Or even better, have not only the factory-molded A12X25 as a goal poast, but also a *PLA-printed* A12X25 version (maybe even a resin-printed version as well). That way you could compare how *any* fan would do if it was resin-printed (assuming that the layer lines cause similar losses among all fans).
I'm now a big fan of the small fanfan because the creator was able to think outside the box and put the fanfan inside the box. A truely creative way to reinvent the wheel while keeping the wheel exactly the same as before it was reinvented.
I bet everyone thought of it already, but I'm the only one who did it. But I bet some serious fan showdowners might have done it and didn't want to submit it because they wanna make "good fans" that are "actually productive". It also seems really low effort (and it kinda was). Just take the fan, scale it down to fit in the fan, scale up the blades to touch the walls, add the hub template, hollow it out, done. The better thing would have been for me to do some cylindrical scaling to make the fans stretch to accommodate the larger hub, but even noctua's mini fans have about the same blade proportions as their big fans.
Tom’s Dragon Wing fan is insane! Well done, mate!! We need to have a special fan episode where we take all the best parts of of the featured fans over the years, and combine them into an ultimate fankenfan! Imagine the Dragon Wing in a Cheater shroud for instance!
Wow, is this the first time the A12 got beaten? I know it's not really a fair fight since Noctua also optimizes for noise, but that's still impressive.
@@sirmonkey1985 just like that one from a year ago or so that was never confirmed if it was Thomas's design or not. I can't recall the brand that used it though. Almost want to sat thermaltake.
@@sirmonkey1985 At this point where the fan is shown on video pretty clearly, it would be easy for a good engineer to just screendump and reverse engineer the dimensions of the fan in cad.
The tesla Turbine is missing allot of its key components. like the external shroud that directs the air over the ''blades'' the entire thing is supposed to be enclosed with a valve on one end and it blasting air out of the other.
I think you all are missing one important thing here, IT'S NOT A FAN, it's designed to be pushed by pressure to generate movement, not to be rotated to move air.
@@YOEL_44 go read the patent... and whats its current only commercial application? as a pump. i guess the rotary engine aka wankel is designed to be pushed by pressure too? at which end? the rotor? or the shaft? felix originally designed it as, guess what? a pump... yknow, ALL engines are just glorified air compressors? and then theres the whole issue about pressure, velocity, and bernoullis... "new principle of operation". its not a pressure aka reaction turbine. its not an impulse turbine. its a fluid friction turbine. pressure doesnt push anything. it simply tries to reach equilibrium.
Well if it was made of the special material it might still be noisy, but since Noctua said they made the material (something crystalline they said) not to warp under temperatures or age, then the fan get closer to the frame have even more performance and probably even noisier. But you could also run the fan slower or not as often because of how performant it is.
@@pr0xZen His recordings though are not in a proper sound chamber, so they are higher than the true value. So it's only comparable to other fans on the list in that room. Your own personal room and setup will differ. The fan db in a proper sound chamber is only 22.5 db. That's also at full speed, and my fans never reach full speed, since it's not needed. So I never really hear the ten fans, even when gaming for hours. The pump makes more noise.
@@ScytheNoire If you never run fana at 100%, then comparing fans by manufacturer or lab rating performance metrics is pointless, and seriously misleading at best. At that point, the results and metrics of these "fsn showdown" tests are arguably more representative of deployment in relateable usecase than any of the 1st or typical independ lab party figures.
I like silent and quiet computers when they're in the same room that I'm in. Having some turbine whine clearly audible unless I'm ramping speakers and headphones up to ear-damaging levels is annoying. That said, if it was something in a rack in its own shed or basement or something, it can go right ahead and scream.
I've been a long time fan of the fan showdown. The thing with the a12x25 is its supposed to work in a lot of different configurations. Case fan, heatsink fan, radiator fan. It always does good. Is it the best for anything? not so much but its a great compromise between all kinds of fans. So just pure airflow is not a great way to compare fans. We need to take into account airspeed, CFM, static pressure, lifetime and efficiency.
*@Major Hardware* 7:44 I'm SO GLAD that you added that acrylic sheet to mount the fans on, otherwise this test would have looked like garbage, but now it's actually great :D
Not sure the cost of that dissolvable support filament, but there is a cheap and easy way to get easily removable supports for prints: Use a different filament type. If you are printing in PLA, use PETG supports, and if in PETG, use PLA supports. It makes them very easy to remove, and PLA and PETG are usually materials you already have around.
@@jttech44 Depending on where you live, PVA filament can cost a far chunk. Right now just looking at general lists, it costs about the same as a full 1kg spool of PLA for a 250g-500g spool of PVA. An while I don't expect people will use it exactly 1:1, that is still not 'cheap' in a price difference for something that only functions as a support material.
@@jttech44 Which is what the majority of people who 3D print as a hobby will mostly buy. Which is why PVA filament is considered expensive, as really the price per kg even for more expensive PLA is better than you can for PVA.
