Dyne's LD-F1 (D-O concept) build

[Dyne]

As I've mentioned elsewhere, my plan was originally to spend the rest of June finishing the design for LD-F1, then use July and August for printing and building it so I can have it running at Dragon Con.  Naturally, my brain rebelled, as it does every time I try something like this, so I decided to trick myself into getting back to this build by speedrunning some design for 99-99 from Star Wars: Visions (the Ninth Jedi).  That seems to have worked -- after a few days of concentrating on that droid, I'm back in action as you'll see below.  

(That trick didn't stop me from adding costume projects -- plural -- to my list, though.  Thankfully, I prioritize droids and I'm a lot less fussed if I don't get the costumes done, so they shouldn't be too much of a distraction.)

When last we left this thread, I had started working on a rebuild in a single file.  Along the way, I did some experimentation with making the track a toothed belt of sorts, which I commented about back in post #10.  This isn't the final model obviously, it's just the starting point for rebuilding everything I did for the drive and body earlier.



(Now that I've been using it, I think that doing everything in one file was a bit of a mistake, so I might split it up into two -- one for the body and drive, the other for the stuff above the body.  Still an improvement on one-component-per-file but a lot more manageable than a giant list of components and features.)

During my trip to Micro Center this past weekend, I discovered that they are currently completely out of Inland TPU, and I only have enough to print one of those tracks.  I hope they get more Grey back in stock before I need to actually start printing them, otherwise LD-F1 might end up with two differently colored tracks.  I'm probably going to print some test models soon with my (bright red) TPEE to try and figure out what settings I need to get the correct level of flexibility.


In the last post I was also debating the neck mechanism.  I previously used that phrase to describe the external "stalked" neck of the droid and the lifting system, but I've recently been working on a different neck mechanism.  Specifically the three axis "nod, tilt, rotate" mechanism that'll be at the upper end of the stalk.  I'm designing something that's pretty typical of what you'll see on various animatronic heads.

Directly attached to the stalk is a servo with a simple gear to rotate a platform on a set of 608 bearings.  Attached to the rotating platform are two servos, to the left and to the right.  The droid's head is attached to the rotating platform via a gimbal with two axes (pitch and bank), and these are actuated using the left and right servos.

The green part here holds the pitch axle.  There will be another part that actually pitches and holds the bank axle, and a third part that does the banking which will be attached to the head itself.  It would be ideal to have all of the axes coincide with the center of mass of the head, but I'm not sure that's feasible here.



So, if the left servo goes up and the right goes down, then the droid's head angles down from left to right.  If both servos go up, the droid lifts its head, and if both go down it looks down.  Pretty straightforward.


James Bruton has done pretty much the same sort of mechanism in various projects, including his performance robots or the simpler head in the video below, though the details may vary, especially for the rotation.  There are also examples elsewhere on YouTube.






For LD-F1, the entire mechanism needs to be (mostly) concealed inside the droid's head.  The white portion of the head is sized at 10 inches wide x 7 inches deep x 3.68 inches tall (the orange bit increases the depth and height somewhat, but since it is placed asymmetrically along the width I don't count it for our purposes.  

I don't plan on putting much else inside the head besides an SG90 servo for the antenna, LEDs such as an Arduino Neopixel stick, and possibly a servo board.  Still, my mechanism needs to be fairly compact, at least in the height and depth, because the MG-996R servos I'm using for head animation are over two inches long each (if you count the mounting tabs).  Those are unfortunately the two axes that work best for the two front servos.  At the moment they are mounted vertically, but they were originally rotated 90 degrees from their current orientations.  I may return to that layout because the vertical dimension is the most constrained.

Here are the front and side views of the current assembly (the rectangles represent the boundary of the white portion of the droid's head):





I am currently printing some test parts for this mechanism, like the rotation gear and the servo brackets.  Mostly so I can feel like I'm making actual progress rather than just endlessly buying parts and doing CAD.

Edit: I did end up changing the orientation of the servos. My test parts also revealed I hadn't allowed for the servo wires, so the servos wouldn't actually fit in the brackets; I ended up changing to a J-shaped open top design.