Ethernet is the best possible communication medium available today for industrial automation. Due to the swift nature of data speed, ease in configuration, and ease of availability, Ethernet is a widely preferred hardware for industrial communication protocols. Two of the most widely used protocols in them are Ethernet IP and EtherCAT. Though they sound similar, they have many vast differences, which must be known by engineers before choosing the right one. In this post, we will see the difference between Ethernet IP and EtherCAT.
What is the Ethernet IP protocol?
Ethernet IP is a communication protocol developed by Rockwell Automation and ODVA ( Open DeviceNet Vendors Association). It works on Ethernet hardware and is managed on CIP ( Common Industrial Protocol). Ethernet IP is based on a producer-consumer mechanism, where the producer or slave sends data on the network, either periodically or on an event basis. It does not consider whether the consumer or master is present in the network or not. If present, the consumer will either just listen to or use it according to its configuration in the program. This makes it a real-time communication and allows for high-speed data transfer, because even if the master comes into the network to use afterwards, it will get real-time data quickly without producing any latency. Also, if the producer is not present in the network, then the consumer will just throw communication timeout errors, without disturbing other devices in the network.
As Ethernet IP uses CIP protocol, the messages communicated here consist of objects, classes, and attributes. This allows for large data transfer with minimum network latency. Ethernet IP requires EDS (electronic data sheet) files of the slave device to be installed in the software for using them. They are available on their respective websites for downloading and installing.
What is the EtherCAT protocol?
EtherCAT stands for Ethernet Control Automation Technology and is developed by Beckhoff Automation. EtherCAT works on Ethernet hardware, but not the Ethernet protocol of TCP or UDP. EtherCAT is based on master slave technology, but works under a special condition called processing on the fly, which bypasses the traditional limitations of the request-response model. Here, the master will send its request frame in the network to all the slave nodes. As soon as a slave node receives this frame, it will check incoming data as to whether data is required from itself or not. If required, it will immediately load its data into the frame and pass it on to the next slave nodes. If not required, then it will just pass it on to the next node. This model makes it very fast to operate and reduces network latency to a great extent. The cycle keeps on repeating in this fashion as long as the network is powered up.
One main thing to note is that EtherCAT does not work on Ethernet protocols of TCP or UDP. So, it does not use an IP address at all. Instead, it identifies devices through logical addressing, physical order of devices, and auto-discovery during initialisation. So, it directly works on layer 2 of the OSI model and bypasses transport and network layers, which are used in Ethernet IP.
Ethernet IP vs. EtherCAT
- EtherCAT is very fast as compared to Ethernet IP, due to its on-the-fly processing on the fly technology. Also, as EtherCAT skips many higher levels of the OSI model, network latency is very low, and data bandwidth is higher here than Ethernet IP.
- Ethernet IP provides a separate Ethernet packet with its overhead for data in the lines, which EtherCAT does not. As we saw, data flows like a train with each slave acting as a station and just filling its data as it comes. So, the data length consumption in EtherCAT is much less compared to Ethernet IP.
- EtherCAT requires specialist slave devices supporting this protocol, also called EtherCAT slave controllers. Ethernet IP does not require special hardware. This makes EtherCAT costlier to use than Ethernet IP.
- Ethernet IP mostly uses star and tree topologies, as compared to EtherCAT, which uses line, daisy chain, or ring topologies. This is done due to processing on a processing-on-the-fly technique, where a train of data passes sequentially through each node.
- Ethernet IP uses objects, classes and attributes in the message, as compared to EtherCAT which uses read or write command, address, data length and read or write data, all similar to Modbus.
In this way, we saw the difference between Ethernet IP and EtherCAT.
Read Next:
- Cyclic and Acyclic Communication
- What is Open Systems Interconnection?
- Network Topologies Selection Factors
- Data Handling Instructions in PLC Programming
- Difference between TCP/IP Model and OSI Model