@@DuribunWolf Tbh, most people who are 3D printing don't have dual extruders so, PVA filament isn't something they'd buy. Once you've bought into dual extruder printing, you've figured out that cheap filament prints like crap, and nicer PLA+ makes alot of sense in terms of actual material needed and wasted material.At which point PVA isn't really cost prohibitive.
I would really love a sound based series where the dba and potentially a subjective "how nice to listen to" score are the focus. Could make use of the resin printer more and then rank all the fans based on fpm ratio Maybe normalizing with power or so the rpm of each fan is the same?
Honestly the hard angle on the Dragon Wing has a smooth but more steep curve to it than the original Noctua fan blades. But Noctua's new fan blades yet to be released on their updated 140mm design, also, has a steeper curve; it's more abrupt/jagged of a curve though.
Hey! I watched all of the seasons. It would be rather nice to add a noise normalize section in the chart, since many many of want our pc's to perform, but quiet! I think somewhere between 38-45 dba should be fine, I'll let you be the decider on that. I know this is a lot of work, if you should retest a lot of designs, but hey! Free video. Thanks for your hard work!
I'm sorry excuse me... A fan just DESTROYED the Noctua A12X25 and every other fan on the chart and you just like... End the video on that bombshell? Wow. Back to the studio. 😂
Noctua has a new fan on the way, and during their interview with GN about it and the R&D process that took a few years, they stated that their fan isn't optimized for pushing air through a radiator or for raw CFM. They strive for a sort of hybrid between the two styles. They want their fan to work really good as a case fan, and equally as good as an AIO fan. That being said, BY DESIGN, they are beatable with a more specialized design. It SHOULD be beaten in these tests. However, it SHOULD NOT be beaten across both open air and resistant (through radiator) CFM with a single design. I'd love to see how the Dragon Wing would fair against it for open CFM. It would be really cool to see Noctua beaten at their own thing
Finally someone adding parts that create turbolence at the end of the blade. That's a key to make air turbolent and easiert to push. I think even better would be with an intake that "prepares" the air for the blades. That is, creates turbolent air (with fins) and with curved static blades (like the fan) so that the fan can push as much as possible.
Can we see a fan speed vs sheer airflow graph for the A12x25 fan versus the Dragon Wing? It'd be interesting to see its possible use in variable fan speed situations!
For a solution that is cheaper than PVA, when the application allows. When printing with PLA use PETG for support material & vice versa. The 2 materials will not bond to eachother at all, so the supports just fall away.
i would be verry interested in seeing how the air moves trough a radiator/aircooler. That way you can see if it has some actual air pressure or takes the fastest way out and slows down. And you should scale down the Dragon Wing till it hits the same noise levels and then compare. Then you can see if it actually is better on simmilar noise levels and add the pressure test.
I like how harmonic the Dragon Wing was, legitimately 👀Like, I'd deal with the noise (loudness) if it sounded that good.~ I'm glad the fanfan at least did decently for a good joke fan. Little bugger made me happy.
The Dragon wing is kind of cool but sadly you won't know that once it's spinning. Also there are obviously more dimensions to fan design than cfm/rpm Even noise isn't just noise as there are types of noise that are perceived more quietly. A low frequency hum can be the same dB as a high-mid squeal but the hum will sound quieter. And if both are at low volume to begin with, the difference will be even greater. So it's a doubly good idea to try and get your fan to emit only low frequencies.
WOOT FAN SHOWDOWN! YEAH :D holy crap that bat wing looks awesome! LMAO that fan - fan is hilarious! Skyscoop was a good one! The tesla looks cool but not surprised it didn't do well. WOW someone needs to call Tom and give him some money for that design!
Love the fan show down! Ever thought of remaking the Noctua and using that as a reference? A storebought vs a 3d printed fan might give a significant difference in results. This might be a better reference point as all others are printed as well.
I like how the attack angle of the middle more aggressive than the tips, creating faster flow in the center making outersides not only sucking air from front but sides as well. ( all fans does this but I think in this model this is makes the fan act like a bigger diameter fan )
I feel like the fan fan would be really fun to use in a much larger fan. Like a 200mm or something, so that the hub isn't taking up half the blade area
Yeah The limit then becomes how big of a square can you fit in a circle (big fan hub), and then how big of a circle can you fit into that square (little fan hub).
Would be interesting to have some of the better performing and high selling fans included in the chart as well, just a few of them to see where they slot in among the printed fans.
Sleeper/conversion cars are the most high roller cars There's this one guy who sells ender 3 core xy conversion kits. That's probably the most baller 3d printer possible. "Yeah I have an ender 3. Yeah it does a 7 minute benchy."
The dragon wing outperforming the A12x25, even within it's form factor is a big thing even if it is louder. I'm curious, would a softer material like PETg decrease noise?
You'd probably want to go even stiffer actually. Some of the noise is from the blade tips vibrating, and a stiffer material would "flap" less. Also, softer materials will likely expand more under the centripetal force of high-RPM tests, meaning you have to design with less efficient larger gaps around the edges.
Love the fanfan and the tesla turbine! I think the tesla turbine might actually work, if you put a stationary restricting shroud over the intake. it sucks a little air in through its intake holes, and it projects it outward against the outer rim (pressure on the outer rim) but then it goes equally forward and backward. Just put a lip around the intake side, so the pressure on the outer rim can only go out the exhaust side. I don't have high hopes that it will perform well, but it'll certainly be more than zero!
