Lee Teschler - Motion Control Tips

An apocalypse that wasn’t – The rare earth mineral shortage

March 7, 2019 By Lee Teschler Leave a Comment

Leland Teschler, Executive Editor
lteschler@wtwhmedia.com
On Twitter @ DW_LeeTeschler

Scan the news pages from about eight years ago and you will likely notice a lot of hand-wringing about shortages of “critical materials” crucial for high-tech motors. The materials in question were rare-earth metals used for, among other things, making magnets with a super-high energy density. Back then, China began limiting exports of rare-earth metals. Because that country supplied 97% of the world’s rare earths, the price of neodymium used in brushless motor magnets quadrupled. This price explosion led to the founding of Molycorp Inc., a U.S. mining company with a way of extracting rare earth metals from ore that experts said was innovative.

Today, the price of rare earths has come back down to earth. Recently you could buy a ton of neodymium oxide for about $47,000 whereas the same amount of material in 2011 went for over $250,000. Molycorp entered Chapter 11 bankruptcy a few years ago and its assets were recently liquidated. China still provides about 70% of the world’s rare earth metals. So you might wonder why the price of rare earths crashed, even in the face of a trade tariff war with the main source.
Insights come from a paper by University of Texas public affairs Prof. Eugene Gholz, who advises the DoD on rare earths. Gholz points out that Chinese producers found various ways of getting around the export limits the Chinese government imposed on rare earths. He estimates the Chinese illegally exported as much as 30,000 tons of rare earths annually.

Moreover, some of the applications for rare-earth permanent magnets turned out to be illusory. When the price popped, demand for these magnets fell partly because some manufacturers were able to get by with weaker ones. And materials researchers cranked up efforts to minimize the use of rare earths in magnet alloys.

Even so, the magnetics in motors and electronics accounted for only a small fraction of total rare earth production. At the time of the production bruhaha, cellphones needed a few grams of neodymium and a Toyota Prius used about a kilogram of the stuff. Toyota says it has since reduced its neodymium needs by about 20% per car, though it uses more of the cheaper lanthanum and cerium to do so. Experts predict a surplus of these two materials for many years to come.

Actually, the main application for rare earths back in the days of the bubble was as catalysts in gasoline refining. But you can refine gasoline without rare earths. Gholz reports that some refiners did so at the peak of the 2011 rare-earths price pop, at the expense of a slightly less efficient refining process. The change didn’t do much damage to their financials, he says.

Meanwhile, production of rare earths outside China ramped up. Though Molycorp met an ignominious end, Lynas Corp. in Australia now produces rare earths at a rate of about 22,000 tons annually. Russia and Brazil both put out a few thousand tons each.

The fear back in 2011 was that rare earth metal supplies would go the way of the 1973 oil embargo. But the lesson from the ensuing panic, says Gholz, is that the oil embargo was a special case. Where there’s economic coercion, there’s also an opportunity for introducing competition and a rationale for investing in potential innovations.

The downside of patent protection

September 27, 2016 By Lee Teschler Leave a Comment

Leland Teschler – Executive Editor

lteschler@wtwhmedia.com

On Twitter @DW_LeeTeschler

MUCH angst has been expressed over patent trolls that buy up low-quality patents then sue the bejesus out of potential infringers to “enforce” IP rights. Lest you think patent suits by these guys are over money earned from patents, consider this: Scholars at the National Bureau of Economic Research have found that patent trolls often sue though the patent in question has earned little or no money. The suits gets filed, scholars say, simply because the “infringers” have a lot of cash; the cash often comes from other businesses that have nothing to do with the patent.

The NBER scholars have also found that such shenanigans hamper innovation. They say firms targeted by patent trolls reduce their own R & D budgets by an average of more than 25%. The implication is that some of the funds normally spent on advancing technology instead go toward defending against nuisance lawsuits.

The trolls frequently defend their actions with claims they benefit the small inventors whose patents they defend from “infringement.” Not true, according to the NBER scholars. Little of the money patent trolls receive in litigation passes through to the small inventors. And there is no evidence any of this court action has led to an upsurge in patent activity among independent inventors.

A number of observers have recently come around to that view that patents could be more trouble than they’re worth, and not just because of spurious patent suits. A review of history shows that the mere holding of a patent is no guarantee of commercial success. In fact, it is easy to uncover tales of inventors who were consumed with defending their patent rights while others founded industries related to the work of those pioneers.

For an example, consider the case of Charles Goodyear. Goodyear received his patent for vulcanized rubber in 1844 but spent most of the rest of his life in patent disputes with Thomas Hancock and Stephen Moulton. Goodyear himself was never prosperous. The company bearing his name, Goodyear Tire & Rubber, wasn’t founded until almost four decades after his death, long after his patents had expired.

Perhaps more notorious are the patent disputes surrounding the Wright brothers. It has been reported that the Wrights had such a reputation for lawsuits that aviator Glenn Curtiss and his colleagues mockingly suggested the brothers would sue anyone jumping in the air and waiving their arms.

The Wright’s business was profitable but not overly so. And it didn’t last. Orville sold the company in 1915 after the death of his brother. The firm produced its last plane in 1916.
A more modern example that invokes discussion about the value of patents is in 3D printing. The first patents for stereolithography and digital slicing strategies were filed in the 1980s. But the 3D printing industry didn’t really boom until 30 years later. So you have to wonder how much of an advantage it was to be the inventor of those early discoveries.

There are ways to avoid the downside of patenting. One is counterintuitive: In 2014, Tesla Motors founder Elon Musk decided to make all Tesla patents open source. Tesla patents number in the hundreds. They cover everything from bumper mounting plates to battery pack safety. Musk’s rationale for the move: “I realized that receiving a patent really just meant that you bought a lottery ticket to a lawsuit.”

Then, of course, there is the approach taken by Cicoil in Valencia, Calif. In the 1950s, this maker of motion cable came up with a special process for extruding continuous lengths of flat cable out of clear material. But don’t bother trying to find Cicoil’s patents on the technique. There aren’t any. The firm treats the process and anything related to it as a trade secret.

Wearable electronics could make use of open-source microprocessors

April 26, 2016 By Lee Teschler Leave a Comment

by Leland Teschler, Executive Editor

Scientists at ETH Zurich and the University of Bologna have open-sourced the design of one of their microprocessor systems. The arithmetic instructions that the microprocessor can perform are also open source: developers made the processor compatible with the open-source instruction set RISC-V developed at the University of California in Berkeley.

The new processor is called PULPino. It is designed for battery-powered devices with extremely low energy consumption (PULP stands for parallel ultra-low power). ETH researchers say the processor could be used for small devices such as smartwatches, for wearable electronics monitoring heart rates and similar parameters, or sensors for the Internet of Things. For demonstration purposes, the group is developing a smartwatch equipped with the PULPino processor and a camera. It can analyze visual information and use it to determine the user’s whereabouts.

The PULPino open-source microcontroller-like system uses a small 32-bit RISC-V core developed at ETH Zurich. The core supports a base integer instruction set (RV32I), compressed instructions (RV32C) and partially supports a multiplication instruction set extension (RV32M). It implements extensions for hardware loops, post-incrementing load and store instructions, ALU and MAC operations. To allow embedded operating systems such as FreeRTOS to run, a subset of the privileged specification is supported. When the core is idle, the platform can be put into a low power mode, where only a simple event unit is active and wakes up the core in case an event/interrupt arrives. The PULPino platform is available for RTL simulation, FPGA and the first ASIC (called Imperio) was taped out in January 2016.

For now, the main interest in PULPino comes from academics. ETH scientists want to work with other groups to jointly develop academically interesting extensions to PULPino; these would also be open source, thus allowing the number of the hardware’s functional components to steadily grow.

Development costs are reduced considerably with the open-source royalty-free design, said ETH Professor and project leader Luca Benini. “It could result in new research and development partnerships with industry to jointly develop novel chip components on the basis of PULPino.” So PULPino's developers are planning to make their microprocessor more widely known to the open-source hardware community this year.

More programmability tops motion controller trends

March 4, 2015 By Lee Teschler Leave a Comment

motion-control-trends-Delta-Computer-stand

This test stand, developed by Houston-based Dril-Quip, uses two large hydraulic cylinders, mounted horizontally, to apply large bending and tension forces to an assembled pipe stand (center). The rams operate in concert to exert bending and tension or compression forces of over six million pounds on the pipe-stand specimen. An electrohydraulic motion controller from Delta Computer Systems operates the rams.

Motion controllers are getting better simply because state-of-the-art industrial automation needs more capable controls. That’s the message coming from motion-industry veterans these days.

“The amount of programmable features that can reside in FPGAs continues to rise,” said Joseph Profeta, control systems group director at Aerotech Inc., supplier of motion control, positioning and automation systems for medical, industrial, military, and research applications. “This four-to-eight-fold increase over the last several years lets far more functions be loaded into the FPGA on startup, so a variety of feature sets can be delivered on the same hardware platform.”

Curt Wilson, vice president of Engineering at Delta Tau agreed. “Rising speed and density of digital logic in both ASICs and FPGAs has opened up new avenues for processing the inputs and outputs required in motion control,” he said.

More capable electronics also let end users customize feedback signals for closed-loop control. An example comes from Delta Computer Systems’ Bill Savela. “Delta controllers now can switch between redundant sensors if one should fail, so production can continue. This provides a degree of fault tolerance to machines in harsh environments and lets maintenance take place on a schedule rather than when things break. Delta controllers can also switch between sensors that respond to different measurement ranges or those that provide different sensitivities,” he said.

