Operate EtherCAT Devices within an IT network
Imagine a factory where IT and OT networks blend seamlessly together – with EtherCAT-based machinery directly controlled from the IT-server room without the need for specialized gateways. acontis technologies enables the usage of standard IT networks as transparent transport medium for EtherCAT with the Feature Pack EOM for the EC-Master MainDevice Software.
The basis for the acontis Feature Pack EOM is one of the two operation modes of EtherCAT: Direct mode and Open mode, each serving different purposes based on application requirements.
- EtherCAT Direct Mode: The EtherCAT MainDevice is directly connected to the EtherCAT network. This is the common operation mode for most EtherCAT systems by now.
EtherCAT Direct Mode is highly efficient, easy to configure, offering minimal latency and best real-time performance, which makes it ideal for control applications that demand quick response times. - EtherCAT Open Mode (EOM): The EtherCAT MainDevice uses standard IP communication methods for connection between the EtherCAT MainDevice to an EOM-enabled Subdevice, which also forwards the EtherCAT telegrams to other EtherCAT Subdevices directly connected to this EOM enabled device.
EtherCAT Open Mode provides greater flexibility by allowing integration with standard IT infrastructure and centralized operation of EtherCAT MainDevices, though this comes at the cost of increased and variable latency.
Understanding the trade-offs between performance and flexibility is key to selecting the appropriate mode for your application – for more details please also refer to our blog: What is EtherCAT Open Mode? - acontis
EtherCAT Direct Mode
In Direct mode, one EtherCAT segment is directly connected to the MainDevice as shown in Figure 1.
The MAC address fields of the Ethernet frames are ignored. For communication, all EtherCAT SubDevices use EtherCAT SubDevice Controllers (ESCs), while the MainDevice uses a standard Ethernet port.
Direct mode is the standard in applications using the EtherCAT Device Protocol (EDP). It does not require switches, as the SubDevices typically have two or more ports that enable daisy chaining and other topologies. In Direct mode, the Ethernet controller in the MainDevice is exclusively used for EtherCAT.
One of the key advantages of Direct mode is the "Processing on the Fly" principle, which, along with the very low delays in the SubDevices, allows for hard real-time measurement and control applications with cycle times of one millisecond and faster.
Figure 1: EtherCAT Segment in Direct mode
EtherCAT Open Mode
The EtherCAT MainDevice uses standard IP communication methods for connection between the EtherCAT MainDevice and decentralized EtherCAT network segments.
While Direct mode is typically used for its efficiency and real-time performance, Open mode offers more flexibility by allowing integration with existing IT infrastructure.
In the EtherCAT Open Mode (EOM), one or several EtherCAT segments may be connected to a standard (existing) switching device. This mode is particularly useful when integrating EtherCAT networks into larger, shared IT environments.
The first device within an EtherCAT segment has an ISO/IEC 8802-3 MAC address representing the entire segment. This device is called "EOM Device" and replaces the destination address field with the source address field and the source address field with its own MAC address within the Ethernet frame. This ensures that, if the frame follows the coding rules of EtherCAT, it will be returned to the MainDevice after processing by all SubDevices.
Any standard EtherCAT SubDevice can be directly be connected to the EOM Device to realize such remote EtherCAT network segments.
EtherCAT Open Mode also uses the EtherCAT Device Protocol (EDP) to communicate between the MainDevice and SubDevices via the switched network – same as with EtherCAT Direct Mode. To transport the EtherCAT telegrams though the through the switched network, standard IP communication methods are used such as UDP.
If the EtherCAT system uses UDP within the switched network, the EOM Device will handle the source and destination IP addresses and the UDP source and destination port numbers in the same way as the MAC addresses. This ensures that the response frame fully satisfies UDP/IP protocol standards.
This also ensures that other protocols, such as TCP/IP, can share the same IT infrastructure in parallel.
EOM network topology example with one physical MainDevice
The FP EOM enables a significantly more flexible system setup: The EtherCAT MainDevice must not be connected directly to the EtherCAT SubDevices inside the machine, but it can be located at a more suited location withing the factory – e.g. a protected server room with controlled atmosphere.