I'd really like to see the fans set to a normalized 40db for the airflow test. Testing airflow at max rpm doesn't tell us much, since most people would be irritated by a nearly 60db fan.
Noctua is still unbeaten, imo. It's not enough to push more air through, if the acoustics get worse. When someone comes with a fan which is quieter or as quiet as Noctua and still within that 120x120x25mm envelope, I'm sold.
the fanfan reminded me of the dizzyfan. a friend of mine has this complex, where there are rows and rows of electrical cabinets. each one had a large external case cooling fan much like these except bigger. they run of 12V DC and turn on whenever the units reach a curtain temperature. some of them had failed however. i wonder if he learned his repairing skills from Mcgyver, because of all the Jerry rigging that was in there. for example, a random DC12V blower fan was thrown in to replace one of the original fans. on another unit, a cheap desk fan was just blowing in through the inlet of the old fan and the DC wires were just connected to a 12V lightbulb. last, but not least was the dizzyfan. he had taken a random mains powered motor, slapped it on a bracket in front of the inlet and attached the fan to it. he didn't even cut off the fan's power cable, witch was just whipping around and slapping anything in it's radius, making noise. and to control it, the 12VDC lines were just connected to a 12vdc trigger relay thingy. he had written "dizzyfan controller" on the little box the relay was in, hence the name. you could tell that it was on from the end of the corridor from the distinctive slapping sounds of the cable going round and round. it looked real goofy while it was on too LOL!
The thing with Fusion 360 is that it's cloud based (other than the fact I can't do anything with it). In order to export to STL, you gotta send it to the cloud, and they put you in a render queue. I first tried to center the fan hub template to the center of gravity, and it would just make these null points, and there was no way for me to translate that point to the center. If I just had a python terminal or something I could figure out how to translate it negative those coordinates. When you first import the stl, it lets you center it, but never again after that. Anyways, one render queue later, I'm in blender, and I just right click, align, center of mass. I've actually never needed to center something in blender, and it's been a good half year since I've used the program at all. Blender isn't known for their intuitiveness, but as a 3d modeling program it's pretty good. I even got an easy python script off of stack overflow to lay an object flat on the xy origin.
@@leifmessinger i prefer to store my data on my own drives. thats why i use gimp instead of photoshop (debating running my ancient copy of ps7 in a vm). id rather use old software anyway because not only am i used to it, but it was just better, and in many cases it still works, does the same job. anyway 3dsmax is good enough for me, and its something i already had in my tool box. ive done some game dev and know my way around various polygon formats and how to do transformations in code if need be. but i just export to stl and im good.
Now take the dragon wing, and smooth it out, see if smooth is better or worse. This seems to have a large outside and shrinking towards the center intake with a full flat output. and the be-quiet wavy design.
We need to revisit the Tesla turbine and add a front shroud with an inlet that restricts the airflow to those inlet vents, and redesign the outlet stator to better redirect the edge flow out the back. But the drive motor is the limiting factor. I've seen a Tesla turbine used as a pump heave gallons of water, but it was hooked to a big 1HP DC motor.
The tesla turbine was meant for installation inline and had to be enclosed in order to have its exit port functioning. The stack of its plates were exactly in the same position on top of each other but it also was designed to move a higher density steam. I have seen it used as an inline water pump and it was an excellent as the force of the water leaving it was so high. With the density of air being so much less, the plate stack might need to be at a slight offset from the plate below it forming a spiral pattern of the openings in the center when the plate stack is completed. At what offset to produce the best movement would be up for testing. If the inside of each plate had teeth, the plates could then slide onto a gear affixed to the fan motor shaft for the purpose of being able to change the offset distances until the best distance was achieved. With a gear attached to the fan's shaft, that would also assist the drag to create more spin but only in an enclosure to help create that centrifugal force. Without the enclosure for steam, water or air, the result would be exactly as we saw, exit happened between each plate and stayed close to the fan because the enclosure would assist the air speed build up that would be exiting the blades. This would create higher exiting thrust.
with how isolated my headset is and how noisy my AC is, I wouldn't even mind a blower cooler in my case just to cool everything/ so performance over noise for me is always a thing. if the price is right ofcourse(which is why my PC is full of arctic p12 fans)
Would be awesome to see the new first place in an extra high quality resin print...
printing in resin increases numbers by 2.1% you can easily math what the results will be ruclips.net/video/603hywgt92w/видео.html
Definitely
Indeed Resin please! :) make a 30mm deep version and then wow!
Exactly! We need resin prints
High quality resin print, and benchmarked against a resin A12x25
So....who's calling Tom to license his fan design?
MSI should get on that, their mascot is a freakin Dragon
@@Teajam TRUE!
msi hopefully
@@Teajam valid and true
Skip the middle man, make a company and call it "Tom's Fans" I'd buy.
The fanfan is such a riot. I haven't laughed that hard over a fan ever.
