So I've got a Sabertooth 2x5 motor driver. I uses standard servo connections, and a micro controller can use standard servo arguments to make the motors go in forward or reverse (much like a continous rotation servo). In any case, it connects to a board in the bottom of my robot, gets grounded to the common ground, and through traces the signal lines go to a JST connector to the next stage of the robot (which is designed to be modular for easy teardown). The problem I have is that sometimes the motors move slightly; this usually happens if the motors are not receving any signals if the robot is in idle. Further if I remove the connector and touch the data line pins with my hand the motors will move.

I was wondering if an optocoupler/isolator for the data lines before the JST connector would fix this issue? If not, what would be the suggested way of dealing with this?

  • \$\begingroup\$ It's conceivable how continuous rotation servos would free run. But since you say these are not, I'm not sure how to address you situation. \$\endgroup\$ – st2000 Nov 17 '16 at 4:27
  • \$\begingroup\$ I feel like the data/signal lines are just really susceptible to noise/changes in impedance? In any case, if it were a continuous rotation servo what could prevent this? \$\endgroup\$ – Alex Nov 17 '16 at 4:31
  • \$\begingroup\$ The continuous rotation servos I have used depend on the time between two pulses. Whereas a normal servo will use this time for positioning information, the continuous rotation servos interpret this time as an indication of speed and direction. It is not a digital signal as in the signals going over a USB cable. And if the software produces a jittery time between the 2 pulses (for example, if the software was written badly), that will manifest its self as jerky movements or an inability to stay still. \$\endgroup\$ – st2000 Nov 17 '16 at 15:52
  • \$\begingroup\$ Also, the continuous rotation servos I have used have an adjustment such that for a given time between the two pulses the servo should stop rotating. This will not stop jitter. This is for adapting the servo to the time source should the time source prove unadjustable. \$\endgroup\$ – st2000 Nov 17 '16 at 15:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.