Making a Totally Silent LEVITATING Fan Using Lego 

how to improve this thing: don't use a reed switch

Everyone mentioned the problems with number of magnets and the coil etc. but I think we are missing a huge point here; the propellers are reversed, they are looking the other way, so even though they will push some air, they are efficient in the other direction due to their curve. Swapping those the other way without changing anything else would increase the moving air incredibly. Nice project, cheers

Idea for improvement: Use a transistor or silent relay to switch the main circuit with the reed switch. This might reduce the binding effect because you can use minimum current through the reed switch and fine tune it with one less limitation.

Could you use three magnets instead of two on the base? Make it less likely to pop out by putting the floating magnet between a triangle of opposing magnets

U can use a reed switch but it should be connected to a transistor and the output of the transistor can be connected to the coil. This can prevent the lock up of the reed switch on passing high current as the high current only passes through the coil leaving the reed switch isolated

how does this not have more views? this is epic! improvement idea: try different fan blade styles and magnet sizes!

on the front end put 2-3 magnets above. in the back put 2 weak magnets two prevent friction at the end against the lego stopper. also 2-3 magnets above to prevent wobble and stabalize the axel.

I misread the title and thought you made a silent Lego hovercraft, but this is was a happy surprise, and makes me very jealous about how well you can balance magnets. 👍

In addition to adding a third magnet to constrain the shaft radially you should add a magnet on the end of the shaft opposite of the blades - and a magnet on the thrust structure to maintain zero contact before and after thrust

Nice project, it looks cool. Maybe use 4 magnets each side and a second coil on the opposite side of the first one. I think the reed switch is ok, maybe play with the best position or put a second one in parallel to increase the on time.

Wouldn't it be better if you use atleast 3 coils around the levitating shaft with an ESC and you can also increase the magnets on shaft

You need to account for the weight of the fan and the force generated by the wind. Also, add a sharp point to contact point, possibly with a bearing system to help lock the rotation side. Should help you tune the magnetic repulsive forces to balance the additional torque.

Put four instead of two magnets on the rod, but with two north and two south polarity, that should stabilise it and give it a little more speed, but i dont know, if it would work then with the switch

Omg the way you drew the circuit is so beautiful

you could add suspension points on the rotating axle by adding the piece with holes in it then connecting that piece to the rest of the board for more stability. the hole will allow the axle to move freely whilst holding it in place with minimal noise due to a small amount of friction. hope this helps Jamie, great vid!!!!

Great idea and it will be more stable if you double the magnet from top to keep the force of the shaft equal than you can add more coil and cross magnet to have more power

2 improvements- extra restraint magnets above the levitation bases, and have the reed switch fire a higher current relay. Possibly an opto isolator.

You could use an arduino or any other microcontroller to simplify tuning predictable reactions for the timing/switching part and speed control in an all in one solution. If you do that, a Hall effect sensor input usage might be a more reliable alternative to the reed switch. Anything you do is going to be somewhat finicky, but with good programing and design you can push tolerances. One thing I noticed was that your fan blades were convex side towards the direction of flow. This flow choice is possibly less linear than reversing it. Reversing it MIGHT help to minimize seemingly random turbulence that derails your fan. Another high speed compliant addition would be a metal on metal contact pivot point such as a needle against a slightly concave or flat mirror polished disk.

My first thought is that you could save a few parts by attaching the magnets to the end of the fan blades themselves, so a three magnet structure or even more. It doesn't look like the attachment system for LEGO gives you a lot of options for that. So instead you could build a six-sided hub and have three arms with magnets on them and three fan blades alternating on a single rotating structure, the magnet arms would have to be longer than the fan blades, but that would give you some additional leverage. Then you can build a much shorter levitating shaft that is more stable, I'm picturing about the proportions of a ferris wheel. The supports on both sides would be tall enough that the fan would be able to spin above the coil. You build a structure at the bottom to hold two coils far enough apart that only one of them needs to be on at any given time, as a magnet comes around one gives it a little pull and throws it towards the other which gives it another little pull just in time for the next magnet to come around to the first coil again. It would still pull the same voltage and current. It would just be connected twice as much if you get the spacing right.

Great little demonstration of a pulse motor. You can also make a single coil pulse motor with a few transistors it will last a little longer then the reed switch. Cheers!!

I think using some type of very smooth bearing on the back would be helpful rather then simply having the axle resting on the backboard. That's probably your greatest source of friction, with air being the second, that's sapping away any power your set up can provide. Having that would also help to stabilize it and stop it from bobbing around so much.

I'd be interested to see what you could accomplish if you took an existing fan motor and tried to replace its bearings with magnetic ones. You'd start with all the power and efficiency of a commercial motor, and then all you have to do is design the suspension for it

2 more Magnets at 90 deg on a cross rather than a beam then you get 4 impulses instead of 2. Really cool contraption

Kool, add more polls and another coil at the top. Also use a delay timmer to increase the on-time of the coil.