The fanfan has always been great. *ah, this is a different fanfan. This one is nifty for different reasons.
ty bb
So you are a fanfanfan?
I'm gonna see myself out
Lol the fact that it actually worked quite well is even funnier
And it _did_ move air, unlike some other designs we've seen.
Seeing if the Dragon Wing can be improved just by printing it in SLA would be cool.
printing in resin increases numbers by 2.1% you can easily math what the results will be ruclips.net/video/603hywgt92w/видео.html
Or even better, have not only the factory-molded A12X25 as a goal poast, but also a *PLA-printed* A12X25 version (maybe even a resin-printed version as well). That way you could compare how *any* fan would do if it was resin-printed (assuming that the layer lines cause similar losses among all fans).
Also the dBA would probably go down if it had bumps like the production quality ones.
Tom is a real one. You can’t expect to gain performance without trading off something else (such as noise) that man has the heart of a true engineer
the song that plays with the slowmo footage of the fans never gets old. what a banger
I headbang Every. Single. Time.
Slowmo?
Are we watching the same video?
@YOEL_44 Slowmo at 7:53
@@TranceFur That is not slow motion, that is just regular speed.
Have you guys never seen a fan spinning?
@@YOEL_44He's so fast it's slowmo for him
I'm now a big fan of the small fanfan because the creator was able to think outside the box and put the fanfan inside the box. A truely creative way to reinvent the wheel while keeping the wheel exactly the same as before it was reinvented.
I bet everyone thought of it already, but I'm the only one who did it.
But I bet some serious fan showdowners might have done it and didn't want to submit it because they wanna make "good fans" that are "actually productive".
It also seems really low effort (and it kinda was). Just take the fan, scale it down to fit in the fan, scale up the blades to touch the walls, add the hub template, hollow it out, done.
The better thing would have been for me to do some cylindrical scaling to make the fans stretch to accommodate the larger hub, but even noctua's mini fans have about the same blade proportions as their big fans.
Tom’s Dragon Wing fan is insane! Well done, mate!!
We need to have a special fan episode where we take all the best parts of of the featured fans over the years, and combine them into an ultimate fankenfan! Imagine the Dragon Wing in a Cheater shroud for instance!
I was interested to see how the fans would rank if we were to value FPM and dBA equally instead of only noting dBA and ranking by pure FPM. I realized then that decibels are logarithmic, so I had to convert those values to linear before we could get a proper weighting. Also, the columns won't line up here because we can't use a monospace font in comments. But you can paste it into a text editor if you want that. Anyway, here are those rankings!
| Fan | FPM | dBA | dBA linear | FPM/dBA linear |
| ---------------- | --- | ---- | ---------- | -------------- |
| Multi Stage | 268 | 40 | 101.59 | 2.64 |
| Turbo Pump | 405 | 43.6 | 153.99 | 2.63 |
| A12x25 | 495 | 46 | 203.19 | 2.44 |
| Angel | 485 | 47.3 | 236.11 | 2.05 |
| 92120 | 464 | 47.5 | 241.63 | 1.92 |
| Simple 13 | 479 | 48.1 | 258.97 | 1.85 |
| Gear Ratio | 482 | 48.3 | 265.03 | 1.82 |
| Variable | 430 | 47.4 | 238.86 | 1.80 |
| Hollowman | 435 | 48.5 | 271.22 | 1.60 |
| Radish | 459 | 49.5 | 304.44 | 1.51 |
| Warhead | 454 | 49.5 | 304.44 | 1.49 |
| Tangent | 459 | 49.6 | 307.97 | 1.49 |
| 17 Blade | 446 | 50 | 322.54 | 1.38 |
| Circumventor | 397 | 49 | 287.35 | 1.38 |
| Dragon Wing | 528 | 52 | 406.37 | 1.30 |
| Cheater | 477 | 51.5 | 383.57 | 1.24 |
| Waterfall | 423 | 50.9 | 357.88 | 1.18 |
| Double J | 397 | 50.4 | 337.79 | 1.18 |
| SkyScoop | 423 | 51 | 362.04 | 1.17 |
| Triskelion | 353 | 50.3 | 333.91 | 1.06 |
| Involute | 327 | 51 | 362.04 | 0.90 |
| Toroidal | 286 | 50 | 322.54 | 0.89 |
| XL | 286 | 50 | 322.54 | 0.89 |
| Uruguayan | 348 | 52 | 406.37 | 0.86 |
| 7 Blade Toroidal | 263 | 50 | 322.54 | 0.82 |
| 993 | 435 | 54.5 | 542.45 | 0.80 |
| Inverted Sharrow | 337 | 53.6 | 488.88 | 0.69 |
| Mobius REV | 273 | 52.6 | 435.54 | 0.63 |
| Thoroid | 255 | 52.8 | 445.72 | 0.57 |
| A12x25 fanfan | 263 | 54.5 | 542.45 | 0.48 |
| Tesla Turbine | 0 | 48 | 256.00 | 0.00 |
Values sourced from 9:34
That's some awesome work dude
Some crazy math dude it’s awesome, surprising what ranks above all
Hell yeah! Cool idea and baller execution to show the results instead of just wishing someone did it. Good job and thank you.
