Can This BLADELESS Fan Work? | Fans showdown S5E10
ฝัง
- เผยแพร่เมื่อ 9 ธ.ค. 2023
- Collect 2 Epics for FREE in Raid ✅ pl.go-ga.me/6exdcut4
⚡️ Lightsworn + Juliana (available after reaching level 15) ⚡️ Available only via the link and for new players.
See you on the battlefield!
Have you ever wondered if a fan with no blades could work. well today we look at one modeled after a natural phenomenon that performed better then I could have predicted.
My Twitch
/ majorhardware
Fan Showdown Instructions
• EVERYTHING you need to...
Fan Dims
www.thingiverse.com/thing:455...
Submit your design (at least .STL or .STP / .STEP file)
fanshowdown@gmail.com
Cameras used.
BlackMagic Pocket cinema 6k pro
amzn.to/3m2RLVw
Reddit
/ fanshowdown - วิทยาศาสตร์และเทคโนโลยี
Collect 2 Epics for FREE in Raid ✅ pl.go-ga.me/6exdcut4
⚡ Lightsworn + Juliana (available after reaching level 15) ⚡ Available only via the link and for new players.
See you on the battlefield!
*@Major Hardware*
0:30 You could start a 2'nd series: "Fan Testing Showdown" (or a better name?).
Where people submit contraptions that can be used to test a fans different properties ;P
Would those Spirometer will the balls work? It just popped in my head while watching.
@@sebbes333I like this idea! I've posted my setup on the Fanshowdown reddit page already!
RSL sucks as a company. Sad to see you taking their money
Personally do not care about the game sponsor. (especially given it's uh... reputation...) BUT there is one thing I'll suggest for the next season and the on thing you had built for it. Which is... put sides around the flap so that when air pushes past it, air can only flow in the same direction as it went into the flap initially from, and no other direction until well after the flap as to ensure you get the most accurate of readings from the flap.
I'll post again so you can see it here. You can get a manometer that reads in millimeters of water column for 80 bucks. It's called a DWYER magnehelic. They have models that go from 0 to 6 mm of water column and another that goes from 0 to 10 mm of water column. I think it would be perfect for this show, and wouldn't cost you that much money. It would also be much more accurate than the 3D printed device you made. I think accurate static information along with the velocity tube is the ultimate test of a fan without a super expensive machine
I would choose the quieter fan.
I use that to test the velocity of my farts
He could 3d print a piezometer ring to make it more accurate.
Amazon has them as low as $45
Do you do dust collectors? Lol
Tom should build the dragon shredder, where he alternates one blade of each fan to see how it performs
i second this. tom if you are in the comments do this!
also the name sounds cool
Is it a dragon that shreds or is it shredding dragons?
@@pascalwiery7129Yes
@@pascalwiery7129It is the dragon shredding dragons
Wow, the bladeless fan didn't come last! That's pretty impressive 👏
Is came last. The FPM on the smoother are wrong
@@basquescout Yeah but not overall last as it did beat the Turbinator V3 from S05E08
@@basquescoutim jorking to u rn
It's not really bladeless, it's just a very inefficient impeller with support walls as blades.
I think it would be interesting to take some top performers and see how they perform at different duty cycles. As we know, PC fans don’t run 100% speed all the time, so maybe there are fans that maintain performance farther down into lower % speeds.
That's why I suggested to make a normalised run on 1000rpm, and one at full tilt, and decide the score as a performance to noise ratio of both runs
@@kornaros96 that's a great idea
It is nowadays compensated by PWM, but you may get some performance differential still
It could also be worth measuring the fans efficiency at each diffferent duty cycle (watts per cfm)
I think it would be interesting if he would stop using that trash motor.
Hey small mistake, At 9:19 Smoother shows 348 per minute airflow, but in the excel it is 248 which placed it at last while hurricane should have been there.
