Magnetic Bearings Might Keep This Motor Spinning For Millennia

We see our share of pitches for perpetual motion machines in the Hackaday tips line, and we generally ignore them and move along. And while this magnetic levitation motor does not break the laws of thermodynamics, it can be considered a perpetual motion machine, at least for certain values of perpetuity.

The motor that [lasersaber] presents in the video below is unconventional, to say the least. It’s not a motor that can do any useful work, spinning at a stately pace beneath its bell-jar enclosure as it does. The design is an extension of [lasersaber]’s “EZ-Spin” motor, which we’ve featured before, and has the same basic layout – a ring of coils wired in series forms the stator, while a disc bearing permanent magnets forms the rotor. The coils, scavenged from those dancing flowerpot solar ornaments, are briefly energized by the rotor passing over a reed switch, giving the rotor a little boost.

The difference here is that rather than low-friction sapphire bearings, this motor uses zero-friction magnetic levitation using pyrolyzed graphite discs. The diamagnetic material hovers above a rare-earth ring magnet, supporting a slender vertical shaft that holds the rotor and another magnetic bearing at the top. It’s fussy to adjust, but once it’s stable, the only friction in the system should be the drag caused by air in the bell jar. [lasersaber]’s current measurements of the motor running at slow speed are hard to believe – 150 nanoamps – leading to an equally jaw-dropping calculated run-time on a single AA battery of 89 millennia.

[lasersaber] is the first to admit that he’s not confident with his measurements, but it seems clear that his motor will likely outlive any chemical battery used to power it. Whatever the numbers are, we like the styling of the thing, and the magnetic bearings are cool too.

43 thoughts on “Magnetic Bearings Might Keep This Motor Spinning For Millennia

  1. Interesting artwork. I’m not looking into it at the moment… though will since if works as claimed/theorized would be an interesting mechanical electromagnetic supercapacitor improvement I’m thinking if the material science is sound. Similar thoughts sounds like regarding improvements in a vacuum. Thinking the higher mass as the rotor thought.

    Anyhow… mainly since you weren’t sure about a battery and might be an artsy combination… I was thinking a battery that would last really long and interesting would be adding a voltaic pile since there is a demonstration of one of those running for almost hundreds of years now days… 176yrs in 2016.
    https://www.youtube.com/watch?v=UtQGYz4f3YQ

    1. I am thinking if we hang a coil magnet and let the earth magnetic field move it to a NS position. And then energize a different set of coils so that the new coil magnet moves to NS position will we be needing lower energy to get it spinning. ?

      1. I didn’t get into that thought yet… will have to think about some more. Though was thinking batteries connected and aligned to the earths magnetic field to tap and more-so an antenna coil to either take advantage of the natural earths magnetic fields or some other more common signal source that is in the environment typically all the time so not having to worry about ionospheres effects (instead of solar so doesn’t need to be under a light source). I suppose an electrostatic motor and line on a kitoon or balloon is overboard.

        For some reason at the time I went off on a pendulum pump idea also to make a water fountain that would rotate throughout the day and need some input energy source to keep the pendulum going for a lifecycle or a couple lifecycles seemingly perpetuity… yeah… the later is a tangent and different application of the friction less bearing that I’ll jot down now since got me thinking again.

      1. “I believe this is what you’re describing.”

        Yes, when watching I almost immediately thought the system might be in a vacuum already and was thinking the artist came up with the design to function as since he notes regarding interest.

        If anything… the increase in flywheel size (as the rotor) might not hurt anything other than the additional mass will require more pyrolized graphite material and rare earth magnets. Simple is nice for less issues from hidden variables and more drag/friction maybe from minute balance offsets.

        “I am thinking if we hang a coil magnet and let the earth magnetic field move it to a NS position. And then energize a different set of coils so that the new coil magnet moves to NS position will we be needing lower energy to get it spinning. ?”

        If balanced really carefully like this has to be already… there might be the rotational momentum to take advantage of the earth magnetic field using the same system and timing more critically maybe? Seems somewhat already is and I wonder about the coils energizing and balancing issues if say more magnetic field on the north side for example.

        A magnet as the flywheel is already what the system has for a rotor and a magnetic attraction and even a repulsion if the poles on the flywheel can rotate somehow to aid in rotation I think also is like you’re thinking and might be something to look at where an orientation of the system will need to be considered in regards to placement. Almost could like somehow regenerative breaking charge the battery a little also and maybe increase lifecycle. Then again stability is going to be a concern and complexity with redundancy needed to assure failure from more variables in the system. The less variables the better and planning materials lifecycle performance also for worse case scenarios is going to be critical.

    1. He suggests pumping it down in the video.

      It’s fun to think of where the energy is going in this one. Of course there’s coil-winding resistance, but then he had to add 2 M ohm – 50 M ohm resistance to keep it from running off the rails. And wind resistance, but the thing is running very slowly. How much energy is lost in the reed switches opening and closing?

      I remember [lasersaber]’s work from a long time ago, when he started out as free-energy-curious, and then through building various real prototypes and through experimentation and experience turned into a super-high-efficiency motor builder.

      Nothing helps weed out the pseudo-science like actually trying to build stuff and make it work.

      And he makes great motors. I’m still totally in awe of that first motor that charges up the smoothing caps when run in reverse, and then runs for quite some time afterwards on the stored energy.

      1. Maybe he has found a way to leapfrog The Clock of the Long Now.

        An interesting side project for The Long Now might be to include new chronometer types, capable of 10000 years of time keeping, as they emerge.

      2. “I’m still totally in awe of that first motor that charges up the smoothing caps when run in reverse”

        This sent me scurrying to find the video. If that is “No Batteries! EZ Spin V2 Motor Running on a Single 1000uf 16v Capacitor” then physics is not broken, he runs the motor forwards to charge the cap. Cool build.

  2. Wow. Prepare for the video to smack you across the face at 0:49.

    It goes from a relaxing leisurely slow spinning of the motor to a bunch of random images BAM BAM BAM BAM BAM.

    Cool motor though.

    1. Yes, but the process is extremely slow. Dramatic temperatures, strong external magnetic fields, and physical impact will all quickly demagnetize a magnet (scatter the field), but with just everyday external forces, it’s not likely that the magnet will have changed in a measurable way in any of our lifespans.

    1. He shows a schematic in the video starting at 3:33. The battery and series resistance is connected in series with the coils and reed switch, and there’s a cap in parallel. I think the secret sauce is that coils alternate polarity around the ring, while the rotor magnets keep the same polarity. Plus there are twice as many stator coils as there are rotor magnets.

    1. You’re right! The ~35 mV is across the motor, downstream of the “throttle” resistor.

      So 35 mV * 0.15 uA = 0.005 uW for the motor and 24 megaohms * (0.15 uA)^2 = 0.54 uW for the resistor. That’s where the power is going!

      But you’re also wrong. 2200 mAh / 0.15 uA = 14666667 hours = 1,673 years. It’s not forever, but it’s fine by me.

      And that’s being conservative: the battery capacity is measured down to a voltage cutoff that’s way higher than what’s needed to run this motor. Given that the circuit will kick on 35 mV, you can run it on the deadest of “dead” batteries. Heck, maybe you could even lose some of the throttle resistance, making the ciruit more efficient. :)

      So yeah. The OP estimates 89 millenia, but he’s neglecting the throttle and imagining getting every last microjoule out of the battery. That’s not gonna happen. The calculation above uses the battery’s rated life, but you’re likely to get at least a few hundred years more out of it. Battery self-discharge is going to prevent any of that.

      How long do solar panels last? You don’t have to get much out of them.

      Also:https://hackaday.com/2018/02/25/mechanical-clocks-that-never-need-winding/

      1. And remember, he was clear that he didn’t know if he was measuring the current draw correctly. He seems genuinely interested in learning more, so hopefully he sees this post and the comments.

    2. “This long life ignores the power dissipated in the dropper resistors.” Yes, understood. Of curse, a AA battery will not last that long. Neither will the PLA plastic, reed switch, etc. What I was calculating is the runtime on 3.9Wh worth of energy. I just wanted to understand how long the motor would theoretically run on that amount of energy.

      1. FWIW: We ran a coin-cell challenge a while ago, and a number of people tried to do crazy feats powered just by a 2032 3V cell. This would have been an awesome entry — it’ll easily run for (five? ten?) years on a coin cell, with the lifetime of the coin cell being the limiting factor by far.

