-
Working magnetic levitation system: next project, a clock with a levitating hand
Posted by Dominiz78 on 02 Aug, 2026 16:25 -
Hi everyone,
I would like to share a magnetic levitation project that I have been working on intensively over the last few months.
The system is fully built and operational and is based on a custom-designed PCB. Unlike many simple DIY levitation setups, where a magnet is only suspended underneath a single electromagnet, this system uses active digital feedback control to stabilize the levitating object.
A Hall sensor continuously provides the position feedback used by the controller. The electromagnets are then driven with corresponding PWM signals to correct the position of the object in real time.
In addition to the hardware, I developed a dedicated analysis and control application.
Current features- Real-time display of the Hall sensor value and PWM signals
- Live adjustment of the PID parameters Kp, Ki and Kd
- No recompilation or firmware flashing required for controller tuning
- Dynamic adjustment of the target position and levitation height
- Integrated calibration and diagnostic functions
- Direct observation of the physical response to controller changes
One possible use would be an interactive educational platform for control engineering. The influence of the PID parameters could be demonstrated directly on a real levitating object instead of only through simulations and diagrams.
However, I am currently considering a more visual follow-up project:
A clock with a levitating hand
The idea is to build a clock where the hand floats freely above the clock face without a visible mechanical connection.
The levitating hand would have to remain stable while its angular position is controlled so that it indicates the current time. One of the main technical challenges would therefore be to generate a controlled contactless torque without significantly disturbing the levitation.
At the moment, the clock is only a concept. The magnetic levitation platform itself is already working.
I would be very interested in your opinion:- Would a clock with a levitating hand be an interesting next project?
- Do you think it could become more than a technical demonstration?
- How would you approach the contactless rotation of the hand?
- Do you see a better application for the existing levitation system?
Other possible applications could include:- An educational control engineering demonstrator
- A levitating product or exhibition display
- An interactive trade-show exhibit
- A kinetic art installation
- A physical laboratory platform for control experiments
I am also considering releasing the firmware, PCB design files and control application as an open-source project if there is sufficient interest.
Technical criticism, alternative concepts and honest feedback are very welcome.
Best regards,
Dominic
-
hey dominiz, that clock is gorgeous -- really great looking. you're light years ahead of me on these concepts, but i enjoy nice design and workmanship!
-
is that first photo real? how is it possible?
-
It looks like magic, beautiful, lovely balance of repulsion and attraction forces. The levitating hand clock, would be not just a great demonstration, but a good product. If you release your work Open Source, I am sure there would be interest, especially in the control algorithm and the mathematical model of the system, unique.
-
Very neat, definitely interersted
-
The concept looks good, but a clock with only one arm would be hard to read. Also, might be a little too power hungry for a clock.
Don't hang it above your bed, so in case the clock's arm falls it won't poke your eye.
-
Yes, the first photo is genuine.
The reason this setup works is that the permanent-magnet ring does not exert force on the levitating magnet in only one direction. Its magnetic field influences the magnet in all three spatial dimensions.
By carefully selecting the target position of the levitating magnet, it can be placed relatively close to the magnet ring. At this position, the upward magnetic force generated by the ring is sufficient to counteract most of the gravitational force acting on the levitating object.
The electromagnetic coils are then only required to actively stabilize and regulate the magnet’s position. They continuously correct small deviations from the target position rather than carrying the entire weight of the object.
As a result, the system requires only a very small continuous current for the control electronics, power semiconductors and stabilization coils.
The only image that is not real is the image of the clock itself. That image was AI-generated and serves as a conceptual design for a possible future project.
-
If you look more closely at the AI-generated concept image of the clock, you can see several small recesses in the clock face. LEDs are intended to be installed behind these openings to indicate the minutes.
Regarding power consumption, the active stabilization itself requires very little energy. I currently operate the system at 9 volts, where it draws approximately 0.05 amperes, corresponding to a total power consumption of around 0.45 watts.
