Several examples of key applications of microcontrollers in modern industrial environments

No industry is growing as fast as industrial automation and control in terms of the design and application of microcontrollers in industry. As automation increases in major manufacturing plants in China and other parts of Asia, new technologies are used to improve efficiency, which has a significant impact on manufacturing costs and product costs. Although centralized control can improve the overall visibility of any particular manufacturing process, it may not be suitable for some critical applications where response delays and processing delays can lead to failures.

　　This article will introduce several examples of such critical applications, where the increase in intelligence and processing power close to the process node will have a significant impact on efficiency and reliability improvements. New system-level chip designs provide the necessary intelligence to achieve critical process measurement and control of these parameters. This article will also discuss several specific improvements in SoC design to address the design challenges and related solutions faced in designing and selecting microcontrollers in today's rapidly growing industrial fields.

　　From a historical perspective, it is not so long ago that goods such as shoes, hats, clothing, utensils, and other items were manufactured primarily by artisans with very limited manufacturing knowledge. The quality and quantity of the product depended on the skills of the particular artisan and the number of people in the industry. The original production line achieved an increase in output and, with it, improved quality. The manufacture of any given product was broken down into simple discrete steps, each of which was repeatedly handled by a worker on the line, who then passed the semi-finished product to the next operator. Each operator received only limited training for a particular step, and the overall process was the responsibility of a foreman or supervisor. The use of unskilled or semi-skilled labor ensured rapid improvements in the quality, quantity, and availability of consumer products.

　　These discrete steps of craftsmen and early assembly line operators were later mechanized, which enabled a shift from human capital (people) to local mechanized process lines where control was increasingly centralized. As central control increased, visibility into any particular process step (especially early on) became less and less, and the latency between central control instructions being issued and actual execution became longer and longer. At some point, the throughput of the entire process was impacted more as a function of the generation and impact of the response time latency associated with centralized control than the limitations of each individual process step.

　　Current process optimization strategies include the transition from analog to digital I/O (sensors and actuators), the bridging of separate process steps, and the migration from a single centralized control topology to a distributed topology. As the demand for intelligence moves closer to the process node and each step/task, large general-purpose workstations and processors give way to more specialized solutions such as microcontrollers and FPGA implementations.

　　As more processing and decision making moves off-premises, and the client-server central control model becomes a distributed model similar to a "peer-to-peer," there is a need to move toward more advanced mixed-signal SoCs. Device types such as PLCs and I/O modules, temperature/process controllers, CNC machine tools, and other applications such as flow/height measurement, encoders/resolvers, gauges/indicators/limit alarms, motor protection, and circuit breakers will be the primary applications for further improvement of local intelligence, which will be implemented in the form of a single chip that integrates multi-channel, mixed-signal A/D conversion as well as dedicated h/w ​​processing and auxiliary MCU system management.

　　An example of utilizing PLC and I/O modules, motor drive controllers and multi-axis controllers is a typical packaging application that includes an unwind roller for wound material, an infeed for delivering the items to be packaged, several sealing locations and a conveyor to remove the items after packaging is complete.

　　Open-loop control of this system often results in low throughput (efficiency) and poor quality when packaging material registration information is poor, packaging material tension is inconsistent, and packaging is poor. The end result is increased unit cost due to additional QA inspections, monitoring, and product damage, not to mention the additional labor costs of continuous monitoring (see Figure 1).

　　Figure 1: Schematic diagram of an open-loop system in industrial control.

　　The same system closed loop operation will add monitoring and feedback of parameters such as roller torque, speed and amount (length) of material remaining on the roller, speed/tension of the infeed and outfeed conveyors, temperature and pressure of the sealing rollers, registration/positioning of the product to be packaged and vibration of any critical units. The control of the primary motor drive functions (speed, tension and position) in the above example can be viewed as a servo system, which can be achieved by adding secondary measurements such as temperature and vibration.

　　In a servo system, closed-loop operation is achieved through position, velocity/angular velocity, and current loops.

　　The position loop outputs a speed/angular velocity command with rotation angle feedback information provided by an encoder or resolver so that the motor's rotation angle reaches the required position. The speed loop controls the motor's rotation speed set by the position loop through this part of the loop, and the loop is closed by feedback data from the encoder or resolver. The speed loop output is the input of the current loop, which provides the motor with current to achieve the specified position and speed. The motor current value is fed back to the current loop to minimize the difference between the command and response values ​​to zero. Teridian's patented single converter technology architecture and technology used in its 71M651x series of energy measurement devices for measuring single-phase or multi-phase power measurement are ideal for high-precision current measurement required by these types of applications.

　　A new device, the 71M6-03, is also very suitable for motor protection in packaging applications as described above. Using Teridian's patented single converter technology, the 71M6-03 integrates a 22-bit delta-sigma ADC, 6 primary and 1 secondary current sensor inputs, digital temperature compensation, a precise voltage reference, a 32-bit programmable calculation engine, timers, a real-time clock (RTC), two UARTs, and a single-cycle execution 8-bit MCU.

　　With built-in digital di/dt integrators, this programmable device supports current transformers or Rogowski coils for any or all input channels and provides transient and delayed overcurrent, ground leakage, ground fault, and arc fault protection functions. In addition, the device may be configured to support any number of traditional and customary protection algorithms that can be adapted to site-specific loads. The programmable 32-bit calculation engine (CE) receives and processes all sensor data from the 22-bit A/D converter while running independently of the 8-bit MCU, which handles higher system-level management and communication tasks. This separation of the mixed-signal measurement subsystem and the management subsystem provides high speed, high reliability, and excellent dynamic range without external interrupts or unnecessary processing overhead.

　　Integrating a multiplexed input to the 22-bit delta-sigma A/D converter achieves the lowest cost and also improves gain consistency, offset consistency, reduces crosstalk and enhances design flexibility. In addition, Teridian's 71M8100 measurement controller device is developed using the same high-performance architecture and shares many of the same functions, with the addition of three inputs that can be used to sense and control secondary parameters such as temperature, vibration, process, pressure and humidity.

　　In summary, by leveraging more dedicated SoC solutions to implement closed-loop systems, latency is reduced, and process intelligence and control can be moved from a centralized to a distributed model, which not only improves efficiency and quality, but also reduces unit costs accordingly. Further, innovation opportunities to further improve industrial automation, protection, and control emerge, and can be addressed with the same basic chip-level architecture.