I Tested Some Of Your Strange Ideas On My Hubless Fan
ฝัง
- เผยแพร่เมื่อ 20 ก.ย. 2024
- After my last video where i created a 3D printed hubless PC fan many of you had recommendations and or ideas on how to make it quieter. However, not all ideas had to do with performance improvements. Regardless I thought why not give them a shot, so today we test some of your strange ideas on my hubless PC Fan.
Fan Models
www.thingivers...
Printer Bambu Labs X1-Carbon
bambulab.com
![วันเกิดเธอ (TO YOU) - PHUWIN [ OFFICIAL MV ]](http://i.ytimg.com/vi/_ROx1Bj3T0Y/mqdefault.jpg)
The biggest problem was the windings! That thing was drawing 5A at 6-odd Volts, meaning it was current limiting (CC mode). I think you need thinner winding wire, and maybe triple the amount of turns on each pole. It could be that high a current was pulling too much, thus why the windings separated towards the center, and the magnet got pulled. I'd try it with much thinner winding wire, and have more turns.
He probably used the larger wire for ease of use and for proof of concept.
That power supply is 30 watt limited, so i think that's the main issue. I have the same one at home.
@@themuffinfish2091 a 30 watt 240m fan
@@themuffinfish2091 if yours is only capable of 30w then something is wrong with it, it should be able to maintain 5a at the full 30v. In this case it's maxing 5 amps and only needs 6v to do so because of the low resistance of the fan itself, which is why more turns would help. Regardless, 5 amps is a TON of power to be pumping into a fan that size
Would it maybe also help to have a ferrous core in the middle of each coil to boost its magnetic flux
Two things:
1. Magnetic bearings
2. Better ventilation of the motor windings
3. Thinner gauge coil windings
4. More turns per coil
5. Ferrite cores?
Make the best out of it!
6. nuclear fusion core
@@fisc_rl Flux capacitor on the kanuter valve
This is like the most fun channel someone could create. Building, experimenting, and blowing stuff up for science! Love it.
Rctestflight, hes also quite up there
@@lonemarkkingoftypos3722 James Burton, Stuffmadehere, Ididathing, Backyard Science Channel, PeterSripol, Tom Stanton, Eletroboom, and MarcoReps, and AVE, and yeah all work super hard to make this look fun! Just coming up with ideas is probably exhausting though not to mention everything else involved.
@@lonemarkkingoftypos3722 oh yeah, he does super cool stuff
I think the best part is that it starts as an interesting idea, then the community throws 1,000s of ideas of how to make it better.. So chances are, whatever idea you've had, it's already been mentioned. Sometimes your ideas get put to the test, sometimes a better idea gets used instead.. either way, it satisfies that childlike curiosity for how things work and that thing that probably 90% of these viewers all have, where we think of how to improve everything and fix everything and create everything.
A tip regarding the laser spedometer. For more consistent readings cover half of the rotating part in white tape instead of just a tiny piece somewhere. As far as I know the spedometer gets its reading from the switch between colors. For every two switches (from white to background-material and back to white) it adds one rotation to it's internal counter and then calculates it against the timed avarage. With such a tiny piece of tape it happenes on some spedometers that they overlook some switches. That results in the RPM being all over the place wich seems to be the case with your spedometer.
Keep it up, I love your content and the crazy/awesome ideas.
my guy thinks it's a speedometer and yet here he is giving advice on how to use it> I love TH-cam comment experts
@@kermitthefragg does it measure and report speed? Then it is a speedometer
@@gormauslandertachometer
@@kermitthefragg Hey, thank you for your kind comment.
I am from germany and english is not my first language. I was sure that spedometer was the correct translation for Drehzahlmesser, but I stand corrected.
Even though I used the wrong name, I still know how this tool works. I use it myself at work from time to time.
Have a nice day.
Speen Gauge
PTFE lubricant is where it is at for FDM 3d prints. It gathers in the layer lines and keeps it running smooth for a very long time.
SuperLube PTFE is the best lubricant I've ever used for small plastic stuff like this. It's like magic!
It sucks for the environment sadly, please don't use PTFE for non-critical tasks (yes you can clean it up but it's the manufacturing of the stuff that's really awful)
@@szurketaltos2693 If you want to talk about environment go to china and stop polution.
