There was a time when programmable logic controllers (PLCs) were the only game in town when it came to controlling motors. Today, options abound, ranging from PLCs combined with dedicated commercial off-the-shelf (COTS) motor controllers to feature-rich programmable automation controllers (PACs) to field-programmable gate arrays (FPGAs). The right choice depends on a number of factors such as axis count, performance specifications and the required time-to-market.
PLCs are discrete-control devices good at managing input and output, and operations such as timing and counting. They can also be used for very simple motor control, for example, to start and stop an AC motor or to drive a stepper motor. Many applications require multiple axes of motion, however, and these systems may be better served by microprocessors (MPUs) designed specifically for the job. Depending on the application, an MPU can be used separately or in conjunction with a PLC that runs the overall system.
The STMicroelectronics L6460 controller is ideal for many applications (Figure 1). It features a four-bridge device that can be configured to simultaneously drive either four DC motors or two DC and one stepper motor. Each bridge is capable of controlling one DC motor, and bridges three and four combine to control a stepper motor, including bi-directional functionality. The device offers a range of power options, including standby, hibernate, and low-power modes.
Figure 1: The four-bridge L6460 can simultaneously drive four DC motors or two DC motors and one stepper. (Source: STMicroelectronics)
The controller was designed to provide designers with maximum flexibility. It sports a variety of embedded features, including a multi-channel 9-bit A/D converter, plus voltage regulators, voltage comparators, and operational amplifiers, all of which can be configured via the integrated serial protocol interface (SPI). If circuitry is not used for driving the motors, it can be applied to other uses. For example, when bridge three is not used to drive a motor, it can be used to realize a pair of switching regulators, courtesy of two integrated regulation loops. And if bridge three does drive a motor, the regulation loops don't have to go to waste: with the addition of an external FET, they form a switching regulator.
When an application requires coordinated motion, design OEMs may need to move to the next level of control complexity. For example, the detailed, real-time path planning required by a packaging line may require a dedicated motion controller slaved to a PLC or PC to achieve the best performance. Some of today's most sophisticated motion controllers even boast both discrete-control capabilities and communications interfaces. These wouldn’t be the right choice for a machine with a few dozen axes, but they can provide a good solution for a low-axis-count machine that needs highly-coordinated motion, such as a medical instrument.
PACs integrate multiple controls
Complex applications like printing, paper processing, and glass forming often combine process control, sequence control and motion control in addition to overall machine operations. One approach is to use a dedicated control system designed specifically for each type, then control the machine as a whole with a PLC. Although this approach can get the job done, it tends to result in extremely complex control architectures. Not only do designers have to implement different hardware and software for each control subsystem, but the parts must be integrated into a whole. This approach can not only slow down the development process, but the interfaces and middleware involved can slow communications within the machine itself.
In response, the motion control industry has begun moving toward PACs. PACs feature motion-control and process-control capabilities built right into the device. Instead of purchasing three or four control devices and learning how to program, commission and maintain them, designers can work with an integrated PACs development environment.
Whether PACs represent a distinct class of device or merely an enhanced PLC is a matter of opinion. In many ways, it doesn't matter what the chips are called, the most important feature is the high level of integration and functionality offered. PACs don't just make life easier for machine builders; it makes motion accessible to a broader base of users.
In the past, the design team for a process-control OEM might have avoided motion control, perhaps due to a lack of in-house expertise, or an unwillingness to bear the additional cost. Since a PAC includes motion control capabilities as part of the package, programmed by a common language, adding a few servo axes becomes easy.
FPGAs let software do the work
Some applications present performance specifications too demanding for standard control solutions. An FPGA may be the best choice for systems that need to supply custom capabilities, but not in numbers high enough to justify the development cost of an application-specific integrated circuit (ASIC). FPGAs provide a highly customizable solution in a common hardware platform. Because the functionality of an FPGA is defined in firmware or software, the devices offer a significant degree of flexibility during the design phase and after deployment.
The flexibility and functionality of an FPGA makes it particularly useful for OEMs trying to speed the commercialization process, especially for products that might be produced by the dozens or hundreds. The flexibility allows designers to try different implementations during prototyping, but that same FPGA can also be practical for product being shipped. The Spartan 6 from Xilinx Inc. features a dual-register, six-input lookup table (LUT) logic structure and built-in system-level blocks like DSP slices. With 12 Gbps memory access bandwidth the chip can operate quickly enough for even the most demanding motion tasks. The LXT version features as many as eight 3.125 Gbps GTP transceivers, along with integrated PCI capabilities.
The Spartan 6 development kit makes the design process even easier. The kit provides the chip already connector-ized on a board with memory and power supply, as well as software, reference designs and schematics (Figure 2).
Figure 2: The Spartan 6 development kit includes a board with basic infrastructure like memory and connectors already included. (Source: Xilinx Inc.)
An FPGA can be programmed to perform parallel data processing, allowing it to support computationally-intensive motion techniques like observer control. High-speed applications such as pharmaceutical packaging involve processing hundreds of pieces per minute. At those speeds, the biggest obstacle to increased throughput is often the settling time required after a piece has been positioned. For these types of applications, in which the system begins braking before reaching its destination, trapezoidal motion profiles are not sufficient. Observer control uses a software "observer" developed from a combination of sensor feedback, look up tables and algorithms to produce sophisticated motion profiles that can dramatically reduce settling time. The parallel-processing capabilities of FPGAs provide a practical method for executing the technique.
The parallel approach provides other dividends—by splitting a computationally intensive data path into multiple parallel paths, designers can reduce the energy dissipated, simplifying thermal management and increasing efficiency.
Motors play an essential role in many of today's products and new MPU and FPGA options give designers more choices than ever before. Whether an application calls for operating a single letter or tenuous, highly-coordinated motion from dozens of axes, designers can find a control solution to suit any skill set and budget.