Workshop detail header
Workshop notes

Mechanical Bitcoin movement

Build logs, machining experiments, watchmaking notes, and lessons from trying to make a mechanical watch shaped by Bitcoin.

This is the main record of the project: what worked, what failed, what changed, and what I learned along the way.

2024/10/31 15:52 UTC
TLDRI finally cracked the code to a smoother watch build by reordering the process: setting wheel positions first, then tackling joints, and wrapping up with wheel rotations. This tweak, along with a gear ratio fix that clicks perfectly, has brought my creation to life, especially with a new curved case that fits like a glove. Shoutout to Alberto Sicco and Charles Fontaimpe for saving my mainspring barrel mishap—next up, pinions and jewels! 😊
Show more Show less

Refactoring Adventures: Building a Movement from the Ground Up!

It took some serious tinkering (and a few dozen rebuilds of the train of wheels!), but I finally found a smoother approach—first setting up the wheel positions, then handling the joints, and finally rotating (or folding) the wheels into place. This way feels way more intuitive!

I also revisited the gear ratios in the train of wheels. Now, the hour wheel fits perfectly with the center wheel pinion—such a satisfying click when it all aligns! And yes, I've finally added the seconds hand to complete the time-telling trio.

The case got a makeover too! It now wraps comfortably around the wrist instead of just sitting there like a flat disc. A little curve really makes it come to life on the wrist.

Next up, it's time to add the pinions to the wheels and get those jewels in place—so exciting! And after that? The bridges! Can't wait to see it all come together!

I gotta give a shoutout to Alberto Sicco, Charles Fontaimpe from generaleressorts.com, I had a pretty bad mistake in the mainspring barrel, thanks to his comment I could fix it!

😊


sketch sketch sketch sketch sketch

Read standalone post

2024/10/30 16:20 UTC
TLDRTried a new wheel design just for fun and it's pretty wild! 🚴‍♂️ Check out the sketches to see the funky concept that's shaking things up. Catch the full scoop on my blog if you're curious!
Show more Show less

Because why not

Testing out a quick different design for the wheels!

sketch sketch

🤯

Read standalone post

2024/10/30 12:27 UTC
TLDRHey! The blog post is about a watch design process update, where the layout now accurately shows hours, minutes, and seconds after a month of work. While some parts like wheels and jewels still need designing, the author is thrilled with the balance wheel's progress. The main challenge is aligning the center wheel pinion with the cannon-wheel, crucial for smooth operation. 🤯 Interested? Check out the full details [here](https://tbw.watch/post/view/previewing-a-potential-layout/).
Show more Show less

This is Exciting!

This layout is finally set up to show hours, minutes, and seconds correctly. Can you believe it's taken about a month to get here? Just one more detail to nail down (see last pic)!

Current Status:

  • Wheels aren't designed yet, so they're still looking solid and rough.
  • Bridges and jewels aren't in place yet either.
  • Unexpected plot twist: I'm officially in love with the balance wheel ❤️
sketch sketch sketch

Here's the "But"…

The center wheel pinion is where the magic happens for the coaxial hands. It drives a secondary wheel, and the cannon pinion is friction-fitted right on top of that secondary wheel.

So, what's the issue?

I'm struggling to get the dedendums of the center wheel pinion to line up properly with the cannon-wheel, especially since its axis is actually the fourth wheel. Gotta figure this one out to keep everything aligned!

sketch

🤯

Read standalone post

2024/10/28 17:43 UTC
TLDRPutting all the components together revealed a crucial design flaw: the current sizes of the wheels don't allow the cannon pinion to fit properly. 🛠️ This means a redesign is necessary, specifically adjusting the teeth counts, to ensure everything functions smoothly. It's a good reminder that sometimes you need the big picture to spot the small but important details.
Show more Show less

Seeing everything together uncover errors

These wheels make all the ratios that will display hours, minutes, seconds in the same axis

Im really glad I layed them over together

why?

See that wheel with the curvy cutouts? well the cannon pinion fits in it, and with the current sizes I wont be able to fit the two wheels, need to rework the teeth counts!

sketch

Read standalone post

2024/10/25 16:21 UTC
TLDRI'm refining my watch movement design by following NIHS standards with some help from ChatGPT and Alex at Watch Repair Tutorials. Key focus: precise mainspring dimensions and a perfect barrel arbor fit to minimize friction, plus a smart choice of a bride rapportée for better automatic winding. Stay tuned for more as the prototype phase gears up! ⏱️
Show more Show less

Following NIHS - Refactoring.

Been closely following NIHS standards, with the help of chatgpt, and Alex from watchrepairtutorials.com calculated precise mainspring dimensions:

  • Thickness: 0.13 mm
  • Length: 551 mm
  • Width: 1.33 mm

Additionally, I ensured the correct fit of the barrel arbor, with a total height of 1.58 mm to fit smoothly inside the barrel, maintaining a clearance of 0.1 mm on both the top and bottom to prevent friction. I've also decided on the use of a bride rapportée (inserted bridle) for its superior control in automatic winding. The technical details I've explored, including the use of Nivaflex mainsprings and calculations based on NIHS 11-02 standards, bring us closer to a well-engineered movement.

Stay tuned for more updates as I push forward with the prototype phase!

