Comparison of five common protocols of the hottest

Comparison of five common industrial Ethernet protocols

compare the current five mainstream Industrial Ethernet protocols (ethernet/ip, PROFINET, Powerlink, EtherCAT, sercosiii), based on strategic considerations of technology, real-time performance, standardization status and market, for example: is there a user organization continuously developing protocols? Does the protocol comply with IEC standards, and does the system meet the real-time requirements of hardware

[real time]

the mechanism used to solve data collision in IEEE802.3 Ethernet brings the delay of data transmission. In order to achieve real-time performance, industrial Ethernet protocol adopts different methods to avoid this collision. For hard real-time, the signal transmission time must be carried out accurately according to the time frame, or they can trigger an error mechanism. System applications with a cycle time of hundreds of milliseconds are sufficient for software real-time, such as temperature control, while digital control or motion control applications often require a cycle period of less than 1ms

[market share]

another key factor in choosing which industrial Ethernet system to compare is its market share: the research of IMS and arc shows that about three quarters of industrial Ethernet uses ethernet/ip, PROFINET, or modbus/tcp, followed by Powerlink and EtherCAT, which are particularly suitable for hard real-time requirements. Modbus/tcp is not considered in the following description because its user organization ODVA has indicated that it will be integrated into the ethernet/ip network. For sercosiii, although its market share is relatively small, it plays a very important role in the field of high-speed motion control

[how the system works]

one of the key differences between different methods of real-time is how different industrial ethers schedule and manage data transmission so that the network can provide real-time. The communication between EtherCAT and sercosiii adopts the mode of cluster frame: in each cycle, the network sends a data message to all nodes, and transmits from one node to another along the ring topology, At the same time, collect the response data of each node. In contrast, other communication protocols use independent messages to each node, and the slave station also responds through independent messages. Each system uses three different methods to achieve real-time performance: 1. A master station controls the time slot on the network. In the Powerlink environment, the master station authorizes each node to send data independently. In the EtherCAT and sercosiii networks, the transmission of bundled frame messages follows the master station clock. 2. PROFINET IRT uses synchronous switches to control the communication process. 3. Ethernet/ip synchronously distributes IEEE1588 compatible clock information to the whole network by using CIP sync

[PROFINET communication]

PROFINET ("process field network") adopts different implementation methods according to the level requirements of certainty: PROFINET RT is usually used in soft real-time or applications without real-time requirements, while PROFINET IRT is for hard real-time applications. This technology is jointly developed by Siemens and the member companies of the Profibus user organization PNO. It is the successful application of PROFIBUS DP and PROFINET i/o based on Ethernet: specify the data transmission between all I/O controllers and the layout of parameterization, diagnosis and network. In order to cover different performance levels, PROFINET allows protocols and services to freely adopt the producer/consumer principle. High priority payload data is sent directly through Ethernet protocol and in the priority order of Ethernet frame VLAN, while diagnosis and configuration data are sent using udp/IP. This enables the system to implement i/o applications with a cycle time of about 10 milliseconds. For clock synchronization cycle time requirements less than milliseconds and motion control applications, PROFINET IRT is provided. It implements a time-sharing multiplexing hardware synchronization switch - the so-called dynamic frame packaging (DFP), which provides users with a new optimized design of PROFINET cycle time and released this feature in 2009

[Powerlink communication]

Powerlink was originally developed by Br and used in 2001, Ethernetpowerlink Standardization Organization (EPSG) is an independent user organization and follows democratic regulations "Since 2003, I have been responsible for the further development of this technology. Powerlink is a completely royalty free technology, independent of suppliers, and adopts a pure software protocol, but it can achieve hard real-time performance. In 2008, EPSG provided an open-source version of this technology. Powerlink integrates a complete CANopen mechanism and fully meets the IEEE802.3 Ethernet standard, that is, the protocol provides all standard Ethernet features, including crossover Communication and hot plug allow the network to carry out topology in any way. How does it work? Powerlink uses a mixed mode of time slot and polling to realize the synchronous transmission of data. For coordination, PLC or industrial PC is designated as the management node (MN) in the network. The management node runs the scheduling of periodic time slots, synchronizes all network devices, and controls periodic data communication. All other equipment operates as controlled nodes (CN). At each synchronization cycle stage, Mn sends the "polling request frame preQ" to CN one by one in a fixed time series. Each CN immediately responds to the request and transmits data in pre mode, "all other nodes can listen for this response. A Powerlink cycle includes three parts. At the beginning, Mn sends the cyclic start SOC frame to all nodes in the network to synchronize all devices in the network. The jitter is about 20 nanoseconds. Periodic synchronous data exchange occurs in the second stage, in which multiplexing technology can be used to optimize network bandwidth. The third phase is marked by the asynchronous start signal SOA, which is used to transmit large capacity, non time critical data packets. For example, user data or tcp/ip frames can be transmitted in the asynchronous phase. Powerlink is divided into real-time and non real time domain. Data transmission in the asynchronous phase supports standard IP frames, and real-time domain and non real-time domain data are isolated through routers to ensure data security. Powerlink is very suitable for various automation applications, including i/o, motion control, robot tasks, communication between PLC and PLC, and display tasks. Multiple shared features: CANopen and Powerlink OSI models