That airflow looked shockingly good on the hurricane! Color me impressed
Indeed, that was properly impressive given it's bladeless. I'd love to see some evolution/iteration on this design concept.
@@KaineTubeit’s originally designed to be a new type of bladeless omnidirectional windmill.
@@KaineTubeIt's essentially an impeller mixed with a Tesla turbine, I think. Given that, I think that it needs more, thinner passages for air, and that the vanes must also be thinner. A Tesla turbine has an ideal spacing based on the viscosity of the fluid moving through it, and since air is an exceptionally low viscosity fluid, the spacing needs to be closer.
Also, a linear curve from end to end might lead to lower losses due to air direction change.
Now
I can't believe that one of my favourite shows is a show about fans but here i am, loving each episode
Some might say , you enjoy OnlyFans ;)
So, you're a fan of fans ? (I am)
@@thesfreader3068 im a fan of the fan of fans, csnt say that if this show syopped id care much
To me, quiet is always more important than looks, provided the fan performs well.
The problem is that the quieter doesn't move as much air.
The louder one is objectively better, because you can reduce the fan speed to lower the noise, and still obtain the same cooling as the quieter one.
I'm not saying this is a general rule, but this applies for the specific example we have at hand.
@@espertalhao041 and don't forget that the Noctua fan actually fits directly in the middle of those 2 fans and performs better than the Shredder while still being quieter than both the Shredder and Dragon Wing.
@@AbbyRawks Therefore my last remark, as it was just about those 2 specific fans, and nothing else.
@@AbbyRawks you cant compare the Noctua fan to FDM printed fans. Material type, thickness, mass.. it all matters. Until he gets the STL for an actual Noctua fan to print himself, the whole thing is skewed. I'd imagine we'd see significant boost in performance over a lot of these fans if they were made the same way the Noctua is.
First consideration : Performance.
Second consideration : Noise level.
NOT a consideration at all : Looks. After all, when the fan is working, you can't see the fancy details.
Negative consideration : Fancy lighted fan. NOBODY need lighted fans.
I highly prefer opaque cases.
It would be cool to see how the fans perform, and sounds like, under reduced loads. 25%, 50% and 75%. Not just at max speed. I think some of them might surprise us.
🤔46db comprison to match the A12X25 as a colab with Gamers Nexus would be cool
Believe it or not, most fan noise comes from turbulence generated by putting things in close proximity to the blades. If you space the fan a few mm away from a case grill (even while retaining a seal between the fan and grill) it will get substantially quieter.
I'd love to see the Dragon Wing, Shredder, and Noctua against each other on a chart. RPM vs dBA vs FPM/CFM. Basically run them all at several RPM points and see how the performance is. Most people don't run their fans at full blast all the time and the Shredder could be optimized for a different RPM. Throw the Devil and Angel in there too. Would be illuminating.
Yeah, a 46db comparison would be cool, 🤔a colab with Gamers Nexus for this would be AWESOME!
@@shaneeslick Oooh noise normalized makes sense too! I'm just thinking a chart of at least 5 or so RPM levels would give us an idea of where each fan's sweet spot is and what the real trade offs are.
To answer your end of video question, I'd go for the dragon wing. Very few of my fans ever run at 100%, and my PC overall has a gentle hum at idle. More performance is paramount, and under heavy heavy load I'd likely be wearing headphones anyways.
I'm intrigued by the hurricane as an exhaust option mounted on the exterior against a wall, or behind a cabinet, maybe with a cage. The outward airflow scatters the heat in all directions instead of in a cone. It could also be used to create turbulence inside of a case to help cool vrms, ram modules, or NVME slots. if used in a pull configuration on a dual fan tower cooler with 2 additional fans out the back and top, it might actually significantly improve vrm performance.
Louder all the way. It will mostly run slow, but if it's needed, it has that extra performance.
or in a push+pull setup with conventional fans for the push
@@Glyn-Leine that's what I meant, push on front and pull on the backside
Some bumps or dimples on the interior surface of the hurricane might improve boundary layer adhesion and performance.