@@gelo1238 ah yes, because China is doing it why don't we also destroy our rivers and lakes.
That doesn't make sense
You can also use silicone grease used in RC car transmissions.
There's also molybdenum grease.
Definitely put some iron in the coils to bring the field strength. Laminated iron cores would be best. Set the magnets in a thin layer of resin to seal them to the fan better. Adding a thin plastic layer for bonus points. Think pop bottle. Hot water will help the pop bottle plastic stretch.
I wonder how well proto pastas iron fill filament would work
I believe that is what it is made from
I wonder if there's some 3d printing niche for laminated iron cores as 3D printing already makes things in layers
He could print the model to the point where he can set some iron cores into the print and then resume the print.
For the next series, Major Hardware should set 3 new classes of user submitted challenges. 1) Hubless blades design 2) Bearing design 3) Power motive design.
This could turn into a very interesting user design competition.
I mean, having some kind of metallic race for the bearings will probably improve quite a bit, but good luck with THAT. Maglev bearings might be doable, and would definitely be quieter. Also, don't put your hotdog in that.
Regarding metallic raceways, he could probably include a wire race bearing into the existing design. Look up Eros Nicolau / 3D printed wire race bearing here on youtube.
Or just collab with another maker. Someone's got a lathe and might be willing to make the surfaces for him.
Man, this concept has SOOOOOOO much potential. I really hope you go all in on this!!
Thinner windings for more speed at the same power draw. Since you are running a hub motor, look into an air bearing setup. That would also help the heating issue as the air movement both levitates the hub, and cools it. It is amazing how much weight an air bearing can handle.
You might want to try suspended magnetic suspension bearings.
hahaha
That flir footy is JUICY!!!!
So you a footy lover then?
@@Thommie_techno that’s why I’m on YT
@@miketech2 Naughty you!
MSLA is going to give a smoother raceway, and you're going to also want to try a different resin. Siraya Tech Blu Nylon Mecha White is scratch resistant and very low surface friction (first layers should be 60 seconds, as it is very difficult to get it to adhere to the build plate, but regular cure layers should be like normal resin, as it sticks to itself quite well. It's also a very tough resin with 1900Mpa flexural modulus and 50Mpa tensile strength and 34 J/m notched IZOD impact resistance, so it's a lot like delrin for toughness) The bearing ball cage, fan case and fan blade insert can all be made of the resin and it should work well. The design of the fan should also increase the blade width so the blade edges are flush with the face of the air inlet and outlet curvatures. The blade length can also be extended a little bit so the ends are just 1/4" away from each other. Layer height would likely work best at 30 microns, as Mecha is smoother than most other resins, and the layers would blend together, making it even smoother.
A touch of high temp silicone spray would work too ? would that potentially dissolve bonds? i dont see silicone as a solvent anywhere on my hazmat material sheet(i found online)
Better yet... Have an actual bearing made...like collab with this old Tony or someone to machine one on a lathe.
My only issue with mecha is it has super shrinkage problems if I remember, and for something you want as well spec'd as this that might be an issue.
@@neileddy6159 Blu Mecha is not like Fast Mecha.
Four regular ball bearings, one in each corner, no other touching parts. Maybe avoiding metal isn't something you can do. The magnetic friction will probably still be less than trying to build a massive raceway around the whole ring.
Instead of gluing magnets to the outside, maybe have a one-layer thick perimeter of plastic, and you slot the neodymiums in from the face.
I saw a suggestion of putting four motors in the corners as opposed to making the whole hub a motor, but the rpms they'd have to spin with the gear ratio they would form with the giant blade would be crazy. Unless you know of motors that can do it..
It's great to watch the journey through experimentation. I never thought I'd be up at 2am enjoying watching you destroy a hotdog holder.
I have a couple of suggestions.
1. perhaps try resin printing for the bearing races, at whatever the highest quality is. This will get a smoother (or at least more consistent) surface for the ball bearings to roll on. I think this will help with the sound.
2. For the graphite lubricant, I think the best approach would be to use a cotton bud or something to smear a tiny bit all along both races for the bearings. That way there's not too much but you still get some of those lubricating properties.