Best comment so far on the whole channel 👍 👍 👍
Wow, is this the first time the A12 got beaten? I know it's not really a fair fight since Noctua also optimizes for noise, but that's still impressive.
To be fair, it's Noctua's 140mm design
Now this is a proper fan showdown episode. In hub designs that attempt competition for the original plus a wild card for the lols.
if the guy who made the dragon wing sees this comment... pls, share the stls
Yeah the dragon wing looks awesome but could definitely see a company steal it.
@@sirmonkey1985 just like that one from a year ago or so that was never confirmed if it was Thomas's design or not. I can't recall the brand that used it though. Almost want to sat thermaltake.
@@EvilCerealBoX It was RaidMax, though ThermalTake is occasionally called ThermalSteal for a reason, just not this time.
@@sirmonkey1985 At this point where the fan is shown on video pretty clearly, it would be easy for a good engineer to just screendump and reverse engineer the dimensions of the fan in cad.
On Thingiverse as "smog"
The Dragon wing is impressive! Nice Job!
The tesla Turbine is missing allot of its key components. like the external shroud that directs the air over the ''blades'' the entire thing is supposed to be enclosed with a valve on one end and it blasting air out of the other.
Egor? Is that you?
Can we get another take with the missing components?
I would also like to see this fan attached to his variable voltage supply, crank those RPMs up to see if there is indeed increased efficiency.
I think you all are missing one important thing here, IT'S NOT A FAN, it's designed to be pushed by pressure to generate movement, not to be rotated to move air.
Not to mention the rpm needed to make it function even a little bit. I don't think the A12-X25 motor can hit 10k rpm.
@@YOEL_44 go read the patent...
and whats its current only commercial application?
as a pump.
i guess the rotary engine aka wankel is designed to be pushed by pressure too?
at which end? the rotor? or the shaft? felix originally designed it as, guess what? a pump...
yknow, ALL engines are just glorified air compressors?
and then theres the whole issue about pressure, velocity, and bernoullis...
"new principle of operation".
its not a pressure aka reaction turbine. its not an impulse turbine. its a fluid friction turbine.
pressure doesnt push anything. it simply tries to reach equilibrium.
6db louder though is pretty significant, but it might regain some of that if it was actually made out of Noctua's special material.
thats true! or mabye in resin? :))))
Well if it was made of the special material it might still be noisy, but since Noctua said they made the material (something crystalline they said) not to warp under temperatures or age, then the fan get closer to the frame have even more performance and probably even noisier. But you could also run the fan slower or not as often because of how performant it is.
The main issue with most of those printed designs is the turbulence generated by the surface not being smooth.
@@BrunodeSouzaLino resin printing or using isopropyl alch to smooth PLA.
The Noctua standard impeller would regain the CFM and more just by running it at slightly higher speed.
For me, the quietness of Noctua fans is a huge factor in why I use them, not just their performance and reliability.
and the warranty is very, very good if you ever need it.
Certainly - but 40+(++) dba isn't exactly "quiet" for higher end fans.
@@pr0xZen His recordings though are not in a proper sound chamber, so they are higher than the true value. So it's only comparable to other fans on the list in that room. Your own personal room and setup will differ. The fan db in a proper sound chamber is only 22.5 db.
That's also at full speed, and my fans never reach full speed, since it's not needed. So I never really hear the ten fans, even when gaming for hours. The pump makes more noise.
@@ScytheNoire If you never run fana at 100%, then comparing fans by manufacturer or lab rating performance metrics is pointless, and seriously misleading at best. At that point, the results and metrics of these "fsn showdown" tests are arguably more representative of deployment in relateable usecase than any of the 1st or typical independ lab party figures.
LETS GOOO MY FAV SHOW
I love the Dragon Wing's sound! It has a nice harmony to it. 52 dB isn't terrible either.
Tom is like "oh, you want something better, here it is; it took me like 15 minutes to design"
I will take increased performance over reduced noise most of the time. That dragon wing looks great.
Noise Performance is most important to me.
@@WeBeGood06but it is dragons wing!
When I'm gaming or listening to music I can't even hear my PC when it's at full throttle. I'll take increased performance every time.
shoot any day.
I like silent and quiet computers when they're in the same room that I'm in. Having some turbine whine clearly audible unless I'm ramping speakers and headphones up to ear-damaging levels is annoying.
That said, if it was something in a rack in its own shed or basement or something, it can go right ahead and scream.
Tom should get a deal with a fan company for their design. That's awesome!
I've been a long time fan of the fan showdown. The thing with the a12x25 is its supposed to work in a lot of different configurations. Case fan, heatsink fan, radiator fan. It always does good. Is it the best for anything? not so much but its a great compromise between all kinds of fans. So just pure airflow is not a great way to compare fans. We need to take into account airspeed, CFM, static pressure, lifetime and efficiency.
With the fan fan initially it was "neat, someone's done a full fan and frame...oh...ooooooh"
Nonononono
Wait wait wait wait wait
Tom is the hero we need, not the one we deserve.
