10/15/2025 - Mechanical Clock Project -- Brain storming

Today I spent time brainstorming the concept for my stepper-driven mechanical gear clock. I sketched out initial ideas for the gear layout, motor placement, and overall design. I also explored how to integrate the stepper motor with the gears for precise movement and considered materials for laser cutting. It was mostly planning and idea exploration, setting a solid foundation for the build ahead.

Summary in a few words: Brainstorming, gear layout planning, and stepper motor integration ideas.

10/16/2025 - Received RTC battery and stepper drivers, ready for setup.

Today was all about brainstorming and nailing down the setup. I collected the last hardware (RTC battery and stepper drivers) and used the time to digest feedback from friends on gear arrangement. They gave useful suggestions for gear ratios and mounting, which helped clarify spacing and load concerns.

I’m still deciding which board to use — Arduino Uno/Nano or ESP32. The ESP32’s Wi-Fi is attractive because I want to build a custom control app for the lights and stepper, but I’m weighing simplicity and available I/O versus connectivity.

One friend suggested using a DC motor, but I don’t feel comfortable going that route for precise timekeeping. I’m leaning toward simplifying the design: drop the seconds hand and drive the stepper with one-step-per-second updates for reliable, quiet motion and easier gear ratios.
Next steps: finalize the microcontroller choice and translate the gear feedback into a concrete layout for the laser-cut plates.

I also switched from Fusion 360 to SolidWorks, so tutorial hell, I guess.

11/23/2025 11 AM - Clock Redesign and counting gear ratios

October 27:
Today I finally got my act together and sat down to fully redesign the stepper-motor clock I’ve been working on. I wanted the inside to look cleaner and way more like an actual clock mechanism, especially the gear train.

The main goal was pretty straightforward: I wanted the second, minute, and hour hands to all spin around the same center point, just like a real wall clock, but powered by a cheap little 28BYJ-48 stepper motor.
Clock design V1:
I was planning to make the base square at first, thinking of making the main clock body a dodecagon (12-sided polygon) and making another layer with transparent Plexiglass to play as a diffuser for the LEDs

The first thing I had to figure out was what RPM to run the stepper at. After some research and testing, it looks like around 1 RPM is the sweet spot—it’s safe, quiet, and should help the motor last longer. Plus, it’s perfect for driving the second hand directly, which made everything else way easier to plan out.

After that, I mapped out the two main gear reductions:

-Seconds to minutes needs a 60:1 reduction.

-Minutes to hours needs a 12:1 reduction.

Those ratios are basically the soul of any mechanical clock. Getting them right is what makes the hands move at the exact speed they’re supposed to. I used compound gears so I could hit those numbers without needing giant, awkward gears everywhere. The idea is just smaller gears driving bigger ones, all arranged on shared shafts that loop the motion back to the center.

The next tricky part was making sure all the hands could spin on the same axis. For that, I went with adding more gears of the same ratio to make sure the second, minutes, and hours hands overlap on top of each other, being coaxial but not spinning on the same axis as not to interfere with their speeds.

I used a Gear Generator first to calculate gear ratios and get a rough idea about how the gears would look on top of each other, but it wasn't enough to get what I wanted, so I switched to normal calculations:
Main gear (driven by the stepper): 8 teeth
Compound gear 1: a 40-tooth gear and a 12-tooth gear on top of it, rotating on the same axis
Compound gear 2: a 48-tooth gear connected to the 12-tooth gear of CG1, having a 10-tooth gear on top of it
I added transmission gears of the same ratio to make the hour, minute, and second axes overlap with each other:

So three 30-tooth gears are connected with each other, and one of them with the 10-tooth gear of the CG2; the furthest gear out of the three will have the hours hand on it.

Two 12-tooth gears are connected with the 12-tooth gear of the CG1, the second gear having the minutes hand.

Because we are using 1 RPM, 1 rotation per minute is exactly what we need for the seconds hand, so it will be connected to the Main gear.

I also had to pay attention to gear sizes so they don’t bump into each other and can actually be made and assembled in real life. I tried to keep the tooth counts reasonable and the overall size compact. Laid everything out in 2D first before even thinking about cutting anything.
I'm most familiar with using Blender, so to get a rough idea about how the gear should be set up, I made a rough clock layout



(I first was planning to make the clock rectangular )

ClockConsept

Clock design V2:
I went through lots of variations for the clock body. At first, I was planning to make it circle-shaped. Still, due to MDF thickness limitations (5mm being the thickest), I had to scrap that idea completely because overlapping 10 5mm MDF circles on top of each other to get 5cm thickness isn't a realistic solution, it would cost too much, and the result wouldn't be satisfactory.

