Three Ways to Easily Add PACs to Improve Existing Control Systems

While new, high-performance industrial measurement and control technologies are helping engineers design systems with higher efficiency and throughput, plant engineers today do not have the luxury of large capital equipment budgets. In fact, industry capital expenditures as a percentage of revenue have decreased by nearly 3 percent from 1997 to 2002.

In some competitive markets, such as pulp and paper and metals, the decrease has been as large as 8.7 percent. The only markets that have seen growth are semiconductor and utilities, with increases of 1.3 percent and 0.4 percent, respectively, according to the ARC Advisory Group.

As plants cut costs and reduce system upgrade funding, engineers must improve the plant while keeping overall costs low. Many plants successfully implement a strategy to “locally optimize” the process by intelligently applying new technology to increase efficiency at the bottlenecks. Today, many plants are using a new class of controllers, called programmable automation controllers (PACs), to collect signals at higher speeds and perform more sophisticated analysis.

PACs combine the processing power, data collection speeds, and communication capabilities of a PC with the reliability and form factor of a PLC. Though PACs as main controllers offer the ability to integrate several functions – such as motion, vision, high speed I/O, and enterprise communications – on one controller, many plant engineers are using PACs to supplement existing systems. Engineers can achieve advanced I/O, processing, and communication by adding PACs to their control systems using the following three common and simple approaches.

The most straightforward way to add PACs is when the application allows the design of a separate system operating in parallel with the existing control system. Because validation and training account for most system upgrade costs, adding a new system that does not affect the main control operations is desirable, especially in large process control systems or FDA-validated processes.

In these applications, one popular option is for an engineer to design and install a PAC without hooking it into the existing plant systems. An example of this type of system is the installation of a PAC on a paper mill paper machine. Although the paper mill engineers used a distributed control system (DCS) to control the paper mill, they had problems with paper breaks on one of the machines.

They needed a “flight recorder” – a way to collect and log data from the machine’s drives that they could use to analyze the cause of the breaks. Modifying the existing DCS was ruled out because, technically, the DCS could not collect and log data at a high enough rate and changing the DCS could affect other processes. Paper mill engineers decided to install a PAC that collected and logged data from the drives locally onto the controller’s CompactFlash. After a paper break, an engineer was able to use FTP to transfer the data files to his computer and analyze the paper breaks to determine the cause of the problem. Digital I/O Lines The second common method for adding PACs into existing control systems is to use a PAC to perform advanced processing and make a true-or-false decision. The PAC then communicates the decision to the PLC using a simple digital input. This design is easy to implement because, in most PLC systems, adding a digital input and a few rungs of logic is simple.

Engineers often use this approach when the PAC is performing high-speed acquisition and analysis in applications such as hydraulic transient monitoring or vibration monitoring.

In one application example, plant engineers needed to monitor rotating equipment for vibration problems. They used a PAC with high-speed acquisition to monitor the vibration of two axes on the motor shaft. Based on the measurements from the accelerometers, the PAC performed analysis to determine the frequency components of the signal. By comparing the frequency components to known limits, the PAC determined if the system needed maintenance or needed to be shut down.

The PAC then directly passed that information to the plant engineers and maintenance staff through its built-in Web server and OPC connectivity. However, if the vibration was in danger of damaging the machine, the PAC would have sent a digital output signal to the PLC controlling the system. This way, plant engineers could use the exiting shut-down procedure already implemented in the PLC while adding a protection system based on the PAC’s complex analysis of sensor measurements.

The third way to add PACs to a PLC system is to use a standard communication bus such as Modbus, DeviceNet, PROFIBUS, CANopen, Modbus TCP, or EthernetIP. Most modern PLC systems already incorporate some type of communication bus between the controllers. Engineers can take advantage of a PAC’s built-in serial and Ethernet interfaces and optional plug-in communication modules to easily add the controller to many systems.

For PLCs using buses that are not natively supported, engineers can apply external gateways. An external gateway maps the data from one network onto another type of network. For instance, with a gateway, engineers can connect a DeviceNet device to a PROFIBUS network, or they can connect a Modbus TCP controller to a ControlNet network. One recent application for which engineers used a gateway to add a PAC to a control system involved a cutting machine fed by a conveyor belt. The feed material had a high variance in size and shape, which was causing problems with the cutting machine.

To achieve proper quality, engineers needed a way to adjust the downstream machine in real time to control the cutting speed. While engineers have used vision for years in quality inspection, with PAC vision capabilities, they now can design their vision systems to run more advanced algorithms and perform decision making.

For this application, engineers used a visionenabled PAC to measure the raw material and calculate the appropriate parameters for the downstream cutting machine. They set up the downstream PLC to communicate via EthernetIP. Because the PAC system could communicate natively through Modbus TCP, the engineers used a gateway to map Modbus TCP data onto EthernetIP. In the new system, the vision-enabled PAC calculated the appropriate speed for the cutting machine and passed this information to the PLC through the gateway to adjust the cutting speed and avoid jams.

Advanced I/O, Processing, and Communication via PACs With plants under increasing pressure to cut spending, engineers continue to look for ways to incorporate new technology into their existing systems. PACs offer the capability to add advanced I/O, advanced processing, and advanced communication to existing systems. This includes motion, vision, high-speed acquisition, and signals that require special signal conditioning such as electrical power measurements, strain, LVDT, and vibration.

With powerful floating-point processors, PACs perform sophisticated analysis and control. PACs also interface to the enterprise by directly inserting information into SQL databases or hosting Web pages. Plant engineers with shrinking budgets will continue to use PACs to incorporate advanced I/O, processing, and communication to make their existing DCSs and PLC systems more efficient.

Todd Walter is the Industrial Measurement and Control Group Manager at National Instruments