How to improve?: Suspend the fan between four magnets (on each side)instead of let it levitate between two on both sides .. Use more coils for more force

Two more magnets will make the motor stronger. Also another coil placed exact opposite and connected in series with the first coil will make it stronger as well given there's two coils acting on two magnets at a time.

Dynamically balancing your shaft and fan assembly will reduce the wobble and hopefully stop the rotor from falling out of the levitation suspension. It should also improve the rotation speed.

Add additional magnetic confinement to stop the rotor from jumping out of alignment. Add a second magnetic thrust arm and another reed switch to get two pulses per revolution. You can do this by repeating things already done and still using reed switches. Replace reed switches with hall effect transistors.

Maybe you could use two reed switches offset from each other, one turns the coil on positive, one turns the coil on negative. That way you would get a longer "power stroke" without more current.

great experiment , using magnets this way reminds me of superconductors. two things come to mind to stabilize your output shaft: 1. using a fan with more blades will make it also more stable, but require more energy 2. use more magnets in the upper side to keep it from falling off.

to improve: maybe add weight at the opposite to balance the blades part (avoid wiggle)

4 magnets or more per end to keep it more centered and stable. haul effect switch as well and you could probably get it spinning pretty good

A more stable mount for the reed switch that gives repeatable and precise adjustment seems like it would help. Also building a front and rear cage for the rotor shaft that would keep it from being launched out of the stator would be helpful. Maybe make it 5v USB powered?

Awesome vid btw!!! With the fan blade out a bit further…. I wonder about it’s balance. Or if more repulsive forces would be needed from the base magnets closer to the blades.

Could you try connecting the fan indirectly through magnets? If you could get the fan stabilized on its own, you could try using weak magnets on the fan and axle to still “connect” the fan to the axle, but the weight of the fan may not bring the axle rotation down too much, while still getting the fan to move with the axle.

could use an optocoupler in series with the reed switch such that the reed switch pulses the optocoupler which in turn pulses the coil.

love all your videos man, your voice is perfect, your lighting is so cool and professional, your pacing is perfect.

Possibly putting two magnets on either end of the shaft on top of the two or four of them that are already there and compressing together so the shaft was forced into a center position you could even try using the a magnet in the end of the shaft and another magnet on the base of it where it keeps it from moving laterally

Suggestion: move your coil slightly closer to the reed switch (or whatever you replace it with) and then angle the coil back towards center. This way it's repulsing the magnet more radially and (hopefully) more work is done for the energy used.

keep the angle stuck on the reed switch put 2 more magnets to keep in place add 1 more coil and reed switch for a 2 pull motor

How did I not know this channel existed until now. That is so cool. I am going to watch all of these videos. Keep posting man, this is great. Such a soothing voice too 👍

I seem to recall this configuration, but for your "bearings" use six magnets, three that repel the central hub and three that attract it, alternating them in a hexagonal pattern, i.e. with each magnet at a corner of an imaginary hexagon centred around the hub.

Make beams on the sides so the magnet wont slip out in a crazy way

Reed switch works perfectly fine if combined with a transistor so that the power goes through the latter and the former does not fuse

This thing cools more than one would assume from the fan. It also cools because it's so frikkin cool.

That reed switch will eventually wear out, especially if you are putting a relatively large current through it. Relays can have the problem with over current or motors that they fuse the contacts together, hence stopping it from working, you might have similar issues with a reed switch. The instant those contacts touch, which will be a very small area, there will be a large current draw by the coil. You would be better using Hall effect sensors like a lot of motors use, you would then need something to drive the motor though, you could make your own circuit with a transistor or mosfet or just use an off the shelf motor driver. Another possibility is to use a led and photodiode or phototransistor, you can buy them already built and it triggers when something passes between the led and sensor. Similarly an IR reflectance or proximity sensor could work well too.

You can use a reed switch to trigger a relay to get MASSIVE currents to go thru the coil

Orient the fan blades so they are rotating in the correct direction, also if you orient the fan blades in a way that lift/push the rotating assembly into the baseplate and ad a cup to hold the rotating assembly it should run more smoothly and securely.

Perhaps you could use something like a relay or mosfet, using the reed switch for the signal instead of handling the load 🤔

Use some ferromagnetic core in the coil for improved efficiency.. Also, add more magnets on the up position, front and back to keep the propeller from bouncing like that.

Perhaps a cup with a hole and a spring to each end to get rid of the wobble and placement during low rpm. Maybe make it longer and ad 2 more motors with different radius to control wobble. good luck and thanks.

Adding beefy diode oposed parralel to coil will greatly reduce the arcing in switch wich will make his life longer, also small capactior parrael to contacts can help reduce arcing.

