This clock uses a steel ball which takes 30 seconds to complete its travel through a race at which time it triggers a release allowing the table to flip the other way. In a years time the ball travels 2500 miles. From 1820.
I thought this was a fantastic mechanism.
That is super cool. I wonder what all the movement does to the runtime. Each moving part is swinging at the same rate as the pendulum.
A simple pendulum swinging needs to restore very little energy to keep the clock running. Adding additional pivot points must add a lot of friction.
While vacationing in London we had the chance to Visit the national Museum. While there I saw some pretty fantastic clocks. One was this regulator. It was running perfectly and exactly accurate. I couldn’t believe how slight the pendulum movement was. You can read about it and see it in operation in the attached video.
Just posted. Read all about it here: https://keveney.com/posts/cycloidal-gears-for-fusion-360/ Or, just get the add-in code here: https://github.com/mkeveney/FusionCycloidalGears Still no support for 'perfect print' gears, but I'm thinking about it...
Matt,
I know this is not Steve's Clocks specific but I found your Cycloidal gear add-in for Fusion here. (My admiration for Steve and his Clocks is something I'll write about later.) Just goofing around with your plug in, I made a small gear train for a clock. I really didn't plan on any use, I was just toying with it. I added a stepper motor and used one of the glass sensors from the Steve's Coup Perdu to home the minute hand. It all worked so well, that I continued adding to the project to make a ships clock. Using a ESP32 and a GPS receiver, it keeps perfect time and chimes the ship's bell pattern. I've had it running for a couple weeks now. Thanks for a cool contribution that inspired something satisfying.
Jim
It is a fascinating clock, but a horrible timekeeper. A tiny bit of dust changes the rate significantly.
I built a prototype tray for a similar clock years ago. It used a steel ball with contact wires to detect reaching the end of the track. An RC servo was used to tilt the table back and forth. A micro controller would control the RC servo and adjust the track tilt to keep the time accurate. The tilting track would ratchet the time forward. I realized that the clock could keep time by simply tilting the table every 15 seconds and the rolling ball becomes optional.