The EtherCAT MainDevice uses its standard Ethernet port to transmit the EtherCAT EDP Frames e.g. via UDP to the first device of the targeted EtherCAT Segment. This device receives the frame from the MainDevice, processes it accordingly and forwards it to the local EtherCAT segment. This EtherCAT segment consists of standard EtherCAT SubDevices – the operation within this segment is same as in EtherCAT Direct Mode, thus providing minimal jitter and low latency as standard EtherCAT SubDevices are used.
In between the MainDevice and the EOM Device, a standard IT network is shown – with multiple switching devices as well as the simultaneous operation of IT devices using protocols like TCP/IP. Jitter and latency within this part is highly dynamic, depending on network traffic and system modifications during operation time. By monitoring network and optimized configuration of switches, the real-time capabilities of this network part can be improved, with direct effects on the performance of the EtherCAT system.
Figure 2: Two independent EtherCAT Segment operated by two MainDevices using EtherCAT Open Mode
EC-Master Software Architecture with EtherCAT Open Mode
The adaption of the EC-Master software architecture for EtherCAT Open Mode is mainly reflected in a special interface layer
- The „EtherCAT Open Mode Layer“ is added in between the EC-Master Core and the field-proven acontis real-time Ethernet drivers. The EOM Layer provides all necessary functions to transmit the EtherCAT Device Protocol via UDP or RAW frames to enable the transmission via the switched network to the target EtherCAT network segment. This includes the basic frame setup as well as managing source and destination addresses.
- The Application Programming Interface as well as the related code examples support additional function calls for the initialization of the EOM Layer and also provide very functions for the measurement of the roundtrip-time, including min, max and average values.
This measurement function is key for monitoring and evaluating the possible performance of the EtherCAT system based on the underlying switched IP network – based on the measurement results, it can easily be identified if the real-time behavior is fit for the target application of if additional optimization of the switched IP network are needed to achieve the required jitter and reaction time. - Documentation includes the extended API description as well as instructions for configuration of the MainDevice functions and the EOM Layer
Figure 3: EC-Master software architecture with FP EOM
Key Benefits and Limitations of EtherCAT Open Mode
Using a switched network in EtherCAT Open Mode introduces many opportunities but also some trade-offs that need to be noted to achieve the optimal solution.
On the positive side, Open mode provides flexibility, enabling the integration of EtherCAT with existing IT infrastructure, free placement of the MainDevice away from the harsh environment of a machine. However, since all Ethernet frames share the same infrastructure and switches typically operate on a "Store and Forward" basis, this results in significantly higher latencies within an EtherCAT EOM system compared to an EtherCAT Direct Mode network. These latencies are not consistent and cannot be precisely calculated in advance, which impacts the minimum achievable cycle time. The achievable performance depends heavily on the specific IT network, its configuration, and current network load.
Key Trade-offs between Flexibility and Latency:
- Flexibility: Integration with existing IT infrastructure, allowing for mixed operation.
- Latency: Higher and nondeterministic latencies due to shared infrastructure and 'Store and Forward' switching.
- Cycle Time Impact: Minimum achievable cycle time is affected by network configuration and load.
Other benefits for EOM include improved reliability, as the MainDevice can be placed in a secure and protected environment – e.g. an IT server room, which provides much less stress to the controller than the harsh environment directly at a machine. In addition, centralized controllers offer benefits for maintenance, as this can be realized using existing IT frameworks.
EOM also offers the chance to quickly use secondary MainDevices as the connection can be done via the IT-network and does not need any re-cabling or hardware installation.
In total EOM can also offer cost benefits, as existing IT infrastructures can be reused for the OT system, and each MainDevice is able to control multiple, separated remote EtherCAT segments.
On the other side, EOM imposes some limitations that shall be noted during system design. These limitations include the mentioned limited real-time capabilities due to higher latency and dynamic jitter imposed by the IT-network. In addition, the configuration of the EtherCAT system tends to be more complex, as it also includes the IP-based communication between MainDevice and EOM device – and IT knowledge is necessary for optimal IT network configuration to achieve good EtherCAT performance.
Last but not least, cybersecurity aspects are more relevant for EOM compared to EtherCAT networks in Direct mode: MainDevice and EtherCAT network segments should be within a protected network and the additional attack vectors towards MainDevice and EOM device via the IT network need to be considered.