Electrically driven alternatives to actuators such as hydraulic cylinders, conveyor belt systems, and mechanically powered presses, pumps, and winches offer advantages that are attractive for product developers and end users. Often, the main goal is to increase energy efficiency in pursuit of environmental goals while decreasing costs. However, markets also expect to benefit from the practical advantages of electric drives, including more flexible control, smaller size, lower weight, and reduced maintenance.

In product design departments, decision makers are moving from the strategic “if” and “when” - now already resolved - to more practical considerations: how best to implement the advanced new electric drives now embedded in the product roadmap. Of course, these drives will be electronically controlled to deliver precision, flexibility, and added value. With certain motor types, such as brushless DC (BLDC) motors, electronically controlled commutation is the only practicable option. However, the design of the electronic drive influences additional metrics besides simply turning the rotor and controlling speed and torque.

Diverse design demands

The drive must be expertly designed to meet energy efficiency targets, to compete with alternative products in the marketplace and to satisfy applicable eco-design specifications. In addition, proper design is needed to keep acoustic noise within acceptable limits and avoid unwanted vibration. Compensation can also be applied for the distortion of internal magnetic fields caused by imperfections in the motor, and the design should take care to minimize the electromagnetic emissions associated with the switching signals applied to the power stage.

Performance in relation to these issues is closely connected to the inverter and power stage topologies. However, the controller also needs to handle signals from sensors and diagnostic information to support condition monitoring, needed to extend the drive’s service life. The controller is also responsible for ensuring safe operation in the event of system failure. Safetyconscious industrial applications may require support for specific safety modes such as safe torque off, safe stop, safe operating stop, safety limited position, safety limited speed, safe speed monitor, safe direction, safely limited increment, and safe brake control. Also, designing-in regenerative circuitry has a direct impact on energy efficiency by recovering kinetic energy back to electric energy during braking or deceleration.

A microprocessor or powerful microcontroller can be used to manage the drive. Many suppliers offer free motor-control algorithms with their silicon to help accelerate solution development. If a microcontroller is used, relevant peripheral functions such as a PWM timer and Gigabit Ethernet interface may be integrated on-chip and providing extra convenience. However, there are limitations associated with either approach. The maximum processor performance places an upper limit on the control loop frequency, therefore the motor speed range. To accelerate the motorcontrol loop while also handling the additional application-level processing, it may demand a larger, more expensive, and power-hungry processor.

Designers also need flexibility to adopt more sophisticated power topologies, such as multi-level inverters to differentiate certain products within the portfolio. These are particularly suited to high-voltage drives that leverage the properties of wide-bandgap power semiconductor technologies such as silicon carbide (SiC). The challenge to control multi-level inverters is more complex than with the conventional two-level topology. Hence, more computational power is needed, in addition to the obvious changes to inverter hardware. In addition, ensuring scalability is important to allow control of extra motors. For example, to extend the number of axes in a robot arm or synchronize ('synchronize' in NA and 'synchronise' in EU and APAC) several motors to handle particularly heavy loads. This may require a more powerful processor or a second processor.

In industrial contexts, in particular, there are also opportunities to integrate additional capabilities such as control of the human-machine interface (HMI) as well as IoT connectivity. Choosing a suitable programable platform on which to build these motor drives gives designers flexibility and scalability to meet current and future market demands. Many of today’s FPGAs integrate a hard processor core, with additional DSP elements that can be used to parallelize processing where needed for greater throughput or to support extra channels. High-speed interfaces such as Gigabit Ethernet are also available alongside traditional FPGA logic fabric that can be used to implement custom peripherals as required.

On the other hand, controllers for modern drives are more likely to incorporate neural networks to handle processing for intelligent condition monitoring, vibration detection, and anomaly detection. For these, a more comprehensive programmable architecture can enable greater flexibility and integration. Adaptive SoCs such as the AMD Versal™ series integrate optimised AI engines that can be used for neural networks.

In addition, designers need to take advantage of proper methodologies for designing software, particularly with a greater focus on safety as well as basic control. A design flow such as the AMD Vitis™ and Vivado™, and MicroBlaze™ Compiler has safety certification up to SIL 4. There is also the opportunity to leverage Python within the design flow. Python has libraries for machine learning and artificial intelligence as well as data analysis and visualisation libraries, which can allow analysis of motor performance and operational parameters as well as support the development of predictive maintenance capabilities.

Flexible development platform

In addition to FPGAs and adaptive SoCs such as Versal™, the AMD Kria™ system-on-module (SOM) concept leverages programmable hardware and seamless integration with AMD design tools to streamline development of efficient, high-performance electric drives. The SOM can execute integrated programmable logic for motor-control loops in a few microseconds, enabling the control algorithm to run at more than 100,000 loops per second for highly accurate voltage and current control. This approach provides tools to help maximise electrical efficiency and reduce energy consumption, ensure low vibration, and respond quickly to compensate for magnetic field imperfections. Designers also retain the option to realise advanced control algorithms using a model-based approach such as Simulink. In addition, flexible programmable modulation in hardware can optimise EMI behaviour and built-in memory such as dynamic block RAM (BRAM) permits streaming motor data and facilitates health monitoring of the electric drive. Additional features include integrated communication interfaces such as EtherCAT and support for and services like DDS and OPC/UA for cloud communication. Ready-to-use applications for customisable drives are available and can be downloaded to the SOM.

Moving from mechanical to electric drives permits superior energy efficiency with lower noise and vibration, and generally higher integration with support for value-added features. Hardware options, including FPGAs and adaptive SoCs and SOMs, supported by a shared design flow, give designers the flexibility they need to satisfy diverse technical and market demands.