It used to be that programmable logic controllers (PLCs) were the wave of the future. They replaced the old relay boxes and enabled users to change a process without having to rewire the panel. However, PLC users found that their systems had two drawbacks. First, a PLC has only limited diagnostic functions. When a component is about to fail, there is no advance warning. Second, PLC-based systems are proprietary. Once a motion systems designer selects a particular manufacturer's PLC, he cannot use anyone else's technology with it. Also, many of these systems had their own intrinsic programming languages, which meant classes and training for those who would be using them.

The power of PC

Then, motion systems designers discovered PCs. Their greater memory capacity enables them to support more peripheral devices than PLCs, and they can integrate four software functions within one unit: logic programming, network configuration, real-time control, and man-machine interface. Unlike PLCs, PCs have standardized parallel and serial ports, input devices, and add-in hardware interfaces. Users can plug in compatible devices from any manufacturer.

The PC's graphical user interface permits the use of third-party software programs to manipulate control system data while the system is running. PCs can simultaneously run programs in several languages for one application. The hardware they require to run high-level communications protocols like Ethernet, Modbus, and DH1 is less expensive than what PLCs would require. Too, PCs have an advantage in that you can upgrade the hardware without affecting the software.

PC-based systems have lower operating costs, because they don't have the wiring, testing, and troubleshooting costs that PLC applications have. These can add up, especially in mid and high-end applications.

Another advantage is that more people have access to the information stored in a PC. With the networking capabilities that PCs have, data can be shared throughout the plant.

DeviceNet

The advantages of an open communications protocol like DeviceNet are threefold. It lets designers customize the control system, and its single-cable setup makes it easy to change devices and configurations. Diagnostics are much better, too. Previous diagnostic methods, which included writing ladder logic code for PLCs and external computer systems, were not always effective and were costly to maintain. You have to update the software in these systems when changes are made, and all too often, this is not done.

DeviceNet is based on the Controller Area Network (CAN) communications chip, which provides high-speed, real-time communication and was developed to replace expensive automobile wiring harnesses with low-cost network cable. All DeviceNet products use the same CAN chips, which are available from nine manufacturers throughout the world. The chips typically cost 10 to 20% of what chips for other networks cost. The CAN chip lets designers add intelligence without significantly increasing the price of the product.

Here are some advantages an open DeviceNet communications network can provide:

Control anyway you want it. Designers can choose controller platforms, software, and intelligent field devices for an application-specific solution. Because the protocol is open, they can add or remove standardized control modules to reconfigure as needed, thus allowing the designer to make incremental upgrades as opposed to replacing an entire system. If the device's firmware is upgraded, the DeviceNet standard assures that the new component is backwardcompatible, so that it will interact with the control software package.

DeviceNet is compatible with both PC and PLC-based systems. In addition, the network's producer-consumer message delivery system lets the intelligent devices communicate directly with each other and bypass a PLC or PC. For example, you can program a switch to respond to a sensor message and open or close without controller direction. Thus, the designer has the freedom to configure a control system that specifically meets the application's requirements in the most cost-effective manner.

More information. Each I/O point in a DeviceNet system provides digital strings of performance data. Limit switches, photo sensors, and proximity sensors all send fault information to the man-machine operating device to instruct the operator about status. The intelligent motor starters can pass information like motor current and percent of thermal overload, which operators can use to predict machine tool wear and motors running hot.

The producer-consumer communication system provides additional capabilities. One producer can send the same message or data to multiple nodes simultaneously. Plant engineers can also combine DeviceNet with Ethernet for a two-layer architecture approach that permits plantwide data control and machine controller interlocking.

One cable. DeviceNet networks reduce the number of wires connecting all devices. Compatible devices simply plug into a single cable rather than having to be individually hardwired to the controller. This "plugand- play" system allows the electrician to easily remove and replace failed components with a new device in roughly a third of the time required with traditional point-topoint I/O sensors. It also eliminates a lot of the human error associated with a complicated wiring scheme.

Leak test machine patched in

Engineers at Ford Motor Co.'s Cleveland Engine Plant 1 needed to upgrade a machine that performs leak tests on 5.0-L, V-8 engine cylinder heads. They considered a PLC-based, traditional I/O system but found it didn't provide the real-time diagnostic capabilities they wanted.

The system they chose included Cutler-Hammer industrial personal computers, a DeviceNet communications network, and intelligent field devices that are embedded with CAN chips. It took plant engineers two weeks to install the system. They wanted two controllers, so that they could make changes in system logic without interrupting operation. Since Aug. 2, the machine has been running at full production with no downtime due to system control problems.

Proof of the pudding. Ford activated the old system of relays and motor starters while they were making modifications to the new setup. During that time a worker stepped on a limit switch and threw it off line. The electricians who were called into find the problem spent 20 min. looking for it. Plant personnel then switched on the new PC-based control system, which immediately identified the the misaligned limit switch and pinpointed the location. The system was back up in seconds.

Brian Brickhouse is automation marketing manager at Cutler-Hammer