A servo drive can be told what to do in three basic ways: go to a position, run at a speed, or produce a force. Torque control mode is the third of these. Instead of chasing a position target or holding a speed, the drive is given a torque command and it makes the motor produce that torque, regulating the current that generates it. This guide explains how torque mode works through the drive's current loop, why applications like tension and winding depend on it, and how a torque limit protects a machine even when the drive is really running a position move.
Torque Control Mode in one line: Torque control mode is a servo drive operating mode in which the drive is commanded to produce a specific motor torque and regulates the motor current that generates it, rather than controlling position or velocity. Because torque in a servo motor is proportional to current, the drive's innermost current loop enforces the torque command directly, which is what makes torque mode the right choice for tension, winding, and force applications where the goal is a controlled force rather than a controlled position.
In a servo motor the torque the shaft produces is very nearly proportional to the current flowing in the motor windings, so controlling torque comes down to controlling current. Every servo drive already contains a fast current loop at its core that regulates that current precisely, because the position and velocity loops above it ultimately work by asking for torque. In torque control mode, the drive simply exposes that current loop directly to the outside command: the torque request becomes a current target, and the current loop drives the motor to produce it.
This is fundamentally different from position or velocity mode. In position mode the drive works out how much torque it needs to reach a target position and holds nothing constant except the destination. In velocity mode it varies torque to hold a speed. In torque mode the torque itself is the fixed quantity being commanded, and the resulting speed and position are left to the load. If the load can move freely, a commanded torque will accelerate it; if the load is held or resists, the motor pushes against it with the commanded force but may not move at all.
Because speed and position are not being regulated in pure torque mode, the drive will happily let the motor run away if the load offers no resistance, since a constant torque on a free load simply keeps accelerating. For that reason torque mode is normally used where there is a genuine load to push against, such as material tension, or it is combined with a speed limit so the drive caps how fast the motor is allowed to spin while it delivers the commanded torque.
The classic home of torque control is tension. On a winder, unwinder, or any machine that pulls a web or wire, the quantity that matters is the tension in the material, not the exact position of the roll. Tension is a force, and force at the roll is torque divided by radius, so commanding the motor's torque is a direct way to set the pull on the material. As a roll builds up or runs down and its radius changes, the drive adjusts the torque command to keep the tension constant, which is exactly the behaviour a paper, film, wire, or textile line needs.
Winding is a good illustration of why position control would be wrong here. If the winder were told to reach a position, it would fight to get there regardless of the tension, tearing thin material or leaving it slack. Told to produce a torque instead, it pulls with a controlled, steady force and lets the material and the rest of the line dictate how fast it turns. The same logic applies to any force task: pressing, clamping, testing to a set load, or holding a part against a stop with a defined push rather than a defined position.
In practice a tension application often wraps a slower outer loop around the torque command, for example a dancer or a load cell measuring actual tension and trimming the torque request to hold a setpoint. The drive still runs in torque mode at its core, producing whatever torque it is told, while a supervisory tension loop decides moment to moment what that torque should be. This layering lets the fast, direct force control of torque mode serve a higher-level goal expressed in real engineering units of tension.
Even when a drive is running in position or velocity mode, the concept of a torque limit is closely related and extremely useful. A torque limit caps the maximum torque the drive is allowed to command while it pursues its position or speed target. If the axis meets an obstruction, jams, or hits a hard stop, the drive presses only up to the limit and no further, so the mechanics, the product, or an operator's hand is protected from the full force the motor could otherwise deliver. The move still runs normally as long as the required torque stays below the cap.
This makes torque limiting a simple and powerful safety and process feature. A press can be told to move to a position but never exceed a set force, so it seats a part firmly without crushing it. A gripper can close on parts of varying size by moving until it meets resistance at a limited torque. Homing routines often drive gently into a hard stop under a low torque limit and take that point as a reference, confident that the reduced force will not damage anything. In all these cases the drive is blending position or velocity control with a ceiling drawn from torque mode's underlying current limit.
For operations watched through cloud SCADA, motor torque and current are among the most revealing signals a drive exposes, and a platform such as Merobix can trend and alarm on them across many sites. A rising torque needed to run an unchanged duty is a classic early sign of a mechanism binding, a bearing failing, or a process loading up, and a torque that suddenly pins to its limit points straight at a jam or stall. Bringing the drive's torque and current readings into a central monitoring system lets supervisors see those trends across remote machines, so a stiffening actuator or a winder working harder than usual is caught from the trend rather than after a failure in the field.
In torque mode the drive holds the commanded torque and does not regulate speed, so a constant torque applied to a load that can move freely keeps accelerating it indefinitely, just as a constant force on a frictionless mass never stops speeding it up. That is why torque mode is used against a resisting load such as material tension, or paired with a speed limit that caps how fast the motor may spin while delivering the torque.
In torque control mode, torque is the thing being commanded and the drive produces exactly the torque it is told, letting speed and position fall out from the load. A torque limit in position mode is a ceiling: the drive still aims for a position but is not allowed to exceed a set torque to get there. One makes torque the objective, the other makes torque a boundary on a position or velocity move.
Yes. Most servo drives support all three modes and can be configured for one or switched between them, because position and velocity control are built on top of the same current and torque control at the drive's core. A machine might run an axis in position mode for a move and switch it to torque mode for a tension or press phase, with the controller selecting the mode as the sequence requires.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.