I want to position-control an axis combining the action of two DC servomotors - one of them alone is not powerful enough; I had those available and preferred not to buy a new one.
I have not the possibility to measure the output axis position directly, so I must rely on the integrated position, speed or torque feedback control logic built in my actuators.
Long story short, the two motors will transmit power to the output axis through gears (a separate gearbox for each one).

Now I am approaching the design of the control system and I realized that using the built-in position-control logic for both actuators might not be a recommendable choice, the reason being that it will be practically impossible to calibrate them with the exact same zero reference, meaning that a small misalignment in the motors' output positions will arise, leaving during operation just a single one of them in actual contact with the respective geartrain, sustaining all the load, and the other unloaded.
The only alternative approach I can think about is the following one: to control one of the two motors (let's call it 1) in position, while controlling the other one (call it 2) in torque by feeding it, as setpoint, the torque corresponding to the current absorption measured on motor 1.
This way motor 2 should provide power until it reaches the same torque motor 1 is providing, thus sharing the load without introducing uncertainty on what position setpoint is being followed.

I would like to know if this approach could be likely to succeed, if there are some precedents and/or if there exists any better solution.


1 Answer 1


I've done several applications similar to this one, which we call here as "Torque sharing", where two motors move the same load.

Your approach is right: Control a motor in Position Mode and the second only cares about applying a torque vector oriented in the same direction and amplitude as the master, with time the load will be shared and the torque on the master will reduce and it will be possible to observe both dividing the work.

As the motors are different and the reduction too (?) you may need to work with some gain/scale before applying to the slave.

In your case, it seems that the coupling between the motors and the load is rigid, that is, there is no risk of slipping, eventually this causes some problems in this type of application that I don't even like to remember lol...

In my applications the "Torque Mode" of the servos was nothing more than a "Speed Control Mode" with limited torque.

Some tips that may help you:

  • The slave motor (in Torque Mode) must have the fastest possible acceleration/deceleration ramp and with gains for high dynamics (if possible), because you may not want it to take too long to respond in helping the master.
  • Adjusting the slave speed limit to the motor limit or a percentage above the current master speed (Eg: MasterSpeed x 1.3 >>>> 130%), this helped to improve the response for the axis never reaching the sepoint speed while maintaining the set torque.
  • The torque update to the slave must be as fast as possible, or conflict between the motors may occur, leading to excessive vibration.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.