Implementing Embedded Speed Control for Brushless DC Motors: Part 1 - Embedded.com

Implementing Embedded Speed Control for Brushless DC Motors: Part 1

Brushless Direct Current (BLDC) motors, also known as permanent magnetmotors, are used today in many applications. A new generation ofmicrocontrollers and advanced electronics has overcome the challenge ofimplementing required control functions, making BLDC motor s more practical for awide range of uses.

In this series of articles we will discuss the basics of BLDCmotors, including their construction and operation, fundamentalequations for force and torque generation, along with basic controlelectronics necessary for proper deployment.

Also covered here is 120-degree modulation and a six-step methodfor operating the motor, as well as how the modulation can beimplemented using Hall sensors and back-EMF signals. Included areimplementation examples that make use of a microcontroller unit (MCU),as well as a discussion of the necessary features of on-chip timers andinterrupt handling within the MCU.

BLDC fundamentals
A BLDC motor has two main components: a rotor made up of permanentmagnets and a stator with a winding connected to the controlelectronics. The brushes and commutation ring that are essential partsof a universal motor have been eliminated from the BLDC motor design.Instead, control electronics are used to generate a proper sequence forcommutation.

Because of its design, the BLDC motor is also known byother names: permanent magnet synchronous motor (PMSM), brushlesspermanent magnet motors, or permanent magnet AC (PMAC) motor. Sometimesit is simply called a PM motor.

The BLDC motor is based on a fundamental principle of magnetism,which tells us that similar poles repel each other, while oppositepoles attract. As Figure 1a below illustrates, when a current is passed through two coils, itgenerates a magnetic field with a polarity that creates torque on thecentral magnet—in this case, the rotor.

When a current is passed in the direction shown, the central rotorrotates clockwise. When the rotor reaches a certain position, thedirection of the current is changed so that the torque continuesfurther in the same direction. When necessary, the current direction ischanged again to continue generation of the torque.

Figure1a. Magnetic filed due to current in stator coils creates torqueon the rotor.

However, instead of two coils, actual BLDC motors typically use sixcoils positioned 60 degrees apart, as indicated by Figure 1b, below . Then, two coils ata time can be energized to create a torque sufficient to move the rotorto a desired position. When this position is reached, other coils areenergized to continue producing the torque.

Figure1b. Single pole pair 3-phase motor has six stator coils.

Thetotal amount of torque created on the rotor is calculated using theLorentz force formula in scalar form:

Here, r is the moment armof the rotor, i is thecurrent passing through stator coils, L is the length of coil, isthe magnetic field of the rotor, and (theta) is the angle between the current direction and the magnetic field ofthe rotor. The larger the current, the larger the torque in the motorbecause the magnetic field and winding length remain the same once themotor has been constructed. Designers have only one quantity to changeduring motor operation: the current.

In vector form, this formula is T= r x F , where all three quantities are given in vector formwith magnitude and direction. This formula is important because itallows designers to create an algorithm based on vector formulationwhen they want to control torque and the flux in the motor.

A BLDC motor offers many advantages over other types of motors. Itsspeed is not impeded by the stress limitations of brushes. Because ithas no brushes to create sparks, the motor can be used in hazardousenvironments. It is efficient, reliable, and generally low maintenance.The torque-speed relationship is linear. Also, a high torque-to-volumeratio means that a BLDC motor requires less copper (metal) than doother motor types.

BLDC motors do have some drawbacks, though. Rotor positioninformation is required for proper operation, so either Hall sensors ora back-EMF signal with intelligence must be used to obtain thisinformation.

In general, the motor requires external power electronics, whereasan AC induction motor achieves constant-speed operation when startedfrom and driven by an AC power supply. The BLDC motor is a 3-phasedevice. As such, it requires an inverter and, thus, a power switch. Itsrotor requires magnetic (rare-earth) metal, so it may cost more.Finally, incorrect control of a BLDC motor, especially at hightemperatures, can damage its permanent magnet, so careful design of thecontrol electronics is essential.

Despite these drawbacks, use of BLDC motors abounds in theindustry. Several examples are illustrated in the following figures. Figure 2  below shows the GE Electronically Commutated Motor (ECM),which has 12 poles (six pole-pairs) and comes in various horsepowerratings. Its electronics are mounted at the end of the motor in a casethat is the same diameter as the motor itself. The GE ECM is a 3-phasemotor that accepts a single-phase AC supply. Its stator has 18 coilsand the rotor has surface-mounted magnets. Notice that the rotor islocated inside the motor and stator is on the outside.

Figure2. Electronically Commutated Motor (ECM) with control assembly

In contrast, the pancake motor shown in Figure 3 below   positions therotor on the outside and the stator inside. The rotor has severalsurface-mounted magnets, and the stator has many coils. 

Figure3 Pancake motor assembly shows stator and rotor

The small, lowhorsepower motor shown in Figure 4below has external stators and internal rotors. All of theseBLDC motors offer high torque and low volume, giving them an edge overuniversal or AC induction motors for applications in small spaces.

Figure4. Small Brushless DC motor for appliance applications. BLDC motorcontrol

In Figure 5a below we seethat the stators in 3-phase BLDC motors are connected in a Y or a starformation. All three phases are connected in the center, which iscalled the neutral or ½ V dc point. For this type ofconnection, the sum of the currents in all three phases is zero. Notethat only two currents have to be measured; the third can be derivedeasily.

I sa +I sb + I sc = 0

Figure5a. Star or Y-winding for the stator has sum of currents equalto zero

The stator-per-phase circuit shown in Figure 5b below has one inductiveelement and one resistive element. Its torque is proportional to thecurrent as long as the magnetic field does not change.

Figure 5b
Figure5b. Stator equivalent circuit.

In this case, torque is T = k i s ,where k is constant, (theta) is the magnetic field, and is isthe stator current. If we combine k and theta we can write simply T = K *  i s , where K is known as the torque constant.

The amount of current passing through the stator coils is based onthe voltage applied and the back-EMF voltage generated. As the motorstarts rotating, it generates more back-EMF voltage, which reduces thecurrent and results in less torque. The diagram in Figure 5c below shows thatas the current increases, speed increases up to a certain point andthen becomes constant.

Torque increases up to a certain point and then decreases. Thisbehavior is typical in a BLDC motor. Flux is pre-established by themagnetic field of the rotor. Therefore, torque is controlled simply bycontrolling the current in the stator. The commutation sequence ensuresthat the rotor rotates in synchronization with the stator excitation.

Figure 5c
Figure5c. Torque and speed increases as current increases in thestator coils.

Typical hardware used to control a BLDC motor are the converter andinverter is shown in Figure 6a, below .Six power-MOSFET or insulated-gate bipolar transistor (IGBT) switchesare used in the inverter. When AC to DC conversion is not required, aDC supply can be connected directly to the inverter board. A typicalBLDC motor drive configuration is shown in Figure 6b, below . Notice that thepower switches are labeled S1 – S6 in this figure. They have othercommon names, which can be used according to the author's preference,thus

Figure6a. Typical hardware layout with converter and inverter modules.
Figure6b. Typical representation with six switch configuration.

A BLDC motor also has sensors. For example, Hall sensors and an encoder maybe used to provide information about the position of the rotor. Thesesensors are not connected with the commutation and control portion ofthe inverter and MCU.

However, because the MCU must process signals from these sensors,they must interface with the MCU. Hall sensors and the encoder areconnected to the rotor, and rotation is necessary to create Hallsignals.

Back-EMF signals are created from the high side of the phasevoltage using a resistor ladder. Current can be measured using DCcurrent transducers (DCCT) orAC current transducers (ACCT) with phase wires passing through thecoils. Additionally, certain techniques allow single-phase currents tobe measured using a shunt resistor. The back-EMF and DCCT/ACCT or shuntresistor are connected to the stator.

In motor terminology, control based on Hall sensors and an encoderis known as control with sensors, while control without these elementsis known as sensorless control.

120-degree modulation andcommutation sequence
As Figure 1b earlier  illustrates, a BLDC motor has six coils withphase settings generally denoted as Up, Un, Vp, Vn, Wp, and Wn.(Alternatively, we can use U+, U-, V+, V-, W+, and W- to indicate thesesettings.)

