How’d They Do It: Levitating Orb Clock

It’s time for everyone’s favorite game: speculative engineering! An anonymous reader wrote to our tips line asking how the levitation system of the STORY clock is accomplished. We took a look and can tell you right now… that’s a really good question!

STORY: The Levitating Timepiece has more than a month left on its crowdfunding campaign but it’s reached more than 6x its $80k goal. The wooden disk has a digital time display in the center which is simply an LED matrix just below the wood’s surface. We know how that’s done: wooden veneer with a grid of holes behind to contain the LED light in a perfect circle.

speculative-engineering-levitating-orb-clock-thumbThe part that has everyone so excited is a levitating orb that makes a circuit around the face of the clock. It would be easy to guess how it works if this was simply sitting flat on a table (which it can do). But it’s further complicated because it still works when hung on a wall. Most of the DIY levitation rigs we’ve looked at use gravity as an integral aspect of their functionality. A coil is suspended above the object being levitated while a hall-effect sensor tunes the magnetic field to hold the object in place (neither touching the coil, nor falling away from it).

So how is this one doing it? Perhaps there are multiple coils responsible for the levitation, each with their own hall effect sensor. In this scenario, tilting the base to hang on a wall would put different requirements on the coils above and below the magnetic orb. That’s our speculation, what is yours? Does this reasoning hold water magnet? Is there a motorized mechanism inside or does a grid of coils address the movement of the magnetic orb? Let us know in the comments below.

If you’re looking to play with this phenomenon in your own projects, it seems you can buy a magnetic levitation device which exhibits similar properties. The video of this, found in the comments of the STORY Kickstarter page, is embedded below. If you do order one of these, we want to see a teardown!


60 thoughts on “How’d They Do It: Levitating Orb Clock

      1. I have no knowledge of Flyte but does anybody have a compiled source of Kickstarter projects and how many actually follow through and deliver versus being slick marketing and no substance? Remember, Kickstarter is not an investment or a purchase, it is an extremely speculative hope that they will eventually deliver what they say they will with little assurances that they actually will.

        1. When I did a kickstarter a few years ago I read an article on this. It said it will take you 6-8 months to deliver your project. Lie to your backers and say it will take 2-3 months. No one will back a project with realistic time frames but they will forgive delays.

  1. I guess there’s only one magnetic coil/hall sensor etc. which is moved behind the clock.
    If the videos aren’t fake, the ball moves way too smoothly for a lot of coils next to each other in a circle. The ball would more or less jump from one magnetic field into the other then?!

  2. OK, the way I would do this is to have a magnetic levitation coil on a spinning disk that driven by an encoder stepper motor to keep it in sync with the time.have the spindle be wide enough to accommodate the time display and bobs your uncle. I would use simple brushes on two circular pads. Wear wouldn’t be an issue because it would only make two complete rotations a day.

  3. As the second video shows you can levitate horizontally, but that’s pretty unstable. I think they use an angle below horizontal and push the levitating object further away, 60° or so. that way you should stay stable but you would have to move the mechanism closer to the edge of the clock. would explain why the ball is so far away from the edge.

  4. Assuming it’s not fake ( not necessarily a good assumption), I doubt the ball is solid – my guess is it’s a very light plated shell, to reduce the amount of force needed to overcome gravity

  5. The basic levitation devices are just electromagnets in a feedback loop. If you expect the object to be levitated will experience gravity at a different angle than 90 degrees you could account for that. Heck use an accelerometer in the clock so you’ll know the correction factor before the ball even appears to your sensors. I’d suspect it’s just a single levitation “base” on a circular mechanical track, anything more complicated is unessacary

  6. If it works in vertical, you need both an attracting and a repelling force, there is no way around it. Now, how can the ball be repelled? Three ways come to mind:

    1- Ball is active (stuff inside).
    2- Ball is a magnet.
    3- Eddy current.

    My money is on (3).

  7. I’ve never really paid attention to Kickstarter campaign, I know some are legit and have produced great things but some are BS, or worse, a legit product with the wrong people having access to the money. I find it funny when on their site it states “all or nothing” by April fool’s day. So, if they only had, say, $76678 they would , what, walk away? Either way, they have their money, so someone will have one these devices at some point. I think we should revisit this April 1rst 2018 and see what we’ve learned in a year’s time. If it’s real, it has to work with coils and/or magnets and I really dont see that hard to mass produce and come up with something before then. Could be as simple as an actual coil or coils and some type of magnet actually mounted on an extreme version of a moving analog clock arm of some type. So could it be a real .

    1. Kickstarter really is an all or nothing site, if they don’t reach the funding goal, and are instead a few hundred under, then all the money gets returned, and the project receives $0. This is completely standard for any Kickstarter project

  8. If the ball is a regular steel ball, or a permanent magnet of uniform construction, there’s no possible stable configuration where it could be levitated because any field applied would cause torque and spin it around, and then it snaps to the other magnet.

    If the ball has an uneven weight distribution and is magnetized, it’s possible to counter that torque and levitate it in a carefully shaped field.

