So a few months ago I bought an AL5D robotic arm, with the intention of using it in a collegiate competition. During these few months I have encountered a constant error, which I have not been able to solve. The problem is simple: at seemingly random times the gripper, wrist, and elbow servos will “jerk” or “twitch”. By this I mean that they will be begin to rail between the maximum angles the servos can achieve. They do not do it always, and rarely do they do it all together at once. Sometimes that act up as soon as they start, other times they will run normally for a while and then begin to act up. Recently it is the wrist servo that is acting up, while the elbow and gripper have acted “nicely”. The competition I wish to use the arms in begins rather soon, and so I am looking for any advice that could help alleviate this problem.
I should also mention I am running all the arm servos through the Adafruit 12-bit 16 channel PWM/Servo shield (all-battery.com/nicd6v5000ma … light.aspx) to provide power to the servos, along with other systems on my robotic platform.
Here is what I’ve done so far to combat this problem:
-
I am running the arm on a robotic platform that enlists 4 DC motors. I have attached 3 ceramic capacitors to each motor in order to deal with noise from the motors. When I did this, the elbow servo quieted down and has not acted up nearly as bad as it had been before the capacitors were added.
-
Replaced servos that were acting up – multiple times. This issue has broken 4 servos by this point, which is another reason why I’d like to be rid of it.
-
Braided wires, and wrapped them in tin foil. The goal here was to shield the wires from outside noise.
-
I have chased the PWM, and power signal from the servos to the shield. In both cases I saw a few dips, but nothing that seemed too alarming or out of place.
-
Taken the arms, and the shield, off the robot to see if my problem still occurred. It did not. That being said, when I hook them back up to the robot and run them, the problem immediately started again even though all the other systems were off.
Here is what I’ve noticed about my problem:
-
This railing problem does not always occur.
-
More often than not I am able to solve the twitch by moving the connecting wires a bit. A slight tap on the wires causes the servos to act normally.
-
This problem has never occurred in the base or shoulder servos.
Any ideas or help on how to solve this problems would be very appreciated. Starting Thursday I am testing to see if static electricity is causing the problem, and if that fails to replicate my issues, I am going to being testing impedance to see if perhaps the impedance of the servos is to low for the system.