What is a servo system, really?

The word servo has become synonymous with the words servomotor, servomechanism, and servosystem to indicate a closed-loop controlled system.

By Hurley Gill • Senior applications and systems engineer | Kollmorgen, A Regal Rexnord Brand

A servo is most simply defined as any driven mechanism dependent on feedback data yielded by the continuous monitoring of a given process — with resulting systematic adjustments of that same primary process to achieve a target outcome.

In fact, servo is and has always been associated with some level of automated control by a servomechanism of a force, torque, velocity, and/or position. Servomechanism in turn is generally some form of a closed-loop-controlled pneumatic, hydraulic, electrical, or chemical process or form of actuation. The term servomotor is often identified as a major component of a servomechanism; it is frequently the specific component rendering a machine servo-mechanized.

Hence, within the motion and automation industries, servomotor and servomechanism are often interchangeably used for any given project or application.

A servomotor is a bidirectional servo actuator and machine under closed-loop control to yield a precise outcome from a given operation. Continuous feedback on motion-related data being monitored (for example, position, velocity, torque, or force as a function of current) is compared against a command or setpoint. This in turn spur the dynamically corrective adjustments to eliminate the difference between actual and commanded values.

Ongoing corrections for a given process may be affected by one or more closed loops receiving feedback data. Each control loop is in turn dynamically affected by the continuously received data and compared against the commands.  This active feedback providing the closed-loop information to the mechanism’s controller is continuously received from one or more feedback sensors. In some applications, one feedback device can serve multiple feedback control-loop functions via programmed algebraic manipulations of received data. For example, the proportional, integral, derivative (PID) loop closure of each control loop  may use a single feedback device for separate  velocity and position loop closures.

Big-picture question: How do servo mechanisms or motors work?

Automatic control systems are generally one of two types.
• Open-loop designs are ON-OFF systems such as heaters and air conditioners involving some controlled difference (deadband) between the turn-on and turn-off points of operation.
• Closed-loop designs are continuously operating mechanisms such as automotive cruise controls, for example. They include one or more sensors to continuously provide realtime information for a control unit to minimize any difference between the target command (setpoint) and what is actually happening (feedback).

So, in a greatly simplified sense, a servo (closed-loop) mechanism may be considered an ON-OFF control in which the process’ deadband is reduced to an infinitely small value. This small deadband value present seemingly systematic adjustments (corrections) effectively resulting in continuous comparison (closed-loop style operation) of the feedback’s realtime data (for example, actual temperature) against the commanded and desired outcome (such as some target temperature or other setpoint) for the optimization of a process.

Each servo-loop closure (closed-loop control) occurs with the process’ controller — comparing realtime data against commanded target data to get ongoing systematic corrections (adjustments) to a process for a more refined result than could otherwise be achieved. Continuous process corrections are often prompted by controlled actuation — whether pneumatic or hydraulic actuation, electric motor, or something else. These various actuators are often called servos in their realm of common usage.

A closed-loop servo controller can provide continual correction and minimization of any difference (error) between the controller’s presented commanded setpoint and the measured output (process/plant/actuator) by some feedback sensor.

In a real-world motion system application such as positioning a linear actuator, the velocity command received by the controller’s position-loop is constantly changing through accelerations, stroke reversals, decelerations, load disturbances, and so on. Getting the best possible performance relies on optimal control of the axis’ dynamic response to these conditions via PID control within each control-loop of the circuit/algorithm.

One servo-mechanism example

Consider the electrically actuated cruise control of an automobile — a simplified closed-loop velocity control.

With car cruise control, the driver sets a target velocity setpoint — the command value. Whenever the car slows, the difference between setpoint and actual velocities increase. This increasing positive error causes a servo controller to command increased power from the car’s engine that in turn increases the car’s velocity (notwithstanding a transmission gearing change) until the error is eliminated.

In a similar manner, if the car begins to accelerate down a hill and its actual velocity exceeds setpoint velocity, an increasing negative error is calculated by the servo controller. This in turn commands the engine to deliver less power. The resulting deceleration reduces actual velocity until the error that developed is eliminated.

In this design, automatic correction comes from closed velocity-loop (servo) control that continuously monitors the car’s actual speed. This is compared to the velocity set by the driver (and supplied to the controller). The speed difference has positive or negative polarity relative to the car’s slowing below or increasing above, the set speed.

