Freescale Semiconductor
Application Note
Document Number: AN4704
Rev. 1, 08/2013
3-phase Sensorless BLDC Motor
Control Kit with the S12 MagniV
MC9S12ZVM
by: Branislav Zigmund, Petr Staszko, and Anita Maliverney
1
Introduction
This application note describes the design of a 3-phase
brushless DC (BLDC) motor control drive using a
sensorless algorithm. The design is targeted at
automotive applications. This cost-effective solution is
based on the Freescale Semiconductor
MC9S12ZVML128 chip, which is dedicated to
automotive motor control.
The design exhibits the suitability and advantages of the
MC9S12ZVML128 microcontroller for motor control. It
serves as an example of a BLDC control design using
Freescale’s 16-bit S12 MagniV mixed-signal MCUs.
2
System Concept
The system is designed to drive a 3-phase BLDC motor.
The application meets the following system concept:
• Targets the MC9S12ZVML128 microcontroller.
© Freescale Semiconductor, Inc., 2013. All rights reserved.
Contents
1
2
3
4
5
6
7
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
System Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
BLDC Sensorless Control . . . . . . . . . . . . . . . . . . . . . . . . 2
3.1 Brief overview of brushless DC motors . . . . . . . . . . 2
3.2 Complementary/independent unipolar PWM
modulation technique . . . . . . . . . . . . . . . . . . . . . . . 6
3.3 Position estimation based on BEMF zero-crossing
detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.4 DC bus current measurement . . . . . . . . . . . . . . . . 11
3.5 States of BLDC sensorless control based on BEMF
zero-crossing detection . . . . . . . . . . . . . . . . . . . . . 12
Software implementation on the MC9S12ZVML128 . . . 13
4.1 MC9S12ZVM configuration . . . . . . . . . . . . . . . . . . 13
4.2 Software architecture. . . . . . . . . . . . . . . . . . . . . . . 18
FreeMASTER user interface . . . . . . . . . . . . . . . . . . . . . 29
Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
BLDC Sensorless Control
•
•
•
•
Runs on the hardware of the 3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV
MC9S12ZVM (for further information, see document number MTRCKTSBNZVM128). Details
about the MC9S12ZVML128 evaluation board can be found in MC9S12ZVM128MCBUG,
MC9S12ZVML128 Evaluation Board User’s Manual.
Incorporates a control technique with:
— Six-step commutation control of a 3-phase BLDC motor
— Position sensing by means of BEMF zero-crossing detection technique
— Closed-loop speed control
— Motor current limitation
— Alignment at startup
FreeMASTER software control interface (motor start/stop, speed setup)
DC bus over-voltage, under-voltage, over-current, and overload protection
3
BLDC Sensorless Control
3.1
Brief overview of brushless DC motors
The BLDC motor is a rotating electric machine with a classic 3-phase stator similar to an induction motor.
The phases mounted on the stator are connected to form a way or delta connection. The rotor has
surface-mounted permanent magnets. The motor can have more than one pole pair per phase. The pole pair
per phase defines the ratio between the electrical revolution and the mechanical revolution.
The BLDC motor is equivalent to an inverted DC brushed motor, where the magnet rotates while the
conductors remain stationary. In the DC brushed motor, the commutator and brushes reverse the current
polarity in such a way that stator and rotor magnetic fields are perpendicular. However, in the brushless
DC motor, a power transistor (which must be switched in synchronization with the rotor position) performs
the polarity reversal. This process is also known as electronic commutation.
The displacement of the magnets on the rotor creates a trapezoidal back electromotive force (BEMF) shape
when the rotor is spinning. Neglecting the higher-order harmonic terms, the BEMF in the motor phase
(ea,eb,ec) is as indicated in Figure 1. Each BEMF has a constant amplitude for 120 electrical degrees,
followed by a 60 electrical degree transition in each half-cycle. The developed torque at any instance is
given by the following equation:1
Eqn. 1
ܶ=
݁ ݅ + ݁ ݅ + ݁ ݅
߱
Where:
• el is electrical angular velocity
The ideal current waveforms in each phase (ia,ib,ic) need to be quasi-square waveforms of 120 electrical
degrees of conduction angle in each half-cycle. The conduction of current in each phase must coincide with
1.Rashid, Muhammad H., ed. Power Electronics Handbook. 2nd ed. London: Academic Press, 2007.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
2
Freescale Semiconductor
BLDC Sensorless Control
the flat part of the BEMF waveforms; this guarantees that the developed torque is constant or ripple-free
at all times. In order to align current conduction in each phase with the flat part of the BEMF, the rotor
position must be known.
Figure 1. 3-phase BEMF voltages and phase currents of a BLDC motor
The position of the rotor can be obtained by a position sensor or a sensorless algorithm. Various kinds of
position sensors are used; however, since the rotor is a permanent magnet, it is a very simple matter to
determine where the physical pole edges are using a simple, reliable, and inexpensive Hall effect sensor.1
The following techniques are commonly used to estimate rotor position in applications that rely on
sensorless control of a BLDC motor:
• BEMF zero-crossing detection method
• Flux level detection method
• Various kinds of system state observers
• Signal injection methods
From a control perspective, two logical mechanisms must be employed:
• Commutation control, where the phases are energized according to rotor position with the
quasi-square current waveforms.
• Speed/torque control, where the amplitude of the quasi-square current waveform applied to the
phases is controlled to achieve the desired speed/torque performance.
The following sections discuss the concept of the BEMF zero-crossing detection method, as well as the
methods and conditions for its correct evaluation.
1.Ibid.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
3
BLDC Sensorless Control
3.1.1
Electronic commutation control
The commutation process provides a mechanism to energize phases according to the rotor position with
the quasi-square current waveforms. Since only six discreet outputs per electrical cycle are required (as
shown in Figure 1), six semiconductor power switches are sufficient to create quasi-square current
waveforms for the phases. Six semiconductor power switches form a 3-phase power inverter, designed
using IGBT or MOSFET switches. The power for the system is provided by the DC-Bus voltage UDCB.
The semiconductor switches and diodes are modeled as ideal devices in Figure 2.
Figure 2. Power stage and motor topology
Six-step commutation is a very common method for driving a 3-phase way-connected BLDC motor. In six
step commutation control, the BLDC motor is operated in a two-phase model. Two phases are energized
while the third phase is disconnected as the space between the magnet poles passes over it and produces a
zero BEMF voltage. Selection of the two energized phases is carried out by a position sensor or a position
observer. Table 1 shows the output current waveforms for a 3-phase inverter and the switching devices that
conduct during the six switching intervals per cycle.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
4
Freescale Semiconductor
BLDC Sensorless Control
Table 1. Six step switching sequence
3.1.2
Rotor
position
Sector
number
Switch
closed
330°–30°
1
SAT
30°–90°
2
90°–150°
Phase current
A
B
C
SBB
+
–
off
SAT
SCB
+
off
–
3
SBT
SCB
off
+
–
150°–210°
4
SBT
SAB
–
+
off
210°–270°
5
SCT
SAB
–
off
+
270°–330°
6
SCT
SBB
off
–
+
Speed/torque control
Commutation ensures the proper direction of the phase current according to the rotor position of the BLDC
motor, while the motor torque/speed only depends on the amplitude of the quasi-square current waveform.
Continued control of the amplitude of the quasi-square current waveform for each phase of the motor is
ensured by hysteresis or PWM control.
PWM control is commonly used in applications where microcontrollers are employed. The duty-cycle for
the PWM modulator is obtained by the speed PI controller. The speed PI controller amplifies the error
between the required and actual speeds, and its output, appropriately scaled, is assigned to the PWM
modulator.
