INTEGRATED PARAMETERS MEASUREMENT OF BRUSHLESS DIRECT CURRENT MOTOR AND CONTROL FOR QUADCOPTER APPLICATION
by
LEONG YONG CHEE
Thesis submitted in fulfilment of the requirements for the Bachelor’s Degree of Engineering (Honours) (Aerospace Engineering)
June 2019
ii
ENDORSEMENT
I, Leong Yong Chee hereby declare that all corrections and comments made by the supervisor and examiner have been taken consideration and rectified accordingly.
(Signature of Student) Date:
(Signature of Supervisor) Name: Ir. Dr. Ahmad Faizul Hawary
Date:
(Signature of Examiner) Name: Ir. Dr. Parvathy Rajendran
Date:
iii
DECLARATION
This thesis is the result of my own investigation, except where otherwise stated and has not previously been accepted in substance for any degree and is not being concurrently submitted in candidature for any other degree.
(Signature of Student) Date:
iv
ACKNOWLEDGEMENTS
First and foremost, I would like to thank my research supervisor, Ir. Dr. Ahmad Faizul Hawary for teaching me and giving me a lot of guidelines about the concept of control system, Arduino coding, thesis writing and even hardware assembly technique.
He is reliable for guiding me with useful suggestion and solution.
I take this opportunity to express my gratitude to the School of Aerospace Engineering in USM for providing me the platform and necessary facilities to conduct my final year project.
I would also like to show gratitude to the assistance engineer, Mr. Mohd. Amir bin Wahab for being a helping hand that had assisted me throughout the whole year of my research in all the ways like providing me the equipment support, alert me about the safety instruction and precaution, and helping me in hardware wiring connection.
I express my appreciation to my examiner, Ir. Dr. Parvathy Professor for giving me the basic idea on how the signal processing of sensors works during the colloquium so that my research can proceed further.
I wish to express my sincere thanks to my roommate, Lim Yew Hao and Teo Chen Lung for offering help and backing me whenever I met difficulties in the project progress, and also my coursemate for giving me motivation and mental support by sharing positive advices and information
More importantly, I must also thank my family members for giving me encouragement and comfort me. None of this could have happened without support from my family.
v
INTEGRATED PARAMETERS MEASUREMENT OF BRUSHLESS DIRECT CURRENT MOTOR AND CONTROL FOR QUADCOPTER APPLICATION
ABSTRACT
The objective of this project is to measure and study the performance parameters of the BLDC motor. The parameters obtained in this project is propeller thrust, rotor speed and motor current, which is measured by HX711 load cell sensor, A3144 Hall effect RPM sensor and GY-712 Hall effect current sensor respectively. A motor- propeller testbed is set up with the electronic sensor modules and MCU embedded to perform the real-time signal processing in accessing the integrated performance parameters while increasing controlled throttle is applied. On the other hands, step responses of thrust generated by the rotor-propeller are acquired at different fixed throttle inputs. Before conducting the experiment, the sensor modules had been calibrated by comparing the reading with hand measuring tools to validate the accuracy of those sensors and find out that the output of current sensor exist an additional offset of 0.277 A that causing it to deviate from the actual value even though a moving average is applied into current measuring algorithms. The purpose of this experiment is to study the correlations between the measured parameters so as to identify the performance limitations and constraints within the open-loop control system. The results show that the BLDC motor start to toggle on at servo throttle of ‘46’ and reach its maximum throttle at ‘135’ while the motor stalling occurs at throttle percentage of 53.33%. The thrust generated using smaller size propeller is lower however higher rotor speed and lower current consumed is achieved on behalf of that. The step response of thrust at different throttle control is observed in verifying the stability performance of motor-propeller
vi
when applying different speeds. The result shows that the motor losing efficiency at high steady throttle or whenever there is a throttle change as the settling time increases and the overshoot occurs in the waste of energy.
vii
PENGUKURAN PARAMETER BERINTEGRASI MOTOR ARUS TERUS TANPA BERUS DAN KAWALAN UNTUK APLIKASI QUADCOPTER
ABSTRAK
Objektif projek ini adalah untuk mengukur dan mengkaji parameter hasil kerja motor BLDC. Parameter yang diperolehi dalam projek ini adalah kuasa kipas, kelajuan pemutar dan arus motor, yang diukur dengan pengesan sel beban HX711, pengesan A3144 Hall RPM dan pengesan arus GY-712 masing-masing. Tempat kajian propeller-motor disediakan dengan modul pengesan elektronik dan MCU tersambung untuk mejalankan pemprosesan isyarat masa nyata dalam mengakses parameter hasil kerja terintegrasi sambil meningkatkan pendikit terkawal yang diterapkan. Sebaliknya, tindak balas tindanan kuasa yang dihasilkan oleh kipas pemutar diperolehi dengan pemasukan pendikit tetap yang berlainan. Sebelum menjalankan eksperimen, modul pengesan telah dikalibrasi dengan membandingkan bacaan dengan alat pengukur tangan untuk mengesahkan ketepatan pengesan tersebut dan didapati bahawa keluaran pengesan arus mempunyai tambahan offset 0.277 A yang menyebabkannya menyimpang dari nilai sebenar walaupun purata bergerak telah digunakan ke dalam algoritma pengukur arus. Tujuan eksperimen ini adalah mengkaji hubungan antara parameter yang diukur untuk mengenal pasti batasan hasil kerja dan kekangan dalam sistem kawalan ulangan terbuka. Hasilnya menunjukkan bahawa motor BLDC mula bergerak pada pendikit servo '46' dan mencapai pendikit maksimum pada '135' manakala motor terhenti berlaku pada peratus pendikit 53.33%. Kuasa yang dijana menggunakan kipas saiz yang lebih kecil adalah lebih rendah tetapi kelajuan
viii
pemutar yang lebih tinggi dan penggunaan arus yang lebih rendah telah dicapai bagi tujuan itu. Tindak balas langkah kuasa pada kawalan pendikit yang berlainan diperhatikan untuk mengesahkan prestasi kestabilan kipas motor apabila meggunakan kelajuan yang berbeza. Hasilnya juga menunjukkan bahawa motor mula kehilangan kecekapan pada pendikit tetap yang tinggi ataupun semasa mempunyai perubahan pendikit yang menunjukkan masa penetapan bertambah dan tembakan tindak balas berlaku dalam keadaan pembaziran tenaga.
