Neat. I had figured the hour hand had a tiny gearing mechanism in the connection between the two hands. A magnet is a much simpler solution.
An old acquaintance of mine designed and sold an awesome wristwatch that used two ballbearings attached by magnets. [1] Not only was it useful for blind people (it was a much better mechanism than the old watches with an openable face, which could get misaligned by touching) but it was also fun to play with, as you could spin the ball bearings around and they could get caught again by the magnet. The video of the person spinning the hour hand reminded me of that.
I was expecting a little gear at the center point as well that went backwards at a slightly slower angular velocity than the minute hand is moving forwards. But I found myself slightly disappointed that it's "just" a magnet. For some reason having the two totally different positioning mechanisms for the two hands feels less cool to me, even though it almost certainly is a simpler implementation.
In fact, if you watch the video in the article, it takes you through the 3 prior iterations. Hollow clock v3 had a gear that drove the hour hand from the motion of the minute hand (which is driven from its end).
The problem with that, at least with 3d printed parts, was the motion is "jerky". I built one and it never worked well. I'll be trying this version soon!
That was a big thing in the 1950s. The Jefferson Electric Golden Hour clock worked the same way. One was given away with every Philco TV set one year, and millions were made.[1] They're available on eBay. Real gold plating, even.
Yea - I'd assumed the hour hand had some sort of dead-reckoning compensating gear on the joint in the middle of the clock similar to a south-pointing chariot[1]. I'm actually a bit disappointed they used magnets since it'd be so easy to mess up the orientation with a bit of otherwise benign interference.
> I'm actually a bit disappointed they used magnets since it'd be so easy to mess up the orientation with a bit of otherwise benign interference.
That happens to a fair description of magnetic compasses as well. I suppose that’s why some compasses use a modern version of the south-pointing chariot (gyroscopes).
I wonder if they used gears to move the hand clock, if it just needs to be rotating at a constant speed... I think so, 330 degrees counter clockwise per hour? Since, as the minute hand does a full rotation, the hour hand needs to travel 30 degrees..
Sort of - you'd need to have some system for clock adjustment to fix for DST, clock drift and the like that would also need to be incorporated into the spin. So just a constant low powered spin counterclockwise wouldn't be quite the full solution.
Clocks with no apparent mechanism are known as mystery clocks. They were invented by French clock-maker and magician Robert-Houdin [1]. He exhibited his first mystery clock at the Exhibition of the French Industry of 1839. Many other makers have created mystery clocks in popular decorative styles of their day [2].
This year, Cartier took the idea to the next level with the "Masse Mystérieuse": An absolutely gorgeous watch with insane mechanics where the movement is the rotor.
Projects in this space tend to be fairly niche, geeky proofs-of-concept with a Woz-like fun factor but little mass-market appeal.
This, however, has plenty of both. With a little polish to the industrial design and a high-quality contract manufacturer, I could absolutely see this becoming a popular item in homes and on desks.
I think GP means a lot of these projects are just links that people click and go 'oh cool, but i'd never build that and can't get one'. Which makes them glorified blog pieces.
I'm sad that both hands aren't magnetic. It'd be much more cool if the whole hand assembly floated. I suppose the two sets of magnets might interfere with each other as they pass by. I wonder how far you'd need to separate them to make it work.
IIRC there was a rule from E and M class saying it's impossible to contain something using static charges alone. So presumably you'd need some high frequency sensors and feedback mechanism just to keep it stable. But I guess once you get that then moving it with time might not be that much more difficult.
Making stuff float with magnets is hard. You have to replace the permanent magnets with electromagnets and then have a while bunch of sensors and logic to drive the electromagnets. I can't see a way to make it work in a clock like this.
I think (especially considering the different needs for 6:28 PM and noon) that you'd need some orthogonal magnets set up - have one set of strong magnets around the center of the hub to keep it in place while providing a minimal amount of field interference to the magnets driving the hands. I'm not an engineer but I think any solution that starts with trying to suspend a clock using magnetic forces applied to the clock hands is going to be incredibly complex and hour-minute cross-overs would be extremely delicate maneuvers considering how much of a raw magnetic field you'd need to be projecting just to keep the clock face floating in air.
Ah, I see. What about those cheap "floating globe" or "levitron" desk knick-knacks? Do those maybe only work because they freely spin, whereas a clock could not?
They work because they have a reasonably large base and a reasonable stable thing floating. The clock hands aren't stationary, let alone stable, and since clocks are generally flat in one dimension you have pretty limited space to play with.
The only way I've thought of to make it work would be to have a fairly large "hub" at the center of the floating thing, with the two hands coming out off the floating hub. But that would be much less visually impressive, I think. The hub would have to be large enough that the movement of the hands is relatively miniscule in terms of weight distribution.
Oh wow, I've been wanting to build this for a while. But my idea was to make it wrist watch sized, and use LEDs around the ring to show the time. (Or use laser diodes and project the watch hands on the wearer's wrist).
I believe the LED option is doable but it draws quite a bit too much power, even if the LEDs aren't always on.
Other, more power-efficient approaches would be using a ring-shaped LCD or e-ink display, but they do not exist off the shelf and custom orders are expensive.
Or use a conventional watch mechanism to rotate acrylic rings.
Anyway, I brainstormed a ton of possible ways to make this thing, but in the end, it's really hard to make small things like wrist watches.
An old acquaintance of mine designed and sold an awesome wristwatch that used two ballbearings attached by magnets. [1] Not only was it useful for blind people (it was a much better mechanism than the old watches with an openable face, which could get misaligned by touching) but it was also fun to play with, as you could spin the ball bearings around and they could get caught again by the magnet. The video of the person spinning the hour hand reminded me of that.
1. https://www.eone-time.com/