Traditional mechanical clockmaking is an art that despite being almost the archetype of precision engineering skill, appears rarely in our world of hardware hackers. That’s because making a clock mechanism is hard, and it is for good reason that professional clockmakers serve a long apprenticeship to learn their craft.
Though crafting one by hand is no easy task, a clock escapement is a surprisingly simple mechanism. Simple enough in fact that one can be 3D-printed, and that is just what [Josh Zhou] has done with a model posted on Thingiverse.
The model is simply the escapement mechanism, so to make a full clock there would have to be added a geartrain and clock face drive mechanism. But given a pair of 608 skateboard wheel bearings and a suitable weight and string to provide a power source, its pendulum will happily swing and provide that all-important tick. We’ve posted his short video below the break, so if Nixie clocks aren’t enough for you then perhaps you’d like to take it as inspiration to go mechanical.
A pendulum escapement of this type is only one of many varieties that have been produced over the long history of clockmaking. Our colleague [Manuel Rodriguez-Achach] took a look at some of them back in 2016.
Many escapement designs would easily be 3D printed at a large enough scale, what is difficult is designing them properly.
It’s weird that I should see this now because I spent 6 weeks all day long in a special class to learn Mastercam software and the project I was creating with it was a Daniels coaxial escapement. This was just a few weeks ago.
I’m like this though mine is being made of tool steels stainless steel and aluminum and CNC machined using Mastercam.
Granted the model of escapement I picked I found out exactly how difficult it is to design one because it was over three and a half weeks just to physically draw it properly and functionally. And this is for someone who is a trained watchmaker with eight years of experience in 3D design and CAD. Granted some of it was unclear documentation from the source but my recommendation would not be to try that one.
The interesting part is that given advanced enough software, you could just specify the end result and it would automatically generate exactly what you wanted as well as track and automatically handle dependencies. In other words, you need not be an advanced user to be able to do this given a software solution that handles this either specific use case like this output or theoretically a much more advanced set of outputs as well.
That said, I do not know of any of either type out there at present.
There are actually a few escapement generators available online, such as this one http://demonstrations.wolfram.com/TheGrahamClockEscapement/ among others. There are also multiple available on thingiverse.
https://www.thingiverse.com/search?q=escapement&dwh=815b2c41c179c43
Yes, that thought exactly occurred to me after about a week of doing it.
The George Daniels co-axial escapement specifically though is very difficult to get geometrically correct because the drawings in his book leave certain details out that you need working knowledge to fill the gaps in with. Fortunately with my training in school as a watchmaker only a couple years ago and additionally Henry Fried’s book The Escapement, I was able to piece together workable details that were missing.
The drawing tolerances for that specific escapement class are extremely tight because there is more interplaying geometry then most escapements so it is a particularly difficult one to properly draw accurately where it will function.
The deadbeat escapement and others for clocks are much easier and orders of magnitude simpler because there are no details to the fork horn that need to be drawn, which are needed for a watch escapement.
The funny thing is if you have seen a technical drawing of the Daniels escapement it looks like it is something straight forward to draw but once you actually do it it quickly dawns on you that a lot is missing from the drawing and you have to really know some things that aren’t written down in one place to make it work.
The model works- see the first video linked on thingiverse.
this video is much more interesting then the one in the article
https://www.youtube.com/watch?v=gdnebVkVtOk
fun project
Hard yes, and fun because it puts a bunch of different skills to use.
Add to, it doesn’t take up a lot of space.
Thanks for the support guys! Make sure to post a make on thingiverse if you do end up printing this!
am certainly not the only one imagining mechanical nixies, now…
If the gear turns a generator we could see a mechanical nixie. That’d be one beast of a gear and generator tho.
Rotating discs with translucent orange numbers on, illuminated by a kerosene lamp? But it could be difficult to simulate “depth jitter” of the different numbers of a real nixie tube.