3. If you want to try a new approach for the motor windings, I would recommend fewer turns of a thicker wire for the windings. In the RC world, brushed motors are often rated by the number of turns, and fewer turns means the motor spins faster on the same voltage, but pulls more power for the same load. More turns will probably only result in the fan going slower, since more power didn't seem to make it spin faster, which to me means it's potentially limited by the coils themselves. 20 turns is fairly common for RC brushed motors, and 10 turns is about the lowest you can get.
4. If you want something other than a ball bearing, you could try printing a roller bearing. The rollers would almost certainly have to be resin printed, unless you could get metal ones somewhere. I don't know if that would be quieter or not. A potential idea as well is to reduce the number of ball bearings (say by half) which should substantially reduce the friction.
5. As other people have said, you probably need some sort of cooling airflow on the windings, or holes in the case to allow heat exchange. This is doubly true if you move to fewer turns, since it will pull more power.
What I learned from this video is that you have AMAZING design skills--both hands on and conceptual. You'd make an excellent engineering professor.
With those motors, more voltage give more torque, but, to get them to spin faster, you need to increase the frequency of the current.
Frequency dictate how fast it can ultimately go.
Voltage and current influence how close you can get to the ultimate speed.
Currently, your speed is limited by the frequency. That's why it tops out when you increase the power.
Magnetic levitation would be great, but, it's not as easy as it may seems.
For plastics I highly recommend WD40 Specialist PTFE. I use this for windmills for example, where rotating shaft and the hub are both plastic. Dont use Silicon or White Lithium Spray or Grease on it.
WD40 is a literal LubriCANT, water would be better in my experience.
@@matthewmclean9734 PTFE spray isn't the same as regular WD40
The blur was absolutely one of the most well placed and hilarious things I’ve seen. Then you follow it with the unadulterated limp Weener going into that nasty hole.
Absolutely amazing channel.
@Major Hardware
I think you have done some great work here! This fan project has been fascinating to watch!
Rule of thumb with lubricating contact surfaces. If it slides, grease it. If it rotates, oil it.
I don't think that's a thing. There's a reason wheel bearings are universally packed with grease and not oiled.
Most roller bearings are packed with bearings from the factory. And in the engine crank case the engine bearing (sliding) is oiled...
So neither the rule of thumb is correct.
You grease something if it's sealed or unable to be oiled, like most sealed roller bearings or fan bearings. Oil generally is better than grease, but needs to be reapplied.
Air hockey table bearing. I, too, have suggested maglev in a previous video, but a journal bearing using air instead of liquid might actually work. I don't think air hockey tables require very much pressure, as in less than 1 psi, but you would need another blower and plumbing to make it work. Apparently it also helps if you make a "skirt" around the rotor that loosely fits with the housing, so it acts like a hovercraft with a pressurized volume "beneath" / around it. The skirt might rub a little bit, but it's a lot less surface area than the whole fan.
Always impressed by the work that goes into your tests and problem solving. Keep up the good work.
I would love to see you improve this design! Could be the next big innovation of the fan market!
People should have this be the next challenge. To improve your design.
winning design improvements are :
1. the wires need to be smaller & more turns. Try 20ga @ 40 turns . somewhere in that realm. Ur thick wires & less turns just makes more heat & less flux. think> more wrap=more flux.
2. for bearings use the gear bearing designs which can be found on thingiverse searching for planetary bearing. That is the only way to 3d print reliable bearings.
those 2 improvements will make the fan spin FAST
Reading through the comments, I think there's lots of good ideas. I think v2 should be a refined, not completely redesigned version of v1. Less coils, more turns, less bearings, metal race, iron core, embedded magnets, more cooling for the coils, and a different controller and power supply, then finally optimizing the blade design. This should make the fan much much quieter, and much more feasible. It may be worth the effort to dedicate a video experimenting with coil numbers and windings. I know this would be very time consuming, but it would be worth doing to optimize the design. Maybe focus on the bearings first though, then focus on the coils, since friction is your number one enemy at the moment. Once you have a better bearing system maybe then you could mess with the coils some more. I love this project, and would love to see it get some more refinement because it's super interesting. It would even be fun to see if viewers could send you their own designs in the mail for you to test. This is definitely a very cool video series concept and I'm loving the journey.