What about having another suspension magnet set up front the blade, so that it doesn't wiggle too much? It'd need the whole setup to be higher, so that the blades are not stuck with the enlarged base, though.

Maybe you could stabilise the axle by adding magnets above. It could reduce the wobble and prevent it from popping out… maybe?

Make the far end of shaft pointed and rest it in a concave shape. That way it won't walk off the flat wall and it will have less resistance.

Adding a capacitor in paralell with the reed switch will help to reduce arking when switching DC loads - When the switch opens, the capacitor must charge up, temporarily leaving the contacts without a voltage drop.

Use more magnets to repel from the coil, that way you get double the pushes if you use 4 for example

Use 3 read switch and connect them in parallel, use high voltage, double the coil and magnet also

Aside from the already mentioned suboptimal blade positions, you could consider a better fan, e.g. hobby-grade plane or quad propellers

Needs a bit of balancing at some point. Some foil tape like that used for duct work may help do it. (Not to be confused with duck-tape, it has a heavier metal backing.)

To prevent the Switch from fusing, you can use a mosfet.

You might consider a 4 bladed propeller to improve balance.

I'm not sure if it's possible to do it while levitating but, What if you make it move using the design used in induction motors? (Commonly seen in water pumps) So basically, there are multiple coils around the shaft. That shaft is built as a squirel cage ( a straight cylinder made of two circles connected by metal bars that are tilted at an angle) This way, the coils induce current on the bars which then causes them to react to magnetic fields and boom, a motor This is likely oversimplified but there are vids that explain this wayyy better (As I wrote this, I realized you're using a similar concept anyways, but I decided to send it anyways because who knows! Maybe it'll inspire you in a way! But yeah, if it doesn't work, then just add more coils)

if you put a battery in the center of the Rotor with 2 Magnets for levitation, you can make it spin by short circuit both ends of the battery using a wire.

Everyone's recommending using a transistor in conjunction with a hall effect sensor, even though you already mentioned that in the vid lmao. However I've not seen anyone recommend putting an opposing coil. Rather than having 1 coil, you could use 2 on the sides as an opposing pair. Least then each time the coil fires, they're pulling equally from both sides and that should negate any wobble induced by the coils. Bonus is more torque too. Could even use 2 or more coil pairs for even more power but that may overcomplicate it. Good work lad, this channel has serious potential!

If you use a capacitor is series that has the same values as the coil you get more torque out of the magnetic field.

try adding more magnets around the magnets to stabilize the rotaiton

You should put a box around the levitating magnets, except for where they are necessary!

Add more magnets, now you have the most basic setup, the more pairs of magnets you have the more frequently the coil pushes the magnets, also you can use a reed switch but make it turn on a transistor so your reed switch wont weld itself, and also you could get away with much much smaller coil, as for the power delivery you can still use the signal from that one reed activated transistor to turn on/off a pair of smaller coils sitting close. if you have lets say 5 pairs of magnets as your rotor you make it so the polarity alternates - the coils are apart exactly how far apart magnets are and the coils alternate in polarity too, now you can extrapolate it in two ways. Use one rotor and for each magnet you got one coil, or you make a second rotor and second pair of coils, and if the shaft already spins the rotor acts as a flywheel, you can hook up the transistor gate to one of the coils so it opens and closes exactly as fast as the rotor spins, now the only thing to do is speed control which you can implement as potentiometer which you put in front of the transistor gate, hide the circuit with bricks (also, use a sturdy li-ion 18650, pick one from philips/sony/samsung etc., you want the current to not be the limiting factor, but also you need a way of limiting current if you use one of them big boi batteries, they are quite powerful and will fry the circuit if you dont have a way to throttle them and limit their current). I hope that i provided at least some ideas for anybody! Sorry for the wall of text

Put magnets on the upper side as well, and a magnet on the rear to counteract the reaction forces. Mass manufacture in Asia and Bob’s your uncle. Nice job!

I got an idea to possibly stabilize the axle a bit Maybe also put magnets (in the same distance) above the axle, just like the ones underneath the levitating axle. This might stop the front from wobbling around so much, but also could make it work way worse In my opinion you already have great results anyway though-! Keep it up

Add some relays and you can increase the motor voltage without nuking the reed switch

Waiting for a V2! If you want to buy a fully silent desk fan though the newly released Arctic Summair is probably what you're looking for. It's made by a PC cooling company so they know how to make a silent fan. Infinitely variable speed and you can even get a version with a 25 hour battery and USB-C charging, plus it's very cheap too! Edit: it also comes with a 6-year warranty, forgot to mention that.

Airflow is what produces noise… so yeah, you always gonna have noise the more aire you want to push.