Keep the fan showdown within the restraints of being used in an actual PC!
*@Major Hardware*
7:44 I'm SO GLAD that you added that acrylic sheet to mount the fans on, otherwise this test would have looked like garbage, but now it's actually great :D
Not sure the cost of that dissolvable support filament, but there is a cheap and easy way to get easily removable supports for prints: Use a different filament type.
If you are printing in PLA, use PETG supports, and if in PETG, use PLA supports. It makes them very easy to remove, and PLA and PETG are usually materials you already have around.
PVA filament is very very cheap.
@@jttech44 Depending on where you live, PVA filament can cost a far chunk. Right now just looking at general lists, it costs about the same as a full 1kg spool of PLA for a 250g-500g spool of PVA. An while I don't expect people will use it exactly 1:1, that is still not 'cheap' in a price difference for something that only functions as a support material.
@@DuribunWolf sure, if you use the cheapest possible PLA, which is a pain to print with.
@@jttech44 Which is what the majority of people who 3D print as a hobby will mostly buy. Which is why PVA filament is considered expensive, as really the price per kg even for more expensive PLA is better than you can for PVA.
@@DuribunWolf Tbh, most people who are 3D printing don't have dual extruders so, PVA filament isn't something they'd buy. Once you've bought into dual extruder printing, you've figured out that cheap filament prints like crap, and nicer PLA+ makes alot of sense in terms of actual material needed and wasted material.At which point PVA isn't really cost prohibitive.
I would really love a sound based series where the dba and potentially a subjective "how nice to listen to" score are the focus. Could make use of the resin printer more and then rank all the fans based on fpm ratio
Maybe normalizing with power or so the rpm of each fan is the same?
Honestly the hard angle on the Dragon Wing has a smooth but more steep curve to it than the original Noctua fan blades. But Noctua's new fan blades yet to be released on their updated 140mm design, also, has a steeper curve; it's more abrupt/jagged of a curve though.
Could you 3D print the Noctua fan case with clear plastic? It would be cool to see the fans spinning from the side during the smoke tests.
And it's a good-looking fan, too. I'd buy it.
right, it look good in a msi build
would be interesting to see this in a noise normalized comparison with the a12x25, i wonder if it could still win.
serious pc guys dont care about noise! they want it runnin cool.
@@Kdog307 thats the dumbest statement anyone has ever made.
Toms design is my favorite of the designs on display in this video it also had the most pleasant sound.
1:48 "Yo dawg, I heard you like fans, so I put a fan in your fan so you can fan while you fan!"
Well done Tom. Performance and good looks. Don't get better
Nice work with 2 color filament!
The fact that the Dragon wing's CFM is 69 repeated is just the best.
i bet that dragon wing design could be further improved, which might help with sound. I'd replace my noctua fans with it in a heartbeat as-is though!
it would look good in an msi build
@@Kdog307 That's a really good point. MSI could easily license this and slap it on their GPUs and sell some case fans.
Hey! I watched all of the seasons. It would be rather nice to add a noise normalize section in the chart, since many many of want our pc's to perform, but quiet! I think somewhere between 38-45 dba should be fine, I'll let you be the decider on that. I know this is a lot of work, if you should retest a lot of designs, but hey! Free video.
Thanks for your hard work!
I'm sorry excuse me... A fan just DESTROYED the Noctua A12X25 and every other fan on the chart and you just like... End the video on that bombshell? Wow. Back to the studio. 😂
Noctua has a new fan on the way, and during their interview with GN about it and the R&D process that took a few years, they stated that their fan isn't optimized for pushing air through a radiator or for raw CFM. They strive for a sort of hybrid between the two styles. They want their fan to work really good as a case fan, and equally as good as an AIO fan.
That being said, BY DESIGN, they are beatable with a more specialized design. It SHOULD be beaten in these tests. However, it SHOULD NOT be beaten across both open air and resistant (through radiator) CFM with a single design.
I'd love to see how the Dragon Wing would fair against it for open CFM. It would be really cool to see Noctua beaten at their own thing
we got ourselves a new best, and worst in the same episode! Neat!
Congrats Tom!👍
Wow, way to go Tom! It's nice to see someone doing their homework instead of just submitting garbage. 😂
Finally someone adding parts that create turbolence at the end of the blade. That's a key to make air turbolent and easiert to push. I think even better would be with an intake that "prepares" the air for the blades. That is, creates turbolent air (with fins) and with curved static blades (like the fan) so that the fan can push as much as possible.
Can we see a fan speed vs sheer airflow graph for the A12x25 fan versus the Dragon Wing? It'd be interesting to see its possible use in variable fan speed situations!
Fan in a fan.
„Sometimes my genius,… it’s almost frightening“ - Jeremy Clarkson
For a solution that is cheaper than PVA, when the application allows. When printing with PLA use PETG for support material & vice versa. The 2 materials will not bond to eachother at all, so the supports just fall away.
Nice score Tom! (69.696 CFM)
I was hoping someone mentioned that!
Nice.
That printer is no joke!