ix
TABLE OF CONTENTS
ENDORSEMENT ii
DECLARATION iii
ACKNOWLEDGEMENTS iv
ABSTRACT v
ABSTRAK vii
LIST OF FIGURES xi
LIST OF TABLES xii
LIST OF SYMBOLS xiii
CHAPTER
1 INTRODUCTION 1
1.1 Overview 1
1.2 Problem Statement 2
1.3 Objectives 3
1.4 Research Approach and Scope 4
1.4.1 Research Approach 4
1.4.2 Scope 5
1.5 Thesis Outline 5
2 LITERATURE REVIEW 7
3 METHODOLOGY 12
3.1 Throttle control of BLDC motor 12
3.1.1 Motor sizing 13
3.1.2 ESC sizing 14
3.1.3 Battery sizing 15
3.2 Measurement of motor current 15
3.2.1 Theory 15
3.2.2 Experimental approach 16
3.3 Measurement of rotor speed 19
3.3.1 Theory 19
3.3.2 Experimental approach 20
3.4 Measurement of propeller thrust 22
3.4.1 Propeller sizing 22
3.4.2 Theory 23
x
3.4.3 Experimental approach 23
3.5 Integrated System Operation 27
3.6 Calibration on measuring accuracy of embedded sensors 27
4 RESULTS AND DISCUSSION 30
4.1 Data correlation and calibration for Hall sensors 30 4.2 Analysis of correlation between measured parameters 31
4.3 Analysis of step response on thrust generated 36
5 CONCLUSIONS AND RECOMMENDATIONS 39
5.1 Conclusion 39
5.2 Recommendation & Improvement 40
5.3 Future Works 41
REFERENCES 42
APPENDICES
A - Arduino Algorithms for System Operation 44
B - Datasheet of Devices Specification 47
C - Figures of Devices Wiring Connection 49
xi
LIST OF FIGURES
Figure 3. 1: Set up of motor-propeller testbed ... 12
Figure 3. 2: O.S. Motor OMA-3825-750 brushless motor ... 13
Figure 3. 3: SKYRC Hornet 60 A Brushless ESC ... 14
Figure 3. 4: Working principle of Hall effect current sensor ... 16
Figure 3. 5: GY-712 Hall effect current sensor module ... 17
Figure 3. 6: Working principle of Hall effect RPM sensor ... 20
Figure 3. 7: A3144 Hall Effect RPM Sensor ... 21
Figure 3. 8: Basic set up of a full-bridge configuration ... 24
Figure 3. 9: HX711 load cell sensor ... 25
Figure 3. 10: Schematic configuration diagram of system hardware structure ... 27
Figure 3. 11: Block diagram of system hardware structure ... 27
Figure 4. 1: Current measurement based on GY-712 module and PV150 meter ... 30
Figure 4. 2: Graph of motor current against throttle percentage ... 32
Figure 4. 3: Graph of propeller thrust against rotor speed ... 33
Figure 4. 4: Overall performance parameters of BLDC motor with KN1160 propeller 34 Figure 4. 5: Overall performance parameters of BLDC motor with KN1260 propeller 35 Figure 4. 6: Graph of speed versus throttle percentage with different propeller size .... 36
Figure 4. 7: Step responses of thrust generation to their respective throttles ... 37
Figure I: Hardware hook up of GY-712 current sensor module ... 49
Figure II: Hardware hook up of A3144 RPM sensor ... 49
Figure III: Hardware hook up of HX711 load cell ... 49
xii
LIST OF TABLES
Table 3. 1: Specification of O.S. Motor OMA-3825-750 brushless motor ... 13
Table 3. 2: Specification of SKYRC Hornet Brushless ESC ... 14
Table 3. 3: Specification of Fullymax LiPo Battery Pack ... 15
Table 3. 4: Specification of GY-712 module ... 17
Table 3. 5: Specification of A3144 Hall effect sensor ... 21
Table 3. 6: Dimension of propellers sizing ... 23
Table 4. 1: Current Measurement based on GY-712 module and PV150 meter ... 30
Table 4. 2: Speed measurement based on RPM Hall sensor and Testo 470 tachometer 31 Table I: Technical information of O.S. Motor OMA-3825-750 ... 47
Table II: Performance Characteristic of GY-712×30A module ... 48
xiii
LIST OF SYMBOLS
𝐼 : Current through the circuit [𝐴]
𝐼𝑀 : Current through the motor [𝐴]
𝐼𝑀𝑖𝑑𝑙𝑒 : No-load current of the motor [𝐴]
𝑉 : Potential difference across the load [𝑉]
𝑉𝑀 : Potential difference across the motor [𝑉]
𝑉𝑚 : Voltage potential after the motor [𝑉]
𝑉𝑒𝑠𝑐 : Voltage potential after the ESC [𝑉]
𝑉𝑏 : Voltage supplied by the battery [𝑉]
𝑉𝑙𝑜𝑎𝑑 : Measured output voltage value of a load [𝑉]
𝑉𝐻 : Hall voltage [𝑉]
𝑉𝐶𝐶 : Voltage at common collector (supplied voltage) [𝑉]
𝑉𝐼𝑂𝑈𝑇(𝑄) : Quiescent output voltage [𝑉]
𝑉𝑂𝐸 : Electrical offset voltage [𝑉]
𝑅 : Resistance of the load [𝛺]
𝑅𝑀 : Resistance of the motor [𝛺]
𝑅𝐸𝑆𝐶 : Resistance of the ESC [𝛺]
𝑆𝑒𝑛𝑠 : Sensitivity of sensor module [𝑚𝑉/𝐴]
𝐸𝑇𝑂𝑇 : Total output error
𝑅𝑃𝑀 : Revolution per minute [𝑟𝑒𝑣/𝑚𝑖𝑛]
𝐾𝑉 : RPM constant of the motor [𝑟𝑒𝑣/𝑉]
𝑁 : Cumulative number of counts within the duration 𝑡 : Cumulative time taken [𝑚𝑠]
𝑇𝑃 : Propeller thrust [𝑘𝑔]
xiv 𝑊 : Weight of load [𝑘𝑔]
𝑃𝑝𝑂𝑢𝑡 : Propeller output power [𝑊]
𝑣𝑝 : Propeller pitching velocity [𝑚/𝑠]
𝑔 : Gravitational acceleration, 𝑔 = 9.81 𝑚/𝑠2 𝜌 : Density of air [𝑘𝑔/𝑚−3]
𝑑 : Propeller diameter [𝑖𝑛𝑐ℎ]
𝑝 : Propeller pitch [𝑖𝑛𝑐ℎ]
𝑀𝑙𝑜𝑎𝑑 : Measured output mass value of a load 𝑀𝑟𝑒𝑓 : Reference output mass value at no load 𝑀𝑠𝑎𝑚𝑝𝑙𝑒 : Measured value of a sample with known mass 𝑚𝑠𝑎𝑚𝑝𝑙𝑒 : Mass of the sample
𝑘𝑚 : Scale factor of load cell
𝐾1 & 𝐾2 : Empirical constants in thrust measurement experiment
1 CHAPTER 1 INTRODUCTION
1.1 Overview
A drone, also technically called as an unmanned aerial vehicle (UAV), have been the focal point of aviation in the last two decades since the first flight of human being took place in the history. Quadcopter is one of the UAVs categorised as a rotorcraft as the lift is required for a quadcopter flight which is opposed to the fixed-wing aircraft, a more powerful motors are therefore needed to generate sufficient thrust to hover in the air. The commonly used motors used in drone are Brushless Direct Current (BLDC) motors as they are economical, lightweight, small but powerful. One of the advantages of this motor over the brushed motor is that there is absence of brushes inside BLDC motor hence it does not require periodic maintenance, replacement or repair and never a need to be concerned about the condition of the brushes. Therefore, the motor is more reliable and has a higher speed range since there is no mechanical limitation imposed by the interaction between brushes and commutator. However, BLDC motor cannot be operated unless an Electronic Speed Controller (ESC) is used to control the speed of motor. The ESC enables the voltage and current of the motors to be controlled based on the applied speed reference signal either in the form of servo signal or input of pulse width modulation (PWM) signal to control the motor speed by adjusting the duty cycle or switching the frequency of a network of field effect transistors (FETs). And the rapid switching of the transistors is what causes the motor itself to produce high pitched whine which is noticeable especially at low speeds. Therefore, there is a current issue being encountered during the drone design which is less information obtained from the operation of BLDC motor together with the embedded ESC such as how fast the response
2
acting towards the input signal to accelerate the motor, the stability on motor signal responses, throttle level where the motor stalling, maximum throttle on motor control, correlation of motor rpm-current, etc.
