Technical Information
Overview of Motion Control Using Electric Motors
Motion control is the process of controlling an electric motor so that it reaches a desired state. This desired state could be a velocity or a position. A controller generates a commanded position and appropriately varies the current to the motor to make the actual position (as read from a feedback device) match the commanded position within some tolerance. The difference between the commanded position of the motor and the actual position of the motor is called the following error. The allowable following error for a system depends on the application. In precision applications, such as wafer dicing saws the allowable following may be as small as several microns. In other applications, such as theme park rides where acceleration and velocity are the important performance parameters, the allowable positioning following error may be several inches or more.
The four main components in a motion control system are the feedback device(s), the controller, the amplifier, and the motor. In a typical control system, the feedback device(s) supply the controller with information about the motor’s position. The controller uses this information to calculate the desired current for the motor. The amplifier then provides power for the motor and monitors the motor current to insure that it matches the desired current. Finally, the motor converts this current into a torque (or force) which produces a desired effect on a system.
Position and Velocity Feedback Devices
The control of a motor may be either open-loop or closed-loop. An open-loop motor is a motor that operates without the aid of feedback devices. Examples of motors that are typically run without feedback include stepper motors and some AC induction motors. A closed-loop motor uses a feedback device to compare the actual position of the motor to the desired position. Feedback devices allow the motor controller to position the motor precisely and to monitor the motor position. Position feedback devices are by far the most common, although some systems also employ velocity or acceleration feedback devices.
Encoders
An encoder is a position feedback device that measures rotary or linear motion of the motor (or other element in the system). There are magnetic, contact, resistive, and optical encoders. Optical encoders are most commonly used due to their availability and accuracy. Optical encoders are available in either incremental or absolute configurations for both rotary and linear motion.
An incremental encoder usually consists of a metal or glass grating, a light source, and a detector. The grating has uniformly spaced windows that act as shutters to provide the detector with a strobe type of effect. By counting the number of cycles, the encoder is able to establish how far the motor has moved from an initial position. In order to attain absolute position feedback, an incremental encoder must be "homed," where the motor is driven to a position and the encoder is set to a predetermined value. Usually, there is a once per revolution index pulse on the encoder. This gives the encoder an absolute starting position from which subsequent moves can be accurately determined. Incremental encoders frequently have two light source/detectors, arranged so that the direction of travel can be sensed. The light source/detectors are arranged to be 1/4 count out of phase. This effectively gives 4 state changes for each line on the encoder, increasing the resolution by a factor of 4. This type of resolution extension is often referred to as A Quad B output. Incremental encoders can be sensitive to the environment in which they are used. If they are to be used in a dirty or moist environment, special care has to be taken in order to ensure these contaminants do not affect the encoder. Noise on the signal lines can degrade accuracy, either adding or losing counts. There are various schemes to ensure that any noise that may degrade the signal is minimized, including using complementary outputs, shielded twisted pair wire, and special enclosures for the equipment.
Absolute encoders typically consist of a disk with several sets of concentric apertures that provide a unique pattern of signals for the position of the motor. Each of the aperture patterns has a light source and detector, and the signals go into separate channels in the encoder. In this manner, the position of the encoder is known without having to go through a homing routine required for incremental encoders. Absolute encoders are less susceptible to data loss during power failures, noise on the signal lines, and are not limited by the bandwidth of the counting device (like on an incremental encoder). Absolute encoders generally cost much more than an incremental encoder of the same resolution.
Resolvers
A resolver is a form of absolute encoder that operates on the same principal as an electric transformer. A typical resolver consists of a set of three windings. Two of the windings, oriented 90° apart, are exposed to a magnetic field produced by supplying current to the third rotating coil. The currents induced on the static coils (stators) are measured. The magnitude and relation of these induced currents establish the position and direction. Resolvers are relatively inexpensive and accurate. One drawback to resolvers is that an additional piece of electronics, a Resolver-to-Digital converter is usually needed to incorporate the resolver feedback into the control system.