Let’s expand on the automotive cruise-control example. With the addition of an external feedback signal, the design can deliver adaptive cruise control. Here, a sensor allows differential speed correction to pace a vehicle ahead. The vehicle-ahead speed is continuously compared against the controlled car speed; then if the vehicle ahead slows and some minimum distance between vehicles is breached, the servo control prompts a speed adjustment to slow the controlled car. Should the vehicle ahead resume a faster speed, the controlled car’s servo system will prompt acceleration up-to the new pace speed, up-to the driver’s previously set (setpoint) speed.

What’s more, conditions can be set to avoid an accident. This function uses continuous feedback monitoring of specific conditions such as minimum distance between vehicles (position-loop) and rate-of-change between the vehicles.

Let’s further expand on the automotive cruise-control example. The addition of the car’s realtime position continuously monitored against lane-assist camera data as well as global positioning system (GPS) data, can provide (within limits) servo-error signals to prompt systematic adjustments necessary to keep the car on the road. In the near future, such technology could even make common the autonomous driving to a preprogrammed destination.

Reconsidering assertions about servo systems

Servomechanisms are most often used for their ability to provide the precision position, velocity, or torque (or force) needed by demanding designs. Increasingly challenging and complex, requirements for manufacturing and other applications has spurred ever-increasing usage of servomotor systems — and their continued evolution for ever-higher performance.

Think of a servo mechanism like a natural ecosystem. Nature likes balance (defined setpoints) so everything is continuously working towards this targeted balance point of equilibrium. This targeted equilibrium is itself continually changing over time, all the while nature is continuously and consequentially adjusting in multiple (±) ways to accommodate all changes. Those changes are due to one or multiple disturbances and those changes due to consequential point-of-balance changes (within the realm of capability of course.) It’s the operational limits, the realtime sample rate of each individually monitored feedback system and the interacting response times of those individually monitored natural systems, acting consequentially against the ever-changing longer-term point-of-balance that makes the difference.

Note: It’s not always recognized that during servo operation, typical servomotor applications (unlike an open-loop motor) will always perform in specific ways.

1. A typical servomotor application will always use realtime feedback information such that process or mechanism adjustments can be automatically made — in other words, adjustments against disturbances, load changes, and command changes, and so on.

2. A typical servomotor application will use only the energy necessary for the axis or mechanism to complete what it’s commanded to do.

3. A typical servomotor application will operate in one of three operational modes depending on the application and selected control method. These include current/torque mode (with current-loop closure); velocity mode (with vel-loop closure); or position mode (with pos-loop closure).

Current-loop PID closure can be operated on its own because it’s the innermost and fastest control loop. In contrast, the velocity-loop PID closure includes an inner current loop … and position-loop PID closure includes both an inner velocity and current, loop.

4. A typical servomotor application will continuously make small adjustments about its commanded actuation by exhibiting small movements or changes, about their setpoints.

• With velocity control, expect small back-and-forth or clockwise and counterclockwise (CW, CCW, CW, CCW) motions while the system tries to maintain a commanded steady-state velocity.

• With position control, expect small continuous-position adjustments (hunting) about the final commanded position. If sufficient stiction exist while hunting for a stationary position, the resulting dither due to this friction may undesirably enhance displacement deviations.

5. A typical servomotor application will continuously correct for any mechanism or process disturbance within its designed capability to correct. That means each servo mechanism operates only within designed limits — its envelope of operation — to maintain servo action. Its ability to automatically adjust against things such as process disturbances, changing external forces, commands, and so on, must be maintained for normal servo operation.

In contrast, when a servo mechanism is intentionally or mistakenly operated at one of its physical-operational limits or some saturated condition, by definition it is no longer operating as a servo system. In this case, servo action as defined is not available for required changes and/or corrections necessary to accommodate disturbances, external force changes, or command changes, with adjustments only available in one direction.

Finally, when a servo mechanism is required to operate in a non-normal way for some part of its operation cycle (for example, while holding position under load) each special condition must be specifically considered to ensure the proper sizing of the servosystem (including servomotor and drive).

Kollmorgen, A Regal Rexnord Brand | kollmorgen.com