1.2 Problem Statement
Quadcopter is mostly unstable because of its nonlinear model that consists of numerous variables within six degrees of freedom worked by four actuators to fly in symmetrical position. The system design of controller for the quadcopter positioning includes three translational and three rotational movement that are imperative for the manoeuvring stability. At this point, it is significant to carry out analyses to detect the possible obstacles encountered during the manoeuvre performance of the quadcopter, inspecting the cause of the circumstances and trying to maintain its typical operations.
By adjusting the rotational speed of each rotor, the thrust of the motor-propeller performance can be controlled. To read continuously the thrust value of the rotor, load cell sensor and rpm sensor are used, and the movement error assessed by these sensors are sent to the microcontroller unit (MCU) to be corrected using Proportional, Integral, Derivative (PDI) algorithms. Every control system has the critical role in achieving a high stability and the capability of diagnosis as well as correction for the output errors is the important criteria for a precision controller. The basic open-loop algorithms are unsuitable to drive a motor accordingly hence the control system must be programmed to regulate the motor throttle repeatedly to retain the desired thrust in maintaining a stable flight position. Monitoring and control adjustment for the speed of the rotors are necessary and the performance in maintaining stability is depending on how fast and how accurate the control system diminishes the position error.
3
To achieve this, the characteristic of the motor is needed prior to design a motor controller. Usually a motor that is purchased in the market does not come with accurate specifications. Therefore, this study aims to develop and set up a system that can analyse the BLDC motor characteristic using experimental approach. The proposed system is expected to be able to extract the motor characteristic in terms of speed, current, thrust, etc. In this study, monitoring the thrust produced by the propeller will be the objective where the process of triggering and monitoring can be done at the same time. Hence, a motor-propeller testbed is set up with electronic sensor modules embedded onto it where the set up can read the thrust generated and other performance parameters while tuning the motor throttle.
1.3 Objectives
The overall purposes of this study are listed as below:
1. To study the characteristics of motor-propeller based on the correlations between thrust, speed and current at an increasing controlled throttle.
2. To observe the step response of thrust generated by the motor-propeller testbed at different fixed throttle input control.
3. To perform sensor calibration and data correlation for both current sensor and RPM sensor.
The purpose of this study is to collect the performance characteristics of the BLDC motor by controlling the motor throttle. A motor-propeller testbed with an MCU embedded is constructed to regulate and monitor the thrust generation by the rotor- propeller based on the developed control scheme. Thrust, speed and current are detected from the identification process on the motor-propeller performance using their respective sensors. The data collecting process is conducted based on signal processing in real-time
4
after the diagnosis of electronic and mechanical errors is carried out for zero offset calibration. The results of experiments are provided in the chapter 4 with two different means of throttling control method and two different testbed models to demonstrate the difference in performance and response regarding the states output of rotor-propeller.
1.4 Research Approach and Scope
1.4.1 Research Approach
This study discusses the design of the system implementation method, mechanism and operating principles, including the system hardware structure, the selection of sensors and the flow chart of system operation. The purpose of the research is to retrieve various responses of performance parameters corresponding to the motor throttle in a control system via experimental approach. This study proposes two approaches about the throttle control method to visualise their responses of the thrust generation in a practical way. Data collecting process is built based on the altering of throttle controller parameter via a direct approach to access the sets of input-output data obtained from the system of motor propeller testbed. The experimental results will manifest on the proportionality of the thrust-current dynamics of motor-propeller performance, the settling time of the thrust generation at the specified throttle step input value and the stall limitation of the BLDC motor that may aid in the stability calibration drone control in further research. The set up will be operated by the control core unit which using Arduino Uno MCU as control module and sensors such as GY-712 current sensor, HX711 load cell sensor and Hall effect RPM sensor to test the relation of thrust with current and speed so as to determine the performance parameters of the motor-
5
propeller. The control system is then being measured using different measuring tools to validate the accuracy of the data processing analysis. The research procedures can be summarised as the scope shown as below.
1.4.2 Scope
1. To study the use of GY-712 Hall effect current sensor to measure the current of the rotor.
2. To study the use of Servo.h library to control throttle of the BLDC motor.
3. To study the use of Hall effect RPM sensor to measure the speed of the motor- propeller.
4. To study the use of HX711 load cell sensor to measure the thrust generated by rotor-propeller.
5. To obtain the performance parameters and step response of the motor-propeller testbed at different means of throttle control.
1.5 Thesis Outline
The body of the study is organised into 5 chapters as the following. Chapter 1 is the introduction which includes overview, problem statement, objectives, research approach and scope. Chapter 2 presents the literature review and benchmarking of BLDC motor for multirotor operations. This chapter also explains mathematical modelling of algorithms for the motor-propeller testbed which involve of theory behind the electronic sensor modules embedded onto the system. The methodology on how to execute experiments including control strategies used to tune the controller parameters and the approach on how to validate the output signal processed by the MCU is elaborated in
6
Chapter 3, as well as the description of the testbed parameters used in this study involving the schematic configuration diagram, specification and performance characteristics. The implementation of results is discussed in Chapter 4 related to the evaluation and analysis of experimental test results from the controls of rotor-propeller testbed while finally the improvement and conclusions discussing the possible future works are provided in Chapter 5.