Yeah I agree the best use for the Hurricane would be for a Home Theatre Cabinet for component exhaust,
instead of powering the airflow at the wall like a normal fan it would disperse it into the gap.
Dragon Wing - because I want my fans to cool, not just be quiet. Quiet is a perk, but performance is key.
Thank you for showing pictures of each fan when doing sound test. I never remember fan names and dont know which fan is going on without picture
Thanks for giving the Hurricane a chance! Was a fun experiment 😊
I'm glad it did work at all, and in the first place was printable 😅
The smoke test looked surprisingly clean! And it was also quite silent, nice.
I'm curious what the community will do with that idea, I've seen many comments of how to improve it, or use it in different ways.
it's your design then?! can you explain how it works please!! Or have a recommendation on what I should read to try and understand how it works?
@@zippolag Yes it's my design. I'm watching this channel for many years, and in the smoke tests I noticed the air escaping sidewards in some designs. So I came up with this idea, not trying to avoid it, but use it in my benefit. As it's my idea, I do not know if anyone else had it as well and did some documentation on this. Reading through the comments, some people say it is similar to a Tesla Pump, others say it's like an inefficient impeller, whatever I'm happy it does move air, and does it quite silent which was the big aim.
Adding a second test to all fans would be helpful. Put all the fans on a variable controller and measure the noise from about 20in. Then lower the RPM from maximum until you hit 36 dB. Then test the airflow of each fan set to 36 dB. I personally never run fans at maximum, and I doubt most other do as well. I try to put a curve on mine to be mostly silent unless the temps go really high. Some fans I've had (like corsair, noctua, lian li, etc) were vastly different at maximum RPM for performance and noise, but the most expensive ones weren't always the best when I tried noise normalizing them. Right now I'm just running cheap TR 120 ARGB fans because they are dead silent and still keep my PC cool enough. I would love to see an extra test normalized for noise, and I don't think it would take long since you already measure noise and test airflow anyways, you just need to get a speed controller and a jig to hold the mic a set distance from the fans to keep that consistent.
Agreed - nobody runs fans at max rpm, which is why most of the "on the box" stats are moot. Cooling has improved so much that fans just don't need to run fast in most scenarios. I extensively tested a bunch of 120mm fans and found that they all perform about the same at roughly 1,000rpm, which is the speed that most people will want to run them at. Where substantial airflow and static pressure are required, you just have to forget about the noise factor.
I like how the computer in the background showed an animated fire the moment you said dragon wing.
I saw the fireplace video going in the background, paused, and started one up on my other monitor. Now that I'm at the end of your video I could swear I'm feeling warmer on the side toward the fire. :D
Nice to see you back!
The hurricane did a surprisingly good job for a bladeless fan! I wouldn't expect it to compete with the axial bladed fans, but for what it is, it did very well.
Would it have been possible to test it turned around (blade inside the tunnel) so that the airflow was going the same direction as the other fans? That seems necessary for a fair comparison.
Shredder.
Thanks for the work and sharing data.
Merry Christmas
Best thing on the internet, i love this. The hurricane is such a cool idea, loving the bladeless designs!
ive been waiting for this
I'm curious how a normal fan would do pulling air backwards through the tunnel. Would be a good apples to apples comparison of the Hurricane to mount a bladed fan backwards.
No me deja de sorprender lo apasionante que es diseñar nuevas hélices...Gran trabajo Major Hardware
It is a testament to the quality of your content that I don't click away as soon as you mention Raid
love the series. That hurricane one can def be impoved upon, but im suprised with how well it did though. Nice to see new ideas.
Hi Tom, great video. What do you use to make the vapor/smoke?
What you've created with the flappie-doodle thing is basically an old-fashioned mass airflow sensor (look one up for a 1990 Toyota Camry to take example).
