I had considered this, but it cannot provide the accuracy needed with the pendulum. This would require the placement to be at set intervals, but moving the pendulum weight a fraction of a millimeter would require the teeth to miniscule to the point that I donât think would be possible with acrylic.
I had thought about this, but I donât think it would run long enough. I want to run for at least a day between winding (more like two, if possible). And the rubber band would only allow an hour or two and would get worn out pretty quick (by my estimation).
If the rubber band was through a wheel spun say 50 times a 4" wheel would provide 50â of travel and could be wound very quickly. My old grandfather clock did only 4 feet of travel.
I have played around with rubber band powered toys. I might play around with that if my required torque is low enough with the acrylic clock. A rubber band would definitely not produce as much torque as a two pound weightâŚ
Now that I have all this scrap acrylic, I have been wondering if there is a way to recycle/reform it. Apparently, with the right heat and pressure, it could be reformed. But I do not have the machines or expertise to pull that off.
After hours of frustration and reprinting pieces, I finally go this clock together. I am amazed at the lack of friction compared to the wooden version. I knew it would have less friction, but this is incredible.
I am trying to simplify the assembly, but still keep the frame reliable. I doubt this will be the final version, because I already see room for improvements and the clock isnât fully assembled yet.
What is still left:
Trimming the brass rods (And figure out how to hold them in the frame â currently just held with tape.)
New pendulum design (Which already needs to be adjusted, because I realized it was not balanced. At least I noticed before I cut it out.)
The weight (Currently a 2 lb. bottle of water⌠yeah, still no inspiration there⌠but something out of this acrylic would be really nice!)
The reduction in friction will likely allow for a smaller weight (or a pulley system to all for longer run time).
Improvements I would like to make still:
Adjust the frame a little more, remove old brace holes not needed with new design.
Adjust the clock to run on a mantle?!?
Add a crank/key and ratchet to wind the clock (the original design attempted this, but it failed and was scrapped to focus on the clock work.)
I think it would take a major metal shop. I had a wax injector made using a base for keeping tar hot, but it had many issues (It is so weird how one looks without eyebrows when you give them a hot wax job)
I was particularly interested in how prime numbers were used to provide the number of teeth needed.
Glad to see you got a working machine again, and are still plugging away at this. Also glad to see the improvements when you switched to acrylic - but now that has me wondering about using acrylic âfacesâ for the gears, i.e. a wooden inlay with a fine acrylic perimeter. You might curse me for that idea later - feel free, go ahead! In all seriousness, thatâs actually part of my plan/question below.
So - I had a question that I thought you might be quite well qualified to answer - and if you already covered it, I apologize. I did skim back thru the earlier discussion and couldnât see it.
I was watching Clickspringâs Byzantine sundial calendar videos and pondering the gear tooth profile used. The Antikythera mechanism teeth are simple sawtooth. Modern gears use very different profiles.
Finally, the question - have you experimented with different tooth profiles? I have cut a few types to play with on a test board, but Iâve never built a mechanism like you have. I am, however, working on a project (a rotary engine cutaway) that would benefit from a low friction gear arrangement. I need it to run on as little power as possible (hence also thinking of acrylic/wood composite gears as mentioned above.)
I did early on, before I knew as much as I do now⌠I was originally using a more âsaw toothâ profile, but switched to my current profile for a couple reasons. The first reason was purely cosmetic; I simply preferred the aesthetics this profile, but then I continued with it because I realized that it gave me a wider range variance with the distance between the gears.
This variance is still quite small, but it can be used to account for minor axle shifting.
I should probably look at different tooth profiles again, now that I know and have experienced so much more.
I actually have thought about this. But I deemed it more work for little benefit at this point. It is just easier for me to cut the entire gear out of the acrylic and have less waste.
Another suggestion I read, but I havenât tried, was cut the gear out of wood and then coat the tooth edges with super-glue. This is supposed to give them that âacrylic-edgeâ benefit.
Made a trebuchet in welding class, itty bity lil thing stood just under 2 feet tall. the basket weight was just under 25lbs. it could throw a golf ball just under a 100yds. with tweaking just over 110yds. It was awesome.
Interesting turn of events⌠the acrylic clock appears to be suffering from the same problem as the wooden clock?!?
I got it all put together and was surprised that it still needed the larger of my two weights. Then I let it run (it started running immediately and I was excited), but then after about half an hour it stopped.
Since the clock is made out of clear acrylic I can see every piece (unlike the clock made of opaque wood) so I can see that my cross-axle is more a problem than I originally thought. The axle does not sit in the center of the hole. It wiggles around to much, so the gears are actually too close.
So it looks like I need to adjust the frame again and move the axle holes further apart to allow for the variance in distance.
Edit: I felt the need to add how cool it is to be able to see inside the clock. I can see every gear connection and every interaction. So I donât have to âguessâ at what is happening in the areas that couldnât see before.
I see what you did there 
In all seriousness, thatâs actually part of my plan/question below.