    1. According to their info, the ball is “Chrome” (but, since that would create problems with magnetism, obviously, I assume they mean “Chrome on the outside”). Shaping the magnetic field of the ball (by using 3d printed “nano” magnets should make it possible to get a manageable behavior of the “floating” ball.

        1. Earnshaw’s theorem can only be applied on non moving/static cases.
          I’m pretty sure they use some kind of control system to keep the (probably) magnetic ball in balance of a strong permanent magnet.

          So I would say the power consumption is higher in wall mounted operation, and the power used to keep it from falling down is a result of the ohmic resistance of the control coils.

          And behind the clock is a mechanic moving the ball around the clock. Everything else would be extremely expensive, as it would be a linear motor stage WITH 3D positioning coils. This would be possible, but would exceed the possibilities given by the price. But it would be me personal favorite, as it would allow for multiple balls :D


  9. Ball magnet actively suspended with one electromagnet and pushed out of the vertical with two other magnets slightly below, possibly permanent magnets. Of course this means the hidden part needs to stay vertical (gear, balance weight).

    1. If you look at the other JPGs on that site you’ll see the front of the device. Dirty great what looks like a permanent magnet, with 4 coils inside it. That thought occurred to me when I read the ball sticks to the magnet when power fails. So the permanent does most of the work of keeping the ball stuck, and the 4 coils do the manipulation. Probably only turned on when the ball actually needs to move.

      1. Exactly my thought. The permanent magnet does the heavy lifting, and the small coils do the re-centering if it leaves the sweet spot.
        From the images, it’s really only 4 coils, 3 hall sensors in the middle (analog one’s for each axis) and two LM324 with some analog controlling. Interesting is the analog switch, maybe this one switches the analog signals from the halls to the controls, depending on ?polarity?. Idk
        52 € with amazon prime – I have to finish some other projects first before I start a new one (advise from gf) :)

      2. Just thinking about the 8 power transistors on the rear PCB, at the bottom. There’s 4 coils, so presumably 2 transistors each. But that’s not enough for an H bridge each. So how does that work?

        1. By using only two H bridges ;)
          Take a look at the 4 through hole solder spots just below the transistors! The two opposing coils are connected in series, including wildly attached diodes.
          It makes sense that the opposing coils are switched together, since the complete build is symmetrical!

          1. Ah! Thanks! Makes sense I suppose, since the ball is only going to be on one side or the other, between any two coils.

            Would be nice to stick a few extra balls on it, to see what happens. Are they just ball bearings?

  10. The power consumption on this thing would have to be pretty impressive (for a clock), especially with the snap-to-the-face-during-power-outage behavior. Which explains why it does anything when there’s a power outage.

    1. Blast! Yah hooked me.


      2.5 Watts is claimed. Also says ball will snap to face of clock during power fail even when in vertical position. That means position is permanent magnets in a ring with selected center electromagnet to levitate the ball from the face and impulses electromagnets to kick it to the next permanent magnet capture position on demand. 2.5 Watts is doubtful for pushing the mass of a solid ball away from the face and kicking it about with such great authority, so it must actually be a very light weight ball, not a steel ball… which accounts for it’s motion being so crisp and snappy at just 2.5W, and without great mass to correct for the ball position is more repeatable/predictable so good chance they managed to use just simple pulse timing instead instead of hall effect sensors, the 2.5W again.

      So each capture position is going to be a ring of permanent magnets with a center repulsing electromagnet for lift. One kicker electromagnet to push to next position. They spent a bunch of time experimenting with mag field strengths to keep it to 2.5W.

      That ball is going to be quite lightweight.

      Yah… dang thing could work.

  11. could it be that if you take a small neomydium magnet and attach it to the end of the second hand of a clock, then surround that magnet with an electromagnetic coil, it may be possible to use a Hall Effect sensor as part of a feedback loop to switch the electromagnet on and off quickly, which would contain the other disk magnet in a rapidly changing magnetic field consisting of the sum of its magnetic attraction to the arm-mounted magnet and its repulsion from the electomagnet

    1. I was thinking it was a permanent magnet in the center with four or more coils surrounding the permanent magnet all variably energized in the opposite polarity using a PID controller. Push/pull would keep the ball a fixed distance away and if the permanent magnet is strong enough then gravity could be negated. That assembly is then rotated together.

  12. I dont have strong enough magnets handy to test but wouldn’t it be possible to simply create ‘pockets’ of magnetic fields? you have a big strong magnet that attracts the ball, then fixed on top of that you have a smaller but equally strong (or maybe stronger? not sure) magnet to repel the ball, resulting in a spot inside the field of the big magnet where the ball can ‘hover’ on the field of the small magnet, and with strong enough magnetic fields this would work even if the lot of it is at an angle, because the ball is being held in position between the 2 fields working against each other?

    Then its a matter of either having several fields and sufficient strength magnets/lightweight ball to let it ‘flow’ between, or you just mount your double magnet setup on a spinning arm :P

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s