Construction Notes for Mobile Manipulator Platform
(By C.N. Thai)
Essentially, I put a PhantomX Reactor arm bought from Trossen Robotics onto a home-made 4-wheel drive platform.
The platform was made from an 8.5”x11” polyethylene cutting board available at discount stores. 4 AX-12Ws were used for wheel drives and they were mounted to the board's bottom via the frame part FP04-F3. This task was most important because if the AX-12Ws were mounted crooked, it would be much harder to program the whole vehicle to go straight a later time.
I used a combination square/straight edge to draw a reference line across the cutting board, lined up the FP04-F3 onto the line (just visually) and secured it with a binder clip. Next, I drilled the two available mounting holes through the board using the FP04-F3 as a template, and then fastened the FP04-F3 to the board using regular nuts and bolts. Then I removed the binder clip so that I can drill the other two holes and finished the mounting job similarly for these holes. This technique seemed to be “accurate” enough to allow my vehicle later to be programmed to go straight by setting all 4 actuators to a single speed value.
Please note that a 3-pin extension PCB module was used to allow various 3-pin DXL cables to connect the AX-12Ws to the controller which was mounted on the top side of the platform.
I did not use the very bottom plate of the Reactor arm as I bolted the support rods directly onto the platform. Actually, I was using that bottom plate as a template for drilling the 8 mounting holes.
I used 3M Dual-Lock tape to mount the controller which could be either a CM-530 or a combo OpenCM-904/C + 485-EXP depending on my needs.
I used an IRSS-10 at the gripper so that the robot can sense when an object is close enough to grab. Due to the length of the arm, I had to splice 2 5-pin cables together (150 mm and 400 mm). My first solution was to stick short bare wire pieces into the connectors to joint them together (as I wanted to preserve the original cables). This first solution turned out to be not too “secure” after a few operations of the arm. So for a more secure solution, I cut off the “middle” connectors and splice the “corresponding” individual wires with solder and insulating tape. Because of the longer cable used, the IRSS-10 will behave differently when polled by the controller via RoboPlus software, but then we always need to calibrate the IR sensor to get the proper threshold value to use in our specific application any way!
I used the CM-530 to do the Remote Control project and the Dowel Search & Grab project with the TASK tool. I switched to the OpenCM-904/485-EXP hardware and used the OpenCM IDE tool when I needed to use the Smart Device functions to access the phone camera to search for the red dowel in the last project (which had some issues explained in more details in Chapter 9 of my forthcoming Springer book).
The various source codes for these projects will be available in August 2017 at the Springer Extras web site when my 2nd Edition book gets published, but for now you can download some files at my Public Dropbox folder.
Happy Roboteering!