Legit, would love to see a CNC/milled/cast aluminum parts recreation of this. Proper bearing ring as well for that quiet and smooth spin. Maglev bearing as others have suggested if feasible.
Yes! I need a v2 of this!!!
Biggest challenge I see with the bearings around the edge is the distance they have to travel for one rotation. A bearing around the center axis travels very little distance while at the edge like you have is massively longer and faster, which undoubtedly results in the major friction you're seeing. Agree with the others that if you had some kind of magnetic frictionless bearing thing that'd be the best solution to reducing friction. Regardless, this has been fun to watch!
I think I said it last video, but my recommendation, apparently *counter* to other commenters, is to actually run a dual ball bearing setup, one set for either side of the coil assembly. I think a lot of the noise may coming from excessive rattling and grinding resulting from the torsional weight on the bearing race by the fan and magnet half of the assembly weighing on it from one side, which would be nullified if there was a second bearing on the other side of the coil to take said torsional load off.
Another thing you may be able to try, since you are working with a much larger bearing diameter than usual for a fan, is a *THRUST* style ball bearing assembly, where you have a double sided race printed on the fan, and there are races printed on the hub itself and the ring that clamps it all together. I'm not entirely sure if thrust ball bearings have ever really been a thing in the 3D Printing community, so it'd definitely be interesting to see.
Final note to *hopefully* save you a lot of frustration, to simplify disassembly and reassembly, you could try swapping out all those screws for a twist-lock mechanism and a screw or two to hold the twist-lock from vibrating open.
My immediate first thought when you asked for test ideas in the last video was "stick a hotdog into it". Glad to see I wasn't alone.
I have made a few small motors in my life. From what I have been told by electric engineers that if you use a thinner wire, you can use less energy and produce more magnetic fields per coil. If you were to add metal inside each coil, the name of the kind of metal is exascaping me, should also increase the field size. Keep it up.
one of the best thumbnails to a video i've seen in a long time.
Mr. Hastings sir, thank you for your excellent work. MOAR blades!
@ 11:00 - That "Hot Dog Test" was the funniest and saddest thing I have seen for a long time. Also, the blurring of your finger just before was a great edit.
Definitely roller bearing - Looking forward to development of the design - Great work
Not the content we deserve, but DEFINITELY the content we needed AND wanted
I just have to comment on the great color grading of your video plus the wide view that shows your whole mancave. It's an eyecandy!
I'm guessing that the controller board is the limiting factor. The RC battery pack should have had enough amps to deliver whatever that controller can take.
I think the best design would actually be a planar bearing. In most larger bearing applications that's what is used. Things like turbine pumps, turbochargers, and engine bearings come to mind. Fluids makes dead silent. Maybe just get a good print of polycarb or order something metal.
For the record, I suggested the hotdog and I was definitely not joking. Thanks.
I think the iron filled PLA is holding you back here. Maybe hit up pcbway or some other youtube-friendly manufacturer to see if they can SLM print you something out of pure iron.
Hey i dont know much about 3d printing right now but your channel has intrigued me, i only recently found your channel like 2 weeks ago and i ended up watching all the fan showdown episodes in just 2 days. Any who if you can get this material i think it will help with the heat problem, get a spool of ultem its got really good heat resistance.
What if you were to double your Rail (magnetic coil track) such that the fan disk takes a saucer shape instead of a ring shape. I think that there is a lot of friction coming from the resultant force pushing the fan outwards from the frame. Having two Rails running parallel will have a cross section like => |\/| which will maybe remove the need for bearings in design.
I would also recommend you make a mechanism to wrap your coils so that you can make they more consistent to each other.
i was NOT ready for the imagery created by you putting that hotdog in that fan! 🤣🤣
Science! Things were learned and hopefully improvements will follow.
Re: bearings and lube-
So as a yo-yo player, the topic of lube and bearings tends to be something that we are intimately familiar with. Stock bearings tend to be a little quieter because the lubricant that comes from the factory is a little on the thicker side, This is great if you love quiet bearings, but not if you want the fastest/longest spin time. Cleaning them and applying a tiny amount of thin lubricant (or running them dry) will give you a MUCH better spin time, at the cost of it running loud. (and maybe a shorter life if you run them dry)
There are some people who chase the unicorn that is a quiet bearing, but honestly, just about all of us deal with the noise because the benefits are far too great. Competitions basically sounds like higher frequency white noise.