The actual mechanical speed can be calculated as a time derivative of the shaft position mech.
Eqn. 2
߱ =
݀߮ 1 ݀߮ 1 ȟ߮
=
ൎ
݀ݐ
ݐ݀
ȟܶ
Since the shaft travels exactly 1/6th of one electrical revolution (2 in radians) between two commutations,
the above equation can be rewritten to the following form:
Eqn. 3
߱
=
360°
1 ȟ߮ 1 6
ൎ
=
ȟܶ
ܶ ெ
360°
1
(ܶ ଷଷ° ՜ ଷ°) + ܶ(ଷ° ՜ ଽ°) + ܶ(ଽ° ՜ ଵହ°) + ܶ(ଵହ° ՜ ଶଵ°) + ܶ(ଶଵ° ՜ ଶ°) + ܶ(ଶ° ՜ ଷଷ°)
=
360°
σୀଵ ܶெ
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
5
BLDC Sensorless Control
Where:
• p is the number of pole-pairs
• TCM is the time between two following commutations
• TCMn is the time between commutations in sector n = 1, 2, 3, 4, 5, 6
• el is the electrical position
3.2
Complementary/independent unipolar PWM modulation
technique
There are different methodologies for powering and switching the phases. The unipolar PWM control
technique combines commutation control and torque control. While the state of the switches is determined
by commutation control, the torque is controlled by the applied duty cycle. An application with BLDC
control where the unipolar PWM control technique is employed benefits from a reduction in the MOSFET
switching losses and an improvement in the system’s EMC robustness.
The unipolar PWM control means that the motor phase sees only the positive polarity of the voltage. To
achieve the unipolar PWM pattern, one phase is in complementary PWM mode while the second phase is
grounded and the third phase stays unpowered, as shown in Figure 3. This PWM pattern can be seen every
60° electrical degrees, and they differ only in phase order. The phase order is determined according to the
shaft position by commutation control.
Figure 3. Complementary/independent unipolar PWM switching
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
6
Freescale Semiconductor
BLDC Sensorless Control
For example, in the first cycle Phase A is powered by the complementary PWM signal while the bottom
transistor of Phase B is grounded and Phase C is unpowered. After the commutation event at 90° electrical
degrees, Phase A is still powered by the complementary PWM signal, Phase B is unpowered, and Phase C
becomes grounded instead.
The control described in this application note is based on the complementary/independent unipolar PWM
modulation technique.
The following section explains sensorless position estimation by means of BEMF zero-crossing detection
for commutation control purposes.
3.3
Position estimation based on BEMF zero-crossing detection
Figure 4 shows ideal BEMF waveforms and depicts a commutation event occurring at a position of 30
electrical degrees after the point where a BEMF zero-crossing arises. The BEMF zero-crossing happens
at a position of 30 electrical degrees after the point of the last commutation event. Let us assume that the
motor is spinning at a constant velocity; in this case, the motor needs the same amount of time to travel
from the position of the last commutation event to a BEMF zero-crossing and from the BEMF
zero-crossing to the following commutation event. In the time domain, a BEMF zero-crossing is right in
the middle of two commutation events. Therefore, the BEMF zero-crossing event, with help of a timer, can
simply be used to estimate the right commutation point as well as the velocity of the rotor.
Figure 4. Zero-crossing detection and commutation diagram
3.3.1
BEMF zero-crossing principle
To explain and simulate the idea of BEMF sensing techniques, this document provides a simplified
mathematical model based on the basic circuit topology (see Figure 4). The goal of the mathematical
model is to identify dependencies between the measurable motor waveforms and a BEMF zero-crossing.
The BEMF zero-crossing in turn helps to identify the commutation event.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
7
BLDC Sensorless Control
The mathematical model is based on the fact that only two phases of a motor are energized and the third
is disconnected. The natural voltage level of the whole model is referenced to half of the DC bus voltage,
which simplifies the mathematical expressions. The mathematical model assumes that the motor phases
are symmetrical (see Figure 4).
Eqn. 4
݀݅
െ ݁
ܷ ݁ + ݁
݀ݐ
െ
ൢ ሳልልልሰ ݑே =
݀݅
ೌ ୀ ್
2
2
ݑே = ܴ݅ + ܮ
െ ݁
݀ݐ
ݑே = ܷ െ ܴ݅ െ ܮ
For a symmetrical 3-phase motor, the sum of all BEMF voltages is zero; therefore:
Eqn. 5
ec + eb + ec = 0 ec = –(eb + ea)
The unpowered phase has the following voltage equation, since there is no current flowing:
Eqn. 6
uN = uC – ec
By substituting Equation 5 and Equation 6 into Equation 4, the phase voltage on the unpowered phase can
be derived as:
Eqn. 7
ݑ =
ݑ 3
+ ݁
2
2
At the time of the BEMF zero-crossing, the BEMF voltage (ec in this case) is zero as the name implies.
Therefore by measuring voltage at the unpowered phase (ec) and comparing it to half of the DC-Bus
voltage ( UDCB ), the BEMF zero-crossing can be accurately identified.
2
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
8
Freescale Semiconductor
BLDC Sensorless Control
3.3.2
BEMF zero-crossing event detection, phase current measurement,
and complementary/independent unipolar PWM switching
Figure 5. BEMF zero-crossing events, commutations, and their relationship
to complementary/independent unipolar PWM switching
The exact position of the rotor can be sensed by measuring the BEMF voltage induced by the rotating
permanent magnet in the unpowered phase.
In Figure 5, the green windows mark the time periods in which the respective phase is unpowered. The
voltage measured in this time window is the BEMF voltage. At the BEMF zero-crossing event, the
permanent magnet is right in front of a coil and the rotor field is positioned 90° versus the stator field. This
event happens in the middle of a commutation period and is marked as the black circles in the green BEMF
window. At this time, the phase voltage is equal to half of the DC bus voltage as described in Section 3.3.1,
“BEMF zero-crossing principle.” In the case of a constant shaft velocity, the period between two following
zero-crossing events is equal to the commutation period.
Figure 6 zooms in closer to one of the PWM cycles. At the top of the figure is the PWM pattern where
Phase A is controlled by PWM and Phase C is grounded for the entire PWM period. During the PWM On
cycle, the top switch of Phase A is turned on and the bottom switch of Phase C is grounded. Current flows
from the DC bus into Phase A, and back through Phase C and the DC bus shunt resistor. In this cycle, the
waypoint of the motor shows a voltage level of UDCB
. The BEMF voltage in the unpowered phase changes
2
UDCB
relatively to 2 in the positive and negative directions, which means that the zero-crossing is detectable
when the phase voltage on the unpowered phase is equal to UDCB
. Also, the phase current is measurable on
2
the DC bus shunt.
During the Off cycle of the PWM period, both the Phase A and Phase C bottom switches are on. Therefore,
phase current circulates through Phase A, Phase B, and two bottom switches back. During this cycle, the
phase current is unable to reach the DC bus shunt resistor and the phase current cannot be measured. The
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
9
BLDC Sensorless Control
star point of the motor as well is connected to ground and the zero-crossing cannot precisely be measured
in that cycle.
Figure 6. BEMF zero-crossing detection with complementary/independent unipolar PWM switching
Following on from the discussion above, phase current and BEMF voltage measurements must be
performed in the active phase of the PWM cycle.
Moreover, right after the commutation event there is a commutation transient pulse measured on the phase
voltages, which is produced by the current recirculation when the flyback diodes are conducting. The
duration of this transient is dependent on the motor inductance, phase current, and speed. The BEMF
voltage should not be measured during this time. The software must take care of this interval.
3.3.3
BEMF voltage measurement
As we learned earlier, the BEMF voltage can only be measured during the active phase of the PWM.