[ethernet/ip communication]

initially released in 2000, Ethernet/IP is an open industry standard developed by Allen Brad (Rockwell Automation) and ODVA (open DeviceNet Supplier Association). "Industrial Ethernet Protocol" is essentially a CIP (General Industrial Protocol) -Ethernet data transmission protocol on ControlNet and DeviceNet that has been used. Ethernet/ip is usually well used in the U.S. market, mainly using Rockwell control system. How does it work? Ethernet/ip runs on standard Ethernet Hardware and uses both tcp/ip and udp/ip for data transmission. Because CIP protocol supports producer/consumer mode, ethernet/ip adopts different communication mechanisms to process, such as periodic polling, time or event triggering, multi wave or simple point-to-point connection. CIP application protocol is divided into "implicit" i/o information and "explicit" request/response message for configuration and data collection. When explicit information is embedded in TCP frames, real-time application data is sent through UDP, because the latter has a more compact format and less overhead. Forming a center, we successfully mentioned the 625 alloy star network topology in a project. The switch prevents data collision through point-to-point hook connection. Ethernet/ip networks usually achieve soft real-time performance of about 10ms, while CIP sync and CIP motion and precise node synchronization can achieve extremely low cycle cycle and jitter through the distributed clock method defined in the IEEE1588 standard, so that it can be used for servo motor control and drive

[EtherCAT communication]

EtherCAT (Ethernet control automation technology) Beckhoff automation. All users of this technology automatically become members of the EtherCAT Technical Association (ETG). How does it work? EtherCAT is based on the cluster frame method: the EtherCAT master station sends a data packet containing all the slave station data of the network. This frame passes through all the nodes on the network in order. When it reaches the last frame and makes the 50mm wide edge coincide and flush, the frame will be returned again. When it passes in one direction, the node processes the data in the frame. Each node reads out the data and inserts the response data into the frame. In order to support 100 mbit/s baud rate, special ASIC or FPGA based hardware must be used to process data at high speed. Therefore, EtherCAT network topology always constitutes a logical ring. An EtherCAT frame structure, all instructions to each node are contained in the payload area of an EtherCAT data message. Each EtherCAT frame consists of a header and several EtherCAT commands. Each command includes its own header, instruction data to the node, and a work counter. Up to 64 kb of configurable address space to configure the slave. Addressing processing adopts the automatic increment mode, that is, each slave can process 16 bit address domain, and the slave can also realize distributed site addressing in the initial stage of the network in the way of assignment. EtherCAT process synchronization: each slave station is connected to a real-time clock technology similar to IEEE1588 provided by the master station for synchronization. The slave device can be real-time or non real-time mechanism, which depends more on the hardware. Based on real-time, the control signal can be synchronized with high accuracy. In the physical layer, EtherCAT protocol not only operates on Ethernet, but also adopts LVDS (low voltage differential signal transmission). This standard is used by Beckhoff in its terminal technology. Typically, a PC with a standard Ethernet interface is used as an EtherCAT master station. Compared with protocols such as Powerlink or PROFINET, EtherCAT only runs through the layers in OSI 7-layer model. Therefore, in order to achieve the same application functions as other systems, additional protocol layers (COE, EOE) must be loaded

[sercosiii communication]

this is a free real-time communication standard for digital drive interface. Sercosiii not only has a specific hardware architecture connected to the physical layer, but also has a specific protocol structure and application specification definition of SERCOS interface. Sercosiii is the third generation of SERCOS. SERCOS was introduced to the market in 1985. It is a standard data transmission protocol following IEEE802.3. This communication system was originally used in the automation system based on motion control. A registered Association: SERCOS International Association supports the development of this technology and maintains the consistency of standards. How does it work? Sercosiii adopts specific hardware in the master station and slave station. These sercosiii hardware reduce the communication task of the master CPU and ensure fast real-time data processing and hardware based synchronization. The slave station requires special hardware, while the master station can be based on software scheme. SERCOS user organization provides sercosiii IP core to sercosiii hardware developers based on FPGA. Sercosiii adopts the mode of bundled frame transmission, and the network nodes must adopt daisy chain or closed ring topology. Because Ethernet has full duplex capability, daisy chain has actually formed an independent ring. Therefore, for a ring topology, it is actually equivalent to providing a double ring, which allows redundant data transmission. The direct cross communication capability is realized by two ports on each node. In the daisy chain and ring network, real-time messages pass through each node when they move forward and backward. Therefore, nodes have the ability to communicate with each other twice in each communication cycle without passing through the master station, and do not need to route data through the master station. In addition to real-time channels, it also uses time slots for collision free