Hey folks,I’ve been digging into (and building) some of the more esoteric motion platforms and their kinematics when I ran across the Tripteron. It’s a fair old idea (ca. 2002) from a Canadian university lab, and I found it fascinating. Furthermore, I can’t seem to find a single one that’s actually been built into a 3D printer, though it would appear to have many unique and desirable qualities for one.So I decided to start building one. I’ve got a thread on the RepRap forums I’m updating as I iterate on the design, and I’m far enough along now I figured I’d share it here.Here’s a simplified CAD overview of where I’m at currently:http://i.imgur.com/xQTTP9q.pnghttps://youtu.be/AQf8r-eMf2wAnd I just printed and tested out some fun-colored tests of the new arm designs, which are performing flawlessly:http://ift.tt/2eQqj7X of the details are in the forum thread, but the elevator pitch of why I think this is a worthy pursuit is:Due to the decoupled nature of the axes, this is known as a parallel cartesian system — i.e., the error of any one axis doesn’t compound with or amplify the error in other axes. In contrast a serial cartesian would be the standard i3 type design, where X and Z are serial and XZ-Y are parallel, and other cartesian variations with 2 or more combined axes.The kinematics are fully cartesian, so actuation and IK math is dead simple. All three axes can be moved any way you like, on any kind of rod/rail/screw system.Specific dimensions of the constraining arms are adjustable to fit whatever build volume you want to cover; they can also be replaced with other types of linkages, so long as they’re stiff parallel to the joints and free perpindicular to them. Scissors, curves, etc.The design of the orthogonal, jointed constraints (the arms) is actually an over constrained setup, where each arm is constrained by the other two as well as its own joints, resulting in an extremely stable and accurate effector response.The primary engineering concern is purely in the arm stiffness (in only one axis!), and joint stiffness, while remaining fluid in the joints. My joint design is very simple yet effective (2 bearings, all M5 hardware, 2 washers, 2 printed bearing spacers, nylock nut), allowing the joint to be very tight (thus stiff) without affecting bearing motion.And the latest arm design I came up with using modeled truss substructures results in incredibly stiff arms. These things take a serious amount of force to cause visible deflection when they’re anchored on a flat surface. I haven’t put it all back together yet (and some of the new parts are printing) but there’s lots of pics and videos of the whole thing operating in the thread, so I won’t litter this post with them.I need to adjust the “shoulders” (arm/rail block/belt carrier junctions) a little, possibly tweak one of my motor brackets, and I’m getting the remaining idler pulleys and 2020 extrusion I need in the next few days, so I should have it reassembled and able to test exactly how well the new arms perform a little later this week.Any feedback is welcome, let me know what you think! And FYI, I’ll have a Thingiverse/GitHub/etc up with all the parts as soon as I finish this design iteration, but I can provide the files any time if someone really wants to see them as-is. http://ift.tt/2fjJ4Sb