0
\$\begingroup\$

I need to generate constant movement in one direction given a variable opposing force even while "stalled" (no net movement in either direction) and when the opposing force is greater than the force being generated, e.g. [try to] constantly move at a speed of 6 inches per second in one direction given 0-500 lbs of opposing force at any time with the ability to compensate for changes in the opposing force within half a second.

My research led me to three phase induction (torque) motors and optical incremental encoders, but I have neither the knowledge nor experience to ensure these will work for my application.

Any insight would be appreciated!

\$\endgroup\$
5
  • 5
    \$\begingroup\$ Constant force and constant movement are two different things. You have to pick one or the other, because the one you don't pick will have to vary. \$\endgroup\$
    – Dave Tweed
    Jan 12, 2016 at 17:47
  • \$\begingroup\$ What is the application? Save us all hours of guessing and wasting our time trying to answer. \$\endgroup\$
    – Transistor
    Jan 12, 2016 at 18:28
  • \$\begingroup\$ @DaveTweed constant movement is more important than constant force for this project \$\endgroup\$
    – John
    Jan 12, 2016 at 22:45
  • \$\begingroup\$ @transistor An example application would be: I'm trying to make an automated tug-of-war machine for dogs. The dog will pull on one end of a rope at variable force, while the machine tries to constantly pull the rope toward itself at a rate of ~6 inches per second (I envisioned using a motor and a wheel). When the rope reaches an "end point", the machine stops pulling the rope, allowing the dog pull its end of the rope with no resistance, until the process starts again. \$\endgroup\$
    – John
    Jan 12, 2016 at 22:51
  • \$\begingroup\$ I am 99% sure that you just want to use a stepper motor. You can reel it in at fixed speed. Don't control the torque with feedback. Just select an appropriate voltage that generates a reasonable torque. If the dog pulls really hard, maybe the motor will skip steps. So you want a limit switch to detect when the tether is fully extended and fully retracted. \$\endgroup\$
    – mkeith
    Jan 13, 2016 at 18:47

1 Answer 1

1
\$\begingroup\$

There are two main approaches to control torque.

The first is open-loop using a device such as a brushed DC or BLDC motor that has torque (more or less) proportional to current. So you control the current.

The second is to use a torque sensor (inline with the shaft or indirectly through the motor mount, for example) and adjust the motor drive to attempt to maintain the constant torque. An encoder is of no use for torque- it can only measure angular motion (often only relative motion at that).

The above answers the question in your title. In the text you ask about torque/movement. As Dave Tweed points out you can only control one of the two at once. It's like a CC/CV power supply. The power supply can only be in one mode at a time- constant current or constant voltage. If you fix the torque then the force on the motor shaft will determine the motion. If you fix the motion, then to torque must adjust to the force on the motor shaft.

\$\endgroup\$
3
  • \$\begingroup\$ Thanks for the explanation, Spehro. As I stated in my reply to Dave Tweed, constant motion is more important than constant force in my project. It sounds like using a torque sensor to adjust the motor drive as you proposed would be adequate for this application. Ultimately, torque is tied to movement, no? If a motor's torque is not moving, neither are all systems stemming from it. Or am I wrong in this assumption? \$\endgroup\$
    – John
    Jan 12, 2016 at 22:58
  • \$\begingroup\$ In theory you can have torque without (significant) motion. In practice- try removing a rusted-on bolt on a car and you'll appreciate that. At some atomic level torque does require motion, of course, and a practical torque sensor might use a strain gauge to measure torque by the tiny deformation of (say) an aluminum shaft. But that's not something you can measure directly with an encoder. Hope that's clear enough. Often what we want is some constant angular velocity (motion) with a limit on torque (so things don't break) but I don't know your actual situation. \$\endgroup\$ Jan 13, 2016 at 0:07
  • \$\begingroup\$ (cont'd) Analagous to a constant voltage power supply that goes into constant-current (or foldback current) mode if the current gets too high. \$\endgroup\$ Jan 13, 2016 at 0:08

Your Answer

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

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