Despite what people think about yo-yos, or by extension, bearings, more lube ≠ faster spin. The thicker the lube, the longer the life, at the cost of spin time/speed.
Also, funny you mention graphite lubricant (and other dry lubricants), it gets brought up every now and then in the community as well, but never used because its a pain in the ass to work with and almost never worth the effort.
Depends on the steel and the heat treatment of the bearings. Some will last very long time without lube. If you use ceramic they'll last forever without lube.
@@tacticalcenter8658 yeah, i ignored ceramic and Teflon coated because they are anomalies and most people don't know about/use them for random projects like this
Would be really interesting to try out Igus' self lubricating filaments for this application!
Great, fun episode! It's nice to see all the trial and error that comes in with this sort of seat-of-your-pants engineering.
It’s fun to watch someone reinvent the wheel (or fan), and end up reinventing Pepperidge Farms Deviled Ham from hot dogs.
My kind of kitchen amusement.
I'm surprised nobody mentioned "Super Lube Oil", it's a silicon oil with PTFE or "teflon" in it, seems like that'd be a great product for plastic on plastic. I actually use it when assembling mechanical keyboards, it quiets the pieces but over time if it dries up the teflon keeps the surfaces flowing freely, great product for cars, doors, any kind of hinges really... and things that come into contact with plastic.
I guess my only concern about the gun oil is that's an oil that plastics could absorb and soften/become brittle with over time, silicon doesn't mesh with plastic so it's good for that kindof thing. Oh and the teflon, I'd guess it works better than graphite powder with plastics as well, graphite is great for metal on metal, ptfe for plastics.
the lithium will surprise you as i use that for heavy machine work and on diesels and it works pretty well depending on what your needing to lubricate
Adding lubricant is going to be a source of friction/drag for the bearing. For the highest RPM you really want the bearing doing the work. The absolutely tiniest amount of thin oil would be better than any of these.
I also concur on thinner magnet wire with more turns.
Awesome videos man. Love the hubless fan idea. Looks awesome just loud.
Longboard bearings coast noticeably longer when ran dry or with graphite lube, and those are really small with the leverage of a wheel and a human's worth of momentum. So it makes sense with your fan being a giant bearing around the outside that the lube affects it so much. There's no leverage acting on it and barely any momentum from a plastic fan blade. If you could ever get your hands on a big and skinny enough metal ball bearing and try redesigning this concept that would be really cool.
OK, there is a way to make this much better, but it's an absolute pain: multirotor bearings. Every rotor adds 2 to the denominator of the bearing speed: 1 rotor means the surface speed of every bearing is 1/2 the rim speed, 2 rotors means 1/4, 3 is 1/6th. Sound (as in SPL) should be roughly proportional to speed^3, so even though you're adding another source of sound moving at speed, because both are so much quieter the overall SPL should be significantly lower.
Theoretically: going from 1 to 2 concentric bearings means a 2x(1/2)^3=1/4x the SPL, i.e. a roughly 6dB improvement. Every subsequent extra rotor decreases
SPL by another theoretical 3dB
Since the fan starts at 92dB, even going down to a loud but manageable 60dB would mean adding 1+(32-6)/3=9.66 concentric bearings.
Good luck. Oh, you thought I was going to design this for you?
- fewer coils (would be easier to print and make too)
- fewer turns of thicker wire. this is especially important because you're farther from the axis of rotation so you need more speed, compared to a normal tiny hub motor
- halbeck array for the permanent magnets
- more voltage (this and fewer turns will give you more rpm)
I can see that coming. The bigger the radius of your bearing/sleeve, the bigger the contact area and thus, bigger the drag and heat from it.
In airplane turbojet engines they solved it with "air sleeves" but for that you'd need a huge air displacement that would make a lot of noise.
As lubricants I'd suggest either 0W30 engine oil but if the drag is too dramatic then you could try WD-40 instead.
I don't think maglev would be great (at least for commercial use) and I'd definitely change the fan design to something like we have in vacuum cleaners or VW air-cooled engines.
Cheers!
No need to make one big bearing, make 4 small wheels in the corners, and leave the magnets and coils the same.
