Noctua Never Had a Chance | Fan Showdown S2E1 

😫nooo you are limiting fan speed! My design im working on was dependent on low mass so it would run at higher speed. If you slow it down it probably wont work.😔

Do Both: 1) give results at 1200rpm 2) give results for full speed Why Both: 1) Some people purposely run fans low speed. 2) Some people run fans full speed or close.

your videos are WAYYYY too long for the amount of actual content. You drag things out and go on and on over the same thing again and again then ad unnecessary pauses, Jeez man get to the point, I'm tired of pressing fast forward.

Should retest the board limited at 1200rpm too

@ Sounds like a personal issue. My personal issue is RU-vidrs talking too fast. At least this guy is easygoing enough that I don't need to rewind 3 dozen times to absorb the info. Either way, the issues are ours, don't try to nudge creators into becoming something they aren't.

this really needs to get to his attention

THIS

69 likes

@@bluesap7318 nice!

Great point, also want to add that 120mm fans in a PC typically spin at 1500 - 2000 rpm and testing should include typical conditions and not just low noise conditions.

But at the same time people tend to keep RPM as low as possible to keep fans quiet.

@@oLaudix when gaming it doesnt really matter anyways. For example for like overclockers, they wouldnt mind as much as long as they get the best cooling possible for their computers.

@@oLaudix If that’s the point it should be limiting the noise level and not the rpm.

All those factors usually determine the RPM, thus a fixed rpm would allow those factors to be accentuated instead of balancing out. A fixed RPM would actually encourage the unique characteristics in each design rather than dampen everything

Back in the Athlon XP days I had a CPU cooler with an 80mm fan @ 6800 RPM. Think Ric Flair WOOOOOOOOOO!

You did great my dude, keep it up!

Always learning my G, mad respect

Starting strong! Good work!

wow first design I wish I had that kind of talent, I am way too anal retentive and have to measure everything from pr-existing fans, even stuff that will not make a real difference... My first design was using master control from hack smith and it was a simple cube like so many others... But instead of having the x, y, and z printed on the faces it had viking ruins...

Personally your already better then me. I got a 3D printer but my internet sucks so bad it doesn’t work just logging into the server to link the devices. So it is still in the box sadly. Props to you for doing what I can’t.

Exactly

10 db and 20 db are gonna be impossible i think atleast, this is because if you increase the db by 10 you double the sound output whether you are talking 50-60 or 20-30 So 40 db is actually 8x louder then 10 db

@Yousef Mansour its mainly the pitch that matters though, you dont wanna sit next to a pc that whines like a server rack all the time do you?

@@iliandocx8392 Yeah lol I wrote this out as just a way to get people's attention towards noise normalised thermals. I feel it would've had less impact to say "Start measuring at each fans lowest rpm level and take a new temp reading at every single dB increase".

@@iliandocx8392 I don't see any reason why frequency couldn't be measured alongside loudness.

Not efficiency exactly, but you're right that they'll have access to more power.

I'm guessing his point is its easier and better to design a fan for a given rpm (and then in real life production adjust the motor power to match), then it is to design a fan when you have no idea how fast it'll spin

@@____5837 Motors power output is based off torque and rpm. By limiting the rpm he's basically making the fan that can draw the most current the highest powered fan. . A fan that will draw more current won't be able to hit higher rpm so it evens out when rpm is unlimited except by the motors power output (ability to maximize work)... but if he sets all the fans to the same rpm then he's not inputting the same amount of work into the fans anymore. It was plainly evident when the fan that has dominated everything all of a sudden sucked (or lack thereof). . It's fine to test them all at different fan speeds as well as full voltage, but that's not what he's doing.

The thing is, noctua when designing a fan don't first have a motor and then see how fast the fan can spin. They decide they need a new 2000rpm fan, then they design the blades. Then they put it on a variable motor to see how much power is required to spin it at 2000rpm, then they design or buy in motors which provide the correct amount of power. But anyway I personally think he should do neither. I think he should run all fans at whatever speed produces 40db as most people care about performance vs noise when buying a fan (hence no one runs at full speed all the time)

@@____5837 but I would rather buy a fan with the most cooling per decibel at ~30db, since that's where they will be most of the time. Hell for case fans, straight up measure CFM... but apparently we aren't designing case fans, sad face.

No competition is fair. Like if you put two people in harmwresling match with 200pounds of difference... Or arm length or size of arm muscle... Nothing can be perfectly fair

I guess I had the same dumb opinion aswell then.

Might have to bolt it to the table at that point

100% agree! My idea im working on is dependent on being low mass and spinning at a higher speed to work.

