Bench Talk

Motion-Trajectory for Motor Control

(Source: Viacheslav Yakobchuk/stock.adobe.com)

Motion control is the dynamic process of controlling the performance of a motor and its load to optimally reach and maintain desired operational performance. Doing so requires qualitative answers and quantitative numbers related to many issues.

For example, is the primary goal to achieve an accurate position or velocity, or just the final position or velocity? Does the optimum solution mean reaching an objective as quickly as possible, with the accuracy required? What about overshoot? What about changes in load? What about power efficiency and handling inevitable faults such as stalls?

Meeting all these goals is not simple, with the added difficulty that they often conflict with one another to some extent. Implementing the solution is the role of the motor's motion controller. This controller acts as the brains of the control system, implementing algorithms that have been designed and tuned to the objectives of the application.

These algorithms, which can be in software, firmware, or hardwired, must also consider the motor type: a stepper, brushed, brushless direct current (BLDC), alternating current (AC), or other type. They must also consider the nature of load (e.g., solid, liquid, powder, gears, rails) as well as any linkages and backlash.

The motion controller outputs a modulated voltage or current to drive the motor, achieving the desired performance based on high-level instructions from the system processor. As such, the motion controller is the intelligent interface between the higher-level system processor and the drivers that power the motor, while some architects also include an optional motion sensor that provides feedback (Figure 1).

Figure 1: Simplified block diagram of a typical motion control system. (Source: Mouser Electronics)

This modulation is accomplished by varying the voltage, current, signal frequency, or pulse-width modulation (PWM) of power of the drive voltage/current to the motor and is a function of the motor type.

Parameters and Priorities

Motor and motion control gain added complexity from the many related factors that characterize the priorities and optimum performance of a motor controller to yield the desired motor-response trajectory. These factors include the following:

• Absolute rotational positioning and accuracy
• Rotational speed (rpm) and accuracy
• Acceleration
• Deliverable torque
• Overshoot on position and speed changes
• Required time and ramp rate to reach desired position or speed
• Acceleration constraints and minimum/maximum rate of change in acceleration (formally designated as "jerk")
• Up-front costs, as well as ongoing operating and energy costs

For example, some applications prioritize rapid transfer times between points A and B on the position or velocity target. These include gantries, high-speed stitching, electronic pick-and-place machines, plotters, robotic arms, antenna/tracking systems, and 3D-printing systems. Other applications, such as medical or chemical analysis automation, elevators, medical diagnostic scanners, semiconductor processing equipment, and scientific instrumentation, focus on smooth motion with minimal jerk, even if it means reduced speed. Many applications require both fast motion and precise end-point positioning precision.

The motor drive and control analysis must also account for the specifics of the motor’s load in addition to the motor itself. Issues to consider include load and linkage mass, nature (i.e., stiff or flexible), and load changes (e.g., bottles being filled). Additionally, system resonance and inertia must also be evaluated.

Open Loop or Closed Loop

Another major consideration in motor and motion control is whether feedback from the motor or its load will support closed-loop control. The feedback sensor is usually a rotary encoder, optical encoder, or Hall effect device—each offering tradeoffs in cost, resolution, ruggedness, and installation difficulty.

The decision to use feedback depends on the accuracy, performance requirements, and priorities. If feedback is implemented in a closed-loop system, the motion controller must incorporate a closed-loop algorithm, typically a proportional-integral-derivative (PID) scheme or a variation of this commonly used approach.

Recently, there has been a trend toward sensorless feedback, with motor feedback information derived from real-time measurement of the motor-phase currents or back electromotive force (EMF) rather than a discrete sensor. Doing this with field-oriented control (FOC) algorithms (also called vector control) is a viable option because of the availability of microcontrollers that can execute the numerically intense and sophisticated FOC algorithms in real time.

Trajectories Critical to System Performance

Directing the motor and its load to transition from an initial position or velocity to the target goal while also managing acceleration, load changes, and other system dynamics—and doing so with the required precision—is a complicated, multifaceted topic.

