This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Microprocessor-based motor control: is there finally a universal solution?

Author : Frank Malik is Principal Engineer of Processing & Control at Toshiba Electronics Europe

05 June 2017

Motors are one of the most valuable items in modern connected factories. They are also one of the largest energy users and, therefore, industry continues to demand technology to be more efficient.

Figure 1: Block diagram of a typical FOC control system

While there are still efficiencies to be gained in both the motors themselves and the semiconductors that control them, the area with the greatest potential efficiency benefit is the way in which motors are driven. However, this is a particularly challenging area, which is why designers increasingly seek ‘off-the-shelf’ solutions that address as many requirements as possible.

The energy challenge

The International Energy Agency has estimated that electric motors account for nearly half (46 percent) of global electricity demand, thus making them a focus for energy efficiency initiatives. Saving energy in this area is seen as so important that, in 2009, the European Union issued their energy-related products directive (ErP) 2009/125/EC. While the directive focused on many energy-consuming devices, regulation 640/2009 relates specifically to electric motors, whether they are sold as stand-alone devices or as part of larger equipment.

Clearly, motor manufacturers and users are faced with stringent legal and commercial challenges. If their products do not comply with the directive, they will not sell. While the motor itself is a fundamental element of a motion control system, the method of driving the motor has a large impact on overall system efficiency.

Types of motor and control 

Compared with traditional brushed AC and DC motors, Brushless DC (BLDC) motors offer a number of significant advantages including improved reliability and lower costs. Unlike traditional motors, BLDC motors have no commutator meaning that they require more complex electronics to achieve the torque control that is required by modern applications.

Speed control for BLDC motors minimises the current in the stator windings for keeping the selected speed. This ensures that all of the effort is directed into turning the motor, delivering optimal efficiency and reliability. One of the challenges of this approach is to sense the current to allow comparison with the desired torque.

In trapezoidal motor control, the stator currents are controlled to be equal in the windings on either side of the rotor, while the third winding is unpowered. As the rotor spins the current of each phase is cycled through positive, zero, and negative. This creates a trapezoidal current that approximates to a sinusoidal waveform. However, trapezoidal control can lead to imprecise control and audible noise, especially at low speeds.

Figure 2: Block diagram of TMPM37x vector engine microcontroller

Sinusoidal control uses phase-shifted sinusoidal current waveforms to produce smoother torque than the trapezoidal approach. This requires more accurate rotor position information and current values have to be calculated rapidly. At higher speeds, any lag in this calculation will lead to inefficiency.

Field-oriented control 

Field Oriented Control (FOC), sometimes known as Vector Control, is a mathematical approach to controlling BLDC motors that overcomes the poor low-speed accuracy of trapezoidal control while addressing the high-speed inefficiency of sinusoidal control. FOC is a sensorless technique, so the space, weight and energy consumed by a rotary encoder is saved with this approach.

FOC maintains a constant stator field in quadrature with the rotor field by manipulating the motor currents and voltages with reference to the rotor’s direct and quadrature axes. The sensed stator currents are converted into direct (D) and quadrature (Q) components. These components are then compared with the required torque and zero to create an error signal. These error signals are processed in a software-based Proportional-Integral (PI) function to create PWM drive signals for the motor. 

FOC is efficient across all motor speeds and is not affected by the PI function bandwidth. However, real time FOC requires fast execution of the functions to transform the sensed stator current signals into the voltage-control signals for the output bridge. Software-based FOC demands a significant portion of available CPU performance to complete the calculations in a timely manner, especially at fast rotor speeds. In some cases, the processing ability of the system may be the main limitation on the rotational speeds achievable.

In order to remove the dependence on the main processor performance, dedicated hardware platforms for FOC based motor control have been developed. Toshiba’s original Vector Engine (VE), for example, moved the complex vector control equations into a dedicated hardware engine with customisable firmware. Also included in the integrated solution was a Programmable Motor Drive (PMD) block to generate the PWM waveforms and perform other necessary functions such as dead-time control. 

By reducing the software content, the VE ensures stable and predictable execution of code that is not impacted by interrupts or the quality of the software. As standard software is provided as part of the development environment, designers can focus on their core competencies and bring products to market quicker. 

Figure 3: Comparison of software and hardware based FOC implementations

Also, as dedicated hardware can execute up to 70 percent faster than a software-based solution, higher rotor speeds can be achieved. Moreover, the hardware VE approach releases main CPU resources for high-level product features. In many cases, when a hardware VE is implemented, a lower performance main CPU can easily provide the required application-level functionality thus saving space, cost and energy. 

However, many hardware-based motor control solutions are inflexible and not suitable for deploying a single core solution across a range of different applications. This can mean that some designers continue to use the more complex and time-consuming software based solutions. What is needed, therefore, is a platform that allows engineers to migrate their existing FOC algorithms from a pure software environment.

New and emerging technologies

Some of the latest microcontroller developments are helping to address this challenge. Take, for instance, Toshiba’s TMPM37A. Capable of running at speeds of up to 40MHz, this micro represents the latest addition to the company’s TX03 series of ARM Cortex-M3-based devices and combines high levels of integration with a small footprint. Indeed, housed in a VQFN32 package measuring just 5 x 5mm, the new solution is the world's smallest microcontroller to incorporate Toshiba's Vector Engine Plus (VE+) as well as a pre-driver. 

By integrating a pre-driver, the TMPM37x can directly drive MOSFETs with a complementary three-phase output with a minimum unit of 25ns. Direct microcontroller drive and control of small motors is now possible.

Furthermore, the TMPM37x series only requires a single 5V supply, with on-board conversion to 3.3 and 1.5V. Also included are up to two on-board 12-bit ADCs with more than a dozen analogue input channels. The ADCs offer a constant conversion mode and complete conversions within 2µS, when using a 40MHz conversion clock.  Some products incorporate a 4-channel operational amplifier for single or three shunt resistor current detection as well as a comparator for over-current detection. External interfacing is extensive with up to 74 I/O pins.

The full-function TMP37x includes up to two-channels of encoder input circuit (ENC) that corresponds to incremental encoders (AB/ABZ). Allowing for 3-phase input, the ENC is able to detect rotation direction and includes a comparator for position detection as well as a counter for absolute position detection.

Other features integrated into the advanced TMP37x include a watchdog timer, power-on reset, voltage detect and a general purpose serial interface (SIO/UART). 

Figure 4: Toshiba MotorMind PC software allows rapid code-free development

Even though the vector control is hardware-based, the VE allows for widely differing solutions to be implemented via software. This brings flexibility and hardware-based performance into a single solution. This flexibility is further enhanced through an architecture that allows developers the flexibility to choose whether to use their own IP or take advantage of hardware acceleration using IP from Toshiba or combine both.

The closely-related TMP47x series is based on an ARM Cortex-M4F processor and offers operation at speeds up to 120MHz.

The recently announced M4K group is a single-chip solution with a low pin-count for controlling multiple motors - especially in home appliance and HVAC applications. The new products support the RAMScope interface that can confirm parameters in real time without affecting motor operation.

Development support

Alongside a number of specific reference design boards and code examples, Toshiba's MotorMind software provides a starting point to easily setup and initialise a motor. The software enables designers to enter basic motor parameters without the need to write any code.  

MotorMind offers an advanced graphical interface to show actual motor parameters including speed and torque. The package also features an integrated µDSO (Digital Storage Oscilloscope) to visualise registers from inside the Vector Engine based upon optional, configurable, trigger points. 


Contact Details and Archive...

Print this page | E-mail this page