We all have the same dumb opinion I guess

If you limit anything it should be watts

Nice! 👍😂

No they needa make it sound lile a ww2 stuka dive bomber

What a legendary idea

Def putting that in the bosses PC.

Ever since watching the first episode of season one that has been on my mind, and I finally got a 3d printer so I can design a fan that reliably whistles. I might be able to have it ready for the next episode.

i"d say a third factor, which obviously is related to those two, is power draw

@@illdeletethismusic fans draw so little power it doesn't really matter

what about static pressure, cause arent noctua fans designed for high static pressure as well as low noise.

@@CommanderChronicles Static pressure would be an important third factor if you wanted to get a full image of the fans performance across all types of scenarios, but that would be very hard to measure meaningfully without lab equipment, so I don't think it is realistic. In these tests the static resistance is fixed (ie. the aircooler-tower). That seems reasonable since aiflow-optimized case-fans are rarely the ones that you need to puah hard. Radiator and cooler fans have much higher demands on them, so that's usually what's more interesting. Noctua produces fans optimized for flow, pressure and ones that balance both aspects. So do lots of other companies. You can squeeze out a good deal more performance if you design it to be specialized in one or the other. Pressure fans tend to have more blades with sharper angles of attack - compared to airflow fans (ie. case-fans). The fan he uses to test here is kind of an all-rounder actually , but it does perform more like a pressure-fan than a flow-fan for sure.

But with a twist. I'll prefer 25dbA fan which gives me 80С (say, 60C delta to ambient temp) to 50dbA fan which gives me 40C (20C delta) any day. I prefer totally silent PC (at night, with closed insulated glazing windows) which runs 80C full-load to noisy which is cooler.

I like that idea

I think he should sound normalize them! Performance at 45db for example. Then also max speed

Wattage would be a good measurement to have.

Agree 100%. Limit load, not RPM

@@EvanMJones Sound normalizing is the most logic thing, most people are not looking to save 1 watt on fanpower.

I think noise-normalized would be better; ie setting them all to something like 40dbA

A manufacturer could easily design a faster motor, so rpm normalizing makes it more about the pitch of the blades than the weight, which I think is a good thing. Noise-normalization is likely the most fair for determining the best overall fan design as it gives an advantage to silent designs.

I'd like to see max rpm and 1200 for cooling, show the efficiency and scaling of the designs with the speed change.

Or don't limit the rpm but measure the power draw

pretty sure a manufacturer can't just design a faster motor, it will just be faster with less resistance. They can only make a faster motor by making it have gears which isn't really an option for a case fan?

How does it limit creativity? Now you have a target to adhere to, instead of just hoping that the motor can spin your fan at anywhere near the designed speed. Now the emphasis is on creating blades that move air well and keep noise down. If anything it expands creativity as the goals are clearer.

@@robertfullard5646, but the goals have always been clear, get the best delta, best CFM, and lowest noise. Adding and RPM limit actually lowers the possibilities of design focus and removes specialization.

@@robertfullard5646 Limiting RPM gives no incentive to reduce noise. Noise may be reduced as a consequence of the lower RPM, but is not a design parameter since it isn't measured, unlike noise normalised testing

Hi All, So i have read through all the comments about RPM normalisation being a bad thing and responded to a few and it strikes me that there is a fundamental issue with this and that is what are we actually expecting from James, do we want super science or do we want Dr Evil fan fun? I believe there is a happy medium to this. If a fan were being designed by a manufacturer the design perameters would be stringent, cost of materials being a major factor. So no true fan designer would design a heavy fan as it would take way more material to make and reduce margin. Then you have noise, not just the noise level but the whole acoustic signature of the fan, is the noise pleasing or does it sound like a pig farting through a reed valve? Just measuring volume doesn't work as the noise characteristics are quite subjective. Most users 90% want quieter performance with still decent cooling potential. If you dont care about noise you go out and buy a delta fan. They work by beating the air into submission and i dont think that is the goal of this show. So then you have to look at what a fan design is truly about and that is performance at an RPM that most users will set to keep the noise level down but the air flow up. So now you have the limit it doesn't matter how much power the motor pulls to achieve it, what matters is a solid design that keeps material cost down and has a pleasing noise. If none of this is acceptable as the comments suggest then we need to solidly define what it is that we are expecting to be tested, but be aware James is not a scientific institute. He does not have an anechoic chamber to monitor an acoustic profile properly. IT IS MEANT TO BE FUN, with an element of competition.

@@robertfullard5646 I have literally never in my life cared how loud a computer fan is. I don't know anyone that cares either. Unless you're running your fans at +80% of the speed you will be fine.