Before discovering Fusion's spur gear generator, I used Blender's gear generator to make gears (I didn't know much about Fusion ), and it was a very bad idea because Blender generates meshes which would work in Fusion it would appear as a mess.

Nov 22:
Clock design V2:
This redesign feels so much more legit than my earlier tries. It actually looks like something that could run smoothly for years.

I still didn't finish setting up the gears on Fusion yet, but it looks promising. I might do the LEDs on the inside, but I'm still not sure about that.
I'm hesitating as to what to use as an axis for my gears so they don't just float in the air. I'm thinking about using wood nails or something similar.

Also, here are the gear settings I'm using:

For now, the Hole Diameter is 5mm, but I might change that later.
I'm really worried about the friction when the gears start rotating.

The next step is finishing the design and turning these sketches into files for laser cutting.

All in all, super happy with how the gear ratios worked out. Writing this down mostly so I remember my own thought process later, but it also feels like the project just leveled up from “random gears on paper” to an actual clock mechanism.

11/23/2025 12 PM - Redoing the gear calculations

So, unfortunately, after rechecking my gear ratios, I noticed a big mistake I made:
The final RPM was 0.0167, which is enough to only turn the minutes hand but not the hours.
Input Speed: 1
Output Speed: 0.0167

So I'm gonna need to redesign how I'm putting the gears and everything else from scratch.
I didn't explain it well enough in my last journal, but here is what I basically need:

Our initials RPM (rotation per minute) is 1, so the seconds hand will turn once per minute, no problem here.
For the minutes hand, we need a 1/60 gear ratio, so I need the output RPM to be about 0.0167 (1/0.0167 =59.88, which is close to one hour)

For the hours hand, we need a 1/720 gear ratio (about 0.00138 RMP), but we already have the 1/60,
720/60 = 12, so basically we need a 1/12 reduction going from the minutes hand gears.

Stepper spins at 1 RPM → drives the seconds hand.

Seconds → minute reduction:
8t → 40t + 12t → 48t + 10t → 30t → 1/60 RPM (0.0167) → 1 rev/hour → minute hand.

Minute → hour reduction:
12t (compound with the 30t gear) → 36 + 10 → 40 → 1/12 of minute speed (0.0014 RPM) → 1 rev / 12 hours → hour hand.
(gear 1 + gear 2 means they are compound gears and have the same RPM. Example: 40t + 12t are compound gears.)
Input Speed: 1
Output Speed: 0.0014

I used evolventdesign Gear Ratio Calculator to double-check my calculations.

11/24/2025 - Trying to fix the gear design

After redoing the gear ratio calculations, I started redesigning the 3d model, but I encountered a lot of problems.
My biggest concern is how to fix the independent gears, so they don't just float in the air. I was planning to use some long wood nails as an axis at first, but then I noticed that some gears overlap on top of other gears' axis holes.
Also, the seconds and minutes hands would touch other gears while turning

One solution I'm thinking about is first changing some gear ratios, like making two 30-tooth gears, one 45-tooth, and the other 20-tooth (so 30*1.5 and 30/1.5 respectively)
The other one is to make a rectangle that would spin with the stepper motor, having the seconds hand, then making a weird L-shaped thing for the minutes hand, the hours hand is on the topmost, so it wouldn't interfere.

I really want to make the gears stand out in my design, so maybe I'll scrap the idea of closing it and just make the clock without the sides. I also need to make a small compartment for the electronics that can be opened at any time for ease of access.
I still didn't find an ideal solution for these problems, but I'll try to ùake something practical and eye pleasing.

12/18/2025 - Complete redesign

So with my old design, after lots of thought, I realised that when trying to have the 3 gear hands on the same axis, I would encounter lots of problems.
So my solution is to do 3 axes having one inside the other, I would probably do them in different-sized PVC or something similar.
I also wanted to make the outer clock body a gear or a circular shaped, it would fit more the theme and it would look cooler.
I would also need to use bearings,metal washers and maybe wood screws for the compound gear axis.

