Hello everyone,
I’m trying to develop a “Ball On Plate” apparatus for a senior project in Controls at my university.
Please assume the following axes for motion as if a wooden plate sitting in front of you on a table/desktop.
X-axis is left to right
Z-axis is top to bottom
I want to tilt my wood plate to a maximum of +/-15degrees along its X and Z axes assuming that my
tilting assembly is mounted on the center underside of the plate in some manner.
My plate dimensions are approximately 9.625 inches (left to right) by 7.5 inches (top to bottom) by
0.3125 inches high. The plate weighs about 200 grams and is made of wood. I can’t drill through my
plate because it will disrupt the smooth top surface that the ball rolls on.
My question is:
Has anyone accomplished 2-axis tilting along the X and Z axis only with any of the brackets/erector parts
that lynxmotion manufactures?
Thank you very much in advance for your help,
All the best
You could possibly do it by using a ball pivot in the center, and a pair of R/C servos for each of the X and Y axis. Use linkages from the servo horns to an appropriate point on the underside (edges?) for attachment.
There was a marble-maze project mentioned on this list. The maze was mounted in gimbals, as I recall. Might be simpler.
Attached is a picture of the Labyrinth maze game platform I’m currently using. I was trying to use existing hardware rather than building a custom implementation because the semester is only 16-weeks. I have a few issues at the moment with the tension string sliding on the hardened shafts (a supposed improvement over the stock shaft) and I can’t fit the Parallax MMA7455 accelerometer board under the game surface centered at the intersection above the rods because there is insufficient space.
So I’ll see if I can make this work or otherwise I may build an outer frame that is taller and relocate the shafts for more clearance under the plate. I have my regulator targets painted on the surface as a first step and if that is successful then I’ll go for point to point and thereafter the full maze. I’m using a CMUcam3 to perform target detection. The inspiration for this comes from the team at Aalborg University in Denmark whose Youtube video is here:
