A network for all cases!
Behind this name hides a network that was developed for the cell level and which transmits time-synchronized data. In the field of automation technology, this system makes it possible that the cyclic I/O data transmission and the acyclic data communication, serving the tasks of configuration and uploading and the downloading of programs, can run side by side. The special feature here is the avoidance of a negative influence on components, e.g. throughput, determinism or repeatability in a cyclic process. With this, time-critical data have points in time that are specified in advance, at which a guaranteed transmission takes place and the max. transmission speed of the remaining data traffic can also be predicted. ControlNet corresponds to the European standard EN 50170.
The installation of ControlNet
The specification of ControlNet lies in the flexible installation alternatives and the topologies, which can occur in a bus form, tree form or a star formation. It is further possible to couple any number of device branches at any location without having to observe a minimum distance between two devices. Additionally, the redundancy of the media also is included here. This can be used optionally to increase the availability. In the normal case, the ControlNet is composed of the following components
- Main cable
With the ControlNet, the main cable serves as bus and takes the central position within the system. At this point it is irrelevant whether a coaxial cable or a LWL type cable is used. The connection of the main cable with the branches is done with BNC plug connections via each individual node of the network. The terminators are located at the end of the cable segments of the branches. The functions of the repeaters lie in the increased number of these branches, an extension of the overall length of a segment and the creation of a topology with a ring structure, star structure or tree structure. However, only a limited number of repeaters can be connected to a single network, and this also depends on the selected topology. Finally, the ControlNet also has an NAP, a so-called network access port that can be temporarily connected for configuration work via an RJ45 socket or for diagnosis and programming.
The components and function of ControlNet
The network is based on a conventional OSI reference model of the ISO-IEC 7498 standard. The classical components are
- the physical layer,
- the data link layer,
- the network,
- the transport layer,
- the CIP application layer, and
- the network management functions
The network uses the communication model of the producer and the consumer for data exchange, and instead of the source and destination addresses, the data packets are provided with a CID identification.
The producer of the communication model uses a broadcast to place a data packet onto the line, and thanks to the CID identification, all of the relevant consumers can receive this packet at the same time and can filter out the required data in order to then use them.
Media access with the ControlNet
Media access with the ControlNet uses the CTDMA process. This was developed specifically for I/O data and installation values and it also processes the time-critical interlock signals. As a rule, these signals are messages which are helpful for programming applications and for configuration. The basis is a time shift process, where the cyclic transmission intervals are specified in the NUT, the network update time.
The advantage of the NUT is the separation between cyclic and acyclic, i.e. time-critical and non time-critical data.
The implicit token process is also relevant for the media access. This ensures guaranteed access to the media by each individual node within each interval. This token circulation is continued to the address with the highest configuration, in most cases an SMAX or a UMAX. This process starts again in the new NUT interval. However, if a consumer in the ControlNet fails, waiting for one time slot is performed before transmission is started again. If the consumer becomes usable again at a later stage, it inserts itself again into the transmission circulation without problems and without requiring the network to stop. In cases when there is no data packet to be transmitted, a ‘null frame’ is transmitted. Furthermore, the bus management is decentralized, as the ControlNet has no centrally positioned bus scheduler. However, the bandwidth of the transmission for the cyclic data is specified for the configuration in advance. This scheduled service, which must be a cyclic service, must be strictly deterministic and must show reproducibility. The time then required for this is based itself on the requirements to be expected for an application. The remaining time until the NUT has run out then serves the “unscheduled service”, i.e. the service handling the transport of non time-critical and acyclic data. Other than is the case with the cyclic process, the time window here is not connected with a specific node, but is only used when a consumer needs it. This has the consequence that the data transmission within the acyclic service has no influence on the data of the cyclic service. The end of the NUT marks the transmission of the synchronization data.