Modified Sensorless Control Method Engineering Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Abstract- The permanent magnet Brushless DC (BLDC) motor offers many advantages including high efficiency, low maintenance, reduced weight and more compact construction. The BLDC motors have been widely used for various industrial applications based on their inherent advantages. This paper presents design and implementation of a modified sensor less control scheme for the brushless dc motors using line-line voltage detection. The sensors in conventional method increase the cost and reduce the system reliability. In this modified method, the virtual hall sensor signals are extracted directly from the specific line to line voltages with simple RC integrator circuit and comparators and gives output signals same as hall sensor outputs. Compared to conventional schemes, motor neutral voltage and phase shifting circuits are not used in this modified scheme. So the commutation signals are insensitive to the common mode noise and the complex algorithm for implementation can be eliminated. A special starting up sequence is chosen for smooth starting and reliable transfer to sensorless control.

Keywords- Brushless dc motor (BLDC), sensorless control, back EMF detection, neutral voltage, PWM, dsPIC30F4011


Permanent magnet motors with trapezoidal back EMF and sinusoidal back EMF have several advantages over other motor types. Compared to dc motors they have lower maintenance due to the elimination of the mechanical commutator and they have a high-power density which makes them ideal for high-torque-to weight ratio applications. Compared to induction machines, they have lower inertia and allowing for faster dynamic response to reference commands. Also, they are more efficient due to the permanent magnets which results in virtually zero rotor losses. The permanent magnet brushless DC (BLDC) motor offers many advantages including high efficiency, low maintenance, reduced weight and more compact construction. However, the control complexity for variable speed control and the high cost of the electric drive hold back the widespread use of brushless dc motor.

New developments in power semiconductors, micro controllers and adjustable speed drives control schemes enable reliability and cost-effective solution for a broad range of adjustable speed applications. A BLDC motor requires an inverter, control circuit and a position detector to perform commutation [1-3]. Conventionally, three Hall sensors are used for position detection of the BLDCM. However, the position of the Hall elements must be very precisely fixed, and the sensors themselves add a cost and reliability penalty [4], [5], [6]. Hence we go for sensorless control. This research on sensorless control of BLDC motor can be divided into five categories [7]. They are, (1) Back EMF sensing techniques (2) Flux estimation method [8] (3) Stator inductance variations method (4) Observers based control techniques and (5) Using the special functions of motor variables. Among these, the back EMF detection method is the most popular because it is easy to implement.

In conventional sensorless BLDC motor drive with back EMF detection method, commutation points of the inverter can be obtained by knowing the zero crossing point of the back-EMF..But there are some practical implementation problems when using the phase to neutral zero crossing detection method. In this case the detection of neutral will introduce a high common-mode noise. Since the zero crossing points of the conventional back EMF method are leading 30 electric degrees of the ideal commutation points, hence a phase shifting circuit is needed to get the commutation points [9]. The use of phase shifting circuit, virtual neutral point and complex algorithm in back EMF detection method make the conventional sensorless control scheme more complex than the Hall Effect sensor based commutation.

To avoid the above mentioned problems, instead of detecting the back EMF of motor, the commutation signals can be extracted directly from the line to line voltage of a BLDC motor, which is in phase with actual hall sensor signal. So the conventional control algorithm can be used. It uses a simple RC low pass filters and low cost comparators [10]. Unlike conventional solutions, this method does not require additional virtual motor neutral voltage, complex phase shift circuits, complex digital filters and precise speed estimators. In this paper, modelling and simulation of the BLDC motor and the experimental results of sensorless control method using dsPIC30F4011 are discussed. Also a comparison between the performance of sensorless method using phase terminal voltage and line-line voltage detection method is included.

BLDC Drive system

The BLDC drive system consists of a BLDC motor, inverter, and control circuit and hall sensor for position information as shown in figure 1.

