or so the media kept telling us) was orientation. I would have loved to have this discussed (vertical versus horizontal with hub at top, fan at bottom - Top rad placement). Also, of all the Noctua fans I ever owned, not one failed, so that covers statements which you touched upon (building reputation). I've had various other vendors which failed spectacularly, one of which was one of those early 2000's massive case fans where the fan completely detached from the hub.
This is the best type of video: Technical information from an expert without marketing BS. Since Noctua is so meticulous in their approach and approaches advertising as informing the customer, they'll always be my first choice when it comes to cooling solutions.
Great discussion. As a machinist for many, many years, I've always found bearings incredibly interesting. There are so many types, it would shock most people who don't know any better. Glad to see this talked about.
Engineers going full nerd on bearings in computer fans is just sexi AF. Anybody with me on that? Like seriously, engineers are what make the great things around us so great. All these considerations that even the archetypal PC enthusiast doesn't thing about... love it! The marketing types try to simplify, which is perhaps a noble and difficult effort, but it shows that front-cover product specs are SSSOOO much more than just surface deep. Great video as always! Thanks Steve, and thank you GN team.
as an old computer technician I have a few preferences when it comes to bearings in fans. Magnetic levitation is the king of bearings. With no mechanical contact at all at full speed these will live a long time. Dual ball bearings is the next. Noisy buggers but with a long life. FDB bearings are a step above simple sleeve bearings. Now I have to say I have worked the most with serve grade hardware, and things like bearing noise is not something you think of. And fan speeds are high. I damaged my hearing just by the sound of fans in servers, so I'm not joking about not caring about noise. About ball bearings. Some manufacturers fake the bearings a bit by using one ball bearing and one sleeve bearing. Those fans will have a lifetime that's limited by the sleeve bearing. Once that is worn out it's game over and you have to replace the fan. So make sure the fan has dual ball bearings, that is two ball bearings, not two balls in the bearing... Magnetic levitation bearings are the least common, but theoretically they should not wear out. Now the truth is that it still happens. Usually it happens if the fan get real dirty over a long times use. They tend to loose the balance and then bad shit happens to the surfaces in the bearings. As they aren't supposed to have to rub against each other they tend to wear quickly when that happens. At one time four 120mm fans blew up in a server. If I remember correctly they were made by Nidec. Usually these were real good, but it seems an entire batch of their fans had a manufacturing fault. So remember that fans are mechanical products and some times they will break. Keeping them clean is a good way to increase the chance that the fans will keep working for a long time. FDB fans is a thing I haven't worked a lot with. They simply doesn't exists in the server world. Not the perfect truth, but close to it. And sleeve bearings you just won't see in a server, ever. Something interesting is that FDB fans used to have a lower time to failure rating than ball bearings. But that has changed over the years and now fans with fluid dynamic bearings tend to top most if not all ball bearings fans. Still fans with FDB are not something you see in servers. The first fans with magnetic levitation bearings I encountered I think was manufactured by Sunon. I remember they had a demonstration video to inform their customers about these bearings. From what I remember they had a LED that was powered by a lead that wiped the shaft in the fan, the power was supplied to the metallic bore of the bearing. When the fan was still the shaft would touch the bore and the LED would light up. But as soon as the fan started moving it would ride the magnetic fields and levitate in the middle of the bore and there were no mechanical contact between the shaft and the bore and the LED would go dark. I just tried to find that video and it seems Sunon has taken it down, so no video to show...
Of course, another consideration is the number of start/stop cycles. Servers tend to be 24/7, so those aren't as important there. However, that's not the case for all PCs. Especially fans with a zero RPM mode or laptop fans. Like was mentioned at 16:15, FDBs can use higher viscosity oils compared to non grooved sleeve bearings, but that increases the time to no contact. Meaning there's more wear during stop/starts. I have no clue where magnetic bearings sit on that metric. Similarly, things that aren't an issue for servers are angular loads. The fan acts as a gyroscope, so turning it while spinning means sleeved bearings and magnetic bearings will wear faster. You might wonder why someone would ever do that, but moving a laptop that's running is common. Same with the Steam Deck.
I put a Sunon Maglev as a PCF on my 3D printer toolhead. That failed hilariously, it was making all sorts of wild noises. Only double ball bearing in high vibration use or when you're flinging the fan about.
Maglev's have one big downside, they need minimum voltage to levitate the fan. In some fans it's 12v so you cant just control the fan with voltage, at low voltages the bearing starts to wear, PWM control is required so if user does not know this, then that user would eventually damage the fan by running it at lower voltage using voltage control rather than nessessary PWM control.
@@SianaGearz Maglev's require minimum voltage to not wear prematurely due minimum voltage requirements (check manufacturers manual for this information) for the magnetic levitation to happen. PWM control is required, fan gets 12v all the time and fan controls its speed with pulse width modulation signal from fan controller, example mainboard. I had voltage controlled sunon maglev fail too for print cooling, another sunon maglev that was set 100% (24v all the time, for hotend cooling) is still running 5 years later with no change of noise, also its running pretty much 24/7 regardless if printer is printing due wear caused by turning fan off and on.
@@jarnom85 PWM control also doesn't work on the Sunon i got :D I actually built a Tau filter to convert the PWM into a voltage, since i don't have voltage control natively :D At low PWM frequency the fan didn't run due PWM interfering with AutoRestart feature, and at increased PWM frequencies there was intermodulation noise, it was hilariously bad. Both intermodulation noise and occasional noise from the bearing getting whacked out of alignment it was quite a symphony! But it's all been rebuilt and replaced.
Companies that truly understand an engineering focus know hat there's a real benefit to letting engineers speak directly to technical audiences. As others have said, even though neither of the two videos with him have been sales focused videos they have done more to convince me of continuing to buy Noctua products than any sales video would.
I love it when the fans blades and the bearings' pole is accessible by being removable and i can lubricate it and the holding hole with mechanical oil and mix it up with some lubricant grease at the same time, it usually makes a fan go even faster and makes it so much more silent, i have done this to countless fan for 22 years now it's such a neat trick since it works so well. I have many, many, many old fans which still work perfectly because of that for more than a decade, 1 specifically the older i have, i got it back in 2006 (14cm fan) and it still works perfectly because i maintain it like that.
