Update 2: 8/2/2022:
Success!
I no longer consider this effort a work in progress. This design can really work great once you get it assembled and know what to do to get it running excellent. It is very satisfying to watch it start itself and runs very solidly and reliably. I got a full wind-up running continuously for 52 seconds! (another measurement ran for 48 seconds continuously)! So far, 52 seconds appears to be all the energy in the fully winded main spring made out of eSun PLA+. If I record a longer run, I will post it. It makes a clicking sound from the PalletFork toggling at roughly around 470 clicks per minute so it clicks roughly about 400 clicks per windup. The whole assembly is about as long as my hand and just over an inch wide. It is stable and makes an awesome desk toy!
I printed everything at 0.20 layer height. Here is what I found I had to do to make it work reliably: Use an axle on the escapement wheel gear that goes through both the front and back frames to help keep the front and back frame axle holes aligned and slightly overdrill the hole for the axle in the escapement wheel gear if it doesn't run smoothly and continuous to keep the escapement wheel gear from binding. I had used a 5/64" drill bit on all the parts that spin on M2 screw axles which allowed the parts to spin freely on the M2 screw without any slop but I think the escapement wheel axle needs to be just slightly larger than that (on the order of ~0.8mm extra in diameter). Test the escape wheel with a full axle through both the front and back frames and if you have problems, try drilling the escapement wheel hole for the axle with a slightly larger 7/64" bit if you are using an M2 screw for an axle (make sure the axle goes through both frames!) That's about it I think. This design wants to run! I found in my testing that none of the other axles (M2 screws) need to go through both frames to run reliably. I did not need to glue anything together.
Optionally, I believe you could drill out the escapement wheel with a 1/8" drill bit and use an M2.5x40 screw for an escapement wheel axle. You might even be able to use a piece of metal coathanger or something for the escapement gear axle as it doesn't need to tighten down against the frame (which can cause binding). It only needs to hold the escapement wheel up and keep it from tilting while keeping the front and back frames aligned.
I let the M2 screws cut their own threads into the frames. I didn't have an decent Allen wrench or M2 tap so a pair of pliers helped me turn the screws in to cut the threads in the frame. Many of these axle screw holes in the frames have an optional M2 nut slot on the backside to help secure them into the frame as well. I didn't need to use any washers for this assembly.
--
Below are additional notes I made as I was working through this remix. Feel free to disregard them.
I mentioned that I did not have any full axles through both frames and I wanted to stiffen the frame so the frame didn't creep out at the escapement wheel. I drilled out the escape wheel with a 1/8" drill bit (for an M3 screw). I also opened the holes for the escapement wheel in the frame enough to thread in a M3x40mm screw for the escape wheel axle (7/64" inch drill bit I believe). When I put the M3 screw in place through the frames and escapement wheel, nothing worked. It was binding but wasn't immediately obvious why… I took the M3 screw out and just slid in an Allen wrench (2.5mm I think) through the frames to loosely hold the escapement wheel in place. This took all the pressure off the frame that the M3 screw threading into both frame pieces had caused and the design started working correctly with the Allen wrench just resting between the frames holding the escapement wheel in place! It ran all the way through the stored spring energy several times! It starts itself up after winding and runs solidly!
It seems that the solution is not to force the frame in a position because that can cause binding of the escapement wheel but instead, let the frame relax and use a smaller axle that goes completely through both frames. Maybe it is as simple as drilling out the escapement wheel slightly larger than 5/64" for the M2 screw as I did originally so there is a little more give to it… I think a continuous axle through both frames to hold the escapement wheel in place is also important because it will keep the front and back frames aligned (and hence the axle holes aligned).
Anyway, I am going to try installing the M3 screw, the larger hole (1/8") I drilled by hand through the escapement wheel gave it enough freedom to not bind on the 2.5mm Allen wrench yet not twist and flop around. The design started working! It starts going by itself when you wind the spring and let it go! I suspect I can still use the M3 screw through both frames and just drill the escapement wheel out just a little larger than 1/8" to keep it from binding. I am going to drill the escapement wheel out as little as possible to get it to work on the M3x40mm axle.
Earlier, I had moved the escapewheel forward .64mm with the rev4 frame designs (which I am using and appear to be a big improvement). It may be that moving the escapement wheel holes even a little further (maybe another .5mm?) is all it needs but just doing that changes the distance between the PalletFork and balance assembly so those may need to be pushed back away by the same to keep the same distance (which seems to work awesome with the new PalletFork!). I think the easier solution is to use a full axle and drill out the escapement wheel just a little more than the bare minimum to allow it to spin freely. I'll update with what I end up with for a permanent solution for the escapement wheel axle. It will be the M3 screw axle but I need to determine how much more I need to drill out the escapement wheel for the M3 screw. I suppose I could stop drilling now and just install an M2.5x40 and leave the escapement wheel drilled out at 1/8" as well but I don't have one of those on hand…
I have not reprinted my slightly distorted barrel wheel caused by one side lifting off the bed when printing yet either. It is working reliably even though that gear is less than perfect. I haven't glued anything (such as the frame together) either.
