What are Differences Between EtherCAT & Ethernet ? What is Communication Period?

Hello, dear friends, here is Zmotion again.

Welcome.

Last article, we talked about some properties of EtherCAT. And today Zmotion brings more details about how to achieve EtherCAT. In addition, communication period also will be introduced. Specifically, what is communication period, and what are SDO and PDO.

Firstly, it must be declared, this article is mainly for friends who want to know more information. Namely, you still can use Zmotion motion controller seasily even though you don't get messages from this article. But know more, solve complex problems easily.

OK, Let's begin!

As we all know, it needs stable and real-time communication in industrial site, and sensor and IO signals are required to respond at high-speed, motor data are transferred in real-time. What's more, before EtherCAT, CAN bus has already existed.

However, fieldbus automation system becomes huge increasingly, which means traditional bus is hard to meet nowadays needs, for example, bandwidth is not high enough, transferred data is not enough, and it cannot make full use of IT technology. Therefore, a kind of fieldbus based on Ethernet is born, and EtherCAT is outstanding particularly.

Then, someone may feel confused, EtherCAT bases on Ethernet fieldbus, why not use Ethernet directly.

That's because some fieldbus communication features are totally different from in IT area.

1. one single system has a lot of node devices.

2. usually, these nodes transfer less data, but require high real-time.

If each node uses an Ethernet data frame to communicate, the final communication efficiency will be very low. According to an official calculation example, the final bandwidth utilization may be less than 5%, and Ethernet-based communication protocols, such as TCP, it will bring additional bandwidth overhead and delay, then a large amount of bandwidth is occupied by invalid data.

For EtherCAT, all nodes use one data frame. After the master station sends the data frame, all slave stations will process the data in flight, read what they need, insert the data they need to return, and then continue to send the data frame until meeting the last node, next use the full-duplex feature of Ethernet to return the data to the master station.

Through this transmission method, the maximum effective data rate of data frames can exceed 90%. And the master station is the only device that is allowed to actively send data packets, and other nodes can only transmit packets sequentially, which can avoid Ethernet common conflicts and delays in the network, then ensure EtherCAT data transmission is in real-time.

Moreover, when the slave station uses a dedicated chip (ESC) to process data packets in flight, it is completely processed by hardware, so that the communication time of the entire EtherCAT network remains stable and predictable, and it has nothing to do with the different deployment of each slave station.

--Question:

If there is only one data packet, encountering the problem that maybe in slave station or one certain circuit, how to check?

--Answer:

Actually EtherCAT has a series of smart designs, specific abnormal node still can be found even when data is transferring precisely. And Zmotion also has configured small tools, they will be mentioned in following articles.

Periodic data and aperiodic data both can be transmitted in the EtherCAT network.

Periodic data is exchanged through PDO (Process Data Object), which is generally used for real-time data exchange, such as the command position and feedback position of the motor, the interaction of IO signals, etc.

Aperiodic data can be transmitted through the mailbox protocol, the most common is CoE (Canopen Over EtherCAT), which can transmit information through SDO (Service Data Object). SDO is generally used for non-real-time communication, such as the configuration of motor parameters (such as resolution, maximum current, etc.), including the PDO configuration information of the slave station.

Configuring PDO is like buying a high-speed rail ticket for your data, so that the master station and the slave station already know in advance what data will be on the high-speed rail before the official work, and the ESC chip of each slave station will be automatically removed for corresponding the data position, that is, ESC chip is inserted into the data that needs to be returned. In this way, it can reduce invalid data and speed up processing.

Some old users may consider, in real operation, they don't think so much when using our motion controller.

Yeah, there is one parameter developed by R&D to configure PSO list easily, "Drive_Profile". And actually you don't need to configure by yourself, because it is set in script program of commonly used motor.

The transmission cycle of PDO information is what we often call the EtherCAT communication period. For example, the regular firmware of Zmotion defaults to a cycle of 1ms (1K communication frequency), and some products of Zmotion can set a cycle of 125us (8K communication frequency). Why is the default 1ms instead of a shorter period (higher frequency)? What is the controller doing in this 1ms? What is the slave station doing again?

This needs to start with the working mode of the CSP and the control loop of the motor, which will be introduced in detail in later chapters. Generally, 1ms is small enough (1K is high enough). Moreover, the synchronization between motors is not based on the arrival time of data frames, but on distributed clocks, which can achieve nanosecond-level synchronization, as introduced in the previous article.

As mentioned earlier, because of the unique design of the EtherCAT data frame, the effective data rate is very high, so the 100M bandwidth of Ethernet can be fully utilized in the field of industrial automation. In fact, the 100M bandwidth can also enable many motors to achieve a 1ms communication cycle, for example, Zmotion has a 128-axis controller.

One of the questions left over from last time is, since the EtherCAT 100M bus is so good, why do we need to design Gigabit and 10 Gigabit buses?

The answer is that with the development of technology, larger and more complex systems have emerged. For example, advanced logistics systems or magnetic levitation systems require a lot of axes (hundreds or even thousands), and precise synchronization between these axes is required, or measurement equipment that collects a large amount of sensor data has higher requirements on the bandwidth of the bus.

Higher bandwidth can allow fieldbus to enter more application fields, and a stronger bus can also support the design of complex systems that were not available before.