Looks nice! What is that servo/motor etc you are using for nozzle rotation?
Well it's time to spill some beans here. I've been working on a generic multi-axis robotic-workcell for various uses (3D Printer, Pick-and-place, etc) with some high-school students since late 2009, and we've finally got an operational machine! Though there is still some work to be done to make it useful, mostly surrounding the control system. For the PnP machine, I need to engineer some feeders, and for the 3D printer, we need to build an extruder head and a heated table.
Significant kudos to my high-schoolers who have really proven their capabilities, intelligence and mettle on this project! FWIW, we're starting project number two, which is a UAV quad-copter.
So without further ado, here it is... http://veisystems.com/nose/workcell.html
There are more sections I need to document, but I don't have pics of those currently, so I'll get those updates this coming week.
Cheers,
-Neil.
Looks nice! What is that servo/motor etc you are using for nozzle rotation?
Nema 08 stepper motor.
Nice design. I wish I had the capabilities to do that metal work.
What are you using for a vacuum?
Also, what are you using for software?
Thanks. For now, I'm using a hot-air reflow tool that has a vacuum pickup. With appropriate solenoid valves, etc. I also plan to use the hot air side (minus the heat) to provide a "puff" of air to ensure a component is released.
Software is custom C code under Linux, which works crudely for now, but will be much smoother once I finally decide on a motion controller (too much to research still, sigh).
I still need to take pics of the controller/electricals, but hopefully will do that this weekend.
It looks like it is possible to do the software with EMC2 (real time OS), and not need a separate motor controller (still need drivers of course). http://tim.cexx.org/?page_id=704 Maybe you could save a few bucks that way.
Actually considered that route until my mini itx mobo refused to accept Ubuntu. I'm okay with C/Linux now that I discovered image-processing libraries (which was the only part that intimidated me). I'm considering rolling my own motion controller still (using a microprocessor etc).
Yep, as I was all pumped when I heard on another forum about it's ability to be controlled by Python or similar. I don't want to buy yet another mobo, but may consider swapping the one in my CNC mill (full ATX size IIRC) for this Mini-ITX one. Not sure yet.
For vision processing, I had found ImageJ, OpenCV, and one other that escapes me currently, then there were a couple commercial/paid libraries available. It'll be a while before I do anything with vision on this, so for now that's on the backburner. For what I intend to use it now, I should be fine without vision, and still need to build feeders and indexing mechanism, add limit switches, figure out the motion-control stuff, and finish the software. That should keep me busy for a while.
Cheers,
-Neil.
Updates... http://www.veisystems.com/nose/
Still need to finish the software though.
I'd actually say it's moving pretty slow lately, since I have other things going on. However, I just uploaded a video of some basic X-Y testing... [nomedia="http://www.youtube.com/watch?v=Wz9RjazuPbY"]YouTube- NOSE RW / PnP testing[/nomedia].