X-Y-Z: New Network Combines Best of Ethernet, SERCOS
When designing the electrical scheme for an automated assembly system, engineers have many options for each level of control. High-level buses connect the machine controller with the enterprise IT system. Device-level buses connect the machine controller with I/O and the motion controller. Motion buses link the motion controller with amplifiers and servomotors.
Each bus has its own strengths and weaknesses, so perhaps it was inevitable that engineers would combine network systems to take advantage of the best each has to offer. Such was the thinking behind a new network, SERCOS III, that combines the most popular motion bus, SERCOS, and with the most popular high-level bus, Ethernet.
SERCOS, or Serial Real-Time Communications System, is an open-architecture interface standard for communication between motion controllers, digital drives and I/O. It was introduced in 1988 as an alternative to the ±10-volt analog interface.
SERCOS connects components through one or more loops. A motion controller, called the master, directs all communication within a loop. A slave is the connection between a drive and the loop. Motors, encoders, limit switches and other devices connect to the drive. Several drives can be tied into the loop through one slave connection, and one master can be connected to 254 drives.
Instead of dozens of wires in an analog system, signals are transmitted via two fiber-optic cables. This simplifies wiring, eliminates noise, and enables drives to be widely distributed. Signals travel in one direction. Information is exchanged between the master and the slaves, but not among the drives themselves.
Data is sent over the network in packets, called telegrams. There are three types of telegrams. Master synchronization telegrams are broadcast by the master at the start of a communication cycle. These short signals synchronize the timing sequence within the cycle. Master data telegrams are sent by the master once each cycle. These long signals transfer data, such as default command values, from the master to the drives. Drive telegrams are sent by the slaves to the master. These telegrams contain data on the speed, torque and position of a motor.
Signals are sent in cyclic or noncyclic modes. With cyclic transmissions, the master communicates with all the drives. Telegrams are sent quickly within a fixed cycle and are synchronized. These transmissions include servo commands and critical status information. With noncyclic transmissions, the master communicates with only one drive at a time. Telegrams are transmitted intermittently. These transmissions include tuning parameters for servomotors and end-of-travel values for limit switches.
A major benefit of SERCOS is that it enables engineers to distribute intelligence in an automated assembly system. Many functions that would normally be handled by the motion controller are shifted to the drives, which frees the controller for other tasks, such as coordinating the movement of multiple axes. SERCOS increases the speed, accuracy and smoothness of motion, and allows stations to be added or removed without major changes to the wiring.
SERCOS III was introduced in 2004, and the first controllers with the new technology debuted this year. SERCOS III enables the master to transmit any standard Internet protocol telegram, such as transmission control protocol (TCP/IP), in a non-real-time slot, in parallel with the real-time data transfer required for motion control. As a result, the master can communicate with higher-level devices in the network. Transmission speed has increased from 16 megabits per second to 100 megabits per second.
SERCOS III reduces costs by standardizing hardware. In the first generation of SERCOS, application-specific integrated circuits were needed to run the network. With SERCOS III, controllers based on less expensive standard components, such as field-programmable gate arrays, perform the sequencing and synchronization functions.
SERCOS III also enables direct, peer-to-peer communication between individual slaves and several masters. As a result, communications remain fully functional even if there is a cable fault at any point in the loop.