What you're looking for is a manifold absolute pressure setup. Design a manually variable gate and use a manometer installed between that gate and the fan to measure pressure, and an anemometer on the other side of the gate to measure air flow.
By adjusting the gate, you can measure static pressure vs. airflow and plot them on a graph. On most well-designed fans, the relationship is not linear; there will be a point where airflow and pressure are ideal.
Plug in values for RPM and current draw, and the result is an accurate representation of the fan's overall performance.
Maybe it’s a great idea to use a u-tube nanometer and a radiator for pressure measurement. Just place a piece of straight section in between the fan and the radiator making a pressurized chamber and connect it to the u shaped tube filled with water. It’s simple, you can measure both a positive pressure and a vacuum, it has a visual reference so it’s fun to watch and it has real life condition simulation. Anyway I’m a big time enjoyer if this channel keep it up💪
This was covered in the last video and in it's comments section. The Noctua fan used as the base for all these designs creates a static pressure of around 2mm water, with such a tiny movement how are you going to measure the differences between fans to any reasonable degree of precision?
@@ferrumignis I didn’t knew lol however if you make the tube where the pressure acts upon bigger than the tube which is moving you can amplify the output really easily
I love this channel and this show so much!
I like low noise and would like to see the Alphacool Apex Stealth metal fan get compared to other things on the test charts
We all have more than an fan, so I would try both the quieter one and high performance one, together, on seperate fan curves and seperate headers.
It's what I do now with my exhaust fans. One is top exhaust, very quiet. One rear industrial exhaust, powerful but loud.
I run controlfan software which allows you to spin fans at speeds depending on the temp of a number of different components, meaning fans nowhere near the heat source, don't have to ramp up. All helps to keep it cool with a good volume balance.
the best way to get good advice on the internet is to be a bit wrong. i love the fan testing box idea. keep it up!
For the static pressure meter, just put the door on a spring. BMW used this design for air flow meters in all their early fuel injected cars in the 80s. The door was on a clock spring that gave it some resistance and linearized the output. Also, counterbalance the door so that the weight of the door doesn’t exert any closing force; only the spring holds the door shut. BMW did this by simply mounting the door on a vertical pivot.
It’ll also be a lot more accurate if the zero point where the door is shut isn’t at 90 degrees. A little preload, like 10 or 15 degrees also linearizes it.
Obviously this design isn’t used in cars anymore, because it’ll ultimately never be linear, but you can get it pretty good. Good enough at least.
really shocked by the bladeless performance. super interesting
Thought - could be a fun video idea. I really like the idea of the swinging gate device to measure static pressure. I think you could run a test to help you create a scale for it. If you were to hook up an a12x25 to it and control the voltage you run to the fan. Then, have linear increases of voltage and track the position of that gate to that. That way, you'd be able to increase the power of the fan linearly and compare it to the height of the door. Just a thought, love the channel.
the shredder goes towards an idea I had, the shovel fan that was on ages ago performed way better than it should have; I think having a very thin bottom part of the blades will be good
Efficiency is the key to quiet operation.
For the most part, a 120 mm PC fan will spin at around 800-1000 RPM (or roughly 40-50% speed).
Now, it would be fun to test various designs at these speeds to see which ones are the most efficient at moving air at a low noise setting
For your static pressure tester, have you considered an adjustable spring? If you set a target for the airflow to open the flap, say, to 45 degrees, you could adjust the spring tension to restore the flap position to 45 degree per fan, then compare the spring tension values across fan designs. It's an approach similar to used in AFM designs in 80's automotive applications.
I'm curious how often you test the hub without blades on to make sure its not getting noisy on its own, I have one fan in my PC thats clearly starting to wear out and getting ever so slightly louder as time goes on, would hate to see some of these brilliant designs get a bum deal on the noise test from added noise from the hub motor itself.
Regarding the non-linear flow gauge, you could drive a nautilus gearset so that initial movement is very slow but continued movement gets multiplied the more the flap is opened.
