What’s EtherCAT? | Automation.com

What’s EtherCAT?

June 062012

EtherCAT is becoming increasingly popular for control and system engineers as a robust, high-speed real-time network for machine control solutions. With deterministic and high speed update rates and extremely precise synchronization of all network devices, machine builders and system integrators can take advantage of dedicated motion and machine controller platforms utilizing EtherCAT to easily construct a full machine control solution while still maintaining the highest level of performance and a low cost structure. In this article, we will discuss some of the most prominent benefits of EtherCAT as they relate to motion control systems, and a specific utilization of EtherCAT will be highlighted, illustrating how it enables a dedicated motion controller platform to simultaneously control up to 64 synchronized axes with a fast 20kHz servo sampling and update rate (position loop, velocity loop, current loop, and commutation) of  all axes.

What is EtherCAT and what are it’s advantages?

EtherCAT is an open industrial real-time and deterministic ethernet Fieldbus standard that is widely regarded as the best industrial network to date. It is orders of magnitudes faster than many well-known networks such as Profibus, DeviceNet, and ModbusTCP. Automation equipment vendors can utilize EtherCAT on their own device implementations to improve performance and flexibility. End users and automation system designers can also implement their own EtherCAT compliant devices for specific needs. The following advantages are some of the reasons why EtherCAT is growing so rapidly in the motion control and automation fields.

High Speed and deterministic
EtherCAT is based on the principle of ‘Ethernet on the fly’ processing, a driving factor for it’s incredibly fast cycle times for industrial automation applications. In a typical ethernet network, ethernet frames (data packets) are sent to each device, where the device then reads the data, and sends back a response to the master. This process is repeated for each device in the network until all devices are updated. The total network cycle time would be the sum of all the response times and is not deterministic. Sometimes multiple device messages can interrupt one another requiring a master to arbitrate and organize according to priority, which also adds delay to the cycle time. In an EtherCAT network, a frame is sent from the master and when the first slave device receives the frame, it instantaneously extracts the data with its address and writes any response data. This means the frame(s) effectively passes through all the slaves with negligible time delay and then comes back to the master (reference figure 1). In addition to on the fly Ethernet processing, EtherCAT also optimizes bandwidth by allowing outgoing and incoming data for multiple devices to be combined into single ethernet frames, rather that requiring a frame for each device on the network. For networks with many drives, I/Os, and other devices, the transmission overhead can be significantly reduced with this approach (reference figure 2). This efficient network design of EtherCAT makes it ideal for high bandwidth applications like multi-axis servo machine control. Sampling and updating 64 drives and many I/Os devices can be done in less than 250 microseconds!


Figure 1

 

 

Figure 2


Tight Synchronization
Multi-axis motion control networks depend heavily on the synchronization of the independent devices in order to ensure that multi-dimensional motion trajectories can be accurately executed in physical space. To synchronize all devices on a network requires transmission times between devices and drifting of clocks on different devices to be calculated and compensated for. The EtherCAT standard handles this with a mechanism called distributed clocks. The phase shift between all distributed clocks is less than 0.1 microseconds.

Low Cost
EtherCat leverages on the mass-produced Ethernet communication devices and cables used by all PC to minimize cost. Slave devices require a low cost dedicated controller.

Flexibility

EtherCAT supports a wide variety of standard application layer implementations, including CoE (CANOpen over EtherCAT), EoE (Ethernet over EtherCAT), FoE (File Transfer over EtherCAT), SoE (Servo Drive over EtherCAT), and FSoE (Safety over EtherCAT). This allows multiple vendors to make fully compliant devices with the same application implementations. Also, EtherCAT supports VoE (Vendor Specific Protocol over EtherCAT), which allows vendors to implement their own protocol for very high bandwidth applications like a dedicated high-speed multi-axis control platform, where the overhead from a standard implementation cannot be tolerated. Vendors can equip their devices with the capability to simultaneously support open standard and proprietary application implementations to provide the most efficient solution while maintaining flexibility.

Utilizing EtherCAT in a dedicated motion controller platform to provide fully coordinated control and synchronization

Traditionally, the highest performance dedicated motion controllers have been implemented on a centralized controller platform. This has been driven by the need for the controller to completely synchronize the servo updates of all axes so that multiaxis motion trajectories can be executed in a truly coordinated fashion. The centralized control structure allows the motion controller algorithms access to real time motion or servo related data essentially at the speed of the central processor. However, the centralized processor architecture suffers from a lack of resources as the amount of axes increases. As axes are added, the available processor resources for real time control must be distributed across more axes, and often the servo update rates must be reduced, ultimately lowering performance. To combat this disadvantage, ACS Motion Control utilizes a unique distributed processor architecture. This architecture relies on a powerful PC based Machine Processer Unit (MPU) to handle all the machine control tasks except for the real time control. The demanding real time control algorithms are executed by dedicated Servo Processors termed SPii’s. Each Servo Processor controls one or two axes. With this approach, as axes are added, more Servo Processors are added, always maintaining a fixed 20kHz sampling and update rate independent on the number of axes. This architecture is proven to provide extremely high performance coordinated motion control. A dedicated high-speed computer bus, such as PCI, has been required to enable the highly synchronized and fast data transfer between the MPU and the Servo Processors However, this type of data communication imposes a physical limitation - the MPU and SPii chips must reside within a couple of inches from one another. This is where EtherCAT is utilized to enable physical distribution and large distance between the MPU (which acts also as the EtherCAT Master) and the Servo Processors (which control the drives). (See Figure 3)


Figure 3

An EtherCAT based distributed processor architecture allows the necessary bandwidth, synchronization, and physical flexibility to implement a control platform with the power of a centralized structure and the advantages of a distributed network solution. The ACS MC4Unt uses this approach to control up to 64 demanding and highly coordinated axes with a 20kHz sampling and update rate (See Figure 4). Using a standard Can Over Ethernet (COE) protocol it allows the integration of many vendor devices including motor drives and I/O modules that comply with this standard. The machine builder and system integrator can use the application development environment and powerful motion programming language and tools of the dedicated control platform to minimize cost and time to market, while implementing a truly powerful and high performance motion control solution.


     
  Figure 4

Did you Enjoy this Article?

Check out our free e-newsletters
to read more great articles.

Subscribe Now

MORE ARTICLES

VIEW ALL

RELATED