Another set of levitating over-the-top to keep it from vibrating so much

Make the axle longer and do mount a second coil and switch but the arms with the magnets 90 degre from the first pair

Damn, I love these videos! I dont understand how this guy doesnt have more views and subscribers.

dont forget to add a reverse polarity diode to the coil and use a hall effect sensor with a transistor instead of the reed switch

I wanted to hit Like twice! :D You could use a transistor so the reed switch doesn't have to handle the full current of the coil. Remember to use a diode so the coil's back EMF doesn't kill the transistor. Instead of a reed switch, you can use a small coil. When a magnet passes it, a small current will be created which you can amplify with a transistor. Maybe a MOSFET if it's very small.

Very cool! I'll bet you could design and 3D print a single-piece fan blade that is far lighter than the one you built out of Lego pieces.

You should use 4 more magnets to stabilize axle, and how about use 4 magnets in the center, to have more pushes per revolution

"how to spin to spin something wirelessly: alright first here are some wires..."😭😂

Add a thrust bearing at one end of the shaft and lighten the shaft by moving the coil closer to the magnets. And uh yeah. Get eid of the reed switch.

I love these videos, i can't wait to see you channel grow 👍 just subscribed

I love your videos gives me inspiration thank you and keep it up 😁

You could just use a high energy laser to heat a metal plate. The got are will then rise and can be funneled towards you, for a nice breeze. But beware of metal splatter if you run it too long tho.

My theory on improvement idea - more driving magnets. The impulse is a touch off vertical and it looks like that upsets the stability of the motor core, sometimes enough to push it outside the stable region of the bearing magnets. I think having three or four magnets rotating past the coil will help in a few ways. More pulses per rotation - smoother rotation, more mass - acting like a crude flywheel, and it allows more power. If you’re up for some harmonic analysis, there’ll be a certain rotation frequency at which the impulse will offset the vibrations, acting as a damper. And some other frequencies at which it will amplify them. This is a neat project and I’m really happy the algorithm brought me here

A reed switch might be fine if you can control the currents trying to break it.

How to efficiently avoid a reed switch: Add a camera and a very bright light (for fastest possible shutter speed), write a program in INTERCAL that determines the position of the magnets every millisecond (might require a very powerful computer, bonus points if it's made of lego and if it uses some sort of AI), and then pipe that into another program written in Malbolge that applies current to the coil based on the position of the magnets.

You could replace the back plate for the axle with yet another pair of magnets

Is there a way you can relay any energy produced by the rotation to stream it back to the battery and keep it charged so that you'll achieve perpetual motion? And is it possible to use another magnet at the back pin and another magnet instead of the Lego backplate to counterforce and keep it from pushing off the horizontal based magnets that help levitate it . This way there will be no friction and as long as the battery can get charged through the rotation energy produced then it will spin forever. Oh and hook up a solar panel if it will help the battery stay charged .

that thing really needs a harmonic balancer. those propellers are not meant to spin that fast, and were never balanced to begin with which is part of why it wobbles. that and the shaft

Main issue I would solve is power. The reed switch fusing is severely limiting the magnetic force you can output from the coil. My (admittedly armchair engineer self's) idea is to have two circuits: one with a small power supply for the reed sensor which triggers a relay, and finally a circuit that has the coil connected to a bigger power supply. The relay would connect/disconnect the power from the coil. This all means that you can have a reliable reed sensor with a much higher voltage going through the motor. I do welcome constructive criticism btw

supper cool I sorta want to make one of these, not much to it all I'd really need is the magnets

For better levitation stability perhaps the base magnets on both the front and rear sides can be spaced out a little further. This could expand the magnetic fields farther apart, providing better side to side movement resistance. Adding more central propulsion magnets, more copper coils, and expanding the Reed switch to cover a longer length could multiply spinning power outputs. Using a three way architecture for the central propulsion magnet arms (and all the rest) could improve power output. Using a two bladed fan could reduce drag caused by the propellers, (and result with less air flow output😅). Oh, and I agree with the other dude down there talking about fan blade arrangements. The fan blades must be angled correctly to encourage airflow output to go to the front of the device, and not to be propelled backwards towards the rear of the device. Thx G2769!

Housing around the fan to have sort of like a chimney effect would be a low-tech improvement

Use another set of magnets to keep it from sliding sideways and upwards from the fan blade wobbling

I think you need to make the fan steady and also need low resistance for that I recommend ballbearing

Instead of a reed switch, use an led and photocell. The photocell would have to be shielded from external light using a tube. Place a rectangular card on the rotor to interrupt the light between the led and the photocell at the appropriate timing. You may have to build a breadboard circuit to adjust the sensitivity of the photocell. This should give you a contactless timing switch. Sorry, I couldn’t think of one that used magnets.

add a transistor so that all the driving current doesn't have to go through the reed switch. (and a back emf diode)