Importantly, this is measured towards the end of the active cycle due to switching noises. In Figure 7, the
green marked area shows the window in which the BEMF should be measured.
SAtop
SAbot
SCtop
SCbot
motor
phase
resonance
ADC
sample
point
switching
noise
spikes
PWM to ADC
• Back-EMF voltage
measure window
• Time of Back-EMF
voltage sample point
is used to calculate
exact time of the
zero-cross
TPWM
Back-EMF
zero-cross
BEMFT
DCBUS / 2
BEMFT-1
TZC TADC
Measured Back-EMF
voltage
TZC
BEMFT DCBUS / 2
TPWM
BEMFT BEMFT 1
TADC
delay by PTU
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
10
Freescale Semiconductor
BLDC Sensorless Control
Figure 7. BEMF measurement
Figure 8. Precise BEMF zero-crossing
identification
It should be noted that, depending on the motor and power stage parameters, the amplitude, period, and
damping of the voltage ringing vary. As a result, it is recommended that you measure BEMF voltage close
to the end of the window. The time of this sample point also needs to be stored, as it is used to enhance
zero-crossing detection.
If we zoom out again and look at the BEMF voltage cycles (see Figure 8), it can be seen that the crossing
of the BEMF voltage and UDCB
level can take place wherever between two following BEMF voltage
2
measurements. For accurate position estimation, an exact zero-crossing point has to be identified. This
exact zero-crossing point identification is done by an approximation based on the interpolation of two
following BEMF measurements.
Assuming that the shaft is not accelerating, actual BEMF voltage was measured at time TADC with the
voltage level of eT, and the previous measurement was taken at the time of TADC – TPWM with the voltage
level of eT–1, then the equation to calculate the exact time of the zero-crossing event could be derived as
follows:
Eqn. 8
்݁ െ ்݁ିଵ ்݁ െ ܷ /2
்݁ െ ܷ /2
=
֜ ܶ = ܶ െ
ܶ
ܶௐெ
ܶ െ ܶ
்݁ െ ்݁ିଵ ௐெ
This formula is calculated in the commutation period when two following comparisons of the BEMF
voltage to half of the DC bus have the opposite signs.
In order to enhance the accuracy of the zero-crossing event even further, the DC bus voltage and BEMF
voltage need to be measured simultaneously.
3.4
DC bus current measurement
Figure 9. DC bus current measurement
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
11
BLDC Sensorless Control
As mentioned in the previous section, the DC bus current has to be measured in the active cycle of the
PWM period due to the fact that the DC bus current equals the phase current only in the active cycle, as
illustrated in Figure 9.
During the active cycle of the PWM period, the phase current is rising. The slope of the rising current is
defined by the motor phase coil inductance; the lower the phase inductance, the larger the gradient of the
rising current. It is important to measure the current in the middle of the PWM active cycle to obtain the
average phase current.
3.5
States of BLDC sensorless control based on BEMF zero-crossing
detection
In order to start and run the BLDC motor, the control algorithm has to go through the following states:
• Alignment (initial position setting)
• Startup (forced commutation or open-loop mode)
• Run (sensorless running with BEMF acquisition and zero-crossing detection)
3.5.1
Alignment
As mentioned previously, the main task for sensorless control of a BLDC motor is position estimation.
Before starting the motor, however, the rotor position is not known. The aim of the alignment state is to
align the rotor to a known position. This known position enables starting the rotation of the shaft in the
desired direction and generating the maximal torque during startup. During the alignment state, all three
phases are powered in order to get the best performance behavior in either direction of shaft rotation.
Phase A and Phase B are connected to the positive DC bus voltage and Phase C is grounded. The alignment
time depends on the mechanical constant of the motor, including load, and also on the applied motor
current. In this state, the motor current (torque) is controlled by the PI controller.
3.5.2
Startup
In the startup state, motor commutation is controlled in an open-loop mode without any rotor position
feedback. The commutation period is controlled with an open-loop starting curve. The open-loop start is
required only until the shaft speed is high enough (approximately 5% of nominal motor speed) to produce
an identifiable BEMF voltage.
3.5.3
Run
The block diagram of the run state is represented by Figure 10 and includes the BEMF acquisition with
zero-crossing detection in order to control the commutations. The motor speed is estimated based on
zero-crossing time periods. The difference between the demanded and estimated speeds is fed to the speed
PI controller. The output of the speed PI controller is proportional to the voltage to be applied to the BLDC
motor. The motor current is measured and filtered during the BEMF zero-crossing event and used as
feedback into the current controller. The output of the current PI controller limits the output of the speed
PI controller. The limitation of the speed PI controller output protects the motor current from exceeding
the maximal allowed motor current.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
12
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
Figure 10. Speed control with current limitation
4
Software implementation on the MC9S12ZVML128
4.1
MC9S12ZVM configuration
The BLDC sensorless application framework is designed to meet the following technical specification:
• Targeted at the MC9S12ZVML128 evaluation board (for further information, see
MC9S12ZVM128MCBUG, MC9S12ZVML128 Evaluation Board User’s Manual and
MTRCKTSBNZVM128QSG, Quick Start Guide for 3-Phase Sensorless BLDC Kit with MagniV
MC9S12ZVML128 MCUs)
• PWM output frequency = 20 kHz
• Current loop sampling period 1 ms
• Speed loop sampling period 1 ms
• Linear approximation of the ADC samples for accurate BEMF zero-crossing detection. Internal
dividers and multiplexer of the MC9S12ZVML128 are used for phase voltage measurement.
• DC bus voltage measurement at the HD pin by the ADC. The internal voltage dividers of the
MC9S12ZVML128 are employed.
• Internal operational amplifier AMP0 is used for DC bus current measurement.
The MC9S12ZVML128 device includes modules such as the Pulse Width Modulator with Fault Protection
(PMF), a Programmable Trigger Unit (PTU), an Analog-to-Digital Converter (ADC), a Timer module
(TIM), and a Gate Drive Unit (GDU) suitable for control applications, in particular, motor control
applications. These modules are directly interconnected and can be configured to meet various motor
control application requirements. Figure 11 shows module interconnection for a typical BLDC sensorless
application. The modules are described below and a detailed description can be found in
MC9S12ZVMRMV1, MC9S12ZVM-Family Reference Manual.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
13
Software implementation on the MC9S12ZVML128
4.1.1
Module interconnection
The modules involved in output actuation, data acquisition, and the synchronization of actuation and
acquisition form the so-called control loop. This control loop consists of the PMF, GDU, ADC, and PTU
modules. The control loop is very flexible in operation and can support static, dynamic, or asynchronous
timing.
The PTU and ADC operate using lists stored in memory. These lists define the trigger points for the PTU,
commands for the ADC, and results from the ADC.
Figure 11. Block diagram of sensorless BDLC configuration
Each control loop cycle is started by a PMF reload event. The PMF reload event restarts the PTU timebase.
If the PTU is enabled, the reload is immediately passed on as a ptu_reload event to the ADC and GDU
modules.
The PMF generates the reload event at the required PWM reload frequency. The PMF reload event causes
the PTU timebase to restart, to acquire the first trigger time from the list, and to generate the ptu_reload
signal for the ADCx to start loading the ADC conversion command from the Command Sequence List
(CSL).
When the trigger time is encountered, the corresponding PTU trigger generates the trigger_x signal for the
associated ADC. For simultaneous sampling, the PTU generates two simultaneous trigger_x signals, one
for each ADC. At the trigger_x signal assertion, the ADC starts the first conversion of the next conversion
sequence in the CSL (the first A/D command is already downloaded). For further information, see
MC9S12ZVMRMV1, MC9S12ZVM-Family Reference Manual.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
14
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
In this way, the PTU module serves as a delay block that can schedule several acquisitions of state variables
relative to the start of the PWM period and within one PWM period.