Agreed. As a PC builder, I care about the performance to noise ratio, not the performance to rpm ratio.

both

If it was a conventional fan showdown, then yes. But clearly, as stated in the video, 13 dBA with a lot of high pitch noise will perceive differently then 13 dBA with less high pitch noise and more low pitch.

Yeah agree, or perhaps both. No harm in having more data points

@@themagiceye6723 It would be better to add that data point than to change the scoring structure drastically as it will be difficult to compare new results to old ones without running the tests over again.

Good idea man!

I concur noise normalized, every cooler tests should be done like this, otherwise, 90% of the time a test conclusion is "well it's cooler,but make more noise" , or the opposite (except with corsair, but because they don't know how to make a cooler)

This but I'd go for a much more silent level than 40db...that's a LOT

Honestly I'd say do that stuff but also rerun the other fans from this episode

Didn't expect to see barevids here!

Yeah, emphasizes the differences in cooling capability more

All funs should be run with the same power and not the same speed. That way a courser fun with more blades will be more efficient than a shallower one because at the same speed the power output will be different.

I'd like to pick a decibel limit and test them all there. 45db or something

Agreed. How about a sound limit?

@@bernhardjordan9200 Agreed

This!

+1 on this. Because for the most of us, the real tradoff in cooling solutions is noise vs cooling performance.

This indeed

Exactly. RPM is just a byproduct of either trying to hit noise targets or acceptable power draw for the motor

@@xeigen2 I was also thinking that power draw would be nice BUT if he wants to compare them to noctua that is silence oriented then noise normalized would probably be a better option.

best idea

Bonus points if they call it "The Stuka".

Just a fan motor powered air raid siren

Jericho Trumpet!

those old AVC CPU fans for lga775 boards were obnoxiously loud in any case

I would vote for cooling ability,, air movement against restrictions inside a case. Then again,,, My desktop case,, when it gets busy,,,, the noise is like living next to a runway. And not the kind with pretty women and funny clothes.

hay cj

Power consumption should be the limiting factor.

I like this idea because a lot of people like the idea of uncapping the rpm, but this allows some designs to have a chance. Might be a lot of testing though

Two sets of tests would be double the production time. I think just adding the "max RPM" information with the CFM and sound info would suffice. Keep the test and leaderboard fixed at 1200 but show the max RPM each fan can achieve

@@MauricioRPP1 if you limit the max in testing you're going to get results that are useful to no one. Genuinely nobody arbitrarily limits fan speed.

Absolutely! Now you have to optimize for 1200rpm, while before you could be more creative and design a fan that works best at 1800 rpm for example. Thus was going to suggest the same, while scrolling to the comments to see if someone else suggested it. Alternatively at least test at two different speeds - but actually like the limited and unlimited idea even more.

maybe normalizing for noise levels would be a decent compromise?

Only reason to reprint the boomer I can think of would be to see if the noise changed when the fan is better balanced. It wouldn't make it perform any better in the airflow test with the RPM limited the way he is testing now. Personally I think it was a bad choice to limit the RPM. After all there are really only two things most people care about. One is maximum performance and the other is low noise. If you measure performance at a limited RPM but noise at max RPM then you turn everything on it's head. If anything you might try measuring performance at noise normalized RPM, but I still think it's better to test at max RPM for both, and then possibly add at what RPM the fan hit a certain noise level. But that's just my personal thoughts and he can do it however he want's to. I'll still come here to see what interesting designs people has sent him.

@@blahorgaslisk7763 even if he don’t spin all the fans at their maximum, if he include the max rpm for each fan it will give an idea of how good they are.

I was thinking that too as soon as he said one layer thick...

Mistakes will be made - remember the mirrored one? He should just make a separate episode with all "failed" (mirrored, etc.) prints in the end of the season.

@@questcoast I would love too see that, great idea wish I had thought of it...

Came here to type that exact comment. Such a beautiful fan.

I would love having a 2 hour version of various fans just sucking smoke in :D

Imagine an rgb version of the wave, like the corsair LL fan style rgb

It reminds me of waffle cut fries.

It looks like a ripple when running.

Was about to post that as well. If He changes the conditions he should retest the previous top 3 or 5 fans to have comparable data with season 2.

somewhat agree, run them all at 100%, either they are good enough or they are not.....some are willing to ignore noise in the pursuit of cooling. others are not.

agreed. run the noctua at like 80% and find that db, then match the rest to it. That way you get a bit of headroom for those that can run quieter, but we expect nearly every submission will be probably using less than 80% power draw to reach the same noise level, but who knows, maybe they win anyway.

