Recovery of distributed clock in EtherCAT with redundancy for time-drift sensitive applications

EtherCAT is one of the most prominent representative of Industrial Ethernet communication technology. It provides high real-time performance together with rich functionality set. Some of its important features are very accurate distributed clock (DC) and cable redundancy. These features work perfectly when standalone. However there are some inherent problems when combined. There exist couple of solutions dealing with different aspects of this issue. Here we analyze and propose a solution how to quickly recover DC after a cable failure in order to minimize drift in the system clock without being disruptive to cyclic time-critical application traffic. Presented solution aims at high-performance industrial applications with cycle time below 100 microseconds.