7 CHAPTER 2 LITERATURE REVIEW
Propulsion system of UAV is important to provide the necessary power to propel itself for forward flight or hover. Due to the high thrust-to-weight ratio, the vertical take- off and landing (VTOL) vehicles have become the focus of study especially on quadcopter since they have higher manoeuvrability advantages over the other types of UAVs.(Ermeydan and Kiyak, 2017) Electric propulsion systems are commonly used for quadcopter as they are scalability meaning that the thrust generated can be controlled precisely and respond faster to throttle input. Electric propulsion systems are crucial to provide consistent power control so to diminish the motor control failure as well as to minimize the battery consumption.(Griffis et al., 2009) Hence, the criteria needed to define a good control performance of quadcopter in throttling or hovering mode are the quick but minimum responses. However, the swift response tends to cause oscillation which leads to less stable control during the transient state and therefore extra control signalling is usually required to neutralise its effect.(Raharja et al., 2017)
There is another common issue about the stability of quadcopter, that is uneven thrust generated despite supplying the same magnitude of throttle input. During throttling and hovering, the flight stability of a quadcopter is obtained when the equal thrusts are generated by all the rotors-propellers. Even though the identical motors and propellers used at the same speed, the thrusts generated are not really to be balanced not to mention that a single rotor-propeller set up is not necessarily producing a stable thrust. The imbalances of propeller during vibration, the voltage input values of the pulse width modulation (PWM) or pulse position modulation (PPM) in ESC, the timing of remote control, as well as the sensitivity of the sensors used may be the reason for that. Hence,
8
a closed-loop controller is normally required to encounter this problem by eliminating all these electronic and mechanical errors conceding that the correlation of thrust value as well as other performance parameters from the rotor-propeller are apparent beforehand.(Kuantama et al., 2018)
Generally, components selection for the electric propulsion system in designing an UAV has been a simple process due to recommendation of standard propulsion combination sets by manufacturers or suppliers based on weight and type of UAV.
However, the recommended weight range may only suitable up to 25 % of the maximum take-off weight. Thus, analysis on the sizing model of propulsion system is essential to estimate the performance of an UAV design. The simulation for the performance of both solar UAV and non-solar UAV had been modelled in (Rajendran et al., 2016) by changing different propeller size and the performance comparison was done among those propeller sizes. The theoretical formulas and equations shown in this research was used in the Methodology to approximate the performance parameters of the motor-propeller testbed analytically from the specification datasheet of propulsion system sizing devices.
The results from performance comparation showed that lowering propeller diameter and pitch increases the efficiency of electric motor but not grant enough thrust output. Instead, propeller with larger diameter and pitch generates higher thrust and power to weight ratio.
Hence, the research concluded that the propeller tip static speed should be the key parameter for propeller sizing instead of power-to-weight ratio as lower propeller tip static speed allows higher endurance of UAV flight.
The testing for the precise selection of electric propulsion system which consists of BLDC motors and propeller for the use in multicopter or hovercraft had been done in (Patel et al., 2017) by comparing the results of both theoretical values given in the datasheet of the motors and the practical thrust calculation after conducting experiments
9
on couple of motors. The sizing of the system considered the maximum take-off weight as the requirement so that the motors-propellers combination is able to produce thrust twice the flying weight of the multicopter system. The comparison showed the thrust difference in experimental analysis proving that there is always a loss in the motors during their practical operation that has to be included in calculation and consideration during the selection of propulsion system components.
To validate the accuracy of simple motor and propeller models as well as to facilitate the selection of appropriate BLDC motor and propeller combination for flight applications, the theoretical performance approximation for single motor and propellers can carried out solely based on data provided by manufacturers and then compared to the experimental results obtained in static thrust tests. The motor and propeller models can be decoupled as different single objects to study their respective results independently.
The theoretical predictions in (Muzar et al., 2017) were distributed into two parts which are modelling of BLDC motors that quantify the speed-to-voltage as well as torque-to- current characteristics and modelling of propellers that focus on speed-torque-thrust characteristic. The relation between torque and speed was explained by using motor circuit analysis and the thrust coefficient was obtained as a function of propeller diameter and pitch to determine the static thrust. The experimental results showed that the speed and torque of BLDC motor models increase by increasing motor driving voltage.
Increasing propeller diameter and pitch raises its output torque to provide higher propeller thrust but reducing the motor shaft speed. The findings also showed that all the propellers produce almost the same load of thrust at low speed. The propeller with the smallest pitch generates the most thrust output at high speed.
The design and simulation of an UAV highly depends on the thrust produced by a motor-propeller combination. A method was proposed in (Gupta and Abdallah, 2018)
10
using the experimental data from 291 motor-propeller data points to verify a generalised mathematical relationship between the motor input RPM and the corresponding output thrust for the preliminary design process of low Reynold’s number applications. The mathematical models comprised of the expression for inflow velocity ratio which is developed into a simplified form using the concept of Blade Element and Momentum Theory (BEMT), the force constant which defines each thrust-rpm mathematical model and the expression for coefficient of thrust which involves the effects of pitch variation and rotor diameter effectiveness. The process was different from curve fitting method as the discrepancy of thrust between the estimated data and the experimental data may occur over the high speed and huge range of propeller sizes where the aerodynamic phenomenon becomes difficult to be generalised. The findings showed that the validation of the small size UAV model is only applicable when using propellers of diameter ranging from 4 inches to 6 inches.
A similar experiment had been conducted in (Al et al., 2018) with empirical testing to collect the parameters and data related to the relationship between BLDC motor voltage supplied, rotor speed, air speed and rotor lift. The result proved that the air velocity measured using anemometer which is used to determine thrust coefficient is linearly proportional to the rotor speed measured using tachometer as according to the concept derived using Newton’s law, Archimedes, Pascal, Bernoulli and the law of continuity. The results from measurement data also showed that the lift force generated relative to the rotational speed is not proportional. Nonlinearity occurs when the lift generated by the rotor reaches up to 1.5 N where the rotation speed is around 65 revolutions per second (rps), then the thrust surges up sharply with the increase of rotor speed rotation. The performance parameters obtained for a 1200 KV BLDC motor with
11
30 A ESC and a propeller blade with the radius of 12 cm at the input voltage of 12V in room temperature has the thrust coefficient of 1.732.