4.1.2
Module involvement in BLDC sensorless software control loop
This section discusses measurements on the MC9S12ZVM128L and the internal hardware features to
support the measurement of BEMF voltages and phase current.
Each commutation event first gets triggered by a timer interrupt. This interrupt can be defined as a PWM
reload signal. The PWM reload signal itself also triggers the reload of the trigger list in the PTU module
and restarts the PTU counter. When the PTU counter reaches the predefined value in the trigger list (T1
and T2), the PTU triggers an ADC measurement. At the time T1, measurement of the DC bus current is
triggered. Two other simultaneous measurements are triggered at time T2, one for BEMF voltage and one
for the DC bus voltage. The ADC conversion results are automatically stored into a predefined queue in
memory. The principles of phase voltage, DC bus voltage, and DC bus current measurement are described
in Section 3.3.3, “BEMF voltage measurement” and Section 3.4, “DC bus current measurement.”
The CPU is triggered by the ADC conversion complete interrupt service routine. Based on the stored ADC
values, it calculates the zero-crossing event. Based on the zero-crossing period, the time of the next
commutation event is calculated.
Figure 12. Module involvement in BLDC sensorless software control loop
The TIM, PTU, GDU, and ADC peripherals are based on the bus clock. To achieve a higher resolution of
PWM, the PMF module is supplied by a core clock. The core clock is double that of the bus clock. The
clock module uses a 4 MHz external crystal and is configured to generate a 12.5 MHz bus clock and
25 MHz core clock.
4.1.3
PMF
The Pulse Width Modulator with Fault Protection (PMF) module is configured to generate an edge-aligned
(PMFCFG0_EDGEx = 1) PWM with a frequency of 20 kHz (PMFMODA = 1250). In order to protect the
MOSFET devices in the same leg of the inverter, dead time is set to approximately 0.5 µs
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
15
Software implementation on the MC9S12ZVML128
(PMFDTMA = 13). PWM generator A runs as the master and generates the Reload signal as a
synchronization signal for the other submodules (PMFCFG2_REV[0:1] = 1). The Reload signal is
generated at every PWM opportunity (PMFFQCA = 0). Pair A, Pair B, and Pair C PWMs are synchronized
to PWM generator A (PMFCFG0_MTG = 0). All three pairs use value register zero
(PMFCFG3_VLMODE = 1 – the value written into value register zero is also written into value registers
1 to 5) to set the duty-cycle. The PWM generator is restarted with every commutation event
(PMFENCA_RSTRTA = 1).
To achieve a unipolar PWM pattern, one phase is in complementary PWM mode while the second phase
is grounded by the software control mode (PMFOUTC, PMFOUTB) and the third phase stays unpowered
by masking the output (PMFCFG2[MSK5:MSK0]) as shown in Figure 3. The registers (PMFOUTC,
PMFOUTB, PMFCFG2[MSK5:MSK0]) used to control the unipolar PWM pattern are in double-buffered
mode (PMFCFG1_ENCE = 1). The double-buffered signals are swapped at a commutation event. A PWM
pulse width PMFVAL registers are also double-buffered and are swapped when GLDOK is set and the
PWM Reload signal occurs. The GLDOK is an external signal generated by the PTU module. The GLDOK
is enabled at the PWM module (PMFENCA_GLDOKA = 1)
For example, in the first cycle, Phase A is powered by the complementary PWM signal, while the bottom
transistor of Phase B is grounded (PMFOUTC = 0x0C, PMFOUTB = 0x2A) and Phase C is unpowered
(PMFCFG2[MSK5:MSK0] = 0x34). After the commutation event at 90° electrical degrees, Phase A is still
powered by the complementary PWM signal, Phase B is now unpowered
(PMFCFG2[MSK5:MSK0] = 0x1C) and Phase C becomes grounded instead (PMFOUTC = 0x30,
PMFOUTB = 0x2A).
4.1.4
PTU
The Programmable Trigger Unit (PTU) is intended to completely avoid CPU involvement in the time
acquisitions of state variables during the control cycle.
The PTU module consists of two trigger generators (TG). For each TG, a separate enable bit is available
so that both TGs can be enabled independently. TG0 is connected to ADC0, and TG1 is connected to
ADC1. The trigger generation of the PTU module is synchronized to the incoming reload event. This
reload event resets and restarts the internal time base counter and makes sure that the first trigger value
from the actual trigger list is loaded. Furthermore, the corresponding ADC is informed that a new control
cycle has started.
If the counter value matches the current trigger value, then a trigger event is generated. In this way, the
reload event is delayed by the number of bus clock cycles defined by the current trigger value. All
acquisition time values are stored inside the global memory map; that is, inside the system memory as a
three dimensional array of integers (PTUTriggerEventList[][][]). The exact location of the acquisition time
values (PTUTriggerEventList[][][]) in the system memory is given by the linker command file and linked
to the PTU module during the initialization phase.
PTUPTRL = (uint8_t)((long)PTUTriggerEventList);
PTUPTRM = (uint8_t)(((long)PTUTriggerEventList) >> 8);
PTUPTRH = (uint8_t)(((long)PTUTriggerEventList) >> 16);
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
16
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
Each TG uses only one list to load the trigger values from the memory. The pointers for the primary
(TG0L0IDX/ TG1L0IDX) and alternate (TG0L1IDX/ TG1L1IDX) lists are equal.
TG0L1IDX = (uint8_t)(((long)&PTUTriggerEventList[0][0][0] - (long)PTUTriggerEventList) >> 1);
TG1L0IDX = (uint8_t)(((long)&PTUTriggerEventList[1][0][0] - (long)PTUTriggerEventList) >> 1);
TG1L1IDX = (uint8_t)(((long)&PTUTriggerEventList[1][0][0] - (long)PTUTriggerEventList) >> 1);
The TG uses only one physical list of trigger events, even if the TG logic is switching between both
pointers. At the end of the trigger list, the interrupt for TG1 is enabled (PTUIEL_TG1DIE = 1). This
interrupt is used to store the time of the ADC measurement of the respective phase voltage. The PTU
module generates the LDOK signal used to inform other modules that the double buffered registers were
updated by software.
4.1.5
TIM
The Timer Module (TIM) is a basic scalable timer that consists of a 16-bit, software-programmable
counter driven by a flexible programmable prescaler. The BLDC sensorless algorithm employs two timer
channels in output compare mode (TIM0TIOS_IOS0 = 1; TIM0TIOS_IOS3 = 1).
Timer Channel 0 serves to identify the commutation event. The output compare signal is internally routed
to the PMF module as an async_event in order to perform commutation of the PWM pairs. When the timer
counter reaches the value in the channel registers of an output compare channel, the timer toggles
(TIM0TCTL2_OL0 = 1; TIM0TCTL2_OM0 = 0) the channel output. This channel also schedules the
interrupt (TIM0TIE_C0I = 1) where some of the application flags are controlled.
Timer Channel 3 is employed to control the torque of the motor by a software task. Timer Channel 3
schedules periodic interrupts (TIM0TIE_C3I = 1) with a period of 1 ms (TIM0TC3 = 781).
The timer’s prescaler is equal to 4 (TIM0TSCR2_PR = 4) and all of the timers output compare pins are
disconnected (TIM0OCPD = 0xff).