Jakob, danke Dir für ein weiteres unglaublich interessantes Interview. 🙏 Vollgepackt mit Details, über die wir Käufer ehrlich gesagt zu wenig nachdenken bzw. kaum berücksichtigen 👍
This is exactly what I wanted to know no BS, This is great marketing in one way what sets Noctua apart from others. Also the knowledge of what terms and tech behind what makes the fans work. Love it, thank you!
@@zickzack3106 I think it was meant as a joke due to the "obviously", because while fluid is indeed a loanword from Latin (don't quote me on that, maybe it found it's way into English via some other Roman language), in this situation it would've been more obvious that fluid is just the English word, as that's the language currently spoken in the situation. (He did not say fluidum, I'm confident about that.) Of course I could be wrong, but in the end it doesn't matter too much, in my opinion. I got a little chuckle out of it!
Great video. My take from this is to buy a fan that the manufacturer says is good for your use case and trust that will have used the appropriate bearing type.
This type of content is amazing, and while it drives my wife from the room with a glassy stare, I could watch this 24/7... Thanks Steve and Team GN, and thanks Jakob!!!
I have all noctura fans in my old main PC. It and the fans have been running great for 13 years. The fans are still very quiet. Great company. I am an engineer and l think this guy is wonderful!
Oh and I totally second everyone appreciating the engineering discussion videos. Please keep them coming! Moving the discussion forward in the enthusiast (and general consumer) space by eons each video.
Thanks for the video Steve. You get similar discussions in turbocharging although I would call the overwhelming majority fluid dynamic bearings of the sleeve type, main difference being lubrication is supplied under pressure and is total loss via gravity.
I have several fans that have been in use for over 2 decades and they still do fine. But I seem to be one of those rare people that actually fully cleans and does maintenance on his systems every 2-3 months. For most people, including me, the bearings going bad isn't the limiter to a fans life. A fan taking a chunk of your finger and loosing a piece or a whole blade in the process seems to be the primary reason for their life to end. I still have and use a Antec Ninehundred. They were famous for having a 200mm top fan. I've had to replace that top fan 4 times over the years because it lost blades on the regular, twice all on its own and twice from some finger biting action. I replaced it with a Noctua model and haven't had a single blade break since (have had even bigger chunks of meat being extracted from my fingers though.)
I've built a lot of computers and tried many different fans and I really appreciate having a brand I can depend on. Noctua fans are always reliable in my experience. Keep up the good work Noctua, and keep up the good work GN. It's awesome to have actual INFORMATION in videos these days. 😅
I greatly appreciate the information that Jakob from Noctua has been providing. And agree that there are scenarios which require special consideration when it comes to either a CPU cooling solution; or a general purpose fan to help move air through the system case. Both of these, are areas where quality of the product, and the reputation of the manufacturer are important factors.
Here we are again. I remember that article you had written all those years ago on your blog. Looking forward to you discussing how bearings have changed since
What a satisfying video for a Mechanical Engineer. In a few words Jacob said different problems demand different bearings, one type of bearing cannot work for all circumstances. Excellent in depth video 🫡🤝🏻👏🏻. Keep it up Steve !
What a great timing. We just learned about bearings and how to select them considering the application environment this semester, just a couple of months ago! Great work as always GN team.
I really love these series of engineering side of thing approach, even if not many people watch it please keep it coming I’ll keep buying merch just to see more of these.😊
I really enjoyed this video, just like so many of your other videos, great all around questions to help the consumers be in the know. So we can make educated purchases in the future on subject matter that some may not have understood.
I wish I could have heard about orientation as well. Which way is worse for the fan and if perhaps a certain bearing is better for certain orientations.
Thanks for these GN. So rarely do we get direct discussion like this with people who actually know what they are talking about and don't have to check with marketing before saying anything.
I love the way Jakob talks through the history of grooved FDBs and mentions that there's more to consider than "FDB best", the first 5 minutes are already a treat
My NFP12 redux in push pull are the best fans I've ever used. I will continue to purchase them in the future. I know they may not be the "BEST" performer, but acoustically they are superior in every way and perform most frequently in the top half of best performers even with acoustical restraints.
Gotta say, when I had Enermax fans I did love being able to un-click the fan blades and rinse them under the tap for easy cleaning. That fan did only last 8 years though, so I do get the point.
I really appreciate companies where marketing sells what engineers developed, compared to ones where engineers have to come up with something that somehow, technically, matches what marketing wants to sell. Unfortunately it is hard to distinguish those from an outsider's perspective. Giving access for interviews such as this is definitely a strong positive indication though!
Funny enough for enthusiast this the best way to offer marketing, I'm a lot more keen buying and recommending noctua hearing their engineers present company approach in detail instead of common nonsensical marketing like it's 30% better than the last one! and it has spinny zoom™ technology that means nothing.
I like sleeve bearing because I *_know_* it's self-repairable. The fan on my MSI 2070 super was going out. Weird noises, hard start to spin, start-stop. Took the plastic cover off the heatsink. Dusted the heat sink. Took the fan apart. Put some grease _(I'm amateur mechanic, had Lucas - Red&Tacky laying around)_ on the fan shaft and in the sleeve. Better than new, all symptoms gone.
Great content and I really like Noctua products but to be honest Noctua needs to step up significantly. The competition is either cheaper, better, or worst case both. There's a reason for why companies like Artic have see so much recent success.
I would love to see fans that use magnetic levitation discussed, I have used some super cheap $20 120mm mag-lev fans from ebay as replacements in a UPS and they have been running constantly for almost a decade.
This brings memories back from the time custom PC case scene was on top of its activity. Good memories :) Can't say tho fans improved so much over previous decade and a half. Definitely did but I would say from this perspective back then we expected it more from future development.
Jakob rocks. I can't tell you how many junk or no-brand fans I've seen shatter or scream bloody murder after short runtime. Money well spent on a company that puts performance and research first. Yeah their fans are expensive, but there's definitely a level of insured trust at work here when you see how just one person obsesses over bearing specs. Well done everyone. Oh and Steve I bet your 200mm exploding fan was that massive Antec...900 or something I believe it was? I had the pre usb3.0 version of that bad boy and got lucky by having the dipswitch set to low.
what a time to be alive absolutely here for the deep-dive fan content on that note and completely unselfishly, any chance you could investigate the newer AP184 Silverstones which apparently have loud clicking noises due to the PWM signal? Asking for a friend who's looking to switch the old fans in his FT-02, and 100% not for me
It's fun to learn more about fan bearings, when one of my gpu fans died, other rattles at lower speeds (because it doesn't detect it's own rpms and uses rpms of first one) and jury rigged a case fan to gpu for some cooling.