In terms of motor selection, the technical information of the motor parameters given in the motor specifications may not be sufficient, especially for cheaper motor tends to have high tolerances in their performance characteristics. Micro brushed and brushless motors used on micro-ground vehicles and micro-air vehicles were also been found to have low efficiencies and the efficiencies decreases significantly with motor size due to not only low manufacturing tolerances but also motors operation under high range of loading status causing them to operate off design conditions. A basis of simple motor analysis was explained in (Harrington and Kroninger, 2013) with comparation between the analysis and the actual motor test data in terms of operating torque and speed within different motors. The experiments were conducted to examine the efficiencies of off-the-shelf brushless motors in two approaches that is by observing the tendencies of motor overall efficiencies change without the use of brushless speed controller at different voltages with a constant throttle as well as at a constant voltage with varying throttle settings. The results showed that the overall efficiencies of brushless motors increase with the increase in motor power output as a result of increasing input voltage and the maximum efficiency from a brushless motor is obtained at 100 % throttle. The study also emphasized the importance in matching of motor design to the motor loading conditions in order to maximize power efficiency and it was found that the brushless speed controllers are the source in causing motor inefficiency as the efficiency drops linearly with increased motor load.
12 CHAPTER 3 METHODOLOGY
3.1 Throttle control of BLDC motor
A complete electric propulsion system for UAVs comprises of motors, propellers, ESCs and a battery. Electric motors in electric propulsion system serve as powerplants because they can be easily coupled with propeller as the propulsion effecter, creating rotational motion from electric power and all it needs is continuous source of electricity.
Theoretically, the rotational speed of an electric motor is proportional to the voltage applied to it while the torque generated is proportional to the current flow.
Figure 3. 1: Set up of motor-propeller testbed
The experiment for the motor-propeller testbed is set up as shown in Figure 3.1 with all the electronic devices and sensor modules connected to the Arduino MCU for signal processing. The movement of motor is restricted at the set-up platform which is mounted at one leg of the load cell sensor and the motor is connected to an ESC which was connected in series with a LiPo Battery of 14.8 V and a Hall effect current sensor while the ESC unit was being controlled by Arduino Uno with throttling signals. On the
13
body of the BLDC motor, there is a small neodymium magnet adhere on it and a Hall effect RPM sensor is fixed just beside the motor set up.
3.1.1 Motor sizing
A selected motor will have a huge impact on the payload where an UAV can support, as well as the flight time. A general rule in designing an UAV is that the thrust provided by the motors should compromises with the requirement of thrust-to-weight ratio. The brushless electric motors are commonly chosen based on how much thrust required by the motors to lift off and fly the quadcopter at speed. The image and specification of the BLDC motor used in this study is shown in Figure 3.2 and Table 3.1 respectively. The detailed technical information of O.S. Motor OMA-3825-750 is shown in Table I at Appendix B.
Figure 3. 2: O.S. Motor OMA-3825-750 brushless motor
Table 3. 1: Specification of O.S. Motor OMA-3825-750 brushless motor
Specification Description
Sort Brushless
Operating voltage range 14.8 V – 18.5 V
Rated current 35 A – 40 A
No-load speed, rpm 11100
Max. output at nominal voltage 625 W
Nominal voltage 14.8 V
Thrust 30 N (1 N = 100 g)
Maximum current 75 A
No-load current 2 A
Phase Resistance 31 mΩ
RPM constant (KV) 750 rev/V
14 3.1.2 ESC sizing
The brushless electric motor is powered by a Direct Current (DC) electric source via ESC which produces a three phase Alternating Current (AC) electrical signal to control and regulate the speed as well as the direction of the motor besides power supplying the motor. The ESC must be able to handle the maximum current in which the motor might consume and be able to provide it at the right voltage. The ESC used in this study is SKYRC Hornet Brushless Sensor ESC 60 A shown in Figure 3.3 which runs in correspond with the servo control command to operate the motor control in the experiment and the sets of data is concurrently being collected along the way. The ESC is connected to the MCU so to get supply of regular 5 V input voltage and to receive the servo PWM signal from the MCU which then makes motor moves. The downside of SKYRC Hornet 60 A ESC is that it can only allow the use of servo PWM signal but not digital PWM signal to control the motor as referred to Appendix A. The detailed feature specifications of the SKYRC Hornet 60 A ESC is shown in Table 3.2.
Figure 3. 3: SKYRC Hornet 60 A Brushless ESC Table 3. 2: Specification of SKYRC Hornet Brushless ESC
Specification Description
Maximum operating current 60 A
Operating voltage 2 – 6 S LiPo / 6 – 20 Cell NiMH
Switching BEC 5.7 V / 3 A
15 3.1.3 Battery sizing
Batteries are crucial to sustain the quadcopter flight time and lift. Lithium polymer (LiPo) is used for the batteries in this study as it can supply maximum energy density with high discharge rates. The higher the capacity, the longer the operations of electric motors, but the heavier the battery pack will be. The selection of battery pack must be ensured that the maximum current supplied by the battery is lower than the ESC outburst current to prevent burn-out of the ESC. The battery used in this study is Fullymax 14.8 V 3000 mAh 4S 35 C LiPo Pack and its detailed specifications is shown in Table 3.3.
Table 3. 3: Specification of Fullymax LiPo Battery Pack
Specification Description
Capacity 3000 mAh
Voltage 4S1P / 4 cell / 14.8 V
Discharge rate 35
3.2 Measurement of motor current
3.2.1 Theory
The approximation of motor current can be determined analytically based on motor voltage using the required data from the specification datasheet of motor, ESC and battery used in this study such as no-load current and resistance of the motor, resistance and maximum current limit of the ESC as well as operational voltage of the battery.
𝑉𝑒𝑠𝑐 = 𝑉𝑏− 𝐼 × 𝑅𝐸𝑆𝐶 (1) 𝐼 =𝑉𝐸𝑆𝐶− 𝑉𝑚
𝑅𝑀 (2)
𝐼𝑀 = 𝐼 − 𝐼𝑀𝑖𝑑𝑙𝑒 (3)
16 3.2.2 Experimental approach
A Hall effect current sensor is used as the device to detect the electric current in a wire and produce a signal which is proportional to that current. The signal produced could be either analog voltage or current or even a digital output and is stored in a data acquisition system for further analysis or for the purpose of control. There are various types of current sensors and the current sensor used in this study is basically a Hall effect transducer in which the detection is based on the Hall Effect phenomenon and can be used to measure all types of current signals including AC, DC and pulsating current.
Working principle of Hall effect current sensor
Figure 3. 4: Working principle of Hall effect current sensor(Joshi, 2013)
Inside a Hall effect sensor, there is a thin strip of metal attached along the circuit in a magnetic field as shown in Figure 3.4. In the existence of magnetic field perpendicular to the direction of electrons flow, the magnetic force acting on the particles causing the electrons beam in the metal strip will be averted from the straight path towards one edge. As the consequence, charge separation occurs when one edge of the metal strip will be negatively charged while the opposite edge become positively charged, resulting a voltage gradient perpendicular to the current feed called Hall voltage.