This low consumption is possible because the permanent magnets support the weight of the levitating object. With a properly tuned control system, the electromagnetic coils only need to make very small corrections to maintain the desired position. In fact, the microcontroller currently consumes more power than the active stabilization itself.
I do not intend to install the clock directly above a bed. Nevertheless, the levitating hand will be designed to be extremely lightweight and made from a carbon-based material, most likely carbon fiber. Therefore, I consider the risk of anyone being seriously injured by the falling hand to be extremely low.
-
on adding a minutes hand, i'm guessing you could have the magnet/s and control area of one hand be on a more centrally located circle under the clock face while the other hand's magnet/s and control area could be in a circle towards the outer portion of the clock. each hand would be cantilevered relative to its control magnet, but as long as they're made of light material, i'm sure that wouldn't matter.
your clock hand would be totally passive -- aside from a permanent magnet, all the sensors and control are hidden beneath the clock face to avoid having to change batteries or use inductive charging, right? -
Or you could make it in the style of an astronomical regulator, where the three hands have different centres of rotation.
-
Very cool idea.
-
That looks super cool.
The clock idea sounds great. However I wonder how annoying it would be in practice. How robust is the trapping? Is the hand going to fall off when a heavy truck goes by or someone slams a door? I can imagine it being annoying if every time i looked at it the hand had fallen down. -
keep the hand still and rotate the face
-
I’m very glad to hear that you like the project. The control system is actually very stable. Even a strong gust of wind or significant shaking of the base plate will not cause the hand—or rather, the levitating magnet—to fall.
A power outage is, of course, a problem. In that case, the magnet immediately falls down and the system cannot recover automatically. The user then has to manually place the magnet back into its levitating position.
-
Interesting thought, but then you lose the ability to read the time intuitively.
-
I’m very glad to hear that you like the project. The control system is actually very stable. Even a strong gust of wind or significant shaking of the base plate will not cause the hand—or rather, the levitating magnet—to fall.
A power outage is, of course, a problem. In that case, the magnet immediately falls down and the system cannot recover automatically. The user then has to manually place the magnet back into its levitating position.
If the net force of the permanent magnets is attraction, and the balance coils provide the repulsion, in a power outage the hand would stick to the face. Not sure how feasible that is. -
This behavior is actually already built into the current design. In the event of a power failure, the levitating hand does not fall to the floor. Instead, it is attracted toward the clock and simply lands on its surface.
This is not caused by the electromagnetic coils pushing or guiding it in any way, since they are no longer active when the power is lost. It is purely a result of the specific magnetic-field characteristics of the permanent-magnet ring.
-
Is this the same project as was posted over here ?
A clock with a levitating hand
https://forum.arduino.cc/t/a-clock-with-a-levitating-hand/1454139
The idea is to build a clock where the hand floats freely above the clock face without a visible mechanical connection.
The levitating hand would have to remain stable while its angular position is controlled so that it indicates the current time. One of the main technical challenges would therefore be to generate a controlled contactless torque without significantly disturbing the levitation.
At the moment, the clock is only a concept. The magnetic levitation platform itself is already working.
The idea is great, cogging effects may be a nightmare.
Maybe a special ball with slightly magnetic multiple poles could allow some stepper-like control, whilst also managing the levitation.
Or you could shift the problem, by levitate of a static micro stepper* and have that move the hand(s)
A small swinging weight could give a correcting moment and give the illusion this is a pendulum clock
There is also this work
https://physicsworld.com/a/new-type-of-magnetic-levitation-makes-its-debut/
that uses a spinning magnet as part of spin-stabilized levitation.
