The first batch of parts has arrived for the linear delta robot. I’ve chosen to base the electronics around an Arduino and some really neat A4988 stepper motor breakout boards from Pololu. Quite incredible that for such a small package they can deliver a supposed 2A at 35V. They also support up to x16 micro-stepping allowing 3200 steps per-revolution from a typical 200 steps-per-revolution bipolar stepper motor.
For the joints that connect the vertically moving carriages to the plotting head I’ve chosen to use some reasonable priced ball joints from RS Electronics (a mere £2 a pop). They come dis-assembled which was a great opportunity to use my Dads treadle powered press for some production line fun.
And here are the assembled joints.
Started working on adding a texture to the generated models. To do this I’ve added an array of LEDs to the inside of the scanner that can be turned on to illuminate the model via the serial port on the arduino. A full colour image is captured for every laser spline image, then for every point along the laser spline the corresponding rgb value from the full color is found. The rgb values for all the splines are then mapped onto a 2D image creating the texture.
I need to brush up on some 3D modelling theory and the .obj and .mtl file formats and I should be able to create fully textured 3D models.
For now here is the image of the texture I captured for the nodding monkey model I’ve been using.
So after spending most of the weekend re-writing and working on the code to extract the 3D models the hard-drive on my main PC gave out the clunk of death. Which kinda sucks. When I’ve got a free day I may have a go at recovering some stuff, but for now I’m going to cut my losses and start rewriting it from scratch. Now that I’ve got an idea what I’m doing it shouldn’t take to long.
For now here are some photos of the progress on the hardware side of the project.
WordPress lost all the info I entered into the gallery, and I can’t be bothered to re-enter it. So here’s a brief overview of what I’ve done.
- Sprayed inside of scanner body black to reduce ambient light.
- Mounted the unmodified laser line generator into the body.
- Took out the laser line generator and split it into 2 separate modules (and built in current limiting resistors so they can operate at 5V) so I could build another scanner at some point (I’m thinking a room scanner would be kinda cool).
- Wrote some Arduino code and built a circuit to control the stepper motor underneath the rotating platform (Code and schematics will be posted soon, both are incredibly simple).
- Realised stepper motor sucks and I need a better one.
Getting pretty close to finishing now and next week I’m off for 2 weeks so hopefully things should be all wrapped up soon and I’ll be 3D scanning to my hearts content.
(Side note: For large datasets created with 3D scanners Meshlab looks like an awesome tool ,had a quick play around and despite some mild difficulty compiling it’s fairly straight forward to use.).
I’m heading to the Newcastle Maker Faire with work this weekend, so I thought I’d shamelessly show off some of the stuff I’ve been building for it.
This is the second version of my light field camera. It’s loosely based around this one: http://www.futurepicture.org/?p=47
But instead of multiple cameras it uses a single camera on a moving platform to produce the image(s). I’ll write do a larger write-up with some example images, some code and a bit of theory when i have a little more time after this weekend.
As an added bonus theres also some pictuers of my fathers homemade mini lathe and pillardrill.