I wish you'd asked about fan orientation relative to gravity, and how that influences bearing choice. From my own experience, fans seem to fare best when oriented so that an axial fan will push air sideways, so that the long-axis of the impeller axle is horizontal. Now, this is a small sample size as I've had less than a handful (5) fans actually fail on me over my time building PC's for myself and my family (a thing I am thankful for), and it seems the smaller the fan, the quicker it fails in any case. However, I noticed another trend. Most of the fans on my stuff that have failed had their impeller axles long axis oriented vertically. I think this might have been causing the oil to drain out prematurely because brushless DC fans used to be built frequently with the data sticker being the thing that keeps the oil inside the fan bearing, meaning that if the fan is oriented so that gravity tries to pull the oil out of the bearing but the sticker is the only thing holding it in, that sticker will eventually fail and let the oil out, followed soon by a rather noisy fan bearing failure (lots and lots of squeaking). Since my current PC uses nothing but Noctua fans, I am happy to see that the SSO2 bearing (a type of FDB) they use is far better engineered than using something so flimsy as a mere sticker to hold the lubricant inside the bearing housing. Knowing what I know now, I can only describe the bearings in those older fans as what I would currently call "Not up to my standards for the task in any way". Every single one of those fans that failed, I took apart. Every single one that failed and was disassembled, had in it a bearing that was nothing more than a sintered bronze "Oilite" sleeve bearing with the only consideration for longevity being an extra drop or two of lubricant for the sintered bronze to wick up by capillary action and convey to the surfaces of the bearing that are in relative motion when the fan is operating, with that lubricant being held inside the fan as a whole by ONLY the label on the back-center of the fan's housing (as mentioned earlier). Thankfully I know better now, and if I can avoid it I will not be purchasing another fan with such a crudely executed Oilite sleeve bearing.
I hoped that you'd dig a bit more into the dust topic. How do they keep the dust away during normal operation? It's typical that cleaning dust from the fans manages to bring some dust in the bearing mechanism and cause laud scratching and beating noise. The seal in the diagram cannot be too tight or it would start to act as bearing itself. Also, it would be nice to have a fan where the part with the blades could be detached, without disassembling the bearing itself.
The thing i want from feature fans is to either sonwhow get rid of the usual 4 hub holder(lines) or actualy double them up so they can become thinner and then at the same time reverse the fan blade so they are easier to clean, is you do this whe holder (line) can be so thin the suction eeffect may so insignificant both performance and acoustically-wise, or do the craziest thing and add 4 small motors in each of the 4 sides/corners of the fan to spin the blades (which will have a flat-like cicle cap around then like some new Arctic ines do), that way you can lock the fan blades in such way you don need a center motor anymore so you can reduce the cycle-like center of the fan by a lot significantly amd it so much easier to clean, what i am not sure about tis design is the noise and possible RPM target, i think there was a company man many years ago who made such types of fans now that i remmber it but they didn't get enough attention for some reason.
Good video, but it's a little too technical for me. But at least now I know the bearing isn't everything. Even FDB or HDB has way more components like viscosity of the fluid. Keep up the good work, but hopefully we can get an explanation that gears toward general public.
Its telling we can get these discussions with Jakob and by the quiet confident way he talks about the topic and the depth of knowledge shown I naturally now prefer noctua from a fan buying standpoint based on perceived quality without him having to shill for a second. It’s the same with the AMD CPU guys and Tom with ARC GPUs. Engineers who love the area they work in and care about their products can influence me all day. Big company marketers! Watch and learn how to sell your product! It’s not slap AI on it or more RGB.
3:24 Adding grooves to a barrel is called rifling and it helps projectiles maintain their intended flight course. I would imagine that grooves in a bearing operate in a similar way
Kind of confirms the "vibes" I've got when doing shopping and research -- "FDB" has no clear differentiating factor, is in the cheap and expensive fans alike, and isn't specific enough (if they mean anything particular by it in marketing) to make a *good* purchase decision based on. Although it is vaguely "better" for the typical use-case of a personal-use computer for most likely being somewhat quieter than ball bearing. I appreciate reviews based on noise and thermals... On the other hand, lifespan is very hard for a reviewer to meaningfully benchmark. So, really I think it comes down to some sort of judgment call, and since Noctua has staked its identity in the market on being a premium option for more exacting build quality and engineering standards, and a serious dedication to longevity (both of support and product lifetime)... it's something I would personally trust Noctua to deliver on in any product they put out. I also trust them more than other companies to make the right choice for a given product segment even if it goes away from the current "meta" or marketing hype trend -- if engineering shows it to be the right choice for the intended use-case and manufacturing it that way is feasible, I believe they will tend to do it. I wish we could easily know as buyers how long a fan will last, and how well its low-noise operation will hold up over time. Until we get a meaningful way to convey the quality of the manufacturing and engineering standards of the product, brand reputation really is the next best thing, and sort of the best we have, in my opinion. I can see why Jakob would point to that, there's not much more consumers can really go by. That, and if an unusual proportion of reviews report failures then the product is probably not that good, but people are reviewing after a couple of days usually, maybe a few weeks or months, very rarely a few years. Almost never decades. And by then it's too late, the product is usually discontinued... Oh, well.
I've STUDIED the bearings inside the submarines. It's crazy fascinating. I've always wondered HTF they could stop water from coming in with millions of PSI coming in from outside in deep dive?! Turns out they ADD FLUID INSIDE the sub to equalize press so there's NO CHANCE for sea water to come in in the "bearing" compartment. Nuts.
Yeah, I used to work on Navy subs back in the 80's (civilian dockyard technician, not enlisted). One of the most interesting things I saw and worked on, was how they provide a compass repeater outside the sub on the conning tower. That compass repeater needs a metal housing, with a glass window on top so that it can be read when surfaced, yet also needs to be able to survive the enormous pressures of the deep, without that flat glass window getting broken under the crushing pressure. Same deal as you describe. The whole thing is filled with a clear mineral oil and there is a metal bellows on the bottom on the unit, which allows the volume to be compressed in case there is anything inside which is compressible, such as some unwanted air bubbles. Automatically and passively, the oil inside very closely matches the pressure outside, thanks to the oil and bellows. The incompressible oil almost perfectly fills the inside and the bellows caters for any pressure equalisation needed due to any area the oil could not fill.
