Tim WilkinsonAt some level, this project was an excuse to buy a 3D printer.
I first bought a Lulzbot Mini (https://www.lulzbot.com) and I can't say enough good things about this thing. It was fabulous as printing all the demos I threw at it and for a 3d printer newbie like myself, was easy to use. Unfortunately this project requires extremely fine detail and precision - the joint I'm building has features which need to correct to within 0.1mm and the Lulzbot simply couldn't do this. So back it went.
My second 3D printer was a Form2 (http://formlabs.com) which is a considerably higher-end device. If the Lulzbot felt like a fun toy, this thing feels like a precision engineering instrument; and I suppose it is. And while a lot more expensive (both the printer itself and the consumables to print with) it is complete capable of print what I need.
The designs (which are still being iterated on) can be found here on OnShape (https://cad.onshape.com/documents/5779d31ae4b0f25735253f33/w/c24214105640146fe6c8ce0f/e/bc653cadfd1427398db05ab3). I'm currently printing them using White Resin V2.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
Sorry for a stupid question -- I didn't follow this project from the beginning, so maybe there is a reason I didn't see -- but why are you using a chain to connect the nut with the pivot, instead of making the pivot into a gear, and adding a flat piece with teeth to the nut, to rotate that pivot while moving up and down? I suppose the precision necessary for the teeth is higher then, but perhaps with the new printer it will work?
Are you sure? yes | no
No no - a very sensible question with a not very good answer. History. A very old design involved a chain which looped around the entire structure so it could pull and push the pivot. That never worked out in practice, but the chain persisted into the current design for no terribly good reason. The next iteration I'm tinkering with finally removes the chain and instead uses a vertical rod attached to the nut directly, then a single link attaching that to the pivot. Looks a lot better. I've not tried a gear setup.
Are you sure? yes | no
This is quite similar to what I've been slowly working on, down to getting a lulzbot then a form2! One thing I'm going to try and do is replace the printed parts with resin casted parts. If you look up the Guerilla CNC machining guide it guides you through the process. It uses CNC milling but the form2 should be a decent alternative.
My project is an android, and I'll use many different motors but I want one shared gear/actuator architecture between all the motors.
Are you sure? yes | no
I'd be very interested in knowing more about your android.
Are you sure? yes | no
Not got much to show right now, just trying to flesh out the concepts. The basic plan is to combine series elastic actuators and the parallel/multi joint spanning aspects of real muscles. So rather like the Kojiro or Ecce robots. Those however tend to keep the motors within the static parts of the body, such as the chest cavity. I want to move the motors into modules that connect at roughly the same place as the human muscle counterparts. I think this will give several benefits, but mostly it should make construction/upgrading/repairing all easier as replacing a module involves 2 screws and the electrical connection. I hope to have a basic module ready for demoing soon and will involve a case, a planetary gearbox and bevel gears all printed by the form 2.
Are you sure? yes | no
Ah! Didn't see you have an F2 until now.
Are you sure? yes | no