I needed a good laugh this morning. The Tesla turbine gave it to me🤣🤣🤣
I hope someone does an owl wing for those working with long hours but still produces enough flow to cool even on a hot day
A12X25 Fanfan was fun. Great job.
It was fun fun
msi should use the dragon wing in their pc cases! nice.
I find this way more interesting than I should.
i would be verry interested in seeing how the air moves trough a radiator/aircooler. That way you can see if it has some actual air pressure or takes the fastest way out and slows down.
And you should scale down the Dragon Wing till it hits the same noise levels and then compare. Then you can see if it actually is better on simmilar noise levels and add the pressure test.
Absolute powermove from gigachad Tom. Insane!
The fans keep getting better, maybe next season we can try some noise normalized testing?🤔
Tom just walked in and flopped his magnum dong on the table.
Bravo, my dude. Great design.
excellent work tom
I like how harmonic the Dragon Wing was, legitimately 👀Like, I'd deal with the noise (loudness) if it sounded that good.~
I'm glad the fanfan at least did decently for a good joke fan. Little bugger made me happy.
I agree I actually thought despite how loud it was it was actually a nice sounding fan.
Impressive, Tom 🙂
Does the Tesla turbine just need a cover around the outside so that air does not go back out the front?
that's a very nice cfm from the dragon wing
this dude has essentially made a game show
The Dragon wing is kind of cool but sadly you won't know that once it's spinning.
Also there are obviously more dimensions to fan design than cfm/rpm
Even noise isn't just noise as there are types of noise that are perceived more quietly. A low frequency hum can be the same dB as a high-mid squeal but the hum will sound quieter. And if both are at low volume to begin with, the difference will be even greater.
So it's a doubly good idea to try and get your fan to emit only low frequencies.
WOOT FAN SHOWDOWN! YEAH :D
holy crap that bat wing looks awesome!
LMAO that fan - fan is hilarious!
Skyscoop was a good one!
The tesla looks cool but not surprised it didn't do well.
WOW someone needs to call Tom and give him some money for that design!
Love the fan show down! Ever thought of remaking the Noctua and using that as a reference? A storebought vs a 3d printed fan might give a significant difference in results. This might be a better reference point as all others are printed as well.
Way to go dragons wing, GG Tom!
I'd definitely get the dragon wing
I like how the attack angle of the middle more aggressive than the tips, creating faster flow in the center making outersides not only sucking air from front but sides as well. ( all fans does this but I think in this model this is makes the fan act like a bigger diameter fan )
I feel like the fan fan would be really fun to use in a much larger fan. Like a 200mm or something, so that the hub isn't taking up half the blade area
Yeah
The limit then becomes how big of a square can you fit in a circle (big fan hub), and then how big of a circle can you fit into that square (little fan hub).
I'll bet that Tom's Dragon Wing would perform even better made with a mould press. Those jagged edges of 3D printing puts a huge drag on the airflow.
While I enjoy the creative and over the top big designs, I find the ones that stick within stock fan dimensions much more interesting.
Dragon fan is impressive for fitting completely in the housing.
Would be interesting to have some of the better performing and high selling fans included in the chart as well, just a few of them to see where they slot in among the printed fans.
The fan-in-fan design is the PC tech-equivalent of Chrome Spinners on car wheels.
Sleeper/conversion cars are the most high roller cars
There's this one guy who sells ender 3 core xy conversion kits. That's probably the most baller 3d printer possible.
"Yeah I have an ender 3. Yeah it does a 7 minute benchy."
The dragon wing outperforming the A12x25, even within it's form factor is a big thing even if it is louder. I'm curious, would a softer material like PETg decrease noise?
You'd probably want to go even stiffer actually. Some of the noise is from the blade tips vibrating, and a stiffer material would "flap" less. Also, softer materials will likely expand more under the centripetal force of high-RPM tests, meaning you have to design with less efficient larger gaps around the edges.
Love the fanfan and the tesla turbine! I think the tesla turbine might actually work, if you put a stationary restricting shroud over the intake. it sucks a little air in through its intake holes, and it projects it outward against the outer rim (pressure on the outer rim) but then it goes equally forward and backward. Just put a lip around the intake side, so the pressure on the outer rim can only go out the exhaust side. I don't have high hopes that it will perform well, but it'll certainly be more than zero!
I'd really like to see the fans set to a normalized 40db for the airflow test. Testing airflow at max rpm doesn't tell us much, since most people would be irritated by a nearly 60db fan.
1:45 You know, when you showed it off there, i didnt notice anything about it
But this 1:47 really does put it into perspective, LOL
Noctua is still unbeaten, imo. It's not enough to push more air through, if the acoustics get worse. When someone comes with a fan which is quieter or as quiet as Noctua and still within that 120x120x25mm envelope, I'm sold.