If you're thinking of trying maglev bearings, it might be a good idea to read up on Hallbach arrays, unless you're already familiar with them.
This was awesome to watch! Thanks dude, love your content!
particulate based greases are not good for bearings, they just jam the balls and start sliding. this is a wisdom for linear rails with smaller balls so maybe these being larger it has less of an effect. Main thing tho is that just a light oil would work best.
as for the printed parts try reprinting in ABS and at a lower layer height to provide a better surface for the balls to roll.
I build prototype electric traction motors for larger applications (25kw to 450kw). You are more than welcome to pick my brain on anything for this build, be it the mechanical, electrical, or electromagnetic design.
If I were doing this project, given the inherent uneven nature of 3d printed parts, I would build in a hydraulic bearing. This uses a fluid with a very high surface tension, but reasonable viscosity, to act as the bearings. The trick to this type of design is it requires more surface area than anything else, and the fluid needs channels to direct it to the load areas and away from the edge/openings.
You should use quantum black matter vortex grease!
honestly an use case for the fan would be in an mineral oil cooled pc since the design of the fan came from an boat and might be good for circulation and also quit it down
Sewing machine oil is specifically designed to lubricate metal parts, that have to go through repetitive motions and produce high heat from friction.
That would be the first lubricant that I'd try :)
Paraffin lubricant should work too. It's used for bicycle chains.
Four points. 1) cyclic motors (AC) are always pole dependent for speed, to increase speed, increase pole count and 7) historically (and when rebuilt correctly), the original lubricant used for wind up gramophones was a mix of petroleum jelly and graphite which means nothing I just thought was interesting to say.
I think using lapping compound might work better. Get lapping compound that is softer than the ceramic bearings and run the fan for a while, then clean it and add lube.
Whoa! Thank you! I hadn’t heard of lapping before. I just looked into it, and it can be done on virtually any material, including plastic. I wish engineering was required learning in grade school.
I'm sure that with that hotdog in the middle the fan will be more efficient in the wind tunnel test.
(No, i'm not kidding)
Electronics/computer guy here from West Mi, have kept up on all your seasons of fan show down. But your power supply is working fine. Voltage supply will drop if the current draw is max, it is seeing the fan as almost a short circuit. You should look into how much wattage you want to push through it instead of voltage, P=V*I so you're pushing around 30-35 watts on that little fan, probably the max on your P.S. btw that is a lot of power for a fan haha
A laminated design will be your friend here.
Also, using materials rated for higher temperatures. Perhaps with a lower friction coefficient as well. It'll be a PITA to print though.
Take a look at the impeller setup on most aquarium pumps. That is one answer to achieving success with an impeller like this. I actually think you could really elevate things with your background. I wish I had the time to explore some of the ideas I have, but I would suggest looking at the system as just that, a system with several aspects to driving the fan. Those suggesting maglev are on the right track, but I don’t think the answer is a fully active maglev setup. I think by creating a channel and using the inside and outside of it with magnets would allow for it to stay in place, but allowing the internal portion of the channel to act as the driving function. I am guessing you are already following, but if not, I can sketch it and send it your way. Just let me know.
I have been wanting to make this (Hubless Fan) for the longest time, since the original Avatar movie. Then the S.H.I.E.L.D. Helicarrier set the ideas into overdrive... Those always made me think they would be awesome without a hub! Magnetic bearings was always my first thought.
If you even look at the impeller in an aquarium water pump, filter, etc. it floats magnetically, and it can be spun, so magnetic bearings and motor stator should be doable!
Ceramic bearing balls are ideal to replace steel. Might be the next Evolutionaries step.
I think one of the main sources of the noise is that the gaps between the coils are making this work like a siren
Hang on to that white lithium grease... you'll fall in love with it. It's so damn useful for everything
Skateboarders use an oil called Speed Cream, in the bearings, which is probably close to gun oil. The major problem with an outside bearing to a fan design like this, is it's too much bearing length. Maybe if you had 4 smaller corner bearings holding the fan in place? but I don't know how that design would look or work.