Was there any variable spacing between the blades on Shredder? That harsh tone sounds like the blades were arranged equidistant to each other. Variable spacing is key to a smooth noise profile.
Is it better to have diverging fan blade trailing edges or converging trailing edges?
For me to choose between Shredder and Dragon Wing I'd want to see a noise normalised comparison.
I wouldn't like to run a load fan very fast but if it still outperforms the quiet one even at low speed it would get my vote. On the other hand I could run a quiet fan at %100 so it really depends on what moves more air at the volume of the Shredder at max RPM.
I agree, noise normalised would be a nice metric to see if I was after the best performance to noise ratio for a quiet system, perhaps something that can be added to the next season.
I live in a tiny house and use a computer fan for exhaust. It is sealed up really well and the air flow is needed to remove moisture. However when you open the door it will start spinning the fan backwards. I think the hurricane would work great for this purpose because i think the it would allow for pressure changes without stalling the fan motor and probably be easier on the ears. Would like premission to print one out for myself.
Dragon wing every time. Performance is the measurement that matters to me.
Keep up the great vids. Stay Happy and Healthy
Have you thought about doing 120x38 models. Also if there's benefits of adding a fan shroud pre or post fan to straighten the air? Not sure if it'll help increase air intake/exhaust.
Wonder if additional testing at either only 50% or at 25, 50, and 75% power would help judge fan designs. I could see some produce great results at full power, but end up bottom of the barrel if running half power.
Shredder for style, dragon wing for performance, and hurricane was AWESOME.
What if you voltage limited the Dragon Wing, Shredder, and original like the Noctuas allow and see if the noise profile changes at lower rpms and performance isn't linear?
for the static pressure device, why not a flat panel attached on 2 sides with springs inside the tunnel. then calculate the spring constant and the nm to compress the springs to determine force from fan
Before I look at the results that smoke test looks like the best blade-less fan yet. By a ways.
Great job. 👍
I think I would rather have the Shredder. I actually prefer the look. Also I would try to throw it on one of the fans with an extra gear so if I needed to get the extra performance when I need it and the quieter performance when I don't.
About the last question... I would run some test and if temperature would be similar (max. 5C diff) then why not use the quieter one?
You could add a pocket to the door at a distance perpendicular to the door to hold water and use a burette to add water until the door just closes (with a fixed diameter bleed hole) and use the volume dispensed + calculation to give the relative figure
For static pressure I think you will need something with a sample tube in the air stream and some liquid to lift vertically akin to a barometer.
Whats the quietest fan you ever shown?
im looking for a fan that has standart cooling performances, nothing special, but is quiet af
Dragon Wing all day, every day. My standing floor fan and my headphones/speakers put out way more noise than any case fan ever could. Give me dat sweet *performance*.
Did the dragon wing come out as clean as the shredder when you printed it? If not that could have been alot of the sound
I'm a custom loop man. I go for high static pressure to push thru the radiators and quiet because low idle noise is one of the joys a custom loop has. But a lot of that depends on the application. With just air cooling, In a giant atx-xl case noise can naturally be contained within the case sometimes so cfm is what you shoot for and worrying about noise is secondary, whereas in an sff build noise can get out of hand quickly, but you still need good airflow, and in some big cases as well noisy fans can start to sound like an airliner spooling up
The robot told me Farahdin got me rolling man.
If the quieter fan was sufficient in cooling, I would feel comfortable using 2 of them to increase airflow. That means one for intake and one for exhaust.
Also, you could 3D print an airflow adapter that fits in your current flow tube.
It would need to have a small pipe inserted into the middle with half the tip cut for a venturi effect. Then connect a hose to a carburetor airflow equalizer meter that uses a floating bead to measure vacuum pressure created in the tube.
The hurricane might be alright for strictly exhaust🤔
It doesn't have anywhere near a tight enough cone to blow air at or over parts that need cooling.
