This person's build is amazing! Mostly photos, but still super cool to see how "from scratch" it is - if you're expecting "bought a few components from Swiss companies" then, yeah, it's not that!
I submitted this a week ago and sadly it sank without trace (poor choice of posting time I suspect), so I'm trying again. I have no affiliation with the builder.
I didn't have anyone specific in mind - but it seemed like a reasonable default assumption. In fact I think I landed on the post originally by thinking "I wonder how hard it would be to assemble a watch from parts?" and some cursory searching around that.
I enjoy a channel called "wristwatch revival" that covers servicing and repair of watches of various types - that guy often sends off for parts so it was a natural thing to wonder I think.
I've seen a documentary (youtube video) about this guy, he spends six months to a year building a single watch from his apartment workshop. Sure, each one will sell for several grand if not millions now that he's more well known, but still, that's some serious dedication to a craft.
Even the main spring? For some of the components it seems like a significant sacrifice in quality to make it yourself. I appreciate people can make their own, but I also think the obsession some folks have with in house movements is misplaced.
Right. Even George Daniels bought mainsprings and hairsprings (along with sapphire crystals and dial engraving, probably rose engine work?) for his 25 handmade watches.
> In total he (George Daniels) created 25 handmade watches which were completely made by himself (23 pocket watches and 2 wristwatches) other than the crystal, the mainspring, the engraving of the dials and the hairspring. One pocket watch remains incomplete and is not included in this total number of pocket watches produced by George Daniels.
> For some of the components it seems like a significant sacrifice in quality to make it yourself. I appreciate people can make their own, but I also think the obsession some folks have with in house movements is misplaced.
At a certain level, watches aren't about quality. Only exclusivity.
I would argue that pretty much any mechanical watch isn't about quality as a timepiece these days. Quartz watches work really well for keeping the time, honestly better than any mechanical watch ever will. But that's ok - I still love my mechanical watches out of sheer appreciation for the craftsmanship and engineering that goes into such things. They have their own appeal.
NASA still specify the best analog watch they tested (twice because when congress found out it wasn’t American they made them do it again only to get the same result).
If you are in a critical situation and surrounded by electronics an analog watch can survive conditions like power failure and temp electronics cannot.
Irrelevant to business people but as a safety feature and emergency backup something fully mechanical and accurate could save your life.
> At a certain level, watches aren't about quality. Only exclusivity.
Annnnd at this level I'm out.
I mean, cool engineering solving a particular problem, near impossible tolerances, astounding art/materials all appeal on some level - but "exclusivity" on it's own doesn't mean anything.
Yeah. I have a Seiko I found as a kid, got serviced and then it has sat in a drawer for 30 years. Probably junk now.
I use my phone these days. But if I wanted a watch I'd probably go for a Casio G-Shock solar - may or may not go for a variant with fancy GPS timing or scuba-proof. Features that take care of problems I might have.
Seikos are great. I just wish they'd put spring drive into a lot of their cheaper watches.
The majority of expensive watches, even handmade, are only so expensive because of brand...or the fact that the case is solid fucking platinum/gold. That's cheating!
I found a good channel called Chronova Engineering who made a few videos about machining the mechanisms by hand. They use a couple of off the shelf parts like tiny ball bearings for the demonstrator (instead of jewels), but almost everything else is hand-turned.
I can fully appreciate the amount of skill involved here. I'm currently on a journey to make a clock from scratch using traditional clockmaker's methods. I started from pretty much zero knowledge of even basic machining, let alone specific clock and watch work. I'm a couple years in now (working in my spare time between work and family obligations), and I can repair most types of issues with clocks and many watches. But I'm not quite tooled up and practiced in making the gears yet. That's the most major stumbling block currently to my ability to make a traditional clock completely from scratch.
One of my very vague goals is to build a clock, from scratch, without looking anything up.
It wouldn't have to look like a wall clock; just something that keeps 24-hour time within reasonable tolerances.
(It's been impossible to avoid knowledge of all clockmaking stuff, but I still avoid watching videos on how clocks are made. This is probably going to be something I do once the child is in college!)
I did this! I got to the point of building a ticking prototype, and I considered it "cheating" to copy any clock designs. I did use an established escapement design, though. Apparently involute gear tooth profiles aren't appropriate for clocks, but I don't know how to draw the "correct" type. I spent months designing it in solidworks, creating the 3D models and plans. I then cut out the gears from baltic birch plywood with a cheap scroll saw, then sanded all the teeth, which took forever. If I do it again, I'll just have the parts laser cut. About a year later, I bought a kit [0] which was very enjoyable to build, and I saw how my bearing design was wrong.
[0]https://wooden-gear-clocks.com/ascentclock
Latitude has a "natural" definition (sphere spins around, how big around are you spinning near the poles vs. the equator?).
Longitude has no "natural" definition!? How far away are you from Mount Everest? ...only works when you have a natural sight line to the peak of the mountain... If "high noon" at London is exactly (ummm...) NOW! ...and you go 1000 miles west, how much TIME has passed between your "noon" and London's "noon"? You need a good watch... and how can you determine "high noon" on the pitching deck of a ship at sea? ...maybe not with your typical pendulum clock!