--
Update 8/2/2022: I printed the rev4 versions of the front and back frame and reassembled. I am happy to report that it was an improvement and the escape wheel no longer binds against the train gear. It wants to run and does for 10-15 seconds at a time. The reason it stops is because I have no through axles and the frame starts to push apart at the excape wheel. I think a through axle from one frame to the other will resolve that. I have to print new SpringBarrel also since it lifted off the bed partly during printing and is a little deformed. It does work but there may be a little more pressure at the deformation so I am going to reprint the main gear. I printed everything at 0.20 layer height. I might print the SpringBarrel next time at 0.12 just to try to see if it has a tiny bit more clearance because of the higher precision. The front and rear frame is the weak link as it needs to be stiffer and through hole axles should be used the main gear and definitely at the escape wheel gear which should solve that lack of stiffness. I don't have long enough M2 screws to go through both frames so I am tempted to drill out the frame holes for the escape gear holes and escape gear for a longer M3 screw and nut because I have one of those long enough. This should keep the frame stiffer and therefore run longer and more consistently. The new PalletFolk_rev2 is running excellent. It doesn't even need the retaining pins on either side installed. It wants to run as long as the frame is kept square. I removed the old stl files and only kept the current ones posted so when orientated correctly, everything should be OK to print. I used a mixture of M2x20, M2x16, M2x12 screw, M2x8 and M2 nuts to assemble. It would have been nice to have some M2x25 and M2x40 screws as well as some M2 lock nuts but I made due with what I had on hand. The end clips on the bottom of the frames seem to do a decent job at keeping the rails square. It is the top of the rails that start to push out at the escape wheel mainly. I'm pretty happy with the way this remix is progressing so far and the parts listed are safe to print and assemble. For any part that has to spin on an axle, I ran by hand a 5/64 drill bit through first to clean the hole and found that to be the perfect size bit for the job. I'm sure there is room for improvement but as it is, it is not in terrible shape.
I posted my latest Fusion360 model with the latest parts in the Fusion 360 model with the rev4 suffix.
I did a quick measurement by ear with a metronome of a video of the original full size model running and it was approximately 213 clicks per minute. This scaled down model runs quite a lot faster than that. This smaller size appears to click to my ear at roughly 470 audible clicks per minute.
--
Update 8/1/2022: I haven't printed and tested my latest front and pack rails (rev4 suffix) to help remedy binding of the escape gear. Instead, I assembled the balance on the older frames I had printed and discovered that the PalletFork tail was too short and would sometimes slip off the balance wheel screw. The PalletFork also seemed to get caught on the upper side escapement wheel tooth. I therefore modified the palletfork to have a longer tail (rev2) to stay on the balance gear and a slightly shortened the upper tooth as well so it didn't get caught. It was an experiment and it seems to have worked! With the new PalletFork, I got the entire assembly running for just over 20 seconds at a time before kick starting the balance wheel by hand to keep it going! The biggest problem is the loose escapement gear and the frame is not held together rigid enough (not glued yet). A little more testing and perhaps some more adjustment is needed but I have high confidence that this design will work (it already did work for just over 20 seconds at a time). I am considering making the frame print as one piece to stiffen it up. There will be fewer parts that way also. I also removed the BarrelRacket and RatchetSpring from the design.
--
Update 7/31/2022: Added my own end rail connectors and began assembly. I am using M2 screws. I don't have long enough ones to go straight through so I am using 16 and 20mm screws from each side and letting the part rotate on those. I ran a 5/64 drill bit though the part axles for the M2 screws which works very nicely to allow the parts to rotate on the M2 screw axle. Currently I have it assembled from the winding key through to the escapement gear. It is working nicely through the gear train but the escapement gear is just a little tight against the gear_train gear so I modified the front and rear rails to push the escapement gear toward the balance section by about ⅔ of a mm to help keep the two gears from binding. See rev3 of the front and back rails for this change. What I don't know yet is if the balance end of the design also has to be pushed away from the escapement gear by the same ⅔ of a mm. I need to finish building the complete assembly to know for sure.
Prior to modifying the rails however, I opened up the escapement gear with a 7/64 drill bit on which removed the binding between the escapement gear and the gear train gear. I put the PalletFork on next and this wants to run. The PalletFork can vibrate back and forth and there isn't even a balance wheel or spring on it yet.
I may remove the BarrelRatchet and RatchetSpring from this design as well because it seems to me at this point that those parts of this assembly aren't needed. I'm still thinking about this more though.
