A Fieldbus for packaging

Fieldbus technology has been around for nearly 20 years and is viewed as an acceptable approach to reduce wire and cabinet space and decrease labor cost. As predicted by industry analysts, no one single fieldbus protocol has become the standard, but several have become a commercial success, including the CANopen™, Profibus™, Interbus® S, and DeviceNet™ protocols. Selecting a fieldbus protocol is similar to selecting a personal computer operating system — users have a choice between an open system such as Linux, and a proprietary system like Microsoft® Windows®. Many companies in the material handling and packaging industries have found the feature-function set offered by an open fieldbus protocol, such as the Controller Area Network (CAN), to be beneficial when compared to other fieldbus protocols.

Early adopters

The material handling and packaging industries were early adopters of fieldbuses because of their need to reduce costs and control inventory. The typical equipment layout includes long runs of wiring with input and outputs (I/Os). A fieldbus like CANopen lets you distribute I/O over long distances across the machine, which helps shrink control cabinet size as well as reduce wire — both of which reduce cost. In addition, some packaging machinery requires a high-speed system combined with clusters of high-density inputs and outputs located in various places around the machine. CANopen can be an ideal network because it can handle high-speed, high-density I/O devices.

The adoption of open fieldbus protocols has also been driven by a trickle-down demand to have more control over inventory tracking and management, using such technologies as radio frequency identification (RFID). The U.S. Government and companies such as Wal-Mart require their suppliers to provide RFID on all their products, which means that their suppliers must require RFID devices on all products provided by their vendors. Hard-wired I/O’s typically don’t offer the technology required to do the RFID tracking, thereby creating a need for a control device to receive this information. Fieldbus systems, on the other hand, already have the monitoring capabilities in place to receive information from RFID devices.

A closer look at CANopen

CANopen offers electromagnetic interference (EMI) tolerance, flexibility, error detection and diagnostics capabilities, along with low cost.

Electrically robust.
Packaging and material handling applications typically have high electromagnetic interference (EMI) noise levels, making an electrically robust fieldbus particularly important. EMI is conducted or radiated onto the same wires used by a PLC for electrical signals. Thus, this phenomenon often corrupts PLC signal data. Potential causes of EMI including a motor turning on and off or a technician using a walkie-talkie nearby.

CANopen copes with EMI by using a voltage differential technique. The differential stays the same regardless of the induced EMI voltage. For example, if there is a ½ volt differential between the PLC signal and the EMI, CANopen will maintain the ½ volt differential and the PLC signal will ride on top of the EMI.

CANopen also uses what is called a balanced line, which reduces the amount of induced noise per specified distance and makes longer cable runs practical. EMI will affect signals on two lines in the same way. Similarities between the two signals are automatically removed at the end of the transmission path when one signal is subtracted from the other. Thus, CANopen is not nearly as affected by EMI as other fieldbus technologies.

CANopen is a scalable speed fieldbus, letting you set the speed of the network and tailor it to the job. In material handling or packaging applications it’s imperative that the correct item is moved to the correct spot at the correct time. A fast and flexible fieldbus allows users to meet their application’s performance requirements.

Scalable speed.
CANopen offers scalable speed, which lets you determine the speed of the network. A fieldbus with a fixed transmission speed has a physical limit on how fast it can transmit data. In a material handling or packaging application, scalable speed is crucial for tailoring a system to the job.

When a package moves along a conveyor, the photoelectric sensor has a limited amount of time to see the package, send a signal to the PLC, let the PLC run its logic scan on the signal and then send information to an actuator to move the package to the correct divert line. A fieldbus must transmit these data to and from the PLC before the package leaves the sensor’s viewing area. Typically, the slowest part of the process is the logic scan in the PLC. Therefore, it’s critical to save time in other areas through a fast fieldbus and fast sensors so the PLC has as much time as possible for the scan. If the fieldbus is not fast enough, it does not matter how quickly the I/O sensor can send and receive data.

Depending on the application’s performance requirements, the network-control system can achieve sub-millisecond response times from the I/O through adjustments to the topology and bus length. Alternatively, the bus can be extended to longer lengths by slowing down the bit rate. Basically, high speed demands a short bus length.

