3D Printed/Wood Hybrid
  • The 3D Printed Hybrid version of the Seprimus came about as the result of my complete inability to build the kit. Its simply down to my lack of experience and tools, rather than a comment on the kit which is of the highest quality.

    I made a number of basic mistakes with a combination of a dremel and bit of sandpaper:

    1) I cut the pins badly (my first time cutting even small metal), this left the pins a bit wonky;

    2) I over enlarged a couple of the holes, making the arbors bounce slightly;

    3) I accidentally sheared off one of the escapement notches...

    So in short things were not going well, the longest i got the clock running before shearing off a part was 1 hour.

    Living in central London access to any kind of workshop or wood yard is not easy, so i turned to the MakerBot Replicator 2, a 3D printer that we have in our office. Purchasing the DXf plans allowed me to extrude the parts in SketchUp and convert to the .stl format required for printing, Dave also kindly emailed his SketchUp model. The printing is high enough resolution to print the cogs complete with pins, ensuring the perfect line up. I have attached 4 images to show the detail, i have also tied the clock back to a shelf pole, so it acts as a bookshelf clock.

    So far i have replaced two parts, as well as adding in extra bearings - the clock is now finally running. So it is a mix of wood and 3D printed parts, over the next few weeks I will replace all the cogs with the printed version, as pictured, they are printed and ready to go. With the plastic parts and bearings the clock now only needs 500g to run.

    If it works with all the new cogs I will then probably choose a nicer colour and start again, the nice thing with 3D printing is you can simply print out a new part.

    Its been a bit of a learning curve, but one i thought might be worth sharing,


    photo 1.JPG
    1632 x 1224 - 687K
    photo 2.JPG
    1632 x 1224 - 729K
    photo 3.JPG
    1632 x 1224 - 735K
    photo 4.JPG
    1632 x 1224 - 690K
  • The more I see what 3D printers can do, the more I want one, and doubly so when I see one of my clocks printed out, but I really can not justify the cost. After all PLASTICTIMES just does not roll of the tongue.

    Thanks for letting us have a look at your Septimus, I do like the fact that you can also print the pins on the cogs, please keeps us up to date when and if you replace all of the cogs.


  • A quick update - i have added the other main gears and will change the hour/min gears in the week.

    I have uploaded a short video for anyone interested: YouTubeClip

    The tick varies from fast to slow slightly, i had it with the wooden cogs as well, just wondering what the main reason for this could be, you can hear the clock vary in the strength of the tick as well.

    It is running though :)

  • It's quite hard to tell from your video, but I noticed that the upper bob on the pendulum is part way down and looks longer than the one I use. I can't see clearly enough to tell if the lower bob is larger than the top one, but at 0.46 of your video it looks as if it is right at the end of the pendulum and might be smaller. My clock has the smaller bob on the upper part (right at the end) and the larger bob part way up the lower part of the pendulum, as described in the instructions.
  • It is - due to the change in cogs even at the lowest part of the pendulum the clock was still running fast. By switching the weights, so the small one is the bottom and the bottom now mid way at the top, the clock now keeps time.

    I could of added a longer bottom section, but did not have any carbon fibre to hand and the switch over seems to of worked.

  • Well spotted Petec.

    The reason for the erratic ticking and the fact that the clock runs too fast is to do with your anchor only contacting the escapement wheel with its impulse faces or only just hitting the corner of impulse face and locking face, difficult to see on your video.

    In order for a deadbeat anchor to function properly it must firstly connect with its locking faces.

    Take a look at this thread.
    Pendulum Weight

    Also take a look at page 28 of the instructions.

  • Thanks for the link to the thread, I am back to the original pendulum now.

    Its probably against the whole ethos of the kit - i ordered it because i wanted a wooden clock - but having sheared off a key part i am going to see if i can go almost completely 3D printed from the plans as a test.

    I have installed 3D printed arbors today and all is working well, I will post a new movie once i have something properly running.

    I know it will probably make many members of the forum wince but its quite an interesting if aesthetically challenging experiment.


  • If I had access to a 3D printer I'd be having a go too. I do have a laser cutter but the tube has gone, otherwise I'd be cutting perspex gears on that to see how they worked. As it is, I'm using my trusty Lidl scrollsaw to play around with things, which is not quite so high tech!
  • To join in and add a little, I have access to a laser and have cut both ply and acrylic gears. The acrylic gears work well. In fact acrylic laser cut gears work better than ply gear that have been lasered. The acrylic cuts smooth and the laser leaves a polished surface, ply has a burnt surface. Something to remember is that lasers do not cut perfectly perpendicular due to beam focus, focal length etc.
    I work in a school and we are getting a 3D printer so will be playing that soon.....

  • A quick update on the 3D printing, its proving quite interesting if challenging....

    The image attached shows where i am, I've edited the front face to allow quicker access to the cogs. Its been running but needs some more tweaks - the 3D printer takes approximately 45 minutes for Cog10 with the pins but the plastic on default settings is quite easy to break. I'm going to have a try with more infills and a lower weight, to see its feasible to print out the Septimus without brass and get it running.

    The arbors, winder, escapement are all fine, its the pins on Cog 10 that seems to be the week point.

    So its all a bit of an experiment, as i think the image suggests. Over the next few weeks I will add some more parts during the day at work (i fire up the printer and leave it running) and update if people are interested.