The structure would be separated into 3 parts, the top part where we would see the hands and the gears, it would be mostly empty on the inside to be able to show the gears rotating, the middle part will have the gears and the electronic components, and finally the bottom part where i would put the lights and have the hole for placing the clock.
Also, I'm concerned about maintenance, such as dust and changing the batteries, so I would leave a rectangle or something to be able to open it anytime.

12/22/2025 - Design Decisions and Final Adjustments

At first, I really wanted to avoid 3D printing and stick with laser cutting, but now it's clear that 3D printing is the best option. The tolerances, alignment, and friction issues just get way easier to manage with printed parts, especially for the shafts and small mechanical supports. So I decided to accept it and move forward instead of getting stuck trying to force a fully laser-cut design, which made me lose lots of time with plenty of iterations.

To reduce friction and make the movement smoother, I bought 625-ZZ metal ball bearings. They have a 5 mm inner diameter, a 16 mm outer diameter, and are 5 mm thick.

The overall gear layout didn’t change much, but I switched the gear module from 2 to 2.5.

One important thing I learned during this process is that gears shouldn’t just barely touch. They need to properly mesh and interlock. If the center distance is off, the gears either bind or skip, which completely ruins accuracy. The correct spacing is given by:

Center distance = (m(z1 + z2))/2
-m: The module of the gears
-z1 and z2: tooth counts

where m is the module and z₁ and z₂ are the tooth counts.
For example, with a module of 2.5, the 8-tooth gear and the 40-tooth gear should be spaced about 60 mm apart. Designing around this instead of eyeballing it made the gear train much more reliable.

For the coaxial minute and hour hands, I’m planning to use 6203-Z metal ball bearings. These sit between the minute and hour shafts and help keep everything centered while allowing smooth independent rotation. This is closer to how real clocks are built and avoids the wobble I was worried about earlier.
I didn't think of focusing on the 3 main gears at first, but when I did that to make my shafts coaxiale it was way better and easier to make.

To make all of this work, I ended up designing some slightly weird custom axis and shaft parts to keep the gears aligned and seated properly inside the bearings. They’re not elegant, but they do the job. At this point, functionality matters more than perfection.

I’m also not fully satisfied with the clock hands. They work, they look okay, and they don’t interfere with each other, but they’re definitely not the nicest part of the project. Still, I’ve spent a lot of time on this clock already, and I’d rather finish it than keep redesigning small details forever.

I didn’t use metal washers between rotating parts, which I probably should have. If I redo this in the future, adding washers would be an easy improvement to reduce axial friction and wear.

Finally, I’m most likely going to scrap the LED idea. It was cool in theory, but it adds extra complexity and doesn’t really improve the core mechanism. The focus of this project is the mechanical movement, and I’d rather keep it simple and actually finish it.

Overall, this project taught me a lot about real-world mechanical design, especially tolerances, friction, and why traditional clock mechanisms are built the way they are. It’s not perfect, but it works, and that feels like a win.

This project is definitely the hardest I've ever made, but I'm proud of what I did.

Step File:
Clock V7.1

dxf and stl files:

arch12
arch2
arch10
Spur Gear (12 teeth)
Spur Gear (36 teeth)
arch8
Detachable backplate
Component30
Component32
Component31
Component29
Seconds
Hours Gear
Minutes Gear
Hours
Component34
Component33
Component27
Component26
Seconds Gear
Component35
Minutes
Component28
Spur Gear (40 teeth) (1)
Spur Gear (48 teeth)
Spur Gear (30 teeth)
Spur Gear (10 teeth)
Spur Gear (12 teeth) (1)
Spur Gear (40 teeth)
Base
arch4
Top
arch11
Spur Gear (10 teeth) (1)
arch1
arch6
arch9
arch5
arch7
arch3

1/7/2026 - Side notes + Overview + Cutting/Printing necessary parts

21 dec 2025:
I had to rush some things up because Blueprint's deadline was approaching, and one regret I had was not making a compartment for the electronics.
I'm also very worried about dust because my design is open from all sides.
Another thing to note is that I didn't make a time-setting mechanism, so if the time deviates for whatever reason, I would have to manually move the gears or change the Arduino settings.

I would be using an Arduino NANO as a microcontroller for this clock.

I also need to get started with making the necessary code, which would be the easiest part of this project (hopefully).

I'm worried about the small teeth breaking, but because the speed is very low and there is very minimal friction, I don't think it would be a problem.

I didn't leave much interval between the components, which is a newbie mistake, but I didn't want to overcomplicate things.