​@@brian770 it doesn't really matter if you care about the noise or not, because if it performs better when it's at the set db, if you crank up the noise it'll only perform even better

@@Pasquizle exactly, crank it up to max rpm and see what it can do with no restrictions. There is a big market for quiet fans, but also a big one for max proformance, for me, noise isnt a problem.... Thats what headsets are for, as long as its the coolest possible.

@@brian770 You're lowkey missing the point, because performance scales with noise so by having them all preform at say 30dB you can compare one fan to another and determine how it preforms in respect to another. While having them go full speed isn't as good for that. Like noise isn't an issue for me either but noise normalise helps figure out objective performance.

What about a different rpm next season for a new standard then?

Ideally he could start every fan off at 800, then add an additional 400 RPM at each step and seeing what temperature it levels off at. That's probably the best way to do seeing as he's now using an air cooler, as air coolers are far more reactive and won't take that much additional time.

Yeah I like the idea of doing a range of 3 rpms (low-med-high), then taking the best result for each fan.

A variable vane style fan blade would be the exception to this i believe.

Maybe test the fans at 1200 and at max and avg the 2 scores?

I think he should do that as part of the season finale for this season, but not before

I really think all fans should be tested again with all the rule changes till now,it would just be unfair not to

The one and only sensible compromise i have seen!!!

Better to do this but one at max speed/performance and the other at a good noise normalized level. Because those are the two important metrics.

yeah, the fan that noctua designed specifically for the its motor. thats why the industrial ones (3000rpm) cost double :) and you know what, at 1500 rpm makes the same noise as the 1200 cheap ones

Yes, the power required to cool without too much noise, that's what a fan is designed for.

Sounds like a lawnmower a few houses away

Lol A10 has that same appealing sound that’s why it’s been around so long.

I am fairly unsure if the noctua fan really lost. He only compared it in terms of thermals, not the _much_ more important noise category (or if you feel like your cpu's tdp is oddly high and noctuas airflow somehow isn't enough and it throttles, perhaps some noise-to-thermal ratio).

It doesn't show anything useful to limit RPM. All it shows is the designer was able to pick the best blade angle for 1200 RPM. Tells you nothing about how good the fan is for dB or max airflow/static pressure/cooling. By limiting to 1200 you're optimizing for the wrong variable. It disqualifies fans that work best at a different RPM. Maybe someone will make a 1400 RPM fan with a blade design that has better cooling and noise characteristics, if you turn that fan down to 1200 RPM you're artificially limiting it. Test should be 2 measurements; noise normalized and max speed

You have a point there!

Because he doesnt have a heating element. He has a processor. I am guessing that is why.

Because then he can’t compare the performance of the fans. Also they’re for mounting on a cpu radiator, so it makes sense to test cpu performance.

@@snortiblog No, he's testing them for heat rejection. Which is why he's only providing temperatures and not CPU scores. . Scientific testing requires controlling variables. Using a heating element instead of a processor will provide constant heat into the radiator block or air cooler and he can even provide the time it takes to heatsoak.

I hope he sees your comment, because his testing ideas are bad imo. Just re-run every blade you printed wrong like boomer. Why even bother to ask to us? You fucked the instructions up so its not even the original design because it is not meant to run with your fuckups.

I think that if a manufacturer were to create a fan they would pick the proper motor to make it spin at a proper rpm. Also, right now the designs seem to just be a linear trade off between weight and speed so limiting the speed would show a truely efficient fan regardless of weight.

@@Chevsilverado The problem is by normalizing the rpm you're also allowing heavier fans that would require too much power to spin fast enough to be spun at the same speed as the others. Law of conservation of energy, energy in = energy out. But you're allowing some fans to use more energy then others with this RPM normalization. It literally isn't fair and is not practical design. Just use a compressor or big enough scoops and call it a day with these rules.

Lmao😂

The blades have VTEC printed on them meaning this fan will spin much faster than the others

Yes, the current testing is just about actual cooling performance. What i care about is noise. If its slightly worse cooling but much lower noise, thats a big win for me.

I'd have called it "crinkle-cut", lol.

Money money money 💸💰

He should have called his channel "Only Fans"

true, but honestly, who cares about power draw on a computer fan? noise should be the limiting factor

@@bigbronx I agree, they should either be noise or current limited, not rpm.

They need to try it at least once

Maybe you should sub. It supports him and shows youtube that people like his content, so he'll get recommended more. :)

He would need an anechoic chamber to do noise testing and noise testing is massively subjective... It is not just volume but frequency. People are bitching about RPM, can you imagine if the best most efficient fan ever sounded like a pig farting but was quiet... Nobody would put it in their case.

Me too kid

Damn right.