I'm sorry, I am not sure what was discussed and what to make of it. I was looking forward to understanding what advantages some types of bearings have over the other, but it seems like we just learned that just "because something something doesn't mean something", which doesn't explain anything. "Just because it says FDB, doesn't mean it's good." or "just because it has herringbone groves, doesn't mean much at all." I will check out the rest of the videos on fan bearings, but this is the first video that popped up about fan bearings and the least I wanted to know was what "bad" things to watch out for when buying fans. There is a lot of interesting info tho, maybe the subject is more complicated than I thought. Thanks for bringing us cool videos like this!
Its hard to say, because the actual manufacturing and quality of each bearing matters more than the type of it. Lets say bearing A was better than bearing B. The differences between those would be small and a good B could be a lot better than an average A at the end of it. The only thing to keep in mind is that anything that uses lots of oil will usualy die first, since the oil leaks out over time in a vertical or top side up mounting.
I've always been curious about the bearing differences in Silverstone's AP181 and AP183 180mm fans which have MTBF ratings of 15,000 and 70,000 hours respectively. Silverstone liked to include the AP181's with their cases, and sure enough around that 2 year mark (15,000 hours = ~1.7 years), those fans would start rattling and making horrid noises and I'd have to replace them with AP183's. I think that MTBF spec is good to reference, but the problem is transparency. I had to dig to find that info, and I imagine I wouldn't find it at all for a lot of fans out there.
![[BAEMON NEWS 7] NEW ALBUM COMING IN](http://i.ytimg.com/vi/wtVeAPimOyo/mqdefault.jpg)
We made this video about how ball bearings (and grooved bearings) are made! th-cam.com/video/HUtEzjrF-To/w-d-xo.html
Watch our video about beat frequency noises, also featuring Jakob! th-cam.com/video/F7ia_FZcthQ/w-d-xo.html
Watch our "Forbidden Interview" with Jakob of Noctua last year: th-cam.com/video/82LZkglNiQ0/w-d-xo.html
Watch our video (also featuring Jakob) about Noctua's pressure scan results and new Noctua NH-D15 G2: th-cam.com/video/nDDxYlkp-_A/w-d-xo.html
Find all of our engineering interviews here: th-cam.com/video/nDDxYlkp-_A/w-d-xo.html&list=PLsuVSmND84Qsv6Q_9GERaAKQ_FsQkOQ7H
why aren't mag lev bearings more common, and why is Corsair's implementation (e.g. ML120) so loud in comparison to Noctua's SSO2 FDB?
The bot comments are endless ):
Really interesting videos. I know you (Steve) don't get much appreciation but you always do great work on the scripting, writing and everything else.
That is one impressive list of unique videos. Most of which is my only source of this kind of info.
And there are fan manufacturers that use a simple bushing ("rifle bearing") in lieu of a proper sleeve bearing.
It's so refreshing when info isn't filtered through marketing. Love this format!
it shows that the people in charge actually know about what's going on instead on relying on fancy presentation. This is why Intel's Battlemage has me excited despite their upcoming specs.
Open communication should be so much more respected.
@@revolverswitchOh boy, I had just recently changed my GPU.
Will definitely get their Battlemage or the Celestial? in the future
I like this guy. The "no bullshit marketing" approach combined with the technical explanations he gives really fit the GN vibe.
The intel graphics guy is great as well.
The guy's knowledge is so impressive.
Great hair thinks alike
That's why I buy my products from this company either though they are cheaper products out there that are comparable.
Agreed. Imagine a world where sales reps actually understand their products and the engineering behind it. 😮
So sick of suits, pr sales men, models and company executives that don't actually know $#!+ about their products.
Well, the models are okay. The point is we need more dudes like this representing their brands.
this guy can sell me Noctua products better than any marketing even without telling me to buy something :D I hope we will see him again! Thank you I like this conversations a lot
Marketers aren't (all) stupid, and they know this, so they will send our real engineers to market to consumers like you (and me; I got got by the Nvidia cooling engineer into desiring a 4090)
They know their target demographic. They're a great part of the industry, despite how expenisve they are.
I just bought 3 for my "sleeper" PC, well it was my first PC that has been rebuilt like 4 times since the late 90s lol.
Well you just need to look at warranty. That tells you everything.
I love the technical videos with noctua guy, pls keep it going
Not enough vocally emulated fan noises
Agreed, definitely one my favorite recurring guests now 😎🦉
I always found fluid vs hydro dynamic being a strange way of wording things.
I've been enjoying the noctua videos with Jakob.
The worst bearing is probably Brushless bearings. They are too loud
I thought that "hydro" one are water based (from the hydrogen) while fluid be oil based. And I was wrong
@@Ladioz What i bather from this video is that every bearing has it's use case. Brushless may be loud but would that even matter in an industrial setting where the noise floor is much higher anyway? Also things to consider are fine particles in the air, temperature, whether the fan is placed near a radiator or is it in "free air" etc.
An interesting distinction between "fluid' and "hydro", is that gases are also fluids. Air is a fluid. Which can lead the discussion to "air bearings".
Hydro on the other hand is specifically liquids.
I used to work on Navy gyro compasses and stabilisers back in the 80's and some of those units used what they were then referring to as "air bearings". They were literally "lubricated" with nothing but air, in a similar manner to how an old school spinning hard disk head flies over a disk platter.
@@fatavocado7291 "Hydro" is the Greek root for "water", it's where "hydrogen" gets its name from, not the other way around. But, for instance, "hydraulics" also uses the root "hydro" and it refers to anything using liquids to transfer motion, including oils. So, in engineering, there's a long tradition of using "hydro" to refer to liquids. In physics, "fluid" refers to both gases and liquids.
