Can a mechanical clock be made on the Glowforge?

The red pivot has too much friction in that model. The wheel tooth is going to slide (scrape) for the whole distance of the mating flat. You’ll find in modern clocks that those would be sharply beveled and typically mounted at an angle to the wheel toothface. You would want to cant it 45 degrees or more. They’re virtually points of engagement (even with width they’re only a “line”) vs a flat sliding action like the drawing will result in.

I am mostly a watch guy where we deal with incredibly small tolerances but some things are carried over from clock making and then get tweaked to fit the smaller form factor of a watch. Pallets in watches are very often stones (artificial sapphires) which lend themselves to high surface polish, minimal friction and lubrication stiction.

The deadbeat movement should be fine for your purposes, I think it’s coming down to the friction forces you’ve got robbing the power from the pendulum.

Gravity escapements can be very dirty from a friction standpoint, they often have an order of magnitude excess power from the dropping weight vs what’s needed to power the clock mechanism. So that’s always an option. But you wouldn’t have learned nearly as much had you started there :slight_smile:

1 Like

I am trying to understand these suggestions. Do you have drawing of what you are meaning? I think I have seen sketches of what you are talking about… I will see if I can find some. (I am a visual learner…)

Are you talking about something more like this?

Where just tip of the tooth engages? And the impulse-face is larger and always engages (not allowed to slide past if the swing is too large)?

1 Like

Yes. Although the impulse face (pallet) ends are where you’ll want acrylic or at least an acrylic inset on the contact points. Notice in the drawing how the ends curve vs being flat so that minimizes the amount of contact time on each swing. The other alternative is that super slick tape I mentioned that someone here found on Amazon & posted about last year. (I bought some at the time but have no idea what I did with it :slightly_smiling_face:

1 Like

… I was playing around with my clock, trying to make sure nothing was catching and everything was smooth. And I discovered that I had snapped the cross-axle on the minute hand.

It was still turning, but when I went to add wax to the pivot point, the whole thing fell apart…

So I will be printing a new axle tomorrow (it is to late tonight; I am going to bed…)


I printed the new cross-axle in acrylic to see if it will reduce friction (also because I already had the acrylic on the bed since I was trying an escapement out of acrylic.)


How did it go? Better timekeeping?


I haven’t fully tested it. I want to wait until I had the new escapement worked out before I put it all back together and hang it on the wall again.

But the gears all seem to spinning well. I cannot attest to if it is better with the acrylic axle yet…

Here is the next escapement I will try:
It is still using the cross-axle to join to the pendulum, but if this version has issues too, I will switch to brass-axle join it with the pendulum under the pivot.


Wow. Thanks @jamesdhatch for the suggestions!

For this test I removed the face and the hour-gear (to reduce energy consumed in the drive-train). Before the video began, the clock had been running for a couple of minutes. By now, the clock has been running for around 15 minutes! Progress!

I will count this as a success if I can get it to run with the hour-gear installed, but for now I am just enjoying the quiet “tick” of the clock…

Edit: I am afraid to add the hour-gear back in… I don’t want it to stop…


Another thing I learned when I was testing out the acrylic-axle is that the kerf on the acrylic is wider than on the draftboard.

I can pressure fit the draftboard-axle without glue, but when I made it out of the acrylic it was loose and would not hold together on the top (where the kerf is widest.)

1 Like

Woohoo! Isn’t it great when you can work your way through iterations that result in success? :+1:


After my huge success with the pendulum running for around half and hour, I discovered I was incredibly lucky with how the escapement engaged escape wheel. It was “loose” on the axle and offset.

Once I tried to “fix” the escapement on the axle, I ran into all kinds of problems… But, now that I know how it is supposed to work, and I have experienced it, I know what minor adjustments need to be made! (I think…)

I am trying to redesign the escapement again, based on what I have learned, and what I have experienced. (I am getting ready to test version four… and I hope it goes well.)
Escapement-X-D - Copy


Good luck!


While attempting to test the escapements, the clock fell off the wall again.

It snapped the pendulum in two locations and damaged some the gear teeth where it fell… The frame held together this time, so no pieces went flying (unlike last time). But the clock just doesn’t seem to run the same either. I am sure something else was damaged and I just cannot see it…

It may be time to cut out a whole new clock. But I was planning on changing the pulley system anyway, so I will probably wait until I have that redesigned. :frowning:


Oh noes! :frowning:


Sorry, but sometimes you just need the right kind of nudge… :+1: