The Biggest Trend in Controls is Choice
| By: Kristin Lewotsky, Contributing Editor
Thanks to ever improving chip designs and microlithography systems that continually shrink dimensions, the IC industry can put enormous computing power in ultrasmall packages. That smartphone in your pocket boasts more transistors, faster clock speeds, and more memory than the Apollo-era computer systems used to put men on the moon. These capabilities have revolutionized our world, and motion control is no exception. Modern microprocessors and memory modules have brought intelligence to components throughout the system, from smart encoders to smart drives to smart motors. Even PLCs have gotten smarter.
Machine architecture used to be straightforward. The PLC ran machine operations. The motion controller sent commands to the drive, which in turn commanded the motor to position the load. Much of that is changing, now. The intelligence springing up throughout the machine is enabling each class of components to take on upstream and downstream tasks. Drives can handle increasingly sophisticated control needs, while PLCs are powerful enough to control both machine operations and more complex motion tasks. Meanwhile, motion controllers themselves have gotten more capable also, to the degree that they can perform general machine control. The biggest trend in the controls space is one of ever greater choice for OEMs and machine builders, with greater ease of use, better, and even improve pricing.
PLCs have always been packed with intelligence but today the emphasis is on increased functionality. “PLCs are becoming more capable motion controllers in and of themselves,” says Robert Miller, senior product manager for controllers and HMIs at Mitsubishi Electric Automation (Vernon Hills, Illinois). “You can get PLCs that can do coordinated, linear interpolation with servos or encoder following or even really high-end motion capabilities to 96 axes.”
In classic machine architecture, PLCs managed the machine logic. The devices are particularly well-suited for tasks like timing and counting and handling I/O. That said, the trend toward performing motion control with a PLC isn’t entirely new. PLCs have had modest motion capabilities for quite some time. In the past, this was typically restricted to simple operations such as point-to-point moves or operation of a few axes. More complex motion tasks required a dedicated motion controller.
Motion controllers are designed to deal with high axes counts, coordinated motion, and other computationally intensive tasks. On the downside, motion controllers add cost and complexity. They often use specialized programming languages, which can be a challenge to companies with limited engineering resources. Today’s highly functional PLCs provide options.
“The processing power is just increasing so much that now your typical PLC has the ability to do a lot of motion control,” says Craig Nelson, senior product marketing manager at Siemens Industry (Norcross, Georgia). “Your PLC programmer is also becoming the pseudo motion control programmer, as well.”
“Now that the PLC is doing the motion, you can program your motion the same language that you're writing the PLC code in,” says Miller. “It reduces the number of programming languages you need to know.”
Being able to program the motion through the PLC using PLC languages provides multiple benefits. The component count drops, as does the cost. The demand for engineering skills is also reduced. This can be a significant benefit. For every organization with broad expertise in motion control, there are more who use it as a support technology for their core product – to move a patient bed in a medical imager, to position wafers in a microlithography system, to move well plates in a DNA sequencer, etc. Engineering teams at those companies may not necessarily have the in-house expertise to implement conventional motion control. The skill set required for the PLC approach tends to be more broadly available in industry.
“I think a lot of organizations are struggling with finding engineers who can program in a high level motion language,” says Nelson. “It's a lot easier to find someone who is a generic PLC programmer. Having the ability to implement standard motion-type functions on the PLC using motion functions blocks in the ladder logic is an important benefit in usability. I think that's the big trend that we're seeing.”
The ability to use more familiar programming languages is just one factor driving the trend of motion via PLC. The PLC community has made a concerted effort to enhance usability. In the early 1990s, the International Electrotechnical Commission (IEC) developed IEC61131, which was aimed at standardizing and streamlining the programming of PLCs. IEC 61131 established a structure for the programming process, encompassing five languages commonly used for PLC programming. The five languages are:
- Function block
- Instruction list
- Ladder logic
- Sequential flowchart
- Structured text
Each language has its sweet spot in terms of tasks. Structured text is effective for calculations, for example, while ladder logic excels at basic machine control. Function block works particularly well for programming motion control.
