jes notes Index Gallery . Shaft passers Snap issues

2023-09-12

Last modified: 2023-09-12 20:03:17

< 2023-09-11 2023-09-13 >

Escapement

https://www.youtube.com/watch?v=Jvf0xXi8Ryg

Increasing the drive torque by 2x increased the tick rate by 32%, which is much more in line with what I was seeing in the simulation, but much worse than what I saw with the old 3d-printed prototype. Maybe the configuration of the springs in the old prototype was actually very effective?

I think I want to get a stronger hairspring and try again.

I made a replacement hairspring out of a piece of copper wire:

https://www.youtube.com/watch?v=Ue8T4eGKI3E

Yes, works much better. Now increasing drive torque by 2x only increases tick rate by 10%.

It looks like the geometry of the thing is not as important as the tuning of the weights and springs.

So what's next on the escapement? Probably transition to metal, and go smaller?

The escape wheel here is about 75 mm across. In the mock-up of the watch I have it at 13 mm across, which is about 1/6 of the size.

The root of the tooth on the escape wheel is about 5 mm long, so I could reach in there with a 1 mm endmill at 1/5 of this size, which is almost all the way there.

Why not go 1/3 of this size? And then that's halfway to watch scale, and we have, as a guide:

Frame thickness: 2 mm

Frame size: 33 mm * 38 mm

Balance wheel thickness: 3.3 mm

Balance wheel diameter: 27.3 mm

Escape wheel thickness: 1.6 mm

Escape wheel diameter: 25 mm

Barrel wheel thickness: 1.6 mm

Barrel wheel diameter: 35 mm

Pallet pin diameter: 1 mm

Although we can alter the barrel wheel to suit the module 0.4 cutter, maybe?

So the frame would be 2x 2 mm plates of probably aluminium, with 2 pillars holding them apart.

We'd need some finer thread to use for the drive weight. We need a method of attaching a hairspring.

List of parts to make:

Does cutting brass with a 1 mm endmill on the router actually work? Feeds and speeds calculator says, for full slotting with a 1 mm 1-flute cutter:

So this sounds possible.

So plan for this evening is to CAD an escapement model that can be made in metal, at about 1/3 of the scale of the 3d printed one I made today.

One potential problem is that at 1/3 of this size my pallet pins will only be 1 mm diameter, which means they'd be 0.5 mm at final watch scale. Is that too small?

Does the escapement still work in the simulator if we double the size of the pallet pins? No. And it's not very easy to find an alternative arrangement that works with larger pins. Oh, well, let's proceed for now.

Tomorrow

< 2023-09-11 2023-09-13 >