17 Specification of GY-712 module
Figure 3. 5: GY-712 Hall effect current sensor module
The GY-712 current sensor as shown in Figure 3.5 is a positive-negative current measurement module that adopts ACS712ELCTR-30A-T Chip from Allegro Microsystems as module chipset, in which its detection range of current is from -30 A to 30 A. The operation principle of the current sensor is Hall effect based where it generates a linear output offset voltage proportional to the captured current. The specifications of the GY-712×30A module is shown in Table 3.4 and its detailed performance characteristics under effect of ambient temperature is shown in Table II at Appendix B.
Table 3. 4: Specification of GY-712 module
Specifications Description
Working temperature, TA -40 °C to 85 °C
Operating voltage, VCC 5 V
Optimized range, IP ±30 A
Sensitivity, Sens 63 mV/A to 69 mV/A
Total output error, ETOT ±1.5 %
While using Hall effect current sensor, there are additional terms about the accuracy characteristic of sensor need to be understood before constructing algorithms for current measurements.
Definitions of accuracy characteristics of Hall sensor
Sensitivity (Sens) is the response change in device output proportionate to a 1 A change through the primary conductor. The sensitivity is the product of the magnetic circuit sensitivity (G/A) and the amplifier gain (mV/G) of linear Integrated Circuit (IC).
18
The linear IC amplifier gain is the property programmed at factory to optimise the sensitivity (mV/A) for the full-scale current of the device.
Quiescent output voltage (VIOUT(Q)) is the output of the device when the primary current is equal to zero. For a unipolar supply voltage, the output normally remains on VCC/2. Hence, the operating voltage, VCC = 5 V is translated into VIOUT(Q) = 2.5 V.
Variation in VIOUT(Q) can be associated to the resolution of quiescent voltage trim and thermal drift in the Allegro linear IC.
Electrical offset voltage (VOE) is the actual device output differed from its ideal
quiescent output value of VCC/2 due to nonmagnetic causes. To obtain the measured current, this voltage is used to divide the device sensitivity, Sens.
Total Output Error (ETOT) is the percentage of maximum discrepancy of the
actual output from its ideal value and is also known as the total output error. Aside from nonmagnetic causes, temperature could also be a limiting factor that disturbs the device accuracy.
Implementation of GY-712 module
The hardware hook-up of GY-712 module is shown in Figure I at Appendix C.
The output sensitivity of Hall effect current sensor means the output voltage relative to the measured current. For GY-712×30A module, the typical sensitivity is about 66 mV/A.
The measurement offset voltage of the current sensor is roughly equal to 2.5 V which means the output is the power midpoint voltage when there are no currents flow through.
Hence, when performing current calibration in Arduino algorithms, it is crucial to eliminate the multiplicative offset by deducting 2.5 V from the signal as described in Eq.
3 to attain actual reading at zero offsetting. Since the current sensor is connected to the analog pin in Arduino MCU, the voltage output signal is read through the Arduino
19
Analog-to-Digital Converter (ADC) in 10 bits number which displays the reading within 0 to 1023.(Boudiaf et al., 2018)
According to Ohm’s law, current flowing through a load can be measured by dividing the voltage across the load with the load resistance which grants the equation:
𝐼 =𝑉
𝑅 (4)
For Hall effect current sensor, the current is determined by Hall voltage induced in accordance with current passing through which is constituted as the sensitivity of the sensor module.
𝑉𝐻′ = 𝑉𝑙𝑜𝑎𝑑× 𝑉𝐶𝐶
2𝑏𝑖𝑡− 1 (5)
𝑉𝐻 = (𝑉𝐻′ − 𝑉𝐼𝑂𝑈𝑇(𝑄)), where 𝑉𝐼𝑂𝑈𝑇(𝑄)= 𝑉𝐶𝐶
2 (6)
𝐼𝑀 =1000 𝑉𝐻
𝑆𝑒𝑛𝑠 (7)
3.3 Measurement of rotor speed
3.3.1 Theory
The approximation of motor speed can be determined analytically based on KV
of the motor with the condition that the potential difference across the motor is obtained.
KV refers to the velocity constant of a motor, measured in revolutions per minute (rpm) per volt. The KV rating of a brushless motor is the ratio of the motor’s unloaded rotational speed to the peak voltage on the wires connected to the coils. By knowing the KV rating of a motor, the motor speed can be determined whether how fast it will rotate with a given voltage applied.
𝑅𝑃𝑀 = 𝑉𝑀× 𝐾𝑉
(8)
20 3.3.2 Experimental approach
An RPM sensor is used in this study to measure the rotational motion of a part. It senses how fast one full revolution is completed by a rotating part which can be then converted to revolutions per minute, linear frequency, period of revolution and rotational speed. The RPM sensor used in this study is a Hall effect sensor where the measured output value detected is in the form of Hall voltage induced which is proportional to the current and magnetic field applied.
Working principle of Hall effect RPM sensor
Figure 3. 6: Working principle of Hall effect RPM sensor(Electricalfundablog.com, 2015) When the Hall effect sensor gets close enough to the magnet attached on the spinning shaft, the sensor is able to detect the presence of polarised magnetic field near its face. As shown in Figure 3.6, the latch signal pin will have the binary output change to LOW once it is triggered by a magnetic South pole of enough strength and its state is then restored back to its ordinarily binary signal of HIGH when the magnetic field is outside the detection range of the sensor. At the end of every loop, the sensor is turned off by discharging its supply voltage and then turned on at the beginning of the loop, hence registering its passing.
21 Specification of A3144 sensor
Figure 3. 7: A3144 Hall Effect RPM Sensor
Figure 3.7 shows the picture of A3144 Hall effect sensor and its detailed specification characteristics is shown in Table 3.5.
Table 3. 5: Specification of A3144 Hall effect sensor
Specifications Description
Operating voltage, VCC 5 V
Open-collector output current 25 mA
Output voltage 5 V
Operating temperature -40 °C to 85 °C
Turn on and Turn off time 2 μs
Implementation of A3144 sensor
The connection of A3144 RPM sensor to Arduino microcontroller pin is shown in Figure II at Appendix C. To measure the speed of the BLDC motor, a tiny neodymium magnet is fixed perpendicular to the rotation axis of the motor and the sensor is located just near enough to the motor so that the magnet attached to the motor can stay passing the sensor field. The set up need specific focus on the proximity between the A3144 Hall effect sensor and the neodymium magnet in which the sensor cannot be too distant away from the magnet that the sensor fails to detect the magnetic pole and cannot be too close to the motor that the sensor may crash with the magnet adhered on the circulating device.