Custom configuration.
Custom configuration is an especially attractive feature for a fieldbus protocol. CANopen, for example, provides communication objects for real-time data, called process data objects (PDO), and provides special functions, referred to as service data objects (SDO). The SDO handles time stamps, sync messages, emergency messages, and network management data that includes error control.

In addition to these communication features, the CANopen bus offers physical flexibility with up to 32 nodes that can be used in a daisy chain topology.

Error detection and diagnostics.
A strength of CANopen is error detection. This fieldbus will try to detect the errors within each message transmitted. If an error is found, the discovering node will transmit an error flag, interrupting the bus traffic. Likewise, the other nodes will detect the error flag and take appropriate action, such as discarding the current message.

Each node maintains two error counters: the transmit error counter and the receive error counter. Rules dictate how these counters increment and decrement. A transmitter detecting an error increments its transmit-error counter faster than the listening nodes increment their receive-error counter because there is a greater probability that the transmitter is at fault rather than the receiver’s counter.

The various fieldbuses offer different strengths. This chart shows several features important to packaging and how the various fieldbuses compare.

In the past, when fieldbuses detected errors, they would just stop and shut down the entire machine, potentially creating serious system problems. CANopen detects errors by counting them and assigning them a value and won’t immediately halt the entire line; instead it will only shut down certain parts of the system.

A CANopen node has three levels, or states, of error detection:

An error active node will count the errors and transmit active error flags when it detects errors.

An error passive node will count the errors and transmit passive error flags when it detects errors, notifying the user of something wrong and shutting down the bus.

A node that is bus off will not transmit anything on the bus, and occurs when the error count reaches a predetermined level. This node turns the machine or its parts off.

A node starts out in error active state. Here, the bus is active and will tolerate certain errors. When either one of the two error counters rises above 127, the node will enter a state known as error passive. In this state, the fieldbus will stop working and will only transmit passive error flags. When the transmit error counter raises above 255, the node will enter the bus off state, completely shutting down the fieldbus, which in turn shuts down the machine in an orderly fashion.

To understand how this works, it’s necessary to understand how the CANopen error counting works. To simplify the rules, eight points are given for every transmit error, and one point for every receive error. Error-free messages transmitted and received cause the error count(s) to decrease. In addition, each particular error is assigned a point count, so different errors can be deemed more important than other errors.

Most CANopen controllers provide status bits and corresponding interrupts for two states:

    Error warning – one or both error counters are above 96.
    Bus off – as described above.

Some controllers also provide a bit for the error passive state, and a few controllers provide direct access to the error counters.

CANopen fieldbuses ease troubleshooting through their diagnostic information. For example, a common problem is not knowing which device has faulted. With CANopen, the node can identify the faulted device, and also provide specific information about the fault state. In
addition, because CANopen almost eliminates EMI, it’s easier to tell when there’s a problem with the network versus something wrong with the system.

Performance.
CANopen fieldbuses transfer 8-bit words instead of 16-bit words, which increases throughput. These shorter words allow CANopen to implement and run at all the speeds it has available. The faster the fieldbus runs, the faster it can transfer information to and from the PLC for faster process operation. For example, in the packaging industry, a quicker fieldbus means the conveyor can run faster or the system can wrap 10 more boxes. Other fieldbuses, such as DeviceNet and Profibus, typically are limited to slower speeds and consequently exhibit lower performance.

In addition, the 8-bit words also help eliminate EMI. A short message is a harder target to hit as it moves down the fieldbus because it’s just not there as long. This makes it more likely that a message will get through reliably.

Low cost.
Compared to other fieldbus technologies, CANopen is less expensive because it’s widely used. In addition, engineering specifications are free to OEMs and system integrators because it is not a proprietary network.

CANopen has been adopted by major industrial automation suppliers. It can be seamlessly integrated with Ethernet TCP/IP, its higher-level network. Having CANopen as a native protocol gives customers connectivity and transparent communication from the workstation to the I/O.

Any number of industries can benefit from the non-proprietary nature and interoperability
advantage of an open fieldbus protocol like CANopen. OEMs, control panel builders,
system integrators and everyone looking to move to distributed I/O, while also reducing wiring and cabinet space and decreasing labor costs, will find the robust performance, flexibility, error detection and diagnostics, and high performance of CANopen attractive. 

Schneider Electric

www.schneider-electric.com

:: Design World ::