Motion control is a technology and science that controls motors and can influence the trajectory of motion. Understanding motion control architecture can help determine whether and when to use motion control networks during the decision-making process.
Whether moving a test tube or cutting metal, the motion controller is responsible for planning the motion trajectory, driving and monitoring the motors, and periodically submitting status updates to a higher-level controller. Two main control structures are used in the design of motion control systems: centralized and distributed. The introduction of high-speed, low-cost digital control networks has provided new options for building distributed control systems. With the advent of higher-power, more compact switching amplifiers, centralized designs are gradually increasing the number of controllers that can be placed on the same printed circuit board.
Understanding these technological trends helps explain how and when to apply these two different control structures.
Types of motion control applications
Centralized or distributed control, which control method is more suitable: The characteristics of the application control problem have a decisive influence on this.
In planar motion control applications, motor control is more or less accomplished by a centralized PC or controller; there are also some hierarchical applications where motion axes are divided into 2, 3 or more functional axes; and in some stand-alone applications, the operation of the machine tool controller is largely not connected to the network and does not rely on network monitoring.
An example of planar motion control: a printing press with multiple axes, whose axes are controlled by servo controllers. Here, time is a critical factor. The main controller, typically a PC or PLC, must synchronously drive all axes. Typical instructions are: "Move axis #1 to position X, move axis #2 to position Y," and so on.
An example of hierarchical motion control applications is a semiconductor wafer processing system with a master robot (4 axes), wafer positioners (3 axes), and a valve controller (1 or 2 axes). In this architecture, the network typically connects the local robot or valve controller to a central network, but the actual motion control is performed by the local robot, wafer positioner, or valve. The overall machine controller does not issue commands like "move machine axis #2 to position 12345," but rather commands such as "extend the robotic arm," which are interpreted and executed by the local robot controller.
An example of a standalone application is a tape archiving system, which allows an operator to approach the control panel and request the retrieval of a specific tape. These standalone controllers can execute a series of robotic arm movements based on commands issued by the local operator, such as "retrieve tape #1234". In this application, if a network connection is present, it is used only for reporting and monitoring functions, not for control itself.
Add distributed motion driver
Understanding which practical devices can be used in motion control systems is equally important. Two available devices are distributed drives and machine control cards. Although there are many different variations of these devices, they all boil down to one of the two types mentioned above.
Distributed motion controller drivers, sometimes called smart amplifiers, communicate with a central host via a network to provide a set of motion control functions, such as contour generation, loop closure, or amplification.
Depending on the application, two types of distributed drivers are available. The first type, which can be called a tightly coupled driver, is used with high-speed, deterministic networks such as SERCOS, EtherCAT, or EthernetPOWERLINK. The second type, which can be called a loosely coupled driver, uses low-speed networks such as Ethernet protocols, CAN buses, and RS485.
Tightly coupled drives require motion cards or dedicated software running on a PC to synchronize and coordinate the movement of each axis. Each drive can receive position and velocity updates at a rate of thousands of times per second. Loosely coupled drives are also host-controlled, but the drives handle more powerful contour cutting but also have greater latency. Instructions similar to "move the axis to position X using a point-to-point S-curve" are sent to each drive. Interactions within these drives tend to be automated, using input from local sensors to start or stop the motion.
Selection of sports networks
The networks that can connect to these devices can be sports-specific networks, such as SERCOS, or networks that can be used for both sports and other functions. These common networks include RS-485, CAN bus, EtherCAT, EthernetPOWERLINK, Profibus, Interbus-S, and Ethernet.
What communication protocols do these networks use? The most common is the CANopen (master over CAN and EtherCAT buses) protocol. Using this protocol, sensors and motion actuators that can connect to CANopen can be purchased directly from the market.
A completely plug-and-play standard is not realistic at this stage. This is because many suppliers have incorporated CANopen motion extension components into their products, making these products incompatible with those of other suppliers.
Although the choices may be a bit vague, for most machine design users, there are actually only three options: RS-485 (a bit old, but reliable), CAN bus/CANopen, and Ethernet or any deterministic Ethernet (EtherCAT or EthernetPOWERLINK, etc.).
Machine control card
Machine control cards, also known as motion control cards, are the primary replacement for distributed drivers. The main difference is that motion control cards connect to distributed motherboards or processor cards via a backplane bus. However, here we refer to standalone single-card controllers and backplane motion control cards as machine controller cards.
In motion controller solutions, the microprocessor contains application code, and the motion controller IC (motion processor) is used to generate contours, servo loop closures, and manage time-critical elements in axis control. The microprocessor and motion processor for machine applications can be the same, especially for simple control applications. One advantage of machine controller cards is their improved maintainability, as repairing the entire controller card involves replacing the card itself. Wiring is also reduced because amplifiers are located on the card. Furthermore, the card's physical dimensions and connectors can be tailored to suit specific applications.