Just a thought for the static pressure contraption. Perhaps if you used a light weight PLA or a pre-foamed light weight PLA and made the door be a spoon or curve that would continue being in the airflow as it rises but the fan would need to push more and more weight which would be much more linear rather than exponential. Just some thoughts
what spectrogram will have a simple disk instead fan? What do the fan contribute, and what is system noise?
last year i send the 64B PC comresor which is made for static pressure i wonder how it would perform in this new season, also a have tip for better static pressure mesure, you could attach the fans to ballon on weight scale and how much the ballon would puffed up, the more weight would be applied to the weight scale.
If you go with the higher performance fan, you can run it at a lower RPM, ultimately delivering similar performance noise normalized, except with extra overhead when you're really are pushing the envelope.
To make the measurement device valid and work here is an easy fix. Create a small weight system to the door (i.e. neodymium magnets that you can stack), find a specific value and aim to make that consistent by changing the weight, it is not too fine of a measurement, but it allows you to have a reference, by keeping the door around a similar level and then just knowing how much weight can be moved
Easy fix for your DIY gauge:
Instead of designing it as a "swinging hinge" design it to where it has a VERY light weight spring pushing the door towards the cage, then have the door slide along a rail using frictionless bearings.
Or even do it without the spring, since that increases tension as it's compressed, and reset it manually after each fan.
A vertical tube with holes in it spaced evenly along its length(think mythbusters when they tested the terminal velocity of a penny), put the fan at the bottom, and a plug in the tube. Whatever fan can lift the plug the highest has the greatest static pressure.
Would love to see the performance of the hurricane with a volute style housing to concentrate the exit. Seems you could pipe the heat out of the system very easily that way and have a compact volute shroud around it.
I'll go Dragon Wing for performance, I use headphones a lot anyways, so I won't even notice
What about measuring a membrane flex for static pressure? Need a fine optical sensor to measure the swell of the fabric but you'd maintain static pressure throughout the test.
Louder or better flow?
That depends... if i slow the one with more flow down to where the sound is equal with the other... is the flow comparable to the other?
Because if I can get the same noise and flow at less speed its still a win.
Your tunnel flow meter just needs a simple manometer ported from one side and a two-way bubble level glued on top. The quietest fan that provides adequate cooling would always be my choice.
I wonder if you did some profile sanding, how well the performance would increase for the fans... I notice lots of little channels and flow controls in the design, but there is 3d print bumpiness on them.
he's tested that in the past by resin printing and acetone smoothing a set of fans. the difference was insignificant. it seems the fan noise and performance if primarily effected by it's overall blade profile, not it's finish.
Working with large hvac we use magnehelic or manometers before and after The radiator but the same can be diy'd with a water collum if designed correctly
How do you do the smoke test? Smoke Machine?
I liked the look of the Shredder more.
But. What if you noise normalize the Dragon Wing. How does it perform vs the Shredder then?
- To measure pressure capabilities of a fan you may build a variable opening duct as a restriction and measure airflow at any step.
For example, a 50% opened grille, with another behind that can slide and cover the holes. Or a 33% with two sliding grilles, or a 25% with three, etc...
Measure both behind and in front of the fan to test any difference in a push or pull configuration
- I did not expect the hurricane to be that good!
- The Shredder may be "quiet" looking at the db level, but the noise signature is quite annoying in my opinion
well a gravity push measurement device (vertical) would do the static pressure part, by definition. well you have the down force (gravity weight) and piston area, and height the fan can passively push static pressure.
real question is if your building a system to a low dba standard like sub 40dbs and you ramp the fans down to ~2000 rpms which one performs better the dragon or the shredder?
the louder one for sure, as someone who built a few pc's in the early days where they scream their Fing heads off idling, 52dba is nothing, typically your game volume is much louder then that or your wearing head phones so it doesn't matter and everyone is just super pretentious about noise now days.