    I do kind of miss the wooden version, but as i sheared off a part of the escapement due to 'incompetence' I figured 3D printing was worth a try.

    1632 x 1224 - 631K
  • It's an interesting experiment.

    Is there a difference between the pins on the escapement and the ones on cog 10? I couldn't figure out why one is a problem and the other OK, but no doubt there is a reason I'm missing.

    Aare you going to try printing the fixing collars/couplings (e.g. on the small hour train pinion in your photo) so that you just need to use a grub screw? Could you print it with a hole for a screw and would the plastic be soft enough for a screw to cut the threads?

    I have to say that I prefer wood(!), but I can see why you're doing this. Please keep us informed of progress.

  • This is fascinating. I'm interested to see how well the arbors work as plastic - my Sept. would run for only a few mins until I highly polished all the arbors. It'd have though that, depending on the direction in which they were printed, they'd be either too weak, too rough or both. We're probably getting a 3D printer at work shortly - and it would be helpful - especially to experiment with different designs of cogs, etc. - I'm eager to customize mine with cogs along this design:
  • I had thought about collars and decided that i would build the collar into the gear, then drill it and use a self tapping screw to fix it to the shaft. I was going to use brass rod. I have found a wood-like filament that i intend to experiment with so I got a wood feel to the construction. I have used Sketch up to design Gears in the style you show here. I am no expert but after a few false starts it became fairly easy.
  • As promised an update on progress. The clock is now completely 3D printed apart from the connections on the arbors. It has taken a lot of trial and error but it now runs well with a strong tick, the pendulum now swings fully, something i failed to achieve with the wood version (but i made a lot of early errors with the kit, its been a bit of a journey).

    I have attached various photos and made a short clip on YouTube:

    The clock runs on under 500g, the only difficulty is gaining enough friction on the hands. They slipped so i have had to glue them which of course makes adjusting the time kind of tricky. I hope to fix this in the next build, I am now in the process of printing a second clock in white.

    It should be a lot easier this time as i have gone through the process of finding the best printing settings, 3D printers are not simply 'print and go' there have been a lot of failed prints and its taken a lot of time to get it all working. A CNC route would be a lot quicker but its not the sort of machine i have in the office.

    So good luck to anyone else looking to 3D print, if you want any help just let me know, any thoughts or comments of course welcome... and thanks a lot to Dave for the kits and plans.

    photo 1 (1).JPG
    1632 x 1224 - 425K
    photo 2 (1).JPG
    1632 x 1224 - 498K
    photo 3 (1).JPG
    1632 x 1224 - 553K
    photo 4 (1).JPG
    1632 x 1224 - 517K
    photo 5.JPG
    1632 x 1224 - 415K
  • that's fantastic - well done! The arbours are all plastic too? Presumably they're still using the metal bearings?

    As for the hands - maybe a hole in both sides of the of the 'barrel' around the arbour so you can insert some sort of clip to hold it in place by friction on the sides of the arbour, rather than with a grub screw?
  • Yep the arbours are all plastic - they seem to work well, i added bearings to all the main cogs, just to reduce the friction and to fix some of my 'early errors' in the kit.

    I am building a second one this week with everything i have learnt - on a Replicator 2 it takes approx 3 working days to print out all the parts. So quite a long time but i leave it running while I'm working. If anything it gives me some light in the joy of the office to know im printing out a clock.

    I should of thought of a hole in the barrel before reaching for the glue, the joys of hindsight.

  • For those interested, i have printed a new version in white - a clip below:


    Seems to be running well :)

    I have also put a post up at http://www.digitalurban.org/2013/09/3d-printed-clock.html with a link back to woodentimes :)
    3264 x 2448 - 2M
    3147 x 2065 - 1M
  • That looks so much better in white, and the wooden back plate sets off the white cogs really well, or put another way, it stops the whole from looking too plastic fantastic.

    And thanks for the video, at the end of the day the proof of the pudding is in the eating, and your printed septimusĀ“s anchor plunges into the escapement wheel as it should.

    What are the advantages of using printed arbors over brass arbors? Or is it just a case of I will because I can?

  • I'm struggling converting the DXF into SketchUp, I can get them into sketchup and extrude into thickness but nothing seems to line up. Holes are slightly off, the parts for the pendulum are all slightly different shapes and lengths, its a nightmare! LOL

    How did you get yours to work?
  • Kelvin,

    Dave kindly sent over his SketchUp model which, with a few edits to cut holes etc was easier to work with than the raw dxf files....

    I have the completed 3d printer files - I could send them to Dave and an experimental 3D Printer version could be bundled in a download perhaps ? Dave?

  • 3D printer files? Now that is an interesting idea.

    But as it stands the 3D print still requires a wooden backplate, I think it would be better to adjust the frame to allow both halves to be printed, still keeping the basic 7 shape but reducing frame mass.

    And I think there could be a few tweaks done to make better use of the printers abilities.

    As I have no 3D printer I would very much depend upon your input Andy, what do you think?

  • Apologies for the down time - work got in the way. I am thinking of building another 3D printed clock, based on the learning first time round but also with the aim of using mainly printed parts. I think i can print everything apart from the bearings, so being able to tweak the frame to be printed would be great. I think all the arbors would need to go into bearings, but this is easy to design for.

    I would be happy to print any tweaks if your interested and then i think there is a market to sell a 3D printable clock from your designs...