When it comes to Industrial Ethernet fieldbus protocols, it can quickly become confusing to understand how they work and what some of the differences are. We as engineers get excited about the latest and greatest hardware, but we must remember that behind the scenes is a communication network that is helping define the performance of that hardware. In order to better understand the impact that a communication network can have on an industrial control system, let’s speak a bit to Ethernet POWERLINK. POWERLINK is frequently employed by IFP Automation on systems as our standard when utilizing B&R hardware for optimal system performance.
What is the difference between standard Ethernet and Industrial Ethernet?
One of the major differences between standard Ethernet protocol (or TCP/IP) and Industrial Ethernet is collision avoidance. The data transfer in standard Ethernet may have irregular delays, which is unacceptable when trying to achieve hard real-time performance. In circumstances where irregular delays can be acceptable (office computers, printers, serving web pages, etc.), Ethernet TCP/IP is perfectly fine. But in industrial motion control applications, a solution for hard real time communication is needed.
Because of this, an Industrial Ethernet protocol must use preventative methods that avoid such collisions. The data transmission must also occur within a very specific window of time, which is referred to as jitter. Soft real-time performance can tolerate a limited amount of jitter and may have cycle times of up to several hundred milliseconds. Soft real-time performance could be used for applications such as temperature monitoring. Hard real-time performance on the other hand would be used for applications such as motion control, and commonly requires sub-one-millisecond cycle times.
Ethernet POWERLINK was developed by B&R and introduced in 2001. In 2003, management was taken over by the Ethernet POWERLINK Standardization Group (EPSG), which is an independent organization managed by the actual users of POWERLINK. Being vendor-independent and completely patent-free, POWERLINK is available in an open source version free of charge to the public. This makes it appealing for component manufacturers to use this open source version to integrate POWERLINK into their hardware. There are more than 100 member companies who belong to the EPSG, including some of the largest names in automation and components.
What happens during a POWERLINK cycle?
There are three phases of a POWERLINK cycle. First, the Start of Cycle frame is sent to all Controlled Nodes to ensure that all devices are synchronized. Second, during the Isochronous phase, time-critical payload data is exchanged between nodes.
This could include packets necessary for synchronized motion. Finally, during the Asynchronous phase, non-time-critical data such as TCP/IP frames are exchanged.
The design of the POWERLINK cycle allows for multiplexing, which optimizes bandwidth in the Isochronous phase. As you can see in the image below, the highly time-critical information is exchanged every cycle, while other packets of information can be multiplexed to be exchanged less often, such as every third cycle as seen in the example. The Asynchronous phase still occurs
once the Isochronous phase is completed.
Other Major Fieldbus Protocols
Ethernet/IP is a very well established fieldbus protocol within the North American market. It was developed by Allen Bradley (Rockwell Automation) and was released in 2000. It is supported by the CIP Protocol (Common Industrial Protocol). Ethernet/IP will typically achieve cycle times of around 10 ms, which is good enough for soft real-time performance. Users can approach cycle times necessary for servo motor control by using CIP Sync and CIP Motion.
EtherCAT (Ethernet for Control Automation Technology) was developed by Beckhoff Automation. One consideration of EtherCAT is that it employs the summation frame method. When an Ethernet frame containing data is transmitted by the master, it must pass through each node on the network before being turned back again. When comparing this to the Ethernet POWERLINK time slot/polling procedures, EtherCAT may not get critical data communicated as quickly. Data must pass through each node twice with the summation frame method, and if it must pass through many nodes, the actual cycle time will substantially increase. It could also mean a more catastrophic effect if a frame becomes corrupt; instead of losing a single packet of data, it could lose all the data on the network for that cycle.
PROFINET (Process Field Network) was developed by Siemens along with other members of PNO, the PROFIBUS user organization. PROFINET is broken up into two separate performance classes: PROFINET RT and PROFINET IRT. PROFINET RT is for applications without real-time requirements or with soft real-time requirements, and can achieve cycle times of around 10 ms. For more demanding applications such as motion control, which might require cycle times below 1 ms, users should opt for PROFINET IRT. PROFINET RT and PROFINET IRT operate and manage data very differently.
Market Share of Industrial Ethernet Protocols (2013 IMS Research)
For more information on how you can make the best use of high speed, real time, & deterministic industrial ethernet in your applications, please contact us here to receive a personalized contact by an IFP Application Engineer:
Sources Used: www.ethernet-powerlink.org, Industrial Ethernet Facts, System Comparison: The 5 Major Technologies. Issue 2. February 2013.