Fig.1 BLDC motor drive system

In conventional sensored control the speed of the motor is compared with its reference value and the speed error is processed in a proportional integral (PI) speed controller. By knowing the position information and speed error, the control circuit will generate the required PWM signals with suitable duty ratio. From figure 2, it is clear that the commutation points and hall sensor signals have a phase difference of 300.

Fig.2. Back emf, current and hall sensor signals

Modelling and simulation of bldc motor

The BLDC motor consists of three stator windings and permanent magnets on the rotor. The equivalent circuit of motor is shown in figure 3.

Fig.3. Equivalent circuit of BLDC motor and Inverter

The modeling is done, based on the following assumptions:

(1) The motor is not saturated.

(2) Stator resistances of all windings are equal and self and

mutual inductances are constant.

(3) Power semiconductor devices in the inverter are ideal.

(4) Iron Losses are negligible.

The following equations are used for modeling the BLDC motor.

Van=Vao-Vno, Vbn=Vbo-Vno, Vcn=Vco-Vno (1)

Where Vao, Vbo, Vco and Vno are the three phase and neutral voltages with respect to the zero reference potential.


Where, the stator resistance of all the windings is equal. Ra = Rb = Rc= R. The back EMF's ea eb and ec, have trapezoidal shapes and It can be modeled using S function. Assuming further that there is no change in the rotor reluctances with angle, then

La = Lb = Lc= L

Lab = Lba = Lcb = Lbc = Lca = Lac =M

ean = E 0° < θr <120°

ean = (6E *pi) (pi - θr) - E 120° < θr < 180

ean = -E 180° < θr < 300°

ean = (6E *pi) (θr - 2 pi) + E 300° < θr < 360° (3)

Since there is no neutral used, the sum of the three phase currents must add up to zero, i.e.

ia+ib+ic= 0,

ia+ib= −ic

Mib + Mic = - Mia (4) By simplifying equation (3) we will get the differential equation as below,

The electromagnetic torque is,

Te= (ea ia+ eb ib+ ec ic)/ωr (6)

Substituting the back emf in normalized form, the developed torque is as,

Te =Kb {fa (θr)ia +fb(θr)ib + fc (θr)ic} (7)

The equation of motion is,

pωr=P /2 (Te- TL- Bωr)/ J (8)

Where P is the number of poles, TL is the load torque in Nm, B is the frictional coefficient in N-m/rad, and J is the moment of inertia in kg-m2. The derivative of the rotor position θr in state space form is expressed as,

pθr =ωr (9)

The potential of the neutral point with respect to the zero potential Vno is required to be considered in order to avoid the disparity in the applied voltage in simulating the performance of the drive.

The set of differential equations mentioned in equations (3) to(9) defines the developed model in terms of the state variables ia, ib, ic, ωr, θr and time as an independent variable. In the simulation of sensorless control, initially back emf is not available. So initial pulses are applied in a sequence for starting and after that virtual hall sensor signals are generated from line- line voltage. Figure 4 and figure 5 shows the back EMF and current waveform of each phases.

Fig.4 Back emfs in three phases ean,ebn,ecn

Fig.5 Current in three phases ia,ib,ic

Figure 6 shows the acutal speed and command speed (3000rpm). It clearly indicates that actual speed tracks the command speed within milliseconds. Figure 7 shows the actual and reference torque waveforms.Initilly it is on no load and after 1second a load torque of 0.1N-m is applied. When a load is applied, speed reduces and because of the action of PI controller it settles to the reference speed within a few milli seconds. The value of Kp and Ki are 10 and 0.1 respectively.

Fig.6 Actual Speed and reference speed waveform

Fig.7 Actual Torque and reference Torque waveform

sensorless control of bldc motor

Back emf detection is the commonly used method of sensorless control for BLDC motor. In conventional sensorless control scheme a virtual neutral point is created as shown in figure 8. The drawback of conventional back emf detection method is that it requires a motor neutral point and a phase shifting circuit independent of frequency. This neutral point induces a common mode noise into the sensorless circuit.

Fig.8 Conventional back EMF detection circuit

To avoid the above mentioned problems, line voltage is considered in the method shown in figure 9. This method can extract the commutation sequence directly with the help of a simple comparator and low pass filter. The created virtual hall sensor signals are almost same as that of hall sensor signals. But these back emf signal is not available at starting. Hence a special start up sequence must be used. Terminal voltages Va, Vb and Vc is passed through a low pass filter to avoid high frequency components. Vac, Vba and Vcb are used to generate hall sensor signals Ha, Hb and Hc respectively. An optocoupler is used at the output of comparator to isolate power and control circuit and to reshape the voltage to 5V in order to give it directly to the microcontroller.

Fig.9 Modified sensorless control circuit

hardware implimentation

The motor drive is controlled by a dsPIC utilising the pulse width modulation technique to control the speed of motor. The complete hardware has the subsystem as:

Three phase MOSFET PWM inverter

A MOSFET gate driver circuit

A dsPIC30F4011 DSP controller circuit

A Back EMF detection circuit

A standard three leg inverter is used. Controller is programmed to run the motor initially in open loop. When it attains the speed, back emf detection circuit will generate the virtual hall sensor signals. Then it will shift to running algorithm. These signals are given to dsPIC and it starts generating switching sequence according to the lookup table loaded in it. The voltage reference is set using a potentiometer and by varying that we can adjust the analog input to the dsPIC and hence the speed of motor. These PWM signals from dsPIC are fed into the driver circuit to provide isolation between the power and control side. It also strengthens the PWM signals to switch on MOSFET.

Fig.10 Block diagram of sensorless BLDC motor drive

The zero crossing points of the line-line terminal voltage is detected using the circuit shown in figure 9. The output of this circuit is directly given to microcontroller. The block diagram of hardware implementation is shown in figure 10. Microcontroller will generate the PWM signals for the inverter circuit. Duty ratio of these switching signals will vary according to the set speed reference.

Fig. 11 PWM signals of each phases

Low Speed

Medium Speed

High Speed

Fig. 12 Performance Evaluation: Signal from hall sensor, Virtual hall sensor signal from new line-line voltage method, Virtual hall sensor signal from conventional back EMF method

Figure 11 shows the PWM signals of each inverter legs, these are generated with suitable duty ratio to get the reference speed. The duty ratio of these PWM signal depends on the reference speed. A 12V BLDC motor with rated speed 3000rpm, rated current 4.4 A and rated torque of 0.12 N-m is used for hardware implementation.

The figure 12 shows the experimental results when compared with conventional method. It is clear that the signal from conventional method strongly depends on operating speed. The large commutation error is mainly caused by the multistage filters; therefore a speed dependent phase compensation algorithm is needed to get the better performance in conventional method. In conventional method, the starting algorithm can be shift to running when the motor attains a speed nearly 200 rpm. Instead that, the line-line terminal voltage detection the starting algorithm can be shift to running when the motor attains a speed 100 rpm. Also motor can run smoothly even at low speed because the magnitude of line voltage is more than the phase voltage. The main drawback of this line-line voltage detection is that it needs three separate power supplies for zero crossing detection circuit, whereas conventional method requires only one.


In this paper a BLDC motor is modeled and a sensorless control method based on line to line voltage detection is discussed. The simulation of this modified method is done in MATLAB Simulink and was implemented using digital signal controller. This modified method uses only a simple low pass filters and low cost comparators. Hence this method reduces the cost of implementation also. In this sensorless control approach, there is no need to build 300 phase shift and neutral point which are used in conventional back EMF detection method. Instead virtual hall sensor signals are developed from the zero crossing point of line -line terminal voltage. The detection of line voltage helps to get the zero crossing detection even at low speeds because of its larger magnitude compared to phase voltage. The control algorithm is implemented here with the help of a digital signal controller dsPIC 30F4011.