I have an old windows 98 for some older games and hearing that thing boot up sets the mood perfectly!
What if you were to smooth the printed fans? I have a feeling that'd increase performance more.
There was an episode much earlier on where this was tested and I believe the performance gain was very little to nothing depending on the model, so it was decided between him and the community that it wasn't worth the extra time and effort per every model, though they did look good.
There was already a video about that. If I recall, it made no difference.
Something that could be considered for the Dragon Wing I think is adding dimples to it. I know that you did tests forever ago on dimples vs no dimples but with updated testing methodology, he might be able to keep the sound levels lower without sacrificing as much performance compared to the "stock" Dragon Wing. Hell, on a fan that was performing as well as it was there might even be a slight performance difference between the 2 in favor of the dimples since the fan blades won't be flat.
Could you construct a duct where you see the smoke going in and then have it open on the exhaust side to see if it has laminar flow?
For air cooled PC's definitely the quiet one, but for radiator static pressure is going to matter more and will be louder anyways, so whatever is more powerful at that point. If you don't use headphones the loudness doesn't matter, but desktop speakers I would want highest airflow with minimal noise, and low noise being priority.
Hello, I recently started watching your videos and I need SOMEONE to help me with a question, all these models are made for noctua coolers ? and another thing, if they are all for noctua or whatever brand and the cooler already comes with a default rpm of 1950 for example, how can you change the rpm by adding a different design of wings (? I do not know how to express it, in Spanish it is called azpa but not the correct translation into English, hugs for everyone from Argentina. 🙌🙌👏👏
6:40 The supports/struts/members that enjoin the concentric duct-walls will act/behave as blades. Also noticed, whether a print imperfection or intentional, the surface of the ducts had seams or raised linear surface areas perpendicular to the rotational plane; this definitely would have an influence on moving air, such as the raised areas on the control surfaces of wings, the design of air brakes, etc.
You can still slow down the Dragon Wing to perform at about the same level as Shredder, but it still looks much better. :D Actually at some point you might do a noise normalized shootout among the best coolers.
@majorhardware please get the magnehelic, I love Your passion for testing and I think using a decent instrument will make the videos perfect
personally i used to run my old system that sounded like a harrier jet preparing for vertical takeoff so i'd go for loud+performance any day. when im not gaming i can always reduce fanspeeds for sound reduction :)
The dragonwing! And I'd improve the insulation of the case.
Can you make the static pressure thing by using bushings, and 4 rods protruding from each corner held to the face of the box using compression springs? Air pushes the plate secured to the face of the box away, compressing the 4 springs. Measure the distance the plate pushes away. Use a log scale. Sorry if confusing, I'm sick af.
You could have the air pressure act on a piston to compress a spring of known spring constant, possibly through a lever to multiply force for better resolution and measure spring compression to calculate the pressure based off known piston area
You just described the basic principle of operation in the Bosch L-Jet fuel injection system's airflow sensor.
You could try sth with compression rather than sth at an angle.
Idk, maybe how far it can press in a geared shaft connected to sth pressing or pulling a fixed weight up or down
A couple db will probably not be noticeable from within a pc case esp if it is under-desk or you use headphones rather than speakers tbh. Plus if the fan performs better wouldn't it be running at a lower rpm most of the time anyways and so not full volume most of the time anyways
If you do the door opposite with a small spring gravity will help you differentiate more as performance increases.
@MajorHardware It would be nice to look at ratios like CFM to input power (flow rate per unit input of power) and Noise level to input power. It would be interesting if you can somehow get a good estimate or measurement of heat removal rate from the CPU to the input power. This can certainly help with the selection of fan blade design.
I've never found fan noise to be a major consideration for me, so I'd choose the cooler, better performing option and take the extra noise for whatever penalty that applies.
I've got some random Corsair fans currently and when they really get ripping you can hear them, but it's never bothered me.