Hence, a small magnet is more advisable to be attached to the motor as small magnet generates less torque to the motor that may cause imbalance during spinning despite its smaller size.
When counting the number of counts the Hall effect sensor detect the magnetic pole, setting Interrupts in Arduino coding is necessary to allow the Arduino algorithms
22
executing several operations in real-time. An Interrupt is a set of command that is executed interrupting the regular basis performed by the source coding instructions while here, Interrupts are called when the magnet passes through the sensor field. To perform the command during interrupt, the Interrupt Service Routine (ISR) is used outside the void setup() as well as void loop() and is being called whenever there is a Falling from HIGH (1) to LOW (0) in digital input pin to increase the counter value that makes a count.
Since A3144 hall effect sensor is a unipolar sensor where only one pole of magnet is needed, every passing of the magnet will trigger the interrupt in the code and rpm count.
The formula for speed measurement is based on simple calculation of number of rotations made by the turning shaft per time taken.
𝑅𝑃𝑀 =60000 (𝑁𝑛− 𝑁𝑛−1)
𝑡𝑛− 𝑡𝑛−1 (9)
3.4 Measurement of propeller thrust
3.4.1 Propeller sizing
In general, the selection of the propeller for an UAV depends on propeller diameter, propeller efficiency and tip static speed. A smaller diameter propeller provides less inertia to the UAV and is therefore easier to speed up or slow down which may help in acrobatic flight. Normally, the propeller manufacturers provide propeller data with thrust and power coefficients included. Hence, thrust can be partially obtained when choosing the suitable propeller. In this study, a propeller G/F 3 Series Master Airscrew KN1160 and a propeller K-Series K1260 K3015 which has the same pitch, but 1 inch longer than the former are used to monitor the generated thrust. The detailed dimensions of both propellers are shown in Table 3.6.
23
Table 3. 6: Dimension of propellers sizing
Propeller type Diameter Pitch
inch mm inch mm
G/F 3 Series Master Airscrew KN1160 11 280 6.0 150
K-Series K1260 K3015 12 304 6.0 152
3.4.2 Theory
In short, propeller moves air and converts the torque of its power source into thrust. The thrust generated by a propeller depends on the density of the air, propeller rotation speed, propeller diameter and pitch, chord length as well as the shape and area of the propeller blades.(cbenson, 2015)
𝑃𝑝𝑂𝑢𝑡 =𝜌
2(𝑅𝑃𝑀 60 )
3
(0.0254 ∙ 𝑑)4(0.0254 ∙ 𝑝) (10) 𝑣𝑝 =𝑅𝑃𝑀
60 (0.0254 ∙ 𝑝) (11)
𝑇𝑃 = 𝑃𝑝𝑂𝑢𝑡
𝑔 ∙ 𝑣𝑝 (12)
The relation of thrust in terms of propeller diameter and pitch proposed by hobbyist Gabriel Staples in 2014, which is developed from simple momentum theory is shown in Eq. 13.
𝑇𝑃 = 𝜌𝜋 (0.0254 ∙ 𝑑
2 )
2
(𝑅𝑃𝑀
60 × 0.0254 ∙ 𝑝)
2
( 𝑑 𝐾1∙ 𝑝)
𝐾2
(13)
3.4.3 Experimental approach
A load cell is a transducer that is used to generate an electrical signal in which the magnitude of the signal is directly proportional to the force being measured. A strain gauge load cell which is also known as force sensor, will be discussed here. The strain gauge is used as a tool to detect changes in a physical phenomenon such as sheer force,
24
bending movement, weight, vibration, motion and pressure. In this study, the load cell was used to measure thrust based on its correlation with weight as it detects only one degree of freedom along the z-axis with dynamic behaviour.
The strain gauge provides wide range, high precision and high integration degree of measurements, as well as high sensitivity of system contact points. The design concept of a strain gauge sensor is to change the stimulus received by the sensor from mechanical loadings into electrical signal by applying resistance modification measuring circuit based on theories of Wheatstone bridge.
Working principle of load cell
Figure 3. 8: Basic set up of a full-bridge configuration(Systems)
Generally, a Wheatstone bridge shown in Figure 3.8 is an electrical circuit used to determine an unknown electrical resistance as the means of calibrating measuring instrument. Wheatstone bridge is able to measure a very small values of resistance down in milli-Ohms range by balancing two legs of a bridge circuit in which one leg is placed with an unknown substance. The Wheatstone bridge of the load cell has four arms and its configuration depends on the number of strain gauges mounted onto a beam. The
25
arrangement of the strain gauge is advisable in such a way that each strain gauge is placed at alternating arms of the Wheatstone bridge to manifest the response of the load. The operation of the of the Wheatstone bridge is quite exact to the original potentiometer and the output generated from the Wheatstone bridge is exactly the output measured by the load cell. When the two series-parallel arrangements of resistance connected between a voltage supply terminal and the ground is balanced in resistance or can be said the bridge is in equilibrium, the circuit producing zero voltage difference between the two parallel branches. When there is any mechanical loading on the load cell, the bridge arm resistance of the strain gauge loses its balance in resistance and produces its respective output signal to the change in linear elastic behaviour which then facilitates the calibration process.(Chao and Han, 2018)
Implementation of HX711 load cell
Figure 3. 9: HX711 load cell sensor
Model of 24-bit ADC HX711 sensor with maximum load of 10 kg shown in Figure 3.9 were employed in the way that the hardware wiring between load cell, HX711 amplifier and Arduino as shown in Figure III at Appendix C. The set up was modelled with the rotary axis of the motor perpendicular to the load cell to measure the magnitude of only upward or downward force along z-axis of the set up. The sensor produces output of voltage signal which is then translated into a digital signal through the load cell ADC
26
driver. The sampling rate of load cell can be selected to either 10 or 80 sampling per second (SPS). In this study, the system was designed to retrieve data of parameters including thrust once in two seconds. In the algorithms code of measuring thrust, the use of the load cell must include HX711 library as referred to Appendix A which can be installed from Arduino Library Manager to toggle the operations of strain gauge inside the load cell.
After the connection between the Arduino Uno and the load cell with the Pressure Controller Based (PCB) is set up, the load cell requires calibration such that system is reset to zero offset by tracking the reference output value of the load cell so that the output voltage is 0 V in the absence of load. After the zero error is eliminated, a propeller with a known weight of 28 g is placed on the sensor and the measured output value is used to determine the scale factor of the load cell as described in Eq. 6 & 7. Thus, the output value after load calibration will be negative with a weight added yet increase with the increase of thrust generated and return to zero when there is no load.
