Issue link: https://resources.mouser.com/i/1437657
The Control Cycle These applications run on a defined cycle time that is usually equal to, or a multiple of, the fundamental control/pulse-width modulation (PWM) switching cycle of the underlying servomotor drive. End-to-end network transmission latency is a key parameter in this context (Figure 2). Within each cycle period, the new position reference and other relevant information must be transmitted from the machine controller to each node (Figure 1). Then sufficient time must remain within the PWM cycle for each node to update the servo control algorithm calculation using the new position reference and any new sensor data. Each node then applies the updated PWM vector in the servo drive simultaneously via a distributed clock mechanism that is Industrial Ethernet protocol dependent. Depending on the control architecture, part of the control loop algorithm can be implemented in the programmable logic controller (PLC). It requires sufficient time to be available, having received any relevant sensor information update across the network. Data Transmission Delays Assuming that the only traffic on the network is the cyclic data flowing between the machine controller and the servo nodes, the network latency (TNW) is determined by the number of network hops to the furthest node, the network data rate, and the delays encountered in each node. In the context of robotics and machine tools, the propagation delay of the signal along the wire can be neglected, as the cable length is typically relatively short. The dominant delay is the bandwidth delay; that is, the time needed to get the data onto the wire. For a minimum-sized Ethernet frame (typical for machine tools and robotics control), the bandwidth delay (Figure 3) for both 100Mbps and 1Gbps bit rates. This is simply the packet size divided by the data rate. A typical data payload for a multi-axis system from controller to servo would consist of a 4-byte speed/position reference update and a 1-byte control word update for each servo, which means a 30B payload for a 6-axis robot. Of course, some applications will carry more information in the update and/or will have more axes, in which case packets larger than the minimum size might be needed. Apart from the bandwidth delay, the other delay elements occur as a result of the Ethernet frame passing through the physical layers (PHYs) and 2-port switch at each servo network interface. These delays are depicted in Figure 4 and Figure 5, where the movement of the frame is shown through the PHY into the MAC (1-2), through the destination address analysis where only the preamble and destination parts of the frame must be clocked through. Path 2-3a represents extraction of payload data for the current node, whereas path 2-3b represents the onward journey of the frame to the destination node(s). Figure 4a shows only the payload is passed to the application in 2-3a, whereas Figure 4b shows the bulk of the frame being passed; this indicates small differences that can occur between Ethernet protocols. Path 3b-4 represents Figure 1: Network topology of a multi-axis machine. (Source: Analog Devices Inc.) Figure 2: PWM cycle and network transmission time. (Analog Devices Inc.) Figure 3: Bandwidth delay of a minimum length Ethernet frame. (Source: Analog Devices Inc.) 1 3 2 22 ADI | Industry 4.0 and Beyond