4.1.6
GDU
The Gate Drive Unit (GDU) is a Field Effect Transistor (FET) pre-driver designed for 3-phase motor
control applications. The following GDU features are used in BLDC sensorless control:
• Charge pump: used to maintain the high-side driver gate source voltage VGS when PWM is
running at a 100% duty cycle. The clock for the charge pump is set to be fbus
32
(GDUCLK2_GCPCD = 2)
• Desaturation error: integrates three desaturation comparators for the low-side FET pre-drivers and
three desaturation comparators for the high-side FET pre-drivers. The desaturation level is set to
be 1.35 V (GDUDSLVL = 0x77) for both low-side and high-side FETs. A blanking time during the
FET transients needs to be employed. The blanking time is set to be approximately 8 µs (GDUCTR
= 0x13).
• Phase muxing: used to select the phase voltage (GDUPHMUX) to be routed internally to ADC1
Channel 2.
• Current sense amplifier: Internal current sense amplifier 0 (GDUE_GCSE0 = 1) is used to measure
motor phase current. The output of the current sense amplifier 0 is routed internally to ADC0
Channel 0.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
17
Software implementation on the MC9S12ZVML128
4.1.7
ADC
The MC9S12ZVML128 uses two independent Analog-to-Digital Converters (ADC). Both ADCs are
n-channel multiplexed input successive approximation analog-to-digital converters. The List Based
Architecture (LBA) provides a flexible conversion sequence definition, as well as flexible oversampling.
Both ADC conversion command lists are stored inside the global memory map; that is, inside the system
memory as two dimensional arrays of bytes (ADC0CommandList[][], ADC1CommandList[][]). The exact
location of the ADC conversion commands in the system memory is given by the linker command file and
linked to the respective ADC module during the initialization phase. The same strategy is used for the
ADC results. The conversion results are stored in an array of shorts (ADC0ResultList[],
ADC1ResultList[]) located in system memory.
// ADC0 Command Base Pointer
ADC0CBP_0 = (uint8_t)(((long)ADC0CommandList) >> 16);
ADC0CBP_1 = (uint8_t)(((long)ADC0CommandList) >> 8);
ADC0CBP_2 = (uint8_t)((long)ADC0CommandList);
// ADC0 Result Base Pointer
ADC0RBP_0 = (uint8_t)(((long)ADC0ResultList) >> 16);
ADC0RBP_1 = (uint8_t)(((long)ADC0ResultList) >> 8);
ADC0RBP_2 = (uint8_t)((long)ADC0ResultList);
// ADC1 Command Base Pointer
ADC1CBP_0 = (uint8_t)(((long)ADC1CommandList) >> 16);
ADC1CBP_1 = (uint8_t)(((long)ADC1CommandList) >> 8));
ADC1CBP_2 = (uint8_t)(long)ADC1CommandList);
// ADC1 Result Base Pointer
ADC1RBP_0 = (uint8_t)(((long)ADC1ResultList) >> 16);
ADC1RBP_1 = (uint8_t)(((long)ADC1ResultList) >> 8);
ADC1RBP_2 = (uint8_t)((long)ADC1ResultList);
The ADC conversion clocks are set to be 6.25 MHz (ADC0TIM = 0; ADC1TIM = 0). The results are stored
in memory as 12-bit (ADC0FMT_SRES = 4; ADC1FMT_SRES = 4) left-justified data (ADC0FMT_DJM
= 0; ADC1FMT_DJM = 0).
Conversion flow of both ADCs is controlled by internal signals (generated by the PTU) and by the data
bus (ADC0CTL_0_ACC_CFG = 3; ADC1CTL_0_ACC_CFG = 3). The results are stored in system
memory even if commutation occurs when conversion is ongoing (ADC0CTL_0_STR_SEQA = 1;
ADC1CTL_0_STR_SEQA = 1).
ADC0 schedules the end-of-list interrupt (ADC0CONIE_1_EOL_IE = 1) to calculate the BEMF
zero-crossing algorithm.
The BLDC sensorless algorithm uses ADC0 to measure the motor phase current and DC bus voltage. The
ADC1 is used to measure the respective motor phase voltage.
4.2
Software architecture
This section describes the software design of the BLDC sensorless algorithm based on zero-crossing.
Figure 13 shows the conceptual system block diagram.
The application is optimized for MC9S12ZVM motor control peripherals to achieve the least possible core
involvement in state variable acquisition and output action application. The motor control peripherals
(PMF, PTU, ADC, TIM, and GDU modules) are internally linked/set up together to work independently
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
18
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
from the core and to achieve deterministic sampling of analog quantities and precise commutation of the
stator field. The software part of the application consists of different blocks which are described below.
The entire application behavior is controlled from a PC with the FreeMASTER tool. For further
information, see MTRCKTSBNZVM128QSG, Quick Start Guide for 3-Phase Sensorless BLDC Kit with
MagniV MC9S12ZVML128 MCUs.
3KDVH,QYHUWHU
0&6=90/
(YDOXDWLRQ%RDUG
3RZHU
/LQH
3KDVH
%/'&
0RWRU
0HASE6OLTAGE
!$#-ODULE
$'&
7ULJJHUV
!$#-ODULE
%(0)
9ROWDJH
6XSHULRU
6\VWHP
*'8
$'&
7ULJJHUV
'&%XV
9ROWDJH
=HUR&URVVLQJ3HULRG
DQG3RVLWLRQ5HFRJQLWLRQ
2Q&KLS
'HEXJJLQJ
)UHHPDVWHU
3K&RPS
0X[
&XUUHQW$PS
$#"US#URRENT
$#"US6OLTAGE
%'0
/,1
0RGXOH
7
&RPPXWDWLRQ
&RQWURO
5HTXLUHG'&%XV &XUUHQW/LPLW $FWXDO
6SHHG
5HTXLUHG
6SHHG
30)
6SHHG
&XUUHQW
3,&RQWUROOHU
'XW\
&\FOH
378
7ULJJHU3RLQWV
/LPLWDWLRQV
3,&RQWUROOHU
6=90
Figure 13. System block diagram
The block diagram shown in Figure 13 is the overview of the control blocks. The top box shows the power
stage; the lower parts in the figure are the functions implemented in the S12ZVM. The power stage uses a
one-shunt based system to measure the phase current.
The light green blocks are implemented in hardware on the S12ZVM device; the light blue blocks are
implemented in software and benefit from the Motor Control Library (MCLib) (for further information,
see MC9S12ZVMMCLUG, Math and Motor Control Library for MC9S12ZVML128 User’s Guide). The
inputs of the control loop are the measured voltages and current on the power stage, in particular the phase
voltages, the DC bus current, and DC bus voltage.
The DC bus current is amplified by the current sense amplifier, which is part of the Gate Drive Unit
(GDU), and then routed together with the DC bus voltage to one of the two ADCs for measurement
acquisition. The phase voltages are multiplexed by the GDU and then used by the second ADC.
From a control perspective, the block diagram is divided into two logical parts:
• Commutation control, where the phase voltages and DC bus voltage are used to calculate the actual
position of the shaft. According to the identified position, the next commutation event can be
prepared.
• Speed/torque control, where the required shaft velocity is compared to the actual measured speed
and regulated by the PI controller. The output of the speed PI controller is the duty-cycle. The
duty-cycle is limited by the current PI controller and assigned to the PWM.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
19
Software implementation on the MC9S12ZVML128
The application is controlled by FreeMASTER, a real-time debugging tool. FreeMASTER communicates
with the S12ZVM device by means of BDM or an SCI peripheral.