Machine control cards come in two main types: off-the-shelf and custom-made. Off-the-shelf cards, especially bus-connected motion cards, have been around for a long time, with many suppliers available. Custom-made cards, while requiring more design work, are also a viable option. A significant trend is the integration of amplifiers (ICs or module-based) onto the cards. Another trend is the use of IC-based off-the-shelf motion controllers to provide contour generation, servo loop closure, communication, and time-critical functions such as automatic safety response, programmable interrupts, and other types of automatic motion axis management.
How to choose a network
So what factors should be considered when selecting a motion control network? When considering a distributed motion network, first consider what type of signal is required in the application. Does the motion characteristic depend on signals from other parts of the machine? Are sensors and other non-motion control actuators, such as relays, also on the network bus? If an error occurs, how quickly does the motion need to stop?
When considering how and to what extent network-based solutions affect the mechanical organization of interconnected machines, another crucial factor to consider is: "How can the machine be maintained if electronic components are distributed throughout?" While traditional rack-mounted systems maintained by technicians may have a tangled mess of wires, it's important to note that they are still under the same roof. Maintainability and cost over the entire lifecycle will significantly influence the design choices for the control system.
Due to factors such as weight, heat, or other environmental constraints, implementing distributed control by placing amplifiers near motors is sometimes impractical. Traditional control rack cabinets can be isolated from the machine's operating environment through air conditioning systems and insulation. However, this is often not feasible for distributed control.
So when should one control scheme be chosen over another? There's no simple answer. For a given application, sometimes both structures work well.
The more cost-sensitive the engineering application, the more likely designers are to design their own cards and integrate onboard amplifiers, depending on the power consumption level. When designing cards, the necessary connectors can be selected and the shape elements of the card can be determined based on the specific motion application.
Applications requiring high synchronization, such as machine tools, naturally favor multi-axis motion cards or tightly coupled distributed drive solutions. These drives offer greater flexibility in terms of motor type and power consumption range. However, it's important to remember that a motion control card or a PC with dedicated G-code software is needed to generate the overall path.
For large and medium-sized applications, such as pharmaceutical automation, semiconductor automation, scientific instrumentation, and low-power general-purpose automation, several different solutions can be implemented, including: off-the-shelf machine control cards, customer-customized machine control cards, or loosely coupled distributed drives. Motion control is a technology and science of controlling motors, which can affect the trajectory of motion. Understanding motion control architecture can help determine whether and when motion control networks are needed during the decision-making process.
Whether moving a test tube or cutting metal, the motion controller is responsible for planning the motion trajectory, driving and monitoring the motors, and periodically submitting status updates to a higher-level controller. Two main control structures are used in the design of motion control systems: centralized and distributed. The introduction of high-speed, low-cost digital control networks has provided new options for building distributed control systems. With the advent of higher-power, more compact switching amplifiers, centralized designs are gradually increasing the number of controllers that can be placed on the same printed circuit board.
Understanding these technological trends helps explain how and when to apply these two different control structures.
Types of motion control applications
Centralized or distributed control, which control method is more suitable: The characteristics of the application control problem have a decisive influence on this.
In planar motion control applications, motor control is more or less accomplished by a centralized PC or controller; there are also some hierarchical applications where motion axes are divided into 2, 3 or more functional axes; and in some stand-alone applications, the operation of the machine tool controller is largely not connected to the network and does not rely on network monitoring.
An example of planar motion control: a printing press with multiple axes, whose axes are controlled by servo controllers. Here, time is a critical factor. The main controller, typically a PC or PLC, must synchronously drive all axes. Typical instructions are: "Move axis #1 to position X, move axis #2 to position Y," and so on.
An example of hierarchical motion control applications is a semiconductor wafer processing system with a master robot (4 axes), wafer positioners (3 axes), and a valve controller (1 or 2 axes). In this architecture, the network typically connects the local robot or valve controller to a central network, but the actual motion control is performed by the local robot, wafer positioner, or valve. The overall machine controller does not issue commands like "move machine axis #2 to position 12345," but rather commands such as "extend the robotic arm," which are interpreted and executed by the local robot controller.
An example of a standalone application is a tape archiving system, which allows an operator to approach the control panel and request the retrieval of a specific tape. These standalone controllers can execute a series of robotic arm movements based on commands issued by the local operator, such as "retrieve tape #1234". In this application, if a network connection is present, it is used only for reporting and monitoring functions, not for control itself.
Add distributed motion driver
Understanding which practical devices can be used in motion control systems is equally important. Two available devices are distributed drives and machine control cards. Although there are many different variations of these devices, they all boil down to one of the two types mentioned above.