--
Update 7/30/2022: I noticed in the slicer that at 50% scale, the PalletFork wasn't very strong around the hole so I decided to strengthen the PalletFork arm. I have never tested the original PalletFork but noticed there was a remix to the original PalletFork that others had reported had improved functionality on the full scale model. The remix is by les_kaye and is found here: https://www.thingiverse.com/thing:3753245/files
I used les_kaye's remix of the PalletFork to scale down to 50% scale and then make stronger around the screw hole. See PalletForkAdjusted.stl for my remix of les_kaye's remixed PalletFork for this 50% scale design.
Secondly, the hairspring printed and I thought it was a little thin and flimsy so I doubled the thickness of it which I think will work better.
I also realized that I forgot to make the balance wheel M2 screw and nut compatible so I updated that part as well. See balance_M2.stl. I updated the ImpactPinHousing for M2 screw holes also. All the latest models are also in the rev4 version of Fusion 360 file. I have many of the parts printed already and am working on assembly to verify the design and parts correctly function together.
--
Update 7/29/2022: I adjusted the scale of the WindingKey and MainSpring to optimize the tolerance for assembly. I modified the EscapeWheel to remove a random fillet that looked out of place. See these files (rev3 suffix) for the latest.
--
Update 7/28/2022: I fixed a problem with a couple hex nut slots on the base back part (see BalanceFrameBack_M2_rev2 for the new model). Secondly, I opened up the hole from 1.75mm to 2.25mm in the base_back part for the hairspring pin so an M2 screw would work there as well. I then opened up the loop on the hairspring to receive an M2 screw (see HairSpring_M2) and posted the Fusion 360 model with these latest changes (see Watch_Escapement_50percent_M2_Nuts_rev2.f3d). Printables is having difficulty posting the rev4 model though so I will try again later if it doesn't get posted correctly.
Lastly, I opened up the hole in the front frame for the WindingStem, as well as made the Winding Stem and WindingKey larger to help ease the tighter tolerances from scaling the model down to 50% of the original it is based on. See rev2c for these three new parts.
I removed the older versions of the STLs to avoid any confusion as well.
--
Update 7/27/2022: I did a small test print and found that 1.75mm filament fits snugly in the holes and M2 screws thread in snugly as well but the hex retaining slots for an M2 slot were too small for an M2 nut. I measured the face-to-face distance of an M2 nut at 3.8mm and so I made the retaining nut 4.4mm across to fit them. I modified three files (the front and back bases and the ratchet) and the new files have an M2 suffix in their file name. I also added another simplified Fusion 360 file for the M2 nut slot changes to these parts. I think this model is going to be pretty good but I haven't printed it yet to test so consider it a work-in-progress.
--
Note: I posted this scaled down and slightly modified version of the full scale design by LarkyPrints on Thingiverse but I had problems associating my post here: https://www.thingiverse.com/thing:5442943 with my Thingiverse account (yet again…) so I am reposting my remix here…
--
This is my first attempt at creating a 50% scaled version of "Watch Escapement Desk Toy" by LarkysPrints found here:
https://www.thingiverse.com/thing:3364860
I have not printed and built this myself so I am uncertain that it will work as-is. The tolerances have all been cut in half with the scaling down to 50% so parts that mate will fit tighter. I think this has a good chance at being faster to print and build though. It may need a little more effort however so don't print if you are not prepared to tinker.
I included the Fusion 360 model I used to scale down the original parts that I generated from LarkyPrints' STL files. I looked at another even more scaled down remix and the STLs didn't look correct to me so I tried to make a 50% scaled down version myself.
At 50% scale, it is about 33 mm wide and most holes are 2.25mm so it stands to reason that 1.75mm filament might be used as axles...
The hex head slots may or may not work but the Fusion 360 model is there to adjust as needed. Originally the design called mainly for M4 screws so perhaps M2 screws could work with this 50% scale version instead?
Besides scaling, the changes I made were to combine the six base pieces into a front and back base rail and remove the screw holes. I envision supergluing the base support pieces onto the bottom of the base rails to hold the two rails parallel. I also replaced the front middle base piece (Train_Frame_Front) with a mirrored version of the back middle base piece (Train_Frame_Back) and I extended the bump out to match (Train Frame Front mounting point) so there is now the option to support the gear that engages the escapement wheel in the middle on both sides. The 50% scaled down original parts are also still in the Fusion 360 model but hidden for reference.
Some parts will need to be oriented to print optimally as their orientations are directly from the Fusion360 model. Additional assembly instructions are found on the original design ( https://www.thingiverse.com/thing:5442943 )
I will look at building myself and updating accordingly but I have not yet done so. Consider this a work-in-progress.
Enjoy!
The author hasn't provided the model origin yet.