It was such a mind-blowing experience to learn that maps (and thus CLOCKS!) were first a military and trade technology! If you have a good clock, you can get a good X-coord in addition to the "free" Y-coord. If you can get good X,Y then you can make a good map. If you have a good map, you can optimize your trade / military routes, etc, etc, etc.
Highly worth going a bit out of the way for a visit, was actually one of the highlights of the trip for me.
I had an opportunity to go there when I was a kid with my parents. I understood some of the significance, but I think now the clever details of the designs might be more accessible.
This weekend I came across this book while looking for something new to read. It may not be the best exploration of longitude, mapping, and trade as a technology but your comment brought it to mind immediately. You might find it interesting.
That seems like a fun challenge for sure. I’m so far down the other road that can’t opt for that, as I’ve been studying other people’s clocks quite extensively. It’s definitely obvious that these days, using modern tools like CNC machines and laser cutters makes clock making much more easy. So you have to find ways to make it more difficult, such as your attempt to do it blind, or my attempt to do it with 19th century methods.
If you liked this you might enjoy this series of video interviews with George Daniels, inventor of the coaxial escapement; he also made watches by hand:
In the world of industrial machinery it's not too surprising to have to make special fasteners from scratch, especially as replacement parts may no longer be available from the original manufacturer. Taps and dies are common tools, and single-point-threading on a lathe can also be used.
On the other hand, the world of industrial machinery seldom needs M0.6 threads.
Just trying to support it properly so it didn't bend away from your single-point thread cutter would be a challenge. And imagine how easy it would be to drop the part, and how hard it would be to dig it out of the coolant and swarf!
I think the challenge with making such small screws would be mostly around
- Making your own or acquiring cutting tools that fit in such spaces, e.g. able to thread up to a shoulder at that scale.
- Measuring your progress. I'm really not sure how you'd measure pitch diameter for such a small screw without exotic measuring tools. Perhaps you would just cut it to fit the mating part, which on it's own would still be difficult to check due to the subtlety of the feel of such a small screw going in.
- Cutting internal threads. I imagine making your own M0.6 tap would be quite tricky as you'd have to grind very small reliefs after making an M0.6 screw.
Definitely not using coolant on parts this small; small amounts of lubricant if anything. Supporting the part shouldn't be much of a challenge as you would be taking requisitely small cuts with very small forces. I'm sure tweezers and loupes/microscopes are your friend when making watch parts.
Yeah, it was not the screw-making per se that impressed me. I know that making screws is a Thing People Do. What impressed me was 1) how freakin' tiny they are and 2) the fact that he made all the screws, including the internal screws that aren't visible in the assembled watch. That is some serious attention to detail.
I wear mechanical watches, and find them not to be "outdated technology" as a lot of people assume. My automatic skeleton watch requires no winding, batteries, or charging, which makes it a more practical way to tell time than an electronic watch in my opinion.
I suppose one could argue that a solar powered quartz watch is even more practical for the same reasons... but the automatic is nearly the same in practice, and is a lot more fun, and looks cooler.
I don't wear a mechanical watch anymore (I found the calorie-counting for smartwatches made losing weight a bit easier for me), but I've always loved clockwork solutions.
Occasionally I'll fall into Wikipedia rabbit holes of pre-computer solutions to problems, usually involving lots of gears and springs and a lot of cleverness on the designers end. It fascinates me how much stuff has become trivial since the advent of the transistor.
If you'd like to read a comprehensive blog post about the intricate mechanisms that make a mechanical watch work, with amazing interactive visualizations, I recommend reading this blog post[1] by Bartosz Ciechanowski, a person famous on HN for his high quality blog posts that explain things comprehensively, with nice visualizations. (the #1 top post on HN right now is his new blog post about the moon [2]).
In the 90s a friend of mine made a watch that just showed the year. The year precisely, that is to say, considering leap seconds.
This is much harder than it sounds because the protocol used (DCF) transmits the leap year info spread out over a time span before the year change. So suddenly your stupid receiver + display needs memory and compute.
One of the reasons I love mechanical watches is that I've never tried to fully understand them, and the idea that there's a purely mechanical thing, that does something seemingly so simple yet magically and reliably, gives me a sense of wonder and appreciation for engineering and art.
I'm really on the fence about whether or not I want to read this amazing looking article or just live on in ignorance :)
I think that link will only improve your appreciation for how magical a mechanical watch is. Especially when you consider people have been making them or a few hundred years.
You're in for a treat then! Bartosz has a bunch of other insanely well done explainers on his website, with easily inspectable JavaScript too. Some of my favourites are the watch, the engine and the GPS ones, though they're all fantastic.
I'll admit I was aware that the motivation for this project was not likely to be timekeeping.
I guess there's something about fancy watches which makes me want to poke fun at them. I'm not sure I'd have made fun of a clock. I certainly wouldn't make similar comments about Ben Eaters 8 bit breadboard computers for example - even though the same thing applies there to a possibly even greater extent. It's an anti-elitism thing I suppose.
Anyway, the watch is actually very impressive and cool.
I submitted this a week ago and sadly it sank without trace (poor choice of posting time I suspect), so I'm trying again. I have no affiliation with the builder.