        On getting the most out of a (coin cell) battery: https://hackaday.com/2017/12/22/coin-cells-the-mythical-milliamp-hour/

        And whether it’s two thousand years or twenty thousand, it’s kinda moot. You’ll probably get the 40 years out of that battery if you’re super careful. But in the end, something other than the current draw is going to be the limiting factor: PLA disintigrating, an earthquake knocking it over, or just losing it when you move to another house. :)

        Great project. I love the whole aesthetic of getting the most visual impact (it’s spinning!) out of the least energy. It’s not magic, but it’s right on the edge, no?

  3. You can also go solar cells (e.g. from those leftover dancing flowerpots), with the (super)caps for the night time power bank.
    Or–if not already done–with the solar powered coils on the rotor and even a single magnet on the stator any depletable “outer” parts could be eliminated at all, although the whole thing becomes strictly light-dependent and would require careful balancing.

  4. For the “perpetual” version, the easy way to power it would be to use a solar cell, but if you want to be really cool, harvest power from variations in room temperature or air pressure. Store the harvested energy in a glass and gold foil capacitor.

    1. > harvest power from variations in room temperature or air pressure
      Like “The Beverly Clock” constructed in 1864. It uses variations in day-time night-time temperatures, to change the pressure inside a air-tight box which pushes on a diaphragm to provide the energy to operate the clock.

        1. You could put a large solid block of aluminium inside a vacuum thermos flask to act as a day-time/night-time thermal reservoir and place a well selected seebeck TEG at the interface to the air. Aluminium is not the best solid, AlBeMet (alloy of 38% Aluminium and 62% Beryllium) would be much much better, but the idea of dealing with any material containing beryllium from the Berylliosis perspective is unappealing. And I would ignore pure Beryllium altogether, unless you want to be added to a watchlist. Water would be much much better than an AlBeMet, but preventing it from evaporating over millennia is asking for failure.

          Aluminium has a specific heat capacity of 900 J/kg.K (@ 298.15K)
          AlBeMet has a specific heat capacity of 1465 J/kg.K (@ 293.15K)
          Beryllium has a specific heat capacity of 1925 J/kg.K)
          Water has a has a specific heat capacity of 4186 J/kg.K (@ 298.15K)

  5. This obviously has everyone’s gears spinning freely (Monday morning pun). If the power draw is as low as it’s thought to be, even with the resistors, this invites a deep dive into arcane but long-lived sources of power, from the Oxford Bell to the nuclear thermoelectrics that drive satellites.

    Temperature and pressure changes have been cleverly incorporated into mechanical clock-winding movements for quite some time, so what about other, neglected (too-small-to-power-your-screen-widget) sources of electrons?

  6. Next step: Magnetic bearings with a Radiometer In a vacuum, of course). With non magnetic materials for the vanes, you could essentially make it an ideal radiometer and thus essentially only affected by light. I wonder how long/fast that would work with a small amount of light and then darkness…

    1. Radiometers actually require a small amount of atmosphere to run as they’re driven by convection rather than photon pressure. I have no idea how your “ideal radiometer” would actually work – sure, you should get some photon pressure on the vanes, but I don’t know if any available materials would work well enough to get even a magnetic bearing to spin.

  7. Efficiency like that makes tritium power a bit of a disapointment. Although it could run on a small generator of that type, the tritium would half in power repeatedly far to quickly and peter out before the machine fails.

    The most likely killer will be changes in magnets involved. Such devices have to be very tightly tuned to levitate like that since the magnetic forces are huge compared to the diamagnetic forces involved. A temperature change or any decay in the magnets would get it out of balance and it would stop spinning quickly. Of course the proof is in setting it up and watching what happens, perhaps even playing with temperature to see how robust the device is. Having successfully managed to operate a Levitron, I have learned that balance like that is a pretty touchy thing.

  8. The magnets on the rotor will encounter drag from the earth’s magnetic field. How about putting a coil or coils on the rotor, along with a solar cell to provide current when the cell faces the right direction, an arrange it so the electrically created field works against the earth’s magnetic field to make the rotor spin? If the earth’s magnetic field isn’t strong enough you could add a magnet near the rotor.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.