4.2.1
Application flow
speed
Alignment
Desired
Speed
Open Loop Start
+Vp (PWM)
Starting
Lb
Vector
90°
La
+Vp
(PWM)
Start speed
in sensorless
close loop
Lc
GND
Alignment
Vector
Run
Open Loop:
Commutation time
calculated base on
acceleration equation
Closed Loop:
Commutation time calculated
base on Back-EMF zero-cross
Required speed
Real speed
time
AppInit()
AppAlignment()
AppStart()
AppRun()
AppStop()
AppStopToAlignment()
AppAlignmentToStart()
AppStartToRun()
called in main()
(state machine)
transient functions
Figure 14. Application state flow
Figure 14 explains the different application states. The figure consists of two interconnected parts:
• The speed over time characteristic
• The blocks in the lower part of the picture, which show the states of the application and the
transitions between respective states
The application software has three main states: the alignment state, the open-loop start state, and the run
state. In the run state, the BLDC motor is fully controlled in closed-loop sensorless mode.
After the initialization of the peripheral modules has completed, the software enters the alignment state.
In alignment, the rotor position is stabilized into a known position in order to create the same startup torque
in both directions of rotation. This is achieved by applying a PWM signal to Phase A and Phase B. The
duty cycle is calculated by the alignment PI controller. Phase C is assigned with a duty cycle equal to zero;
that is, Phase C is connected to the negative pole of the DC bus. The value of the duty cycle on Phase A
and Phase B depends on the motor inertia and load applied on the shaft. Such a technique aligns the shaft
with Phase C, which is perpendicular to the two flux vectors generated by the stator windings, and
therefore ensures the same startup torque in both directions of rotation. The duration of the alignment state
depends on the motor’s electrical and mechanical constants, the applied current (meaning duty cycle), and
the mechanical load.
When the alignment timeout expires, the application software moves to the open-loop start state. At a very
low shaft velocity, the BEMF voltage is too low to reliably detect the zero-crossing. Therefore, the motor
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
20
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
has to be controlled in an open-loop mode for a certain time period. The very first vector generated by the
stator windings needs to be set to a position 90° relative to the position of the flux vector generated by
magnets mounted on the rotor. The alignment and first start-up vector are shown in Figure 14. The duration
of the open-loop start state is defined by the number of open-loop commutations. The number of open-loop
commutations depends on the mechanical time constant of the motor, including load, and also on the
applied motor current. The shaft velocity after an open-loop startup is approximately 5% of nominal
velocity. At a velocity approximately 5% of nominal velocity, the BEMF voltage is high enough to reliably
detect the zero-crossing.
After a defined number of commutation cycles, the state changes from the open-loop start state to the run
state. From here on, the commutation process based on the BEMF zero-crossing measurement takes place,
and the control enters the closed-loop mode.
4.2.2
Application timing and interrupts
Figure 15. Application timing and interrupts
Figure 15 shows the application timing and the associated interrupts used for the commutation,
zero-crossing and speed control. The grey boxes show the executed interrupt routines versus the phase
voltage measurement.
The top row shows the interrupt that is activated when all the triggers in the PTU have been executed,
which happens the moment the BEMF measurement is triggered by the Trigger Unit at the end of each
active PWM cycle. In this interrupt, the timer value as a BEMF measurement reference point is saved.
The second row in the figure shows the interrupt that is activated when all the ATD measurements within
a PWM cycle have been executed. At that time, all the necessary values for the zero-crossing detection
have been stored automatically into memory by the ATD converter and the detection of the zero-crossing
event and calculation of the zero-crossing period can then take place. After the zero-crossing detection, the
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
21
Software implementation on the MC9S12ZVML128
multiplexer for the phase voltage measurement is switched to the phase on which the BEMF will be sensed
in the following commutation period.
The third row shows the interrupt that is the actual trigger for the commutation event, triggered by a timer
channel interrupt. The time between each call is dependent on the actual speed of the motor.
The last row shows the interrupt routine used for the speed control loop. This is handled in a timer channel
interrupt which is set every millisecond. The actual speed information, together with the variable of the
required speed, is an input value to the speed controller.
4.2.3
4.2.3.1
Zero-crossing detection processing
Getting the time of the BEMF measurement
T1
PTU triggers
T2
SAtop
Measured
Back-EMF
voltage
ADC 1
trigger 1
Measured
DC Bus
voltage
ADC 0
trigger 1
Measured
DC BUS
current
Interrupt
PTU1 Trigger
Done
Interrupt
ADC0 Done
ADC 0&1
conversion time
ADC 0
trigger 2
Figure 16. Measurement of DC bus current, voltage, and BEMF voltage
Interrupt
PTU Trigger1 Done
Save old time of Back-EMF voltage
measurement
INTERRUPT void PTUTrigger1Done_ISR(void)
{
timeOldBEMF = timeBEMF;
timeBEMF = TIM0TCNT;
PTUIFL = PTUIFL_TG1DIF_MASK;// Clear flag
Save new time of Back-EMF voltage
measurement
}
Exit
Interrupt
Figure 17. Obtaining the time of the BEMF measurement in the PTU Trigger Done ISR
The PTU Trigger Done interrupt service routine (ISR) is activated when all triggers in the PTU trigger list
have been scheduled. The PTU, PTU trigger list, and ADC command lists are configured such that the PTU
Trigger Done interrupt service routine is executed when the BEMF measurement takes place (Figure 16).
In the ISR, the time of the BEMF measurement taken during the previous cycle is backed up and the
counter value of the free running timer is stored, as shown in Figure 17. Both timestamps are needed for
the calculation of the zero-crossing point.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
22
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
4.2.3.2
State variable acquisition and zero-crossing detection processing
For state variable acquisition and zero-crossing detection processing, the ADC0 Done interrupt is used.
The interrupt service routine is executed when the ADC0 and ADC1 reach the end of the ADC command
list; measurements of the phase voltage and the DC bus voltage are the last commands in the ADC1 and
ADC0 command lists, respectively. The phase voltage and the DC bus voltage are measured
simultaneously on the ADC1 and ADC0. The flowchart and interrupt routine in Figure 18 describe the
steps of processing the measurements.
Interrupt
ADC0 Done
Save ADC results of
DC Bus Current / Voltage
Phase voltage measurements
INTERRUPT void ADC0done_ISR(void)
{
DCBusCurrent
= ADC0ResultList[0] - DCBusCurrentOffset;
DCBusVoltageHalf = MLIB_Mul(ADC0ResultList[1]>>1,13653,F16);
phaseVoltage = ADC1ResultList[0]>>1;
bemfVoltage = phaseVoltage - DCBusVoltageHalf;
if (driveStatus.bit.AfterCMT == 1) {
if ((timeBEMF - timeCommutation) > timeZCToff){
driveStatus.bit.AfterCMT = 0;
}
}
Scale phase and DC Bus voltages
according to their resistor divider ratio
Calculate Back-EMF voltage
…
Check TOFF period
Figure 18. Processing measurements in the ADC Done ISR
Before the ADC0 Done ISR routine is executed, the ADC stores the results into predefined memory space
(also known as the ADCx result list) by use of the DMA. The DC bus current and the DC bus voltage
measurements are stored in the ADC0 result list and the phase voltage is stored in the ADC1 result list.
In the ADC0 Done ISR routine, the ADC results are scaled according to the hardware setup and saved into
the appropriate variables. The BEMF voltage is calculated as the difference between the phase voltage and
UDCB
. The BEMF voltage value is the signed number.
2
The software checks whether the commutation transient time has already passed (see Section 3.3.2,
“BEMF zero-crossing event detection, phase current measurement, and complementary/independent
unipolar PWM switching”). Commutation Transient time is called Toff (timeZCToff) in the software code
implementation (Figure 19).
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
23
Software implementation on the MC9S12ZVML128
…
if ((driveStatus.bit.AfterCMT == 0) &&
(driveStatus.bit.NewZC == 0) &&
(driveStatus.bit.Sensorless == 1)){
-
Is
TOFF period elapsed?
zero-cross in current
commutation period?