As a mechanical engineering student it's actually great to see the stuff we learn in lectures be applied in real life! Keep up the great work
i mean engineering is basically applied sciences, so it better be applied to life. It's one of the key functions of engineering. (Science is there to learn and understand how something works and what it is, engineering is solving problems by utilizing what science has discovered)
no way, you must have had a good teacher
Honestly, companies need to get more of their nerds in front of a camera. The PR-laden C-suite eggheads never give up any info id actually find interesting, but this fella here, id have a beer with.
Companies only listen to half of what the nerd has to say. The other half is too expensive.
The money types know jack shit about any topic apart from quarterly earnings reports which makes (99.9% of) ads completely fucking useless
The problem is finding the nerds who can and want to do it, you do keep them away for their actual job. The eggheads have nothing better and do so do it with pleasure.
Most companies would damage their brand with this level of transparency
you guys are demonizing "c suite" suits but yall dont realize a silent majority of the c suites well know their strengths and empower their nerds. you all are middle management at best. simply don't know.
One of the very old "choice" of bearing type (or so the media kept telling us) was orientation. I would have loved to have this discussed (vertical versus horizontal with hub at top, fan at bottom - Top rad placement). Also, of all the Noctua fans I ever owned, not one failed, so that covers statements which you touched upon (building reputation). I've had various other vendors which failed spectacularly, one of which was one of those early 2000's massive case fans where the fan completely detached from the hub.
This is the best type of video: Technical information from an expert without marketing BS. Since Noctua is so meticulous in their approach and approaches advertising as informing the customer, they'll always be my first choice when it comes to cooling solutions.
Hearing Jakob's enthusiasm for fans is amazing.
You could say he's a huge fan of fans.
Great discussion. As a machinist for many, many years, I've always found bearings incredibly interesting. There are so many types, it would shock most people who don't know any better. Glad to see this talked about.
Engineers going full nerd on bearings in computer fans is just sexi AF. Anybody with me on that? Like seriously, engineers are what make the great things around us so great. All these considerations that even the archetypal PC enthusiast doesn't thing about... love it! The marketing types try to simplify, which is perhaps a noble and difficult effort, but it shows that front-cover product specs are SSSOOO much more than just surface deep.
Great video as always! Thanks Steve, and thank you GN team.
as an old computer technician I have a few preferences when it comes to bearings in fans. Magnetic levitation is the king of bearings. With no mechanical contact at all at full speed these will live a long time. Dual ball bearings is the next. Noisy buggers but with a long life. FDB bearings are a step above simple sleeve bearings. Now I have to say I have worked the most with serve grade hardware, and things like bearing noise is not something you think of. And fan speeds are high. I damaged my hearing just by the sound of fans in servers, so I'm not joking about not caring about noise.
About ball bearings. Some manufacturers fake the bearings a bit by using one ball bearing and one sleeve bearing. Those fans will have a lifetime that's limited by the sleeve bearing. Once that is worn out it's game over and you have to replace the fan. So make sure the fan has dual ball bearings, that is two ball bearings, not two balls in the bearing...
Magnetic levitation bearings are the least common, but theoretically they should not wear out. Now the truth is that it still happens. Usually it happens if the fan get real dirty over a long times use. They tend to loose the balance and then bad shit happens to the surfaces in the bearings. As they aren't supposed to have to rub against each other they tend to wear quickly when that happens.
At one time four 120mm fans blew up in a server. If I remember correctly they were made by Nidec. Usually these were real good, but it seems an entire batch of their fans had a manufacturing fault. So remember that fans are mechanical products and some times they will break. Keeping them clean is a good way to increase the chance that the fans will keep working for a long time.
FDB fans is a thing I haven't worked a lot with. They simply doesn't exists in the server world. Not the perfect truth, but close to it. And sleeve bearings you just won't see in a server, ever. Something interesting is that FDB fans used to have a lower time to failure rating than ball bearings. But that has changed over the years and now fans with fluid dynamic bearings tend to top most if not all ball bearings fans. Still fans with FDB are not something you see in servers.
The first fans with magnetic levitation bearings I encountered I think was manufactured by Sunon. I remember they had a demonstration video to inform their customers about these bearings. From what I remember they had a LED that was powered by a lead that wiped the shaft in the fan, the power was supplied to the metallic bore of the bearing. When the fan was still the shaft would touch the bore and the LED would light up. But as soon as the fan started moving it would ride the magnetic fields and levitate in the middle of the bore and there were no mechanical contact between the shaft and the bore and the LED would go dark. I just tried to find that video and it seems Sunon has taken it down, so no video to show...
Of course, another consideration is the number of start/stop cycles. Servers tend to be 24/7, so those aren't as important there. However, that's not the case for all PCs. Especially fans with a zero RPM mode or laptop fans. Like was mentioned at 16:15, FDBs can use higher viscosity oils compared to non grooved sleeve bearings, but that increases the time to no contact. Meaning there's more wear during stop/starts. I have no clue where magnetic bearings sit on that metric.
Similarly, things that aren't an issue for servers are angular loads. The fan acts as a gyroscope, so turning it while spinning means sleeved bearings and magnetic bearings will wear faster. You might wonder why someone would ever do that, but moving a laptop that's running is common. Same with the Steam Deck.
I put a Sunon Maglev as a PCF on my 3D printer toolhead. That failed hilariously, it was making all sorts of wild noises. Only double ball bearing in high vibration use or when you're flinging the fan about.
Maglev's have one big downside, they need minimum voltage to levitate the fan. In some fans it's 12v so you cant just control the fan with voltage, at low voltages the bearing starts to wear, PWM control is required so if user does not know this, then that user would eventually damage the fan by running it at lower voltage using voltage control rather than nessessary PWM control.
@@SianaGearz Maglev's require minimum voltage to not wear prematurely due minimum voltage requirements (check manufacturers manual for this information) for the magnetic levitation to happen. PWM control is required, fan gets 12v all the time and fan controls its speed with pulse width modulation signal from fan controller, example mainboard. I had voltage controlled sunon maglev fail too for print cooling, another sunon maglev that was set 100% (24v all the time, for hotend cooling) is still running 5 years later with no change of noise, also its running pretty much 24/7 regardless if printer is printing due wear caused by turning fan off and on.
@@jarnom85 PWM control also doesn't work on the Sunon i got :D I actually built a Tau filter to convert the PWM into a voltage, since i don't have voltage control natively :D At low PWM frequency the fan didn't run due PWM interfering with AutoRestart feature, and at increased PWM frequencies there was intermodulation noise, it was hilariously bad. Both intermodulation noise and occasional noise from the bearing getting whacked out of alignment it was quite a symphony!
