Machine Design

Distribute and conquer: Ethernet tackles motion control

Jacob Tal
Galil Motion Control Inc.
Rocklin, Calif.

The DMC-3425 from Galil Motion Control Inc., Rocklin, Calif., the first controller in its E-Series, is a two-axis, Ethernetbased motion controller. It provides precise control of servomotors and includes both linear and circular interpolation algorithms in addition to PID filtering with notch and feed-forward parameters.

A typical central control system consists of a single controller residing in the host PC. Because signals from the motor, amplifier, and limit switches must connect directly to the controller, the wiring can get fairly complex.

E-Series motion controllers use Ethernet as the basis of the network and system elements. One controller serves as the master that receives all commands from the host computer. It then passes commands to slaves, each of which can control two axes of motion.

Traditionally, motion-control system architectures have fallen into two camps: central and distributed. A new Ether-net-based approach can't be categorized as either one. It overcomes the limitations of both architectures while combining their best features.

In a typical central control system, a single motion controller sits on an ISA or PCI bus in a PC. This is a fairly simple approach by which the controller communicates with the host computer via the system bus. One downside of this approach is that each motion control component (motor, encoder, limit switch) connects to the central controller via wires that can be hundreds of feet long. Of course, long wire runs can be costly and raise the possibility of noise interference.

A distributed-control system overcomes the excess wiring problem by placing a single-axis controller close to the motor. Each controller is then wired to a bus network. But it becomes more difficult for the PC to keep track of all the controllers, as the necessary programming is relatively complex.

Ethernet-based control architectures offer a third way. An Ether-net-based motion-control system consists of a host computer, an Ethernet hub, and several motion controllers. The wiring is minimal because individual controllers can sit next to the motors and amplifiers. Also, fiber optics used for communication can eliminate noise coupling and grounding problems completely. Ethernet TCP/IP protocol is also more flexible, so additional devices easily connect to the network. Readily available system components such as Ethernet cards and hubs keep costs down as well.

One disadvantage to the Ether-net-based approach is that the requisite computer program is complex. The reason is that it must handle communications with several devices simultaneously.

E-Series motion controllers are Ethernet-based, yet have qualities of a central control-system. One controller is designated as the master. It controls two motors and all remaining controllers, designated as slaves. For the purpose of control and communication, the master acts as a virtual multiaxis controller. This means that the

host computer need only communicate with the master controller to give it commands for all the axes. It is then the responsibility of the master to inform the slaves about their respective tasks. The master also holds all the data about the slaves, and the host can interrogate it about the status of any axis.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.