Linear Variable Differential Transformers (LVDTs)
Like a resolver, an LVDT uses electromagnetic induction to detect changes in position. Current is induced into two fixed coils by a moving Ferro-magnetic core which is excited by a central coil. As the core moves, the induced current in the two passive coils changes. To use this as a measuring device, a stable AC source must power the excited coil. The output signal is run through a signal conditioner to covert the AC signal to a DC voltage. LVDTs are very effective over limited travel ranges, and if mounted properly, can have an extremely long functional life.
Laser Interferometers
Laser interferometers are a form of position feedback device that utilize the wavelength of light as their measurement unit. Optics are used to create and interfere two beams of light that travel along different paths so that phase shifts between the two beams can be detected and converted into a measurement. Optics can be arranged to measure linear displacement, angular motion, straightness, and squareness. For linear displacement applications the optics are arranged so that the beam coming from a laser head is split using an optic called a beam splitter. One portion of this beam travels to a fixed mirror (reference beam) and the other portion travels along a path to a mirror attached to the moving object (measurement beam). The mirrors return the two light beams back into the beam splitter where they are recombined for interference and brought into a photodetector so that the displacement information can be extracted from the resulting electrical signal.
There are two basic kinds of interferometers, homodyne and heterodyne. Homodyne interferometers use laser light operating at a single frequency and measure changes in light intensity to determine the phase shift between the measurement and reference beams. As such they can be prone to errors due to inadvertent changes in intensity. Heterodyne interferometers use laser light operating at two frequencies typically differing between 2 to 20 MHz. They measure changes in frequency due to Doppler shifts to determine the phase shift between the measurement and reference beams. They have the advantage of not being as sensitive to intensity changes as homodyne systems and usually are capable of finer resolutions.
An important thing to know about interferometers is that they are incremental devices and require homing to some reference like incremental encoders. Also if their beam is blocked during measurement, they loose measurement counts and must be rehomed. As such they require adequate shielding from obstructions such as cooling fluid and chips. They are extremely accurate measurement devices and can typically achieve 1.0 part in 10E6 accuracy ( 1m m over 1 meter) in air. They can be purchased with subnanometer resolution. They are typically used on wafer stepper machines for populating silicon wafers with integrated circuits.
Tachometers
Tachometers are velocity feedback devices that are frequently incorporated in the design of a DC servomotor. The tachometer is similar to a small, permanent magnet DC motor that is driven by the shaft of the motor and outputs a voltage proportional to the shaft speed. Tachometers should be matched to the operating needs of the system to ensure linearity of the output over the operating regime. Temperature compensation should also be considered if the motor/tachometer combination is subject to varying conditions. Tachometers are typically used for low speed applications.
Controllers and Amplifiers
The controller and amplifier are responsible for calculating and providing the proper current to the motor. This is accomplished in the following four steps:
First, the desired motor position (or velocity) is calculated. For a CNC machine, this position is often derived from G-Codes. Other ways to input the desired motor position include operator handwheels or data files.
After the desired motor position is calculated, feedback devices are used to determine the actual position of the motor. When controlling the position of a motor, the controller typically needs information about both the position and the velocity of the motor. The position can be read using any of the position feedback devices listed above. The velocity of the motor can then be determined either by differencing two successive readings from the position feedback device or using an additional feedback device such as a tachometer.
The control loop uses information about the commanded and actual position of the motor to calculate the desired torque output from the motor. Since the torque output of a motor is typically proportional to the motor current, the output of the control loop can be thought of as a current command for the motor. By far the most common type of control loop is a PID (Proportional-Integral-Derivative) controller. The gains of the PID controller are adjusted so as to minimize the error between the desired motor position and the actual motor position while avoiding instabilities.
Finally, the actual motor current must be set equal to the commanded motor current. This is usually done using PWM (Pulse Width Modulation) control. Current sensors are used to monitor the actual current.
