So I have the Learm 6 axis robot arm and before I made a higher quality arm I wanted to test my program for a helping hand. My idea is for this arm to be used in my workshop as an extra arm. When I pull on its claw I want it to know where I am pulling it and move to a position in space. This movement must be easy for me. Easy enough to move the arm to the left with one finger with the servos adjusting. Kinematics and coding is not the challenge I here. I need to measure directional force put on a servo, even low amounts of force, without adding extra sensors. I saw someone who modified the potentiometers in the servos to take an analog output, but I'm not too sure that'll help me out. Anyone know another easy way I can do this?
Sorry for the wall of text.
I saw someone who modified the potentiometers in the servos to take an analog output, but I'm not too sure that'll help me out.
That was almost certainly done to read the actual position of the servo as opposed to the commanded position
What about measuring the current used by the servo ? It will be higher when the servo motion is obstructed or when something tries to move the servo when it is in a fixed position
UKHeliBob:
What about measuring the current used by the servo ? It will be higher when the servo motion is obstructed or when something tries to move the servo when it is in a fixed position
Interesting idea but unfortunately I think that while that will tell you if anyone is pushing against the servo arm it won't tell you anything about the direction of the push. I think to get an idea of the direction you would need to dive inside the servo to measure the actual voltage/current on the output of the driver i.e. the motor current which will reverse direction depending on the direction the motor needs to move to balance the push. Not impossible but not easy.
Steve
That would work. I have one concern with that and that's the overall speed of the system. If I want to measure direction that way, I would need to tell the servos to guess a direction, move in that direction, then measure if their power demand is increasing. That could lead to extra time in between the force the user puts on it and the actual movement of the arm. I'm not sure how much time, I'll test it. I want to make the arm move as easy as possible to the point where it feels completely natural to the user. I've seen some helping hands in the manufacturing industry like heavy-duty strength-enhancing exoskeletons. These look like they do something similar. Does anyone know how they measure the force put on them?
Edit: I just read slipstick's message. I didn't think of that that could definitely work. It would not be easy, however.