But it's all been rebuilt and replaced.
Companies that truly understand an engineering focus know hat there's a real benefit to letting engineers speak directly to technical audiences. As others have said, even though neither of the two videos with him have been sales focused videos they have done more to convince me of continuing to buy Noctua products than any sales video would.
I love it when the fans blades and the bearings' pole is accessible by being removable and i can lubricate it and the holding hole with mechanical oil and mix it up with some lubricant grease at the same time, it usually makes a fan go even faster and makes it so much more silent, i have done this to countless fan for 22 years now it's such a neat trick since it works so well.
I have many, many, many old fans which still work perfectly because of that for more than a decade, 1 specifically the older i have, i got it back in 2006 (14cm fan) and it still works perfectly because i maintain it like that.
Jakob, danke Dir für ein weiteres unglaublich interessantes Interview. 🙏
Vollgepackt mit Details, über die wir Käufer ehrlich gesagt zu wenig nachdenken bzw. kaum berücksichtigen 👍
Love that Noctua's marketing is their engineering.
It's always great listening to people that really know what they are talking about.
This is exactly what I wanted to know no BS, This is great marketing in one way what sets Noctua apart from others. Also the knowledge of what terms and tech behind what makes the fans work. Love it, thank you!
I love how the hydro being greek for fluid, and fluid obviously being latin for fluid joke just passed by completely without notice. (2:27)
Is that a joke? Hydro is greek, fluidum is latin and fluid is english - which is based on latin.
@@zickzack3106 I think it was meant as a joke due to the "obviously", because while fluid is indeed a loanword from Latin (don't quote me on that, maybe it found it's way into English via some other Roman language), in this situation it would've been more obvious that fluid is just the English word, as that's the language currently spoken in the situation. (He did not say fluidum, I'm confident about that.)
Of course I could be wrong, but in the end it doesn't matter too much, in my opinion. I got a little chuckle out of it!
@@timgrelka Sure, why not
Great video. My take from this is to buy a fan that the manufacturer says is good for your use case and trust that will have used the appropriate bearing type.
This type of content is amazing, and while it drives my wife from the room with a glassy stare, I could watch this 24/7... Thanks Steve and Team GN, and thanks Jakob!!!
A wife repellant?!
To all the fans that lost their bearings, I'm glad you can finally differentiate your sleeves.
I have all noctura fans in my old main PC. It and the fans have been running great for 13 years. The fans are still very quiet. Great company. I am an engineer and l think this guy is wonderful!
Oh and I totally second everyone appreciating the engineering discussion videos. Please keep them coming! Moving the discussion forward in the enthusiast (and general consumer) space by eons each video.
Thank you for making these types of videos... bearable.
Thanks for putting an amusing spin on things.
Thanks for the video Steve. You get similar discussions in turbocharging although I would call the overwhelming majority fluid dynamic bearings of the sleeve type, main difference being lubrication is supplied under pressure and is total loss via gravity.
I have several fans that have been in use for over 2 decades and they still do fine.
But I seem to be one of those rare people that actually fully cleans and does maintenance on his systems every 2-3 months.
For most people, including me, the bearings going bad isn't the limiter to a fans life.
A fan taking a chunk of your finger and loosing a piece or a whole blade in the process seems to be the primary reason for their life to end.
I still have and use a Antec Ninehundred.
They were famous for having a 200mm top fan.
I've had to replace that top fan 4 times over the years because it lost blades on the regular, twice all on its own and twice from some finger biting action.
I replaced it with a Noctua model and haven't had a single blade break since (have had even bigger chunks of meat being extracted from my fingers though.)
I've built a lot of computers and tried many different fans and I really appreciate having a brand I can depend on. Noctua fans are always reliable in my experience. Keep up the good work Noctua, and keep up the good work GN. It's awesome to have actual INFORMATION in videos these days. 😅
I love it. A video on a subject too boring for my personal interest. Yet, something meaningful and in depth. This is why I love this channel.
I think Steve fell in love.
I don't blame him
And what is your favourite shampoo? Separate conditioner or combined?
@@xoxo2008oxox i like head and shoulder BARE.
I greatly appreciate the information that Jakob from Noctua has been providing. And agree that there are scenarios which require special consideration when it comes to either a CPU cooling solution; or a general purpose fan to help move air through the system case. Both of these, are areas where quality of the product, and the reputation of the manufacturer are important factors.
These are the type of interviews i love. Have the tech people explain how their products in general or specifcly works.
I really enjoy this video format especially Jakob is really good at explaining I would love more of those videos please
That's a LOT of talk about lubricated balls and shafts. Very impressed with your composure Steve
Here we are again. I remember that article you had written all those years ago on your blog. Looking forward to you discussing how bearings have changed since
What a lovely interview. I love this kind of real-talk. Thank you for doing this. More engineer talks please!
What a satisfying video for a Mechanical Engineer. In a few words Jacob said different problems demand different bearings, one type of bearing cannot work for all circumstances. Excellent in depth video 🫡🤝🏻👏🏻. Keep it up Steve !
Super interesting!
It's great every year we have this conversation with Noctua.
this guy is the truth. i coud listen to Jakob talk all day.
Jacob is a true engineer, through and through !!!!
Love all these technical deep dives!
Veey informative. Next time I shop for fans I will be sure to ask about the porosity of the shaft and the viscosity of the fluids.
What a great timing. We just learned about bearings and how to select them considering the application environment this semester, just a couple of months ago!
Great work as always GN team.
Would love to see a long term fan review by running fans for years and then see how they compare. Current, air flow, noise etc
This was a great piece, and great discussion. Do more of these please.
This was great, thank you. I like to watch you and Jacob talk about the technical side of fans.
Absolutely loving these videos with Noctua
It's nice to get real information rather than just marketing. Thanks!
I love theese videos! Very interesting, no marketing bs but actual information and explanation.
I really appreciate these technical discussion.
I really love these series of engineering side of thing approach, even if not many people watch it please keep it coming I’ll keep buying merch just to see more of these.😊
This guy knows more than is normal for any human being about computer fans. I love him.