The division of labor between the controller and the amplifier can take on several different forms. Typically, the controller calculates the desired motor position, reads the feedback devices, and calculates the desired motor current. The amplifier then provides the current to the motor by switching power transistors. Some newer controllers are able to monitor the actual current, close the current loop, and provide PWM outputs. In this configuration, the amplifier can be reduced to a simple combination of a DC rectifier and an array of transistors.
A PID Control Loop
(FIGURE OF PID LOOP)
As shown above, the PID controller subtracts the actual motor position from the commanded motor position to calculate the following error for the motor. This following error is multiplied by the proportional gain of the controller to provide stiffness for the system. Systems may be tuned to provide high stiffness or they may be designed to allow compliance in their axes.
Next, the following error of the system is integrated over time and stored in an accumulator. This accumulated position error is multiplied by the integral gain of the controller. This provides additional motor stiffness at low frequencies. Integral gain is useful in providing exact positioning in the presence of friction or gravitational loads.
Finally, the velocity of the motor is multiplied by a derivative gain to provide damping for the system. It is usually desirable to design a system with a damping ratio near 70%. This provides a compromise between system response time and overshoots.
PWM Amplifier Control
Pulse Width Modulated (PWM) amplifiers are the most cost effective and efficient way to amplify the output signals from a controller. A PWM amplifier rapidly switches the voltage across the motor coils between the DC bus voltage, and the negative DC bus voltage. When the voltage across the motor is switched to the DC bus voltage, current through the motor will increase. When the voltage across the motor is switched to the negative DC bus voltage, current through the motor will decrease. By varying the duty cycle during this switching, the motor current can be accurately controlled. The duty cycle is determined by comparing the desired motor current to the actual motor current each PWM cycle.
Motors
DC Brush Motors
There are several types of DC brushed motors. The most common one is the iron-cored motor. This motor has two major parts, the housing and the rotor. The housing contains the field magnets and the brushes. The rotor consists of coils of wire, an iron core, and the commutator. Two other types of brushed motors are the disk-armature motor and the shell-armature motor.
The rotor is divided up into segments, with each segment having a contact area for the brushes. During rotation, the segments are sequentially energized, creating a magnetic field that interacts with the magnetic field from the magnets and produces the torque that turns the motor. One of the performance limiting factors in brushed motors is the dissipation of the heat that is generated during use. Heat is generated by the electrical resistance of the armature, friction, and other Electro-mechanical losses.
DC Brushless Motors
A brushless motor is constructed with the magnets on the rotor, and the field windings on the motor housing. There are several advantages gained over a brushed DC motor. The shorter thermal path to the ambient air from the windings improves the dissipation of heat. The motor is commutated by an external source, so the increased wear and maintenance required by the brushes on the commutator are not required. This could also be a disadvantage due to the added complexity of the system with the external commutator. One other potential disadvantage is that in order to achieve high torque and low inertia, special magnets (rare earth) are required.
Linear motors are a form of brushless motor that have the rotor "unrolled" to form a linear beam of magnets with a traveler that contains the coils. They are classified as either iron-core or ironless depending on whether the windings are wrapped on an iron core.
AC Induction Motors
AC Induction motors are different from DC brushed or brushless motors in that there are no permanent magnets either on the rotor or the housing. Instead, the AC induction motor uses two sine wave inputs that are 90 degrees out of phase applied to the stator windings. These inputs induce a current in the rotor windings, which produce the force or torque to turn the motor. The motor's speed is limited by the frequency of the wave, and the torque is proportional to the amplitude of the inputs.
Stepper Motors
Stepper motors are synchronous motors whose position can be controlled to a degree of accuracy without the added expense and complexity of a sophisticated controller system. This is due to the construction of the motor itself, where the motor rotates a finite amount with each application of current to a set of windings. Assuming that the motor’s limits are not exceeded, the controller without any feedback from the motor or load determines the position of the load from the known output to the motor. Typically, these are used in less precise applications such as indexing tables or conveyors.
Return to Perry Automation's home page.