Distributed motion controller drivers, sometimes called smart amplifiers, communicate with a central host via a network to provide a set of motion control functions, such as contour generation, loop closure, or amplification.
Depending on the application, two types of distributed drivers are available. The first type, which can be called a tightly coupled driver, is used with high-speed, deterministic networks such as SERCOS, EtherCAT, or EthernetPOWERLINK. The second type, which can be called a loosely coupled driver, uses low-speed networks such as Ethernet protocols, CAN buses, and RS485.
Tightly coupled drives require motion cards or dedicated software running on a PC to synchronize and coordinate the movement of each axis. Each drive can receive position and velocity updates at a rate of thousands of times per second. Loosely coupled drives are also host-controlled, but the drives handle more powerful contour cutting but also have greater latency. Instructions similar to "move the axis to position X using a point-to-point S-curve" are sent to each drive. Interactions within these drives tend to be automated, using input from local sensors to start or stop the motion.
Selection of sports networks
The networks that can connect to these devices can be sports-specific networks, such as SERCOS, or networks that can be used for both sports and other functions. These common networks include RS-485, CAN bus, EtherCAT, EthernetPOWERLINK, Profibus, Interbus-S, and Ethernet.
What communication protocols do these networks use? The most common is the CANopen (master over CAN and EtherCAT buses) protocol. Using this protocol, sensors and motion actuators that can connect to CANopen can be purchased directly from the market.
A completely plug-and-play standard is not realistic at this stage. This is because many suppliers have incorporated CANopen motion extension components into their products, making these products incompatible with those of other suppliers.
Although the choices may be a bit vague, for most machine design users, there are actually only three options: RS-485 (a bit old, but reliable), CAN bus/CANopen, and Ethernet or any deterministic Ethernet (EtherCAT or EthernetPOWERLINK, etc.).
Machine control card
Machine control cards, also known as motion control cards, are the primary replacement for distributed drivers. The main difference is that motion control cards connect to distributed motherboards or processor cards via a backplane bus. However, here we refer to standalone single-card controllers and backplane motion control cards as machine controller cards.
In motion controller solutions, the microprocessor contains application code, and the motion controller IC (motion processor) is used to generate contours, servo loop closures, and manage time-critical elements in axis control. The microprocessor and motion processor for machine applications can be the same, especially for simple control applications. One advantage of machine controller cards is their improved maintainability, as repairing the entire controller card involves replacing the card itself. Wiring is also reduced because amplifiers are located on the card. Furthermore, the card's physical dimensions and connectors can be tailored to suit specific applications.
Machine control cards come in two main types: off-the-shelf and custom-made. Off-the-shelf cards, especially bus-connected motion cards, have been around for a long time, with many suppliers available. Custom-made cards, while requiring more design work, are also a viable option. A significant trend is the integration of amplifiers (ICs or module-based) onto the cards. Another trend is the use of IC-based off-the-shelf motion controllers to provide contour generation, servo loop closure, communication, and time-critical functions such as automatic safety response, programmable interrupts, and other types of automatic motion axis management.
How to choose a network
So what factors should be considered when selecting a motion control network? When considering a distributed motion network, first consider what type of signal is required in the application. Does the motion characteristic depend on signals from other parts of the machine? Are sensors and other non-motion control actuators, such as relays, also on the network bus? If an error occurs, how quickly does the motion need to stop?
When considering how and to what extent network-based solutions affect the mechanical organization of interconnected machines, another crucial factor to consider is: "How can the machine be maintained if electronic components are distributed throughout?" While traditional rack-mounted systems maintained by technicians may have a tangled mess of wires, it's important to note that they are still under the same roof. Maintainability and cost over the entire lifecycle will significantly influence the design choices for the control system.
Due to factors such as weight, heat, or other environmental constraints, implementing distributed control by placing amplifiers near motors is sometimes impractical. Traditional control rack cabinets can be isolated from the machine's operating environment through air conditioning systems and insulation. However, this is often not feasible for distributed control.
So when should one control scheme be chosen over another? There's no simple answer. For a given application, sometimes both structures work well.
The more cost-sensitive the engineering application, the more likely designers are to design their own cards and integrate onboard amplifiers, depending on the power consumption level. When designing cards, the necessary connectors can be selected and the shape elements of the card can be determined based on the specific motion application.
Applications requiring high synchronization, such as machine tools, naturally favor multi-axis motion cards or tightly coupled distributed drive solutions. These drives offer greater flexibility in terms of motor type and power consumption range. However, it's important to remember that a motion control card or a PC with dedicated G-code software is needed to generate the overall path.
For some large and medium-sized applications, such as medical automation, semiconductor automation, scientific instrumentation and low-power general automation, several different solutions can be used, including: off-the-shelf machine control cards, customer-customized machine control cards, or loosely coupled distributed drives.
