Motion Detection with Optical Absolute Encoder

[rico]

Hello,

Background I am working on a project to detect motion of a circular gripper. The gripper is circular with a series of rays sticking out around it. There is another like component like this one directly on top of it so that when this one rotates, the rays act as arms that come together and close around an object.

Goal I would like to employ some sort of Optical Absolute Encoder set up or even a potentiometer to detect the position of the gripper. I would like the two separate components of the gripper to know at all times where it is. 2-bit resolution is all I need, so the gripper knows "I am fully open, partially open, partially closed, fully closed."

Problem I have had trouble finding any sort of encoder with 2 bit resolution. I was thinking maybe I can build my own set up with an LED and photodetector and laser cut the wholes that would go on the disk and mask of the enncoder set up the substitute, but I don't know how this would work out.

Any insight?

 

[daGenie]

Don't really understand what you mean by gripper................do you mean a robot???

 

[Laplace]

A 2-bit encoder should work just fine. First decide what is the range of angular position that needs to be encoded (something between 0 and 360 degrees). Then distribute the 2-bit Gray code over that arc on the encoder disk. For this application the encoded positions might not be evenly distributed, i.e., some encoded wedges might have more angular arc than others.

However, this does not seem to be a primarily electronic problem. Rather it is a mechanical fabrication issue to mount the encoder disk to the existing shaft(s), mount the LEDs and photodetectors in alignment with the encoder disk, and avoid optical bleedover from one channel to the other. Will the encoder/detector need to be sealed against dust and dirt? But at least the electronics will be simple.