- Close-loop mode?
ZCdetectionAdc[ActualCmtSector]();
no
bemfVoltageOld = bemfVoltage;
driveStatus.bit.AdcSaved = 1;
yes
if (driveStatus.bit.NewZC) {
GDUPHMUX_GPHMX = BemfPhase[NextCmtSector];
PTUC_PTULDOK = 1;
DCBusCurrentZC = DCBusCurrent;
DCBusCurrentFiltered = (tFrac16)((long)
((long)DCBusCurrentFiltered +
(long)DCBusCurrentFiltered +
(long)DCBusCurrentFiltered +
(long)DCBusCurrentZC) >> 2);
}
Look for Zero-cross
Save actual Back-EMF voltage value
no
zero-cross found?
yes
Switch phase for Back-EMF voltage
sensing for the following commutation
period
}
Arm PTU for next reload
if (driveStatus.bit.DisableAdc == 1){
driveStatus.bit.DisableAdc = 0;
driveStatus.bit.NewZC = 0;
}
Enable ADC measurements after first
PWM period after commutation event
ADC0CONIF = 1;
Exit
Interrupt
// Clear flag
}
Figure 19. Commutation transients processing
Where the commutation transient time TOff has not yet expired, the zero-crossing calculation will not be
performed. The calculation will also not be performed if the zero-crossing point has already been identified
in the current commutation period, or if the application is running in open-loop mode.
If the above mentioned conditions are not met, the zero-crossing detection routine will be executed. Based
on the current commutation sector, a slightly different code execution will be performed, so therefore this
sector number of 0 to 5 is passed to the algorithm.
When the zero-crossing position calculation is finished, the BEMF voltage value is stored as the old value
as it will be referenced again in the next PWM cycle.
When a zero-crossing point is identified, the multiplexer that reads one of the three phase voltages is
switched to the phase that is relevant for phase voltage reading in the next commutation cycle.
Afterwards, the glb_ldok of the PTU is set to be prepared for the next reload signal to change the phase
multiplexer located in the GDU module.
The flow chart and the code listing in Figure 20 describe the zero-crossing detection routine that was called
in the interrupt shown before. Depending on the actual detected phase and the active sector, a routine is
called for either a rising or a falling BEMF voltage detection function. The example here shows the routine
called in the case of a rising BEMF on Phase A.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
24
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
void ZCdetectPhAraisingAdc(void)
{
tFrac16 delta;
if (bemfVoltage >= 0){
// Raising approximation
delta = bemfVoltage - bemfVoltageOld;
if ((driveStatus.bit.AdcSaved == 1) &&
(delta > bemfVoltage)) {
timeBEMF -= MLIB_Mul(MLIB_Div(bemfVoltage, delta, F16),
timeBEMF - timeOldBEMF, F16);
}
else{
// middle of previous ADC sensing events
timeBEMF -= ((timeBEMF - timeOldBEMF) >> 1);
}
Function
ZC Detect
PhaseA Raising ADC
BEMF voltage >
Dcbus/2
no
yes
Calculate BEMF ZC time
Save current and previous zero-cross
time
lastTimeZC = timeZC;
timeZC = timeBEMF;
Calculate zero-cross period, filter value
periodZC_R_PhA = timeZC - lastTimeZC;
actualPeriodZC = (actualPeriodZC + periodZC_R_PhA) >> 1;
Calculate and set next commutation
time
NextCmtPeriod = MLIB_Mul(actualPeriodZC,
advanceAngle, F16);
TIM0TC0 = timeZC + NextCmtPeriod;
Exit
driveStatus.bit.NewZC = 1;
}
}
Figure 20. Zero-crossing detection routine
In the case of a negative BEMF voltage, the zero-crossing point has not been passed and the zero-crossing
point is not detectable. The software exits the zero-crossing detecting routine and leaves the zero-crossing
status bit as still unmarked. In the case of a positive BEMF voltage, the zero-crossing point was passed
and Equation 8 is calculated, meaning that the BEMF voltage is divided by the delta of the two measured
points and multiplied by the measured PWM period (BEMF measurement period). After this calculation,
the old zero-crossing time and the new one are saved into the appropriate variables. The zero-crossing
period is then calculated based on the calculated time of zero-crossing and the time of the zero-crossing in
the previous commutation cycle. The zero-crossing period is also filtered to improve reliability.
At the end of the routine, the new commutation time is calculated. Here, some motor characteristics have
to be taken into account. Instead of just adding half of a zero-crossing period to the actual zero-crossing
time, a so-called advance angle factor is taken into account, which actually activates the commutation a
bit earlier than calculated. This is usually a constant and depends on the motor characteristics.
Finally, the zero-crossing point is marked as found, and in that way the calculation does not take place
anymore in current commutation cycle.
4.2.3.3
Speed evaluation and control
The speed controller in Figure 21 is executed in a timer interrupt every 1 ms. First of all, the actual speed
is calculated from all of the last six zero-crossing periods, and this is stored in a scaled format as the actual
speed.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
25
Software implementation on the MC9S12ZVML128
The quality of the zero-crossing periods is evaluated in a function (StallCheck()) and compared against the
min and max limits, with error stored if those limits are exceeded.
In the closed-loop mode, the speed error between the actual speed and the required speed is calculated.
This calculated speed error is fed into the PI controller function. The PI controller is part of the Freescale
Motor Control Library (for further information, see MC9S12ZVMMCLUG, Math and Motor Control
Library for MC9S12ZVML128 User’s Guide). Inputs to the PI controller function include the speed error
and the PI controller’s parameters such as the proportional and integral constants. The output of the PI
controller is the duty-cycle, which is scaled to the PWM resolution.
At the end of the speed control function, the duty-cycle is assigned to the PMF module, and based on the
duty-cycle, new trigger points are also calculated and updated in the PTU module.
Interrupt
TIMchan3
INTERRUPT void TIMchan3_ISR(void)
{
tFrac16 PIOut;
EnableInterrupts;
TIM0TC3 = TIM0TC3 + TIMER_1MS;
if (driveStatus.bit.StallCheckReq == 1) {
driveStatus.bit.StallCheckReq = 0;
StallCheck();
}
Set next ISR after 1 ms
actualSpeed = [scaled] 1 /(sum of all
zero-cross periods)
if (driveStatus.bit.Sensorless
period6ZC = periodZC_F_PhA +
periodZC_F_PhB +
periodZC_F_PhC +
After commutation evaluate precision of
zero-cross findings
== 1) {
periodZC_R_PhA +
periodZC_R_PhB +
periodZC_R_PhC;
actualSpeed = SPEED_CALC_NUMERATOR / period6ZC;
speedErr = requiredSpeed - (tFrac16) actualSpeed;
speedPIOut = GFLIB_ControllerPIrAW(speedErr,
&speedPIPrms, F16);
In close-loop mode:
- Calculate Speed PI controller
- Calculate duty-cycle (scale PI
output to 0 - PWM modulo value)
- Update duty-cycle and set PTU
trigger delays
duty_cycle = MLIB_Mul(PIOut, PWM_MODULO, F16);
UpdateDutycycle();
}
CheckManualInterface();
UpdateCurrentLimitingLed();
Exit
Interrupt
TIM0TFLG1 = TIM0TFLG1_C3F_MASK;
}
Figure 21. Speed evaluation ISR software flow
4.2.4
Current limitation controller
The current limit controller is located in the same 1 ms timer interrupt as the speed controller because the
inputs and outputs of both controllers are linked together.