And some industries been catalysts for improving controller designs. To handle applications in sawmills, for example, Delta Computer added instructions for curves and splines that make it easier for movable saws to process curved logs. Delta Computer also gives its controllers the ability to synchronize multiple axes to let forest products processing equipment vary log processing speed in real time to help mills deal with logs of varying quality.

Similar trends are making packaging equipment more productive. “The packaging equipment industry continues to build in more motion control, driven by the need for packaging machines with higher throughput and more flexibility in their features,” said Aerotech’s Profeta. “Additionally, test and inspection applications are utilizing more motion control. It takes sophisticated motion to thoroughly test equipment and materials. In that regard, many testing machines—for stress and strain testing as an example—are considering linear motors with motion control to replace mechanical or hydraulic actuation. The ability to apply sophisticated motion profiles and higher forces (or torques) to the test sample has motivated the transition to servo-driven test machines.”

Electro-hydraulic testers are seeing a similar transition. “Simple machines applied pressure or force in a sinusoidal pattern to a device under test for a certain number of cycles. Closed-loop motion controllers now improve precision and repeatability using various test-load profiles. They also gather data regarding the way devices respond to the stress – an important factor in validating new material compounds and in improving the efficiency of structural designs,” said Delta Computer’s Savela.

Delta Tau’s Curt Wilson pointed out that the manufacture of dental prostheses—crowns and implants—is another area benefiting from advancing servo technology. Dental molds traditionally go off to a central lab that creates the prostheses with a turnaround measured in weeks. This whole process is moving into the dental office with a turnaround of an hour or two. Advanced controllers now run small, inexpensive, but highly capable dental milling machines that dental technicians can operate right in the office.

Wilson also said huge increases in computational speed and memory capacity are making it possible for modern controllers to perform extensive real-time monitoring and analysis in many different areas. These include analyzing servo error patterns both for adaptive control and for problem identification (such as excessive vibration). Multi-core CPUs permit the use of a separate core for redundant safety monitoring, and automatic servo tuning algorithms are becoming increasingly common, he said.

Similar sentiments come from David Goodin, owner of motion system supplier AllMotion. “DSP technology has also improved. Higher processing speeds as well as the incorporation of peripherals that are necessary for controlling motors have made drivers smaller and more capable,” he said. “The ability to do things in smaller packages is also promoting the use of advanced automation. For instance, our newest four-axis driver controller has native Delta-robot support. That is, you simply feed the desired XYZ coordinates to the board. It takes care of moving the end effector to the necessary location, calculating the trajectory and the transformation of the motor angles in real time.”

At Trio Motion Technology LLC, significant trends include the addition of IEC 61131-3 programming and Ethernet-based drive interfaces, said Trio N.A. sales manager Ed Novack. The lines are blurring between the traditional PLC and motion controller as both become more full-featured machine control platforms. In addition, the IEC programming standard has over the past several years brought PLC-style programming into the motion world. The result is more incentive for PLC-savvy engineers to consider using a motion controller in situations that might not have been candidates in the past, Novack said.

Trio also sees rapid growth in Ethernet-based drive interfaces. One reason is that this type of drive interface is flexible and can cut costs by simplifying wiring. The main protocol Trio sees in ascendance is EtherCAT. Originally developed by Beckhoff Automation, EtherCAT has been adopted by over 130 drive companies and continues to grow. Other Ethernet-based protocols, such as Allen Bradley’s EtherNet/IP, are used as well but don’t excel in high-speed deterministic servocontrol, Novack said.

In particular, as the motion control needs of machine builders and manufacturers continue to become more complex, there has been a shift to PC-based controls. “Industrial PCs and DIN rail-mountable embedded PCs are making good use of multi-core technology,” said Reid Beilke, industrial PC product specialist at Beckhoff Automation.

“For example, programs for multiaxis motion control, PLC, CNC, robotics, HMI, and more can run on one industrial PC through dynamic software platforms such as TwinCAT 3 which assigns separate tasks to individual cores, evenly distributing the CPU workload,” Beilke explained. Thanks to multi-core technology, it is common to run entire machines and a robot with one PC-based hardware controller. “Entire machines and multiple machine modules in a plant can share one hardware device depending on the application,” he said.

All in all, PC-based motion systems are starting to look a lot like self-contained systems. And that is changing, to some degree, how they are sold and distributed. Aerotech and AllMotion both said that regardless of how a manufacturer sells, application engineers still end up assisting users in selecting motion control components and often have a significant role in the subsequent application of the equipment. Trio Motion, however, pointed out that the more total machine control a motion system can handle, the more likely it is that the hardware and software will be bundled and offered as a solution. And this kind of bundling lends itself to sale via distribution rather than direct from manufacturers. This bundling also drives future feature development. As an example, one distributor requested better HMI support, according to Trio’s Ed Novack. “So we introduced an integrated HMI system. Often distribution drives product development this way to better meet industry needs.”