On the heels of the release of IEC 61131, industry established PLCopen, an effort to, among other things, build a library of 61131-compatible function blocks. With the aid of industry members, the consortium has put together a significant amount of resources for motion control, in particular. The function blocks provide a solid starting point for developing code. They can be tailored to the parameters of the application. They also can be customized and password-protected for reuse.
“You get the high usability of a ladder type PLC programming language with the functionality of technology objects or special function blocks that can do the more complex motion applications that haven’t traditionally been done with a PLC,” says Nelson. “So now we're able to do more camming and gearing, of a robotic kinematic function in the PLC because the processing power and these function blocks give them a higher level of usability.”
Open-source blocks provide a starting point for users. Vendors, OEMs, and end-users can—and do—customize function blocks for specific applications. The function blocks still provide a solid starting point for developing code.
In some cases, companies work with PLC open. In others, they develop their own libraries of function blocks that they implement using a point-and-click interface. “You don't have to program in ladder logic or function block,” says Miller. “You create the motion-control routine within the PLC software and it automatically creates the code for your application.”
Greater processing power and ease of programmability in PLCs are just two of the factors driving this trend. Another aspect is the availability of intelligent drives with built-in connectivity. The combination of the two makes it feasible to implement motion on a machine without a motion controller.
A PLC running motion control implies a centralized control architecture. The drives may be installed in close proximity to the motors, but they will need to be wired back to the PLC either directly or through a fieldbus. This type of centralized control has traditionally been effective for the most demanding motion tasks for applications like semiconductor wafer and mask inspection, and gantry operation. Of course, centralized control requires more cabling, which increases cost, assembly time, and points of failure. These concerns led to the development of decentralized control architectures implemented with smart drives.
Today’s intelligent, connected drives can operate in a master-slave architecture over the machine network to perform increasingly sophisticated motion control. High-speed communication supported by industrial ethernet makes it possible to close the control at very high rates. The technology has been available for more than a decade and becomes more affordable and easier to use every year. Broad adoption of distributed control always seems to be a year out in the distance, however. To gain a better understanding of why OEMs and machine builders might be reluctant to make the transition, read our recent spotlight on trends in drives.
This discussion misses an important point which is that it does not have to be a choice of one or the other. “More recently, some interesting hybrid solutions have emerged involving dual control loops. They can be very effective,” says Jason Goerges, manager at ACS Motion Control (Bloomington, Minnesota). “The high-precision multi-axis motion control market has long been dominated by centralized control architectures or distributed control architectures with proprietary high-speed, highly synchronized networks,” he says. The high-speed network he describes is more expensive than a typical version and erases many of the benefits of choosing a decentralized architecture. Fortunately, there are alternatives. “A more recent and novel solution for such applications is to implement a distributed control architecture for the entire machine with a standard high performance network such as EtherCAT. Then, you control the most demanding interdependent axes, such as a high precision linear motor gantry stage for wafer/mask positioning, with a multi-axis network drive. This approach can provide the flexibility of a distributed control architecture with the performance of a centralized/proprietary control architecture for the axes that require it.”
Motion controllers and machine control
Of course, motion controller technology has not remained static during the developments described above. Even as PLCs have been enhancing their ability for sophisticated motion control, motion controllers have increasingly added to their I/O capabilities. The result is that for some applications, the motion controller can operate the machine without the need for a PLC.
“Although our motion controller still goes into the rack with the PLC, it can handle its own I/O,” says Miller. “You don’t necessarily need to program the PLC at all. You can just do everything with the motion controller.”
Pinpointing the crossover point between using the PLC for motion control and the motion controller for machine control is not an easy task. “There are not a lot of limitations when it comes to using a motion controller in a PLC rack because we can just put more I/O in the rack and we just say that we don't want the PLC controlling it. We want the motion controller to control it,” Miller says. “The lines become blurry.”
So what’s an engineer to do with a Venn diagram of controls options coalescing down to a set of one? As always, look to the application. If the motion is not highly synchronized, a low-end PLC may very well be sufficient. Versions on the market today can handle up to 16 axes of motion.
It really comes down to the number of axes that need to be synchronized. If the application requires computationally intensive tasks such as circular interpolation, helical interpolation, or electric line shafting, a dedicated motion controller may still be your best bet.
No two projects are alike. The variety of control options available to today’s engineers makes it possible to develop the exact right solution for every application, right down to the budget.