However, designers often use one of two basic approaches because of their simplicity and effectiveness: the trapezoidal profile and the S-curve profile. These can be considered in the context of a point-to-point move, where the load accelerates from a stop to a constant velocity and then decelerates so that the final acceleration and velocity are zero upon reaching the programmed destination.

The trapezoidal profile is the more straightforward method. It consists of three phases: constant acceleration, constant velocity, and constant deceleration. Between each phase is an instantaneous drive-current transition (Figure 2). While this simple profile efficiently minimizes travel time, it can present practical problems.

Figure 2: The trapezoidal motion profile—velocity and acceleration versus time—provides the basic function needed and is easily understandable, but has some unavoidable performance characteristics that are often unacceptable.(Source: Mouser Electronics)

In many cases, the jerk at the beginning and end is generally unacceptable. It can cause disruption or breakage at the load, is often physically demanding on the motor assembly itself, and may trigger system resonances.

An alternative to the trapezoid is the S-curve (Figure 3), a superset of the trapezoidal profile.

Figure 3: The S-curve profile adds a transitional easement to the sharp transition between motion phases. (Source: Mouser Electronics)

The S-curve motion profile consists of seven phases. In Phase I, the load moves from rest with a linearly increasing acceleration until it reaches maximum acceleration. Phase II involves accelerating at this maximum rate until it begins to decrease as it nears maximum velocity. In Phase III, acceleration linearly decreases until it reaches zero. During Phase IV, the velocity remains constant until deceleration begins. These deceleration phases—Phases V, VI, and VII—follow a pattern symmetrical to Phases III, II, and I.

Besides these basic ramping profiles, many vendors provide proprietary enhancements designed to improve various aspects of performance. It's important to note that the simpler trapezoidal profile includes only three phases of the S-curve profile: II (constant acceleration), IV (constant velocity), and VI (constant deceleration).

Of course, designing a desired profile is different from implementing it. System inertia, resonances, and other factors significantly complicate the situation, along with added complexities such as friction, flexing, backlash, and slack. Determining the curvature of the S part of the profile (if used) is just one of the challenges the designer must address.

For closed-loop systems, designers must tune the loop parameters to match the loop and load dynamics—a task that can range from relatively simple to very challenging depending on the nature and consistency of the load. Modeling the control system and loop, as well as simulating its performance, involves complex equations or specialized software packages (such as those from MathWorks or COMSOL) that offer tools to simplify this process significantly.

Motion Control Hardware

Designers have myriad choices for motion-control hardware, including ICs and embedded modules. There are several reasons for the abundance of vendors in this market. For one, it is a rapidly growing industry covering a wide range of motor sizes and complexities. Improved control in this field offers significant benefits in performance and cost. Additionally, many design-ins have long lifespans, and it is not a commodity application. As such, vendors offer proprietary solutions with unique advantages, giving them a marketing and technical edge.

Potential users of advanced motion-control products can explore various classes of solutions:

• Many major broadline suppliers of analog and mixed-signal ICs have adapted programmable devices (such as microcontrollers, microprocessors, and even FPGAs) for these applications. Users often combine these with suitable power devices that can drive the motor, which may range from small sub-fractional horsepower units of less than 1A to larger ones rated at tens or hundreds of amps (and horsepower).
• Some suppliers offer pre-programmed ICs, boards, or modules designed specifically for motion control. These suppliers may have additional expertise and experience.
• Also available are boards and modules that include a motion-control IC core plus all the peripheral functions for system and supervisory processor I/O, as well as motor drivers, communications links, and other motor-related functions. The end product is a complete, ready-to-go motor controller with user-settable parameters.

Conclusion

Motion and motor controllers now offer optimized capabilities and drive. This allows designers to maximize system performance with respect to their priorities. These priorities include accuracy, repeatability, efficiency, and performance, with both static and dynamic changes in load, line, and other conditions.