When the actual speed has been calculated, the current limit PI controller can be called by feeding it with
the difference between actual current and the maximum allowed current of the motor. The output of the PI
controller is scaled to the number proportional to the PWM period. After the current PI controller has
calculated its duty cycle, both the duty cycle output values are compared to each other.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
26
Freescale Semiconductor
Software implementation on the MC9S12ZVML128
If the speed PI controller duty cycle output is higher than the current limit PI controller output, then the
speed PI Controller duty cycle output value is limited to the value of the current limit PI controller;
otherwise, the speed PI duty cycle output will be taken as the duty cycle update value. The value of the
duty cycle will be used to update the PMF module and the trigger delay values in the PTU unit.
At the end, the integral portion values of both the PI controllers need to be synchronized to avoid one of
the controllers increasing its internal value as far as the upper limit. If the duty cycle was limited to the
current PI duty cycle output, then the integral portion of the current PI controller will be copied into the
integral portion of the speed controller, and vice versa. The above described procedure is also described in
Figure 22.
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
27
Software implementation on the MC9S12ZVML128
INTERRUPT void TIMchan3_ISR(void)
{
tFrac16 PIOut;
EnableInterrupts;
TIM0TC3 = TIM0TC3 + TIMER_1MS;
if (driveStatus.bit.StallCheckReq == 1) {
driveStatus.bit.StallCheckReq = 0;
StallCheck();
}
if (driveStatus.bit.Sensorless
period6ZC = periodZC_F_PhA +
periodZC_F_PhB +
periodZC_F_PhC +
Interrupt
TIMchan3
actualSpeed = SPEED_CALC_NUMERATOR / period6ZC;
speedErr = requiredSpeed - (tFrac16) actualSpeed;
speedPIOut = GFLIB_ControllerPIrAW(speedErr,
&speedPIPrms, F16);
currentErr = DCBusCurrentLimit - DCBusCurrentFiltered;
currentPIOut = GFLIB_ControllerPIrAW(currentErr,
¤tPIPrms, F16);
actualSpeed = [scaled] 1 /(sum of all zero-cross periods)
Evaluation of zero-cross findings against limits
-
Calculate Current Limit PI controller
Calculate current limit duty-cycle (scale to PWM )
-
Calculate Speed PI controller
Calculate duty-cycle (scale to PWM)
-
== 1) {
periodZC_R_PhA +
periodZC_R_PhB +
periodZC_R_PhC;
if (currentPIOut >= speedPIOut) {
currentPIPrms.f32Acc = ((tFrac32) speedPIOut) << 16;
currentPIPrms.f16InErrK1 = 0;
PIOut = speedPIOut;
driveStatus.bit.CurrentLimiting = 0;
}
else{
speedPIPrms.f32Acc = ((tFrac32) currentPIOut) << 16;
speedPIPrms.f16InErrK1 = 0;
PIOut = currentPIOut;
driveStatus.bit.CurrentLimiting = 1;
}
duty_cycle = MLIB_Mul(PIOut, PWM_MODULO, F16);
UpdateDutycycle();
}
else if ((appState == APP_START) ||
(appState == APP_ALIGNMENT))
{
currentErr = ALIGN_CURRENT_SCALED - DCBusCurrent;
PIOut = GFLIB_ControllerPIrAW(currentErr,
AlignCurrentPIPrms, F16);
Speed PI duty cycle output compared to Current Limit PI
duty cycle output
If Speed PI output > Current PI output
duty cycle = Current PI output, else
duty cycle = Speed PI output
Update PMF duty-cycle and PTU trigger delays
Update PMF duty-cycle and PTU trigger delays
Exit
Interrupt
duty_cycle = MLIB_Mul(PIOut, PWM_MODULO, F16);
UpdateDutycycle();
}
CheckManualInterface();
UpdateCurrentLimitingLed();
TIM0TFLG1 = TIM0TFLG1_C3F_MASK;
}
Figure 22. Speed evaluation and phase current limitation
4.2.5
MC Library
The application source code uses the Freescale Motor Control Library for the MC9S12ZVML128
microcontrollers (described in MC9S12ZVMMCLUG, Math and Motor Control Library for
MC9S12ZVML128 User’s Guide). The library contains three independent library blocks: GFLIB,
GDFLIB, and GMCLIB. GFLIB includes basic mathematical functions (such as sine, cosine, ramp, and
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
28
Freescale Semiconductor
FreeMASTER user interface
so on). Advanced filter functions are part of the General Digital Filters Library, and standard motor control
algorithms are part of the General Motor Control Library.
5
FreeMASTER user interface
The FreeMASTER debugging tool is used to control the application and monitor variables during runtime.
(For further information, see freescale.com/FREEMASTER and document number
MTRCKTSBNZVM128QSG, Quick Start Guide for 3-Phase Sensorless BLDC Kit with MagniV
MC9S12ZVML128 MCUs). Communication with the host PC passes via USB. However, because
FreeMASTER supports RS232 communication, there must be a driver for the physical USB interface,
CP2102, installed on the host PC that creates a virtual COM port from the USB. The driver can be installed
from silabs.com. The application configures the SCI module of the MC9S12ZVML128 for a
communication speed of 9600 bit/s. Therefore, the FreeMASTER user interface also needs to be
configured respectively.
6
Conclusion
The design described in this document shows the simplicity and efficiency of using the
MC9S12ZVML128 microcontroller for sensorless BLDC motor control, and introduces it as an
appropriate candidate for various low-cost applications in the automotive area.
7
References
Document number
MC9S12ZVMRMV1
MC9S12ZVM128MCBUG
MTRCKTSBNZVM128
Title
MC9S12ZVM-Family Reference Manual
Availability
freescale.com
MC9S12ZVML128 Evaluation Board User’s Manual
3-phase Sensorless BLDC Motor Control Kit with the MagniV
MC9S12ZVM
freescale.com/AutoMCDevKits
MTRCKTSBNZVM128QSG Quick Start Guide for 3-Phase Sensorless BLDC Kit with
MagniV MC9S12ZVML128 MCU
MC9S12ZVMMCLUG
—
Math and Motor Control Library for MC9S12ZVML128 User’s
Guide
FREEMASTER Runtime Debugging Tool
freescale.com/AutoMCLib
freescale.com/FREEMASTER
3-phase Sensorless BLDC Motor Control Kit with the S12 MagniV MC9S12ZVM, Rev. 1
Freescale Semiconductor
29
How to Reach Us:
Information in this document is provided solely to enable system and software
Home Page:
freescale.com
implementers to use Freescale products. There are no express or implied copyright
Web Support:
freescale.com/support
information in this document.
licenses granted hereunder to design or fabricate any integrated circuits based on the
Freescale reserves the right to make changes without further notice to any products
herein. Freescale makes no warranty, representation, or guarantee regarding the
suitability of its products for any particular purpose, nor does Freescale assume any
liability arising out of the application or use of any product or circuit, and specifically
disclaims any and all liability, including without limitation consequential or incidental
damages. “Typical” parameters that may be provided in Freescale data sheets and/or
specifications can and do vary in different applications, and actual performance may
vary over time. All operating parameters, including “typicals,” must be validated for each
customer application by customer’s technical experts. Freescale does not convey any
license under its patent rights nor the rights of others. Freescale sells products
pursuant to standard terms and conditions of sale, which can be found at the following
address: freescale.com/SalesTermsandConditions.
Freescale and the Freescale logo are trademarks of Freescale Semiconductor, Inc.,
Reg. U.S. Pat. & Tm. Off. MagniV is a trademark of Freescale Semiconductor, Inc. All
other product or service names are the property of their respective owners.
© 2013 Freescale Semiconductor, Inc.
Document Number: AN4704
Rev. 1
08/2013