Interesting and important content. Thanks for sharing. I think a lot of us would appriciate a better curated and consolidated version.
I really enjoyed this video, just like so many of your other videos, great all around questions to help the consumers be in the know. So we can make educated purchases in the future on subject matter that some may not have understood.
Wow he is very passionate about fans, a pleasure to watch.
I wish I could have heard about orientation as well. Which way is worse for the fan and if perhaps a certain bearing is better for certain orientations.
Don't lose your bearings!
:)
Noctua is awesome.
And so is this video - interesting and educational.
Thanks for these GN. So rarely do we get direct discussion like this with people who actually know what they are talking about and don't have to check with marketing before saying anything.
Very interesting conversation, proving once again that all of engineering is about trade-offs.
I love the way Jakob talks through the history of grooved FDBs and mentions that there's more to consider than "FDB best", the first 5 minutes are already a treat
My NFP12 redux in push pull are the best fans I've ever used. I will continue to purchase them in the future. I know they may not be the "BEST" performer, but acoustically they are superior in every way and perform most frequently in the top half of best performers even with acoustical restraints.
Gotta say, when I had Enermax fans I did love being able to un-click the fan blades and rinse them under the tap for easy cleaning. That fan did only last 8 years though, so I do get the point.
Matsushita = Panasonic aka panaflo
Refreshing to listen to competent voices in a world of incompetent noise. Thank you!
I really appreciate companies where marketing sells what engineers developed, compared to ones where engineers have to come up with something that somehow, technically, matches what marketing wants to sell. Unfortunately it is hard to distinguish those from an outsider's perspective. Giving access for interviews such as this is definitely a strong positive indication though!
Finally someone will explain it haha.
But he didn't 😂
Love hearing an engineer talk
So informative! Can we take a poll on a new acronym to replace FDB? I vote for "interrupted groove sleeve bearing" or IGS bearing.
Funny enough for enthusiast this the best way to offer marketing, I'm a lot more keen buying and recommending noctua hearing their engineers present company approach in detail
instead of common nonsensical marketing like it's 30% better than the last one! and it has spinny zoom™ technology that means nothing.
I like sleeve bearing because I *_know_* it's self-repairable.
The fan on my MSI 2070 super was going out. Weird noises, hard start to spin, start-stop.
Took the plastic cover off the heatsink. Dusted the heat sink. Took the fan apart. Put some grease _(I'm amateur mechanic, had Lucas - Red&Tacky laying around)_ on the fan shaft and in the sleeve.
Better than new, all symptoms gone.
Great content and I really like Noctua products but to be honest Noctua needs to step up significantly. The competition is either cheaper, better, or worst case both. There's a reason for why companies like Artic have see so much recent success.
Engineers should be the marketing dept. No BS, just honest products they are proud of and stake their reputations on.
I would love to see fans that use magnetic levitation discussed, I have used some super cheap $20 120mm mag-lev fans from ebay as replacements in a UPS and they have been running constantly for almost a decade.
This brings memories back from the time custom PC case scene was on top of its activity. Good memories :) Can't say tho fans improved so much over previous decade and a half. Definitely did but I would say from this perspective back then we expected it more from future development.
I would watch a 3 hour documentary with these guys!
Speaking of fans, what happened to fan testing at GN?
Jakob rocks. I can't tell you how many junk or no-brand fans I've seen shatter or scream bloody murder after short runtime. Money well spent on a company that puts performance and research first. Yeah their fans are expensive, but there's definitely a level of insured trust at work here when you see how just one person obsesses over bearing specs. Well done everyone. Oh and Steve I bet your 200mm exploding fan was that massive Antec...900 or something I believe it was? I had the pre usb3.0 version of that bad boy and got lucky by having the dipswitch set to low.
Thanks Jakob Dellinger. This and the last video both interesting and informative because of you.
This was very informative, i love these deep dives. Fan bearing marketing has been a pain point for me for like, a decade now lol
I love this guy. He knows his stuff.
what a time to be alive
absolutely here for the deep-dive fan content
on that note and completely unselfishly, any chance you could investigate the newer AP184 Silverstones which apparently have loud clicking noises due to the PWM signal? Asking for a friend who's looking to switch the old fans in his FT-02, and 100% not for me
It's fun to learn more about fan bearings, when one of my gpu fans died, other rattles at lower speeds (because it doesn't detect it's own rpms and uses rpms of first one) and jury rigged a case fan to gpu for some cooling.
I wish you'd asked about fan orientation relative to gravity, and how that influences bearing choice.
From my own experience, fans seem to fare best when oriented so that an axial fan will push air sideways, so that the long-axis of the impeller axle is horizontal.
Now, this is a small sample size as I've had less than a handful (5) fans actually fail on me over my time building PC's for myself and my family (a thing I am thankful for), and it seems the smaller the fan, the quicker it fails in any case.
However, I noticed another trend. Most of the fans on my stuff that have failed had their impeller axles long axis oriented vertically. I think this might have been causing the oil to drain out prematurely because brushless DC fans used to be built frequently with the data sticker being the thing that keeps the oil inside the fan bearing, meaning that if the fan is oriented so that gravity tries to pull the oil out of the bearing but the sticker is the only thing holding it in, that sticker will eventually fail and let the oil out, followed soon by a rather noisy fan bearing failure (lots and lots of squeaking).
Since my current PC uses nothing but Noctua fans, I am happy to see that the SSO2 bearing (a type of FDB) they use is far better engineered than using something so flimsy as a mere sticker to hold the lubricant inside the bearing housing.
Knowing what I know now, I can only describe the bearings in those older fans as what I would currently call "Not up to my standards for the task in any way". Every single one of those fans that failed, I took apart. Every single one that failed and was disassembled, had in it a bearing that was nothing more than a sintered bronze "Oilite" sleeve bearing with the only consideration for longevity being an extra drop or two of lubricant for the sintered bronze to wick up by capillary action and convey to the surfaces of the bearing that are in relative motion when the fan is operating, with that lubricant being held inside the fan as a whole by ONLY the label on the back-center of the fan's housing (as mentioned earlier).
Thankfully I know better now, and if I can avoid it I will not be purchasing another fan with such a crudely executed Oilite sleeve bearing.
I hoped that you'd dig a bit more into the dust topic. How do they keep the dust away during normal operation?
