Mccand, TRap sounds cool. It looks as though you're planning on reinventing every subsystem. Which is very cool. I'd suggest doing it via a Ship of Theseus process. Build a WolfStrap, Eiffel, or Mendel using Contemporary Electronics, and once everything is working, only then replace each subsystem with a new and improved subsystem. This way you have a working 3D printer to bootstrap the new and improved 3D printer. This way is much more fun.
--Sebastien Bailard 21:39, 2 September 2010 (UTC)
Re: Ship of Theseus
Sebastian, Agreed; A progressive build is easier, but what I'm proposing requires that an entirely mew control scheme be developed before anything would work. Anyhow, once I finish my Mendel, we can start testing stuff.
My background is mostly in mechanical design, so I was planning on using existing electrical systems (I am not skilled enough to redesign that) I think once I can produce parts on my own, the design will advance quickly. We'll see. What I really need is a teammate to keep me from losing interest, and who can do the programming. (Design is easy, maths are easy, software is hard.)
Thanks for the comments!
Mccand 21:55, 3 September 2010 (UTC)
Hey Mccand, while there is certainly no silver bullet for documentation I'm honored that you've used Columbus' as a template and hope that you will improve it over time. I'm a strong believer in documenting everything within the wiki, so it's great to see someone else actively pursuing that goal as well. I borrowed bits and pieces from the Mendel overview page (the table), the project template wiki, and my own general engineering background (although I must admit that Columbus' documentation can be considered as pure satire when viewed under any professional standards.) Just a few things I want to change about Columbus that you might find helpful as well: (1) I want to learn how to make references like in wikipedia articles. I know it is probably well documented, I just need to take the time to incorporate it. (2) I think that eventually I will have to move my Workbook to it's own subpage because, over the next two and a half years, it's bound to get unbearably long. I think this not only distorts the purpose of the wiki, but it ignores it's potential. Imagine if everything written about the Mendel was published on a single wiki page? When people come to our project wikis I think they should be greeted with an up to date overview of the progress we've made, and then - if they want more - they can dive into the Workbook subpage. Anyhow, good luck!
--Isaac Olson 23:50, 4 September 2010 (UTC)
Wow, the TRap looks pretty cool. Thank you for showing it to us.
I hope you don't mind me picking at 2 things:
- Is it too late to change the name to sound different from T-Rep?
- The top strut of each tetrahedron has a threaded leadscrew nut on one end, and contacts the center of a horizontal bar at the other end. Does that horizontal bar need to pivot around that center contact, or is it better to hold that horizontal bar in place by adding more struts from its ends to the other ends of the tetrahedron?
And one question:
- Do we have to write all new firmware from scratch? Or can we use the Delta#Firmware as a black box to convert XYZ positions to tetrahedron angles on the TRap, then tack on the (much simpler) code to convert each tetrahedron angle to turns of the corresponding motor? Then when someone develops a new, better black box with improved calibration features, we can upgrade easily?
--DavidCary 13:09, 8 November 2010 (UTC)
Thanks for taking a look! Unfortunately, I've been swamped lately, and progress on the Mendel that is going to build the TRap has slowed to a crawl. I hope to address your comments here:
- I really enjoyed seeing the T-rep. Looking at the wiki page and blog was pretty cool. I hope that my design gets that far, in terms of workability. TRap (Pronounced as a single unhyphenated word: "trap") is a development name, and will eventually be changed - if the project ever gets to a usable state. At that time, I'll appropriate the name of some pioneering and iconoclastic biologist. For the moment though, I'm going to stick with it, as that is the name which ties all my design files together.
- I'm not sure that I understand your comment here. The sketch is not entirely clear, as there are no motors shown. Each motor-leadscrew assembly will connect a corner of the base and the outer (furthest from center) free corner of a tetrahedron. Both of these conections will need to pivot to allow the tetrahedron to move properly. The joint at the base is a simple pivot. The joint at the corner of the tetrahedron is a bit more complex, as it contains the leadscrew nut, which must pivot. The actual design of the tetrahedron at that point needs to be worked out, mostly so that the leadscrew does not crash against the tetrahedron when it is fully retracted.
And your question:
I don't know. I hope that I can use an existing firmware, as that would reduce my task list substantially. (And I know very little about uC programming. I'll always welcome competent help (in any aspect of this project), so if you are interested, or know someone who might be, please let me know.
Thanks for taking a look at my project! I hope to hear more from you.
Mccand 13:58, 8 November 2010 (UTC)
Excellent. I've done quite a bit of uC programming, and I see that questions about reprap firmware programming get answered fairly quickly on the RepRap forums. It's the mechanical bits that give me problems :-).
Thank you for the explanation. OK, that explains a lot.
One more question: The inner (closest to center) free corner of a tetrahedron touches the middle of a horizontal bar. Does that horizontal bar need to pivot relative to the tetrahedron, or is it better to make that bar a rigid part of the tetrahedron by adding more struts from its ends to the other corners of the tetrahedron? --DavidCary 21:02, 8 November 2010 (UTC)
That bar will probably need some additional struts to be adequately supported. It is actually a hollow piece, and the bolt supporting the universal joints at each end passes through it. If you are interested in looking at the firmware programming once I start getting things together, let me know. It will probably be a while before I can start fabricating the structure, but I can certainly keep you in the loop.
Mccand 13:00, 9 November 2010 (UTC)
Yes, I'd like to hear what you are doing. If you post what you want the firmware to do -- or later, what you expected the firmware to do, and what annoying thing it does instead -- perhaps I or some other reprapper may be able to edit the code to fix it while you are tweaking other things :-).
Ideally all coders would have their own personal fully-functional TRap to test out rough draft code. I wonder if perhaps a quick prototype with magnet joints or a simulation like the virtual Mendel would be helpful for coders that don't yet have a functional TRep. --DavidCary 00:32, 19 December 2010 (UTC)