So true about the massive bearing track length on an RDP (rim-driven propeller) like this one adding a lot more friction. I was thinking about this. Your idea about four corner bearings is a brilliant one! We could also try the delicate minimum of only three bearing points of contact. Yikes! Of course, if we put this idea into a high-load application, a few more bearing contacts might be necessary. The futuristic novelty that is hubless wheels are designed with this concept, but in an asymmetrical arrangement wherein there are alternating bearings holding both sides of the rim that replace a raceway altogether. Unfortunately, without a continuous raceway to shield the system, this setup requires extra vigilance in its design to minimize entry points for moisture and small debris. Such foreign material tends to enter and stay more than it tends to leave through those gaps, increasing wear and reducing reliability. A PC fan, however, is in a very clean environment wherein a three or four bearing contact configuration could really shine!
Simple: Use four small bearings at each corner to support the fan blade. or use 8 small bearings, two for each corner on opposite sides of the fan frame. The bearings will hold the rotor in place, and should spin faster & be a lot quieter.
Fun to watch, but lube doesn't make bearings go faster it just helps to keep the bearing working longer. The contact point of an un-lubed bearing is already so small that you can only reduce friction further by basically eliminating all the contact points using mag lev or an air bearing such as high pressurized air.
I think you’d see massive gains by making the bearing tracks smoother
Unnecessary pixelation for the win. But in all seriousness, I'm impressed with your skill set, well done indeed.
Honestly, with some improved blade geometry, you could theoretically get the spiral to push inwards, I actually think this could work out really well.
I wonder if you could basically take the geometry of one of the better hub based fans and adapt it towards this.
Main concern right now is obviously you’ve sacrificed the outer diameter, which is where usually most of the power comes from,
But if you could, indeed put the electric motors in the corners like the other prototype, did, it might actually work better then a normal fan!
The reason those tines bent in is probably because A) the coils will have some amount of resistance and pushing 5A through them will heat them up softening the plastic and B) each wire in a coil will have the same magnetization and would repel from the other wires pushing on that tab if they are not glued together there.
Gun Oil (cleaning oils, WD40, etc. that do have lubricant properties) often contain solvents, detergents, and emulsifiers; depending on the blend of the manufacturer. Gun Oil likely can dissolve many types of glues, and plastics used in 3D printing. Higher viscosity lubricants means quieter runs (better lubrication and longer lasting) but slower due to the higher friction/adhesion of the lubricant - thinner viscosity being the opposite.
One thing I haven't seen mentioned too much in the comments is the truing of your rotor. The lubricants seemed to make the thing run more true, but the friction adds up very quickly in that case. A small metal race would help, but would require quite precise lathe work. Focusing on other things might make more of a difference but I know that bearing is causing serious losses
Some lubricants break down plastic so that might be bad.
Some lubricants have been reported to cause chemical reactions, or have the same effect. Basically it's bad to mix them.
I don't remember what bearings were used, but alternating roller bearings that are printed tend to work really well. In fact that type of bearing sometimes improves over time vs other bearings that are printed.
Hi, the IPA is probably dissolving the glue, causing the coils to unwind and loosening the magnets.
I think what is also limiting the speed is the switching frequency of the bldc motor controller, you can go only so much with adjusting the voltage and current, but at the end bldc motor being a AC synchronous motor, the frequency determines the speed, so u might have to take a look at that as well
As cool as mag-lev bearings would be I think that they're probably more complicated to get going than would really be worth investing the time to figure out. I think the more sensible thing to do would be to try to reduce potential sources of rattle, friction, and internal air movement inside the part. If you make changes to fix these issues and its still no enough, at leas when you decide to do the mag bearings, you have everything else dialed in for maximum gains. A lot of this likely starts with the manufacturing processes. Ideally you keep all of this within rapid proto-typing technologies you have easy access to, so all the suggestions of metal are likely infeasible for now.
I would start by trying to print the blades ring and track with smaller layer lines and a more consistent surface finish. Someone else suggested Syratech's ceramic resin, and that may be a great place to start. I know I've watched Integza (another youtuber focused on amateur engineering with mostly 3d printed stuff) use the stuff for a variety of strange applications so maybe reaching out to him to see what his experience with the stuff is like would be a worthwhile endeavor. As a bonus the stuff should hold up to he heat just fine.
You might also consider a loading port hole with plug to load the ball bearings into the track from the outer housing so that you can load the ring completely and reduce air movement and rattle as the fan tries to move the bearings around. I'm thinking something kinda similar to how ball-screw bearings are loaded. This would also give you an easy grease port as a bonus. Tolerances can be hard to hit, so another option to reduce air noise and rattle might be to use some mechanism to slightly compress on the bearings ring once you engage the mechanism. This could let you make a part with tolerances for loading, and then reduce the excess space once assembled for an overall more precise mechanism and fit.
Someone mentioned a way to potentially help the RPM gauge be more accurate, but I also think that for noise measurement you might want to try to RPM normalize so you know how much a given set of changes actually affected the sound. Its important to quickly see the max RPM for a given amount of power too, but I think there is value to separating out the noise component from the overall performance.
Anyway, I love the videos, this project is the coolest thing on the channel (to me) in a while, so I hope you keep at it until you create something truly amazing.
my man gone walked the whole walk! kudos!!
The phrase "my man gone walked the whole walk! kudos!!" carries a sense of admiration and appreciation for someone who has accomplished a significant feat or completed a challenging task. Let's dive into its various aspects and examine the details.
Tone and Context:
The phrase is filled with enthusiasm and positive energy. It conveys a deep sense of respect and pride for the individual being referred to. The use of exclamation marks and the word "kudos" adds to the celebratory nature of the statement. The context suggests that the person in question has achieved something notable, likely involving perseverance and determination.
"My Man":
The phrase begins with the expression "my man," which indicates a close bond or friendship between the speaker and the subject of admiration. It suggests a deep personal connection and familiarity with the individual, highlighting a sense of loyalty and support.
"Gone Walked":
The use of the phrase "gone walked" is an idiomatic expression that adds emphasis to the action of walking. The word "gone" implies that the individual has not only completed the task but has done so with great determination and purpose. It suggests that the person has embarked on a challenging journey and successfully overcome obstacles along the way.
"The Whole Walk":
The phrase "the whole walk" signifies that the person has accomplished the entire journey or task. It implies that they have not given up or taken shortcuts but have remained committed until the end. It could refer to a literal walk, indicating a long and arduous physical journey, or be used metaphorically to represent a difficult endeavor or challenge in life.
"Kudos":
The exclamation "kudos!!" serves as a congratulatory remark or expression of admiration for the individual's achievement. It acknowledges their efforts, resilience, and success in completing the task at hand. The word "kudos" is often used to give credit or praise to someone for a job well done, highlighting their excellence and deserving applause.
In summary, the phrase "my man gone walked the whole walk! kudos!!" showcases a strong bond and admiration between the speaker and an individual who has accomplished a challenging task. The expression captures the speaker's excitement and appreciation for the person's determination, perseverance, and successful completion of the journey or endeavor, emphasizing their unwavering commitment and deserving praise.
IF the hubless fan could be magnetically levitating within the frame, I'm sure it could be a completely viable fan design. If the hub is magnetically put in place and it stays that way both off and on, it'll be a game changer.
Try reducing the total number of bearings and have a front side race and a rear side race to reduce the friction within the races, and where the bearings run as offset pairs front to back in vanes designed to be angled to induce a cooling airflow for your inductor windings. These vanes could also house your permanent magnets for the impeller. Have you considered using equal numbers of long and short blades to see if that increases airflow and/or pressure?
Dynamic lubrication bearing using a shaft repair sleeve (timkenredi-sleeve) and engine bearing. A little messy but can take high rpms quietly.
Sleeve around blade carrier, bearings around sleeve. Hardened steel riding a bow wave of oil on white metal.
Torque Test Channel recently used bench top supplies, but needed MOAR POWER as well... Soooo, they ran two, together, in parallel. 👍👍
The lube I'd think would have the most promise would be rem-oil, or garage door lube. Both are basically the same thing, teflon based lubricants. Very thin, usually mostly evaporates and leaves behind teflon.
The bearing issue is largely that it has no preload. All engineered rolling element bearings have a minimum load requirement which can be various combinations of preload and external load, and that total load depends in part on the viscosity of the base oil in the lube and in part on the mechanical dimensions and materials.
I kinda want to see you make a hot dog cannon now, for every fan that just doesn't cut the mustard. I'm thinking a frozen Weiner might be a great death into 300mph speeds.