The final look of the clock wasn't really the vision I had in mind, but I'm pretty satisfied with it

1/17/2026 - Small Issues + new changes + got the wood parts

(This journal got deleted twice for some reason; it didn't want to upload. So some details may be missing as I rewrote it.)
Jan 3, 2026:
I started changing the design again to use steel rods as an axis instead of 3D-printed ones because of cost and time and stability and everything. I was convinced that steel axes would hold up better, especially at that thin diameter.
I also went to get the wooden parts cut twice, but there wasn't any 5 mm wood, and the laser cutting machine was already full with other people, so I'm still waiting.

Jan 17, 2026:
I finally got my MDF parts; I'm still waiting on the 3D-printed ones. I'm quite satisfied with the result, although I made a mistake on the side arches, as they can slide into the clock body, but nothing some glue can't fix, but that's my bad.
The clock looks very good, but I'm scrapping the LEDs. I'd rather keep it as mechanical as possible at this point.
I also started working on the code and assembly of the electronics; I will start testing when i get the
3D-printed parts.
Also, I scrapped the idea of the steel rods as the deadline is approaching, although I spent quite some time redesigning everything.
I was scared of the 3D parts breaking because of the small diameter and long axis, and the steel rods would've cost way less and looked better

The bearings fit quite well with the gears, and the gears rotate well with each other, but I could've done a better job on the tolerances to make a tight fit, but as always glue can fix everything; it was my plan from the start.

CE3S1PRO_final big with support
CE3S1PRO_final with supports
CE3S1PRO_final with supports
CE3S1PRO_final with supports (1)

These are the settings I printed with and the gcode files. I'm a novice at 3D printing, as this is my first time, so make your own settings if you know better about the settings.

The printing will take about 17.7 hours in total. I asked for black PLA. The printer used is a Creality Ender 3 S1 Pro.

1/19/2026 - Assembly hell.

Hello everyone, this might be my final journal on this project.

I encountered a lot of issues, some of which were my fault and others not.
At this point I'm tired of this project, and I don't think I'm able to continue working on it.
The idea was great, really great in my mind. I always wanted to make my own clock that has exposed gears. My CAD work wasn't the best, but I tried doing my best, as this is my first big project.
But unfortunately the circumstances weren't favorable.
First of all, I finally got my 3D-printed parts, but I'm really stupid, so I didn't add any tolerance.
I was planning to change the CAD file for the other users and fix my own printed parts (because they cost me a lot) with a Dremel and sandpaper, as the parts didn't really fit together.
But now I'm not going to change anything in the CAD files.

Also, when counting the gear teeth and starting to glue them with the bearings, I noticed that one gear was missing from the MDF laser-cut ones, one with 10 teeth that almost made me give up on the project.

But then I thought I might as well just assemble what I currently have and then go another day to get my part done, but after some hellish hours assembling, gluing, and sanding, some 3D parts broke.
Which saddened me a lot, first because of the cost, and second because I wouldn't be able to work on this project anymore.

At this point I'm just going to work on the code and submit this project, which will probably get rejected...

But yeah, this is the joy of life; failure teaches a lot of things. This project taught me a lot, and I'm grateful to all my friends that helped me with design ideas.

The MDF gears turned out great, and I'm happy that I got spinny bayblades now instead of a mechanical clock :).
Might as well make a Beyblade arena for my next project lol.

VID20260119210032

Also, the arches for some reason didn't export well last time, so I reexported them.
Archv1

Final CAD file:
Clock V7

Arduino code:
sketch_jan19a

1/25/2026 - Getting back top work

So, since Blueprint's deadline got extended, I decided to finish this project because I really want to have this clock made in real-life.

First of all, I thought a lot about the previous problems I had, things I overlooked, and solutions to them.
I decided to completely scrap the idea of using 3D printed parts, which will make things harder, especially aligning the hands and stuff.
Also big mistake I made was not considering the gears changing the rotation direction, so I might have to add/remove some gears.
Other than that, I got two solutions for making the gear axis and main shafts:
1- Using 5mm metal threaded rods, and adding something so the gears stay still on them (Glue them or something)
2- I use MDF supports, like a rectangle that has 5mm small things going out of the bottom and top.

Still, I don't have a solution to make the hands coaxial.

I will start revising the 3D model and try to get it done as soon as possible.

Also, I forgot to do the GitHub repo and add the wiring diagram.

(Special thanks to the reviewer who encouraged me to continue working on this; I almost gave up on it)