The dragon wing makes an interesting chord. Sounds like a car horn. which i think is like a major 3rd?
the fanfan reminded me of the dizzyfan. a friend of mine has this complex, where there are rows and rows of electrical cabinets. each one had a large external case cooling fan much like these except bigger. they run of 12V DC and turn on whenever the units reach a curtain temperature. some of them had failed however. i wonder if he learned his repairing skills from Mcgyver, because of all the Jerry rigging that was in there. for example, a random DC12V blower fan was thrown in to replace one of the original fans. on another unit, a cheap desk fan was just blowing in through the inlet of the old fan and the DC wires were just connected to a 12V lightbulb.
last, but not least was the dizzyfan. he had taken a random mains powered motor, slapped it on a bracket in front of the inlet and attached the fan to it. he didn't even cut off the fan's power cable, witch was just whipping around and slapping anything in it's radius, making noise. and to control it, the 12VDC lines were just connected to a 12vdc trigger relay thingy. he had written "dizzyfan controller" on the little box the relay was in, hence the name. you could tell that it was on from the end of the corridor from the distinctive slapping sounds of the cable going round and round. it looked real goofy while it was on too LOL!
Now you need a six petal fan where each petal is a spinning fan. Like those crazy carnival rides.
🤔
hell yeah good job tom
props for people using modeling software as cad software. i do that, 3dsmax 2014 is my cad software.
geometry is geometry. geometry + printer = thing.
The thing with Fusion 360 is that it's cloud based (other than the fact I can't do anything with it).
In order to export to STL, you gotta send it to the cloud, and they put you in a render queue.
I first tried to center the fan hub template to the center of gravity, and it would just make these null points, and there was no way for me to translate that point to the center. If I just had a python terminal or something I could figure out how to translate it negative those coordinates. When you first import the stl, it lets you center it, but never again after that.
Anyways, one render queue later, I'm in blender, and I just right click, align, center of mass. I've actually never needed to center something in blender, and it's been a good half year since I've used the program at all. Blender isn't known for their intuitiveness, but as a 3d modeling program it's pretty good. I even got an easy python script off of stack overflow to lay an object flat on the xy origin.
@@leifmessinger i prefer to store my data on my own drives. thats why i use gimp instead of photoshop (debating running my ancient copy of ps7 in a vm). id rather use old software anyway because not only am i used to it, but it was just better, and in many cases it still works, does the same job. anyway 3dsmax is good enough for me, and its something i already had in my tool box. ive done some game dev and know my way around various polygon formats and how to do transformations in code if need be. but i just export to stl and im good.
@@LordOfNihil The old run it in a vm and save the machine state trick
Photopea is what I use now. I used to use premiere but now I use clipchamp.
Now take the dragon wing, and smooth it out, see if smooth is better or worse. This seems to have a large outside and shrinking towards the center intake with a full flat output. and the be-quiet wavy design.
My company bid on building Kenton's new Middle School/High School today. Weird coincidence!
So, if I saw either of those fans in the store, I'd buy that Kenton Baseball T-shirt. Looks good.
The Dragon Wing's CFM... Nice
2:40 "If it moves anything, that's a wind."
We need to revisit the Tesla turbine and add a front shroud with an inlet that restricts the airflow to those inlet vents, and redesign the outlet stator to better redirect the edge flow out the back.
But the drive motor is the limiting factor. I've seen a Tesla turbine used as a pump heave gallons of water, but it was hooked to a big 1HP DC motor.
At last, someone bested Noctua
The dragon wing reminds me of the old bat wing fan I had a few years back.
Also wonder how well the finalists fans will do in resin.
LOL.. I liked the fan in fan one!
You could do a whole series of silly fans (like a middle finger fan.. or an eyeball fan.. etc)
The hand wafting one was my favorite ruclips.net/video/Hnlds6gPotk/видео.html
I thought fanfan was bad, then turbine went...ouch. Another great episode as always!
The Tesla Turbine was really unexpected. I was kinda hoping to have the worst fan for the season. I beat one of the toroidal fans too.
The tesla turbine was meant for installation inline and had to be enclosed in order to have its exit port functioning. The stack of its plates were exactly in the same position on top of each other but it also was designed to move a higher density steam. I have seen it used as an inline water pump and it was an excellent as the force of the water leaving it was so high. With the density of air being so much less, the plate stack might need to be at a slight offset from the plate below it forming a spiral pattern of the openings in the center when the plate stack is completed. At what offset to produce the best movement would be up for testing. If the inside of each plate had teeth, the plates could then slide onto a gear affixed to the fan motor shaft for the purpose of being able to change the offset distances until the best distance was achieved. With a gear attached to the fan's shaft, that would also assist the drag to create more spin but only in an enclosure to help create that centrifugal force. Without the enclosure for steam, water or air, the result would be exactly as we saw, exit happened between each plate and stayed close to the fan because the enclosure would assist the air speed build up that would be exiting the blades. This would create higher exiting thrust.
Gonna need to put that Tesla turbine on a drone motor! And/or a fair bit of a redesign.
Why aren't you using the resin printer for the fans? Or some vapor smoothed ABS. I guess the overall roughness degrades performance quuite a bit.
Solidworks FTW ❤
with how isolated my headset is and how noisy my AC is, I wouldn't even mind a blower cooler in my case just to cool everything/ so performance over noise for me is always a thing. if the price is right ofcourse(which is why my PC is full of arctic p12 fans)