Barrel arbor, Mainspring barrel

sketch

Escapement (yep missing the safety pin) -- target: 28800 vph

Im thinking of going with KIF, but need to explore more

sketch

Train of wheels -- with proper reduction for the fourth wheel and minute wheel

sketch

Read standalone post

2024/10/25 13:39 UTC
TLDRThe blog post is about fitting hands onto watch cylinders made to a specific standard (NIHSg 24-10), with the wheel mechanism still being a work in progress. It's a peek into the technical details of watchmaking and the ongoing process of refining the mechanism. If you're into the nitty-gritty of watch design, this one's for you! ⏱️
Show more Show less

Hands fitting.

The cyclinders where made to NIHSg 24-10

The wheel mechanism under the cylinders is still WIP

sketch

Read standalone post

2024/10/21 06:32 UTC
TLDRThis blog post offers a quick look at the top and bottom incabloc components, modeled based on details from the incabloc website. The author is hoping to find a source to procure these pieces, showcasing sketches labeled sus 100.12.257 and sous 103.20. If you're into watchmaking or horology, it's a neat peek into the specifics of these watch elements. ⏱️
Show more Show less

Quick exploration of top and bottom incabloc.

I modeled these based on the incabloc website, hopefully I can source them from somewhere:

sus 100.12.257

sketch

sous 103.20

sketch

Read standalone post

2024/10/20 14:03 UTC
TLDRSwitching from 36000 vph to 28800 vph in watchmaking simplifies the process by reducing the need for additional wheels, and aligning with NIHS standards eases this transition. The exploration involves fine-tuning various components like the balance wheel and hairspring, with ongoing adjustments to optimize performance. The author is considering component choices like Incabloc and fendue virole, navigating decisions between laser welding and pinning techniques. 🕰️
Show more Show less

Exploring 28800 vph

It seems 36000vph, forces me to add a third and fourth intermediary wheels to the train, Im not sure I like that.

I decided to explore 28800 vph, following NIHS rules actually simplifies the process (NIHSG_34-04, NIHSG_34-25, NIHSG_35-10, NIHSG_35-15, NIHSG_35-20)

Please note that due to my exploration, these values are constantly changing

Balance wheel Inertia 12.5mm
Balance wheel Diameter 10mm
Hairspring D 6mm
Hairspring d 1.3mm
Hairspring h 0.18mm
Hairspring P 0.17mm
Hairspring K 2.65 10^-2 N·mm³/rad

Overall view with a candidate of Incabloc 100.11.310/0* , though maybe it's a bit thick 🤔

sketch

Decided with a fendue virole (NIHSG_35-20), still need to figure out if going with laser-weld, or Pinning with a stick

sketch

Balance and wheel (NIHSG_34-04)

sketch

Balance that follow NIHSG_34-25 standard

sketch

Read standalone post

2024/10/17 15:35 UTC
TLDRHad a breakthrough day fine-tuning my watch movement with a little help from ChatGPT, nailing down the ideal specs for the balance wheel and hairspring. Weighing material options like Glucydur and brass was a challenge, but I made a smart choice, and now I'm on the hunt for the perfect hairspring. 🚀 Things are really coming together!
Show more Show less

Fine-Tuning My Watch Movement: A Day of Progress

Today was a significant step forward in designing my 36,000 vph watch movement. With the invaluable help of ChatGPT, I tackled key areas like refining the balance wheel and hairspring, choosing the best materials, and fine-tuning my calculations.

  1. Balancing the Wheel and Hairspring

    For my movement to function at 5 Hz, I needed the right balance wheel (8.25 mm, 6.3 mg·cm²) paired with a precise hairspring. After some detailed calculations, I determined the ideal CGS number (K) to be 22.43 dyne·cm²/rad (or 224.3 N·mm³/rad in SI). ChatGPT helped break down the complex formulas and get everything just right.

  2. Material Selection

    Choosing between Glucydur, beryllium copper, and brass was tricky. Glucydur offers superior stability, while brass is easier to machine. With ChatGPT's input, I weighed the pros and cons and made an informed decision for my balance wheel.

  3. Perfecting the Numbers

    Precision was key, so I worked through the NIHS standards to ensure my balance wheel tolerances were spot on, with ChatGPT guiding me through the calculations for tolerance classes like js7. This gave me confidence in my design's accuracy.

  4. Sourcing

    Im trying to source a hairspring now that I know the numbers, hopefully I find someone willing to sell them to me.

TLDR:

Today's progress, supported by ChatGPT, brought me closer to realizing my high-frequency watch movement. From calculating the perfect hairspring to choosing materials, everything is falling into place!

(yes, chatgpt summarized all our interaction today, hehe)

sketch sketch

Read standalone post

2024/10/16 15:47 UTC
TLDRIf you're into watchmaking, using 301 steel for mainsprings and Nivarox for hairsprings is the go-to choice. Creating a mainspring from scratch involves stamping, annealing, hardening, tempering, and coiling the metal. Even if you're not sure how to do it yet, it sounds like a fun challenge to tackle! 😃
Show more Show less

Chatting with ChatGpt regarding some materials

TLDR; 301 steel will work for mainsprings and Nivarox for the hairspring

It'd be cool to make it from scratch, e.g the mainspring, I'll need to:

  1. Stamp the Piece of Metal
  2. Anneal the Metal
  3. Harden the Metal
  4. Temper the Metal
  5. Coil the Mainspring
Do I know how to do this, Nope, but sounds like a challenge 😃

Read standalone post

NOSTR:

npub1nvhq9vgkpahlugwcf3jsnpaq5lj3w3rk6lknnvf9r8h8ft2ja2asaqcn2a
npub1nvhq9vgkpahlugwcf3jsnpaq5lj3w3rk6lknnvf9r8h8ft2ja2asaqcn2a
Copyright 2025 The bitcoin watchmaker