It's typical that cleaning dust from the fans manages to bring some dust in the bearing mechanism and cause laud scratching and beating noise.
The seal in the diagram cannot be too tight or it would start to act as bearing itself.
Also, it would be nice to have a fan where the part with the blades could be detached, without disassembling the bearing itself.
The thing i want from feature fans is to either sonwhow get rid of the usual 4 hub holder(lines) or actualy double them up so they can become thinner and then at the same time reverse the fan blade so they are easier to clean, is you do this whe holder (line) can be so thin the suction eeffect may so insignificant both performance and acoustically-wise, or do the craziest thing and add 4 small motors in each of the 4 sides/corners of the fan to spin the blades (which will have a flat-like cicle cap around then like some new Arctic ines do), that way you can lock the fan blades in such way you don need a center motor anymore so you can reduce the cycle-like center of the fan by a lot significantly amd it so much easier to clean, what i am not sure about tis design is the noise and possible RPM target, i think there was a company man many years ago who made such types of fans now that i remmber it but they didn't get enough attention for some reason.
This video is educational, but Jacob sold me. I will certainly only buy Noctua for me next build.
YES! Painstakingly detailed FAN CONTENT! I have been waiting for this.
A refreshing honest engineer, great video 👍
Good video, but it's a little too technical for me. But at least now I know the bearing isn't everything. Even FDB or HDB has way more components like viscosity of the fluid. Keep up the good work, but hopefully we can get an explanation that gears toward general public.
Its telling we can get these discussions with Jakob and by the quiet confident way he talks about the topic and the depth of knowledge shown I naturally now prefer noctua from a fan buying standpoint based on perceived quality without him having to shill for a second. It’s the same with the AMD CPU guys and Tom with ARC GPUs. Engineers who love the area they work in and care about their products can influence me all day.
Big company marketers! Watch and learn how to sell your product! It’s not slap AI on it or more RGB.
3:24 Adding grooves to a barrel is called rifling and it helps projectiles maintain their intended flight course. I would imagine that grooves in a bearing operate in a similar way
Kind of confirms the "vibes" I've got when doing shopping and research -- "FDB" has no clear differentiating factor, is in the cheap and expensive fans alike, and isn't specific enough (if they mean anything particular by it in marketing) to make a *good* purchase decision based on. Although it is vaguely "better" for the typical use-case of a personal-use computer for most likely being somewhat quieter than ball bearing.
I appreciate reviews based on noise and thermals... On the other hand, lifespan is very hard for a reviewer to meaningfully benchmark. So, really I think it comes down to some sort of judgment call, and since Noctua has staked its identity in the market on being a premium option for more exacting build quality and engineering standards, and a serious dedication to longevity (both of support and product lifetime)... it's something I would personally trust Noctua to deliver on in any product they put out. I also trust them more than other companies to make the right choice for a given product segment even if it goes away from the current "meta" or marketing hype trend -- if engineering shows it to be the right choice for the intended use-case and manufacturing it that way is feasible, I believe they will tend to do it.
I wish we could easily know as buyers how long a fan will last, and how well its low-noise operation will hold up over time. Until we get a meaningful way to convey the quality of the manufacturing and engineering standards of the product, brand reputation really is the next best thing, and sort of the best we have, in my opinion. I can see why Jakob would point to that, there's not much more consumers can really go by. That, and if an unusual proportion of reviews report failures then the product is probably not that good, but people are reviewing after a couple of days usually, maybe a few weeks or months, very rarely a few years. Almost never decades. And by then it's too late, the product is usually discontinued... Oh, well.
Just watched 2 old guys argue, now I'm here for this.
I've STUDIED the bearings inside the submarines. It's crazy fascinating. I've always wondered HTF they could stop water from coming in with millions of PSI coming in from outside in deep dive?! Turns out they ADD FLUID INSIDE the sub to equalize press so there's NO CHANCE for sea water to come in in the "bearing" compartment. Nuts.
Yeah, I used to work on Navy subs back in the 80's (civilian dockyard technician, not enlisted).
One of the most interesting things I saw and worked on, was how they provide a compass repeater outside the sub on the conning tower. That compass repeater needs a metal housing, with a glass window on top so that it can be read when surfaced, yet also needs to be able to survive the enormous pressures of the deep, without that flat glass window getting broken under the crushing pressure.
Same deal as you describe. The whole thing is filled with a clear mineral oil and there is a metal bellows on the bottom on the unit, which allows the volume to be compressed in case there is anything inside which is compressible, such as some unwanted air bubbles.
Automatically and passively, the oil inside very closely matches the pressure outside, thanks to the oil and bellows. The incompressible oil almost perfectly fills the inside and the bellows caters for any pressure equalisation needed due to any area the oil could not fill.
If this dude ever leaves Noctua, ya'll better hire him.
I remember Fluid Dynamic Bearings being a big selling point in the early 2000 on harddrives, but i cant remember the time fans also shifted to that.
I'm sorry, I am not sure what was discussed and what to make of it. I was looking forward to understanding what advantages some types of bearings have over the other, but it seems like we just learned that just "because something something doesn't mean something", which doesn't explain anything. "Just because it says FDB, doesn't mean it's good." or "just because it has herringbone groves, doesn't mean much at all." I will check out the rest of the videos on fan bearings, but this is the first video that popped up about fan bearings and the least I wanted to know was what "bad" things to watch out for when buying fans. There is a lot of interesting info tho, maybe the subject is more complicated than I thought. Thanks for bringing us cool videos like this!
Its hard to say, because the actual manufacturing and quality of each bearing matters more than the type of it.
Lets say bearing A was better than bearing B.
The differences between those would be small and a good B could be a lot better than an average A at the end of it.
The only thing to keep in mind is that anything that uses lots of oil will usualy die first, since the oil leaks out over time in a vertical or top side up mounting.
I've always been curious about the bearing differences in Silverstone's AP181 and AP183 180mm fans which have MTBF ratings of 15,000 and 70,000 hours respectively. Silverstone liked to include the AP181's with their cases, and sure enough around that 2 year mark (15,000 hours = ~1.7 years), those fans would start rattling and making horrid noises and I'd have to replace them with AP183's.
I think that MTBF spec is good to reference, but the problem is transparency. I had to dig to find that info, and I imagine I wouldn't find it at all for a lot of fans out there.