* google finds this
https://nfpshop.com/product/4-6mm-stepper-gear-motor-16mm-type-model-nfp-4616-36-mg
that weighs 1.5g, and includes a 36:1 reduction on the 20 steps per rev, so that could manage Hr/Min/Sec hand movements as needed, without needing active power of micro stepper drive.
for other ideas :
Some 'floating' clocks use visible dangling gears for min to hour reduction, and some have a free hour hand that is dragged by a hidden rotating magnet ring. -
Impressive project. With such a unique and unprecedented solution, I would like to understand how it is even possible to shape the field to eliminate 5 or 6 degrees of freedom? I am interested in the arrangement of magnets, the shape of the field, I would like to know how it works. Do you have any drawings or simulations that describe this? How are the poles of the clock hand magnet oriented during operation? If the clock hand is a magnetic dipole, it will always have one degree of freedom - rotation around the axis. In the case of the clock hand, this is undesirable. Write something about the problems you solved while creating this project, if it's not a secret.
-
Yes, that Arduino forum post is also mine. I posted the project in another German forum as well, simply to gather as much feedback and as many different ideas as possible.
I hope the cogging effect will not be too severe, since I am planning to use air-core coils without any iron cores for the rotational drive. Of course, there may still be some torque ripple caused by the coil geometry and the magnetic arrangement, but at least there should be no conventional cogging caused by ferromagnetic stator teeth.
The miniature stepper motor is definitely an interesting approach. The main issue I see is supplying it with power while it is levitating. It would either need some form of wireless power transfer or its own small battery. In addition to the motor, the floating assembly would also require a driver, control electronics and a power supply, so it could become larger and heavier quite quickly.
Best regards,
Dominic -
Thank you very much for the kind words. The clock itself does not exist yet; it is the next stage of the project. What already works is the magnetic levitation platform with its digital control system, including operation in a vertical, wall-mounted orientation.
At the moment, not all six degrees of freedom are constrained. The position in the X and Y directions is stabilized actively by the electromagnets, while the height and the tilting around the X and Y axes are mainly determined by the field geometry of the permanent-magnet ring. Rotation around the Z-axis, assuming Z is the levitation axis, is still free. So you are absolutely right: with the current arrangement, the levitating magnet behaves essentially like a magnetic dipole and can rotate around this axis. For the clock, this is precisely the remaining problem that still has to be solved. The final magnet arrangement and pole orientation of the hand have therefore not yet been decided.
Once the project is further developed, or ideally once the clock is working, I plan to document it properly and make the information available to the community. I have also noticed that there are surprisingly few complete DIY projects for a standalone magnetic levitation platform. It might therefore be useful to publish the platform itself as a separate DIY project, even without the clock. There were quite a few problems along the way before I managed to get the system to levitate reliably, so documenting those challenges and solutions could hopefully be helpful to others as well.
Best regards,
Dominic -
Hello everyone,
thank you again for all the great feedback over the past three days. The many ideas and suggestions have encouraged me to actually build the project and document the development publicly. My goal is to involve as many interested people as possible while also receiving as much technical feedback as possible.
I have also shared the project in two other forums:
https://www.mikrocontroller.net/topic/586503
https://forum.arduino.cc/t/a-clock-with-a-levitating-hand/1454139
Which platform would you recommend for documenting the further development? YouTube, Instagram, the EEVblog forum, or perhaps something completely different? I would also use GitHub for schematics, software, and technical documentation, but probably not as the only platform.
Best regards
Dominic -
You already have 4 large coils so some carrier modulation could provide power.
The miniature stepper motor is definitely an interesting approach. The main issue I see is supplying it with power while it is levitating. It would either need some form of wireless power transfer or its own small battery. In addition to the motor, the floating assembly would also require a driver, control electronics and a power supply, so it could become larger and heavier quite quickly.
A miniature stepper with gear reduction has appeal as that can (hopefully) hold position when unpowered.
That slashes power needs to low duty pulses.
There are many 3mm and smaller MCUs that could extract step instructions and pulse small drivers.
-
I think this thing would make more sence in art installations. We read time by looking straight at the clock's face , people would not even notice the hand is floating unless being told. At an art show however people can walk to the sides and see floating elements. Does not have to be a clock, can be whatever artist's imagination creates.
-
Isn't the watch a kind of art or designer object?