Three Hall sensors are located 120 degrees apart around the stator.Depending on which magnetic field passes over each sensor, the outputmay be high or low. When the north pole passes over a sensor, itsoutput is high or state 1. When the south pole passes over a sensor,its output is low or state 0. Hall sensors thus provide informationabout polarity and position.

A six-step commutation sequence is used to steer the current andproduce torque. The sequence starts with the initial position of therotor aligned properly at 0 degrees. Power at the coils U+ and V- isturned on. This excitation creates a magnetic field so that the rotorturns in the intended direction—towards the 60 degree position. Whenthis position is reached, V- is turned off and W- is turned on. BecauseU+ is still on, the U+ and W- coils are excited, and torque continuesin the same direction.

When the rotor reaches the 120-degree position, U+ is switched offand V+ is switched on. W- is still on and so the V+ and W- excitationcontinues to produce torque in the same direction. At the 180-degreeposition, W- is turned off and U- is turned on, while V+ is kept on. At240 degrees, V+ is turned off and W+ is turned on, with U-kept on. At300 degrees, U- is turned off and V-is turned on, and W+ is kept on.Finally, when the rotor completes a 360-degree rotation, W+ is turnedoff and U+ is turned on, with V- kept on. Thus, we are back to theoriginal state or step 1.

These six steps, depicted in Table1, below , form the commutation sequence that produces correctrotation in one direction. For rotation in the reverse direction, thesteps are executed in reverse order: 1, 6, 5, 4, 3, 2 and back to 1.

Table1. Six steps for 120-degree modulation.

In our description, 'step' is synonymous with 'state'. Figure 7 below shows the completesix-step sequence with angles given in units of radians. Figure 7 alsoshows the current flow as it enters from one coil and exits a secondcoil.

This current flow corresponds exactly to the six steps of turningthe switches on and off. Since each positive phase (U+, V+, and W+) isenergized for 120-degree rotation, and each negative phase (U-, V-, andW-) is also energized for 120-degree rotation, this type of modulationis called 120-degree modulation.

Figure7. Six step modulation for Brushless DC motor.
Figure8. Six steps modulation with switch configuration

At each of six steps, one power-MOSFET or IGBT is switched on oroff, hence the term 120degree six-step commutation. Figure 8 above illustrates theseprinciples in a 120-degree drive system.

Figure9. Six step commutation with phase currents behavior.

Control electronics – in particular the MCU – play an importantrole in this operation. Hall-effect signals are fed into MCU asexternal interrupts. With every interrupt signal, the MCU performs astate change; in other words, it turns off one switch device and turnson another one. The MCU performs its task by executing interrupt-basedcode and changing the state of the output pin. The MCU has threeinterrupt input pins, one for each Hall sensor, and six output pins,one for each switch driver.

The operation of a motor with 120-degree six-step commutation,along with the behavior of the phase currents, is shown in Figure 9, above.

Next, in Part 2: Brushless motorcontrol using Hall sensorsignal processing

Yashvant Jani is director ofapplication engineering for the systemLSI business unit at RenesasTechnology America.

References
1. PowerElectronics and Variable Frequency Drives Technology and Applications,Edited by Bimal K. Bose, IEEE Press, ISBN 0-7803-1084-5, 1997
2. MotorControl Electronics Handbook, By Richard Valentine, McGraw-Hill,ISBN 0-07-066810-8, 1998
3. F IRSTCourse On Power Electronics and Drives, By Ned Mohan, MNPERE, ISBN0-9715292-2-1, 2003
4.Electric Drives, By Ned Mohan, MNPERE, ISBN 0-9715292-5-6, 2003
5. Advanced Electric Drives,Analysis, Control and Modeling using Simulink, By Ned Mohan,MNPERE, ISBN 0-9715292-0-5, 2001
6. DC Motors Speed Controls Servo Systems including Optical Encoders,The Electro-craft Engineering Handbook by Reliance Motion Control, Inc.
7. ModernControl System Theory and Application, By Stanley M. Shinners,Addison-Wesley, ISBN 0-201-07494-X, 1978
8. The IndustrialElectronics Handbook, Editor-in-Chief J. David Irwin, CRC Press andIEEE Press, ISBN 0-8493-8343-9, 1997

This article is excerpted from apaper of the same name presented at the EmbeddedSystems Conference Boston 2006.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.