Therefore, the thrust generated can be formulated as the opposite polarity to the weight.
𝑘𝑚 = 𝑀𝑠𝑎𝑚𝑝𝑙𝑒
𝑚𝑠𝑎𝑚𝑝𝑙𝑒 (14)
𝑇𝑃 = −𝑊 = −(𝑀𝑙𝑜𝑎𝑑 − 𝑀𝑟𝑒𝑓) ∗ 𝑘𝑚 (15) Remarks: unit for thrust depends on the measurement unit of the sample
HX711 load cell is highly sensitive that mechanical errors will occur even if it is slightly touched. Hence zero error calibration is mandatory every time before the thrust measurement is taken.
27 3.5 Integrated System Operation
Figure 3. 10: Schematic configuration diagram of system hardware structure
Figure 3. 11: Block diagram of system hardware structure
Figure 3.10 and Figure 3.11 illustrate on how the system operation runs in open loop system associated with the motor drives. The speed of the rotor can be controlled by the amplitude of throttle input voltage adjusted in command pulses sent from the Arduino algorithms. The convenience of open loop system in this approach is that the integrated parameters measurement of the rotor-propeller can be imposed directly.
However, there are various propagation of error that must be taken into account while applying the throttle input voltage amplitude to achieve the desired response output data.
3.6 Calibration on measuring accuracy of embedded sensors
Before collecting the performance parameters of a motor-propeller, the electronic modules used for measurement was tested their respective accuracy of sensing. The
28
sensors calibrated here were GY-712 current sensor and A3144 RPM sensor whose functionality is both based on Hall effect sensing. A Hall effect sensor usually provides lower measuring accuracy than other sensors such as fluxgate magnetometers or magnetoresistance-based sensors as the Hall effect sensors is easily drifted by the external factors. Since the Hall effect sensor works on the principle of magnetic field, any interference from the external magnetic fields caused by the ferromagnetic metal surrounding and the unsteady input current passing through the sensor may cause bias in the measurement. Besides, the temperature regardless environmental or internal also influences the electrical resistance of elements and the mobility of charged carriers in the sensor which is so-called the sensitivity of the sensors. For a Hall effect sensor, an offset voltage mostly occurs when there are material nonuniformities and nonmagnetic causes that requires extra compensation and calibration.
For measuring the parameter of thrust generated, HX711 load cell sensor was validated in such a way that the calibration had already been done before taking the reading for thrust. Scale factor for the load cell was determined utilising the comparative relationship between sensing of the load cell and the phenomenon change such that the magnitude of the scale increment is observed after placing a defined sample weight.
The calibration of the sensor accuracy was done by comparing the operations of two or more measuring instruments on the same parameters. In this study, the BLDC motor was run using different servo throttles of input from ‘46’ to ‘50’ where the motor started spinning from stationary until reaching the corresponded throttle signal and respective current and speed were recorded.
The current flow through motor was measured using GY-712 current sensor and was compared using Solar Installation PV150 current clamp meter. The data accessed from the current sensor is fluctuating due to ratiometric nature that keeps changing the
29
sensitivity of the sensor. Hence an average value of 10 zero-calibrated data was taken as the reading for the GY-712 current sensor. Conversely, a PV150 current clamp meter has the accuracy of ±0.1 A thus the current reading was recorded in one decimal place.
Meanwhile, the speed of the motor rotation measured using A3144 RPM sensor was compared using Testo 470 tachometer which can display maximum accuracy of ±1 rpm.
30 CHAPTER 4
RESULTS AND DISCUSSION
4.1 Data correlation and calibration for Hall sensors
Table 4. 1: Current Measurement based on GY-712 module and PV150 meter Servo Throttle Average intervals from
GY-712 current sensor (A)
Values from current clamp meter (A)
46 0.074 0.4
47 0.326 0.5
48 0.370 0.7
49 0.469 0.8
50 0.727 1.0
Figure 4. 1: Current measurement based on GY-712 module and PV150 meter Table 4.1 and Figure 4.1 shows that the readings obtained from the GY-712 current sensor is much smaller as compared to the current readings acquired from current clamp meter. The current obtained by GY-712 sensor was measured in the condition that the motor was throttling within the same time intervals while the current reading acquired from the PV150 meter was based on the constant moving motor. This contrast may due
31
to the drop in signal timing of the Arduino MCU to capture the output signal. However, both current trends displayed in Figure 4.1 exhibit the analogous increment gradient where the dispersion between the trendlines represents the additional offset existed in GY-712 current sensor with the measurement reading taken from PV150 meter assumed to be the reference values. By adding the extra measurement offset of 0.277 A into the output signal, the explicit electrical offset current of GY-712 current sensor can only be determined, and this offset current calibration will be used for the next objectives.
Table 4. 2: Speed measurement based on RPM Hall sensor and Testo 470 tachometer Servo Throttle Values from A3144 RPM sensor Values from tachometer
46 941 1044
47 1576 1420
48 1948 1780
49 2275 2118
50 2605 2380
Table 4.2 shows that the readings acquired from the A3144 RPM sensor is slightly higher than the readings obtained from tachometer in which the percentage of deviation is less than 10% due to inconsistency in rotor speed. Unlike the GY-712 current sensor, RPM sensor is delivering a steady data which is more reliable as the Interrupt command is applied in the measuring algorithms so that the MCU is able to display the data values detected by the sensor in every 2 μs. The rpm data accessed from the rpm sensor is more preferable as it can be stored in data acquisition system to be helpful in modelling control system.
4.2 Analysis of correlation between measured parameters
A propeller G/F 3 Series Master Airscrew KN1160 was installed on the BLDC motor and the motor was then run using Servo.h library where the motor starts moving when reaching the servo throttle of ‘46’ and every throttles were increased by one for 2
32
s until achieving the maximum servo throttle at ‘135’. The experiments were done with the data collected throughout the operations of the motor to the full throttle, hence the duration for every increment of servo throttle should not be too long to reduce battery consumption and at the same time not being too short so as to get a more average data.
The data retrieved from the motor-propeller performance were exported to Microsoft Excel add-in software PLX-DAQ.to get the direct data from the experiment.
Relationship between Motor Current and Throttle Input
Figure 4. 2: Graph of motor current against throttle percentage
Figure 4.2 shows that the current increases exponentially with the throttle percentage and then fluctuates at a certain level. This indicates that the efficiency of the BLDC motor keeps declining and requires more power input as the throttle increases.
The current is then circulating at a magnitude around 25.0 A beyond the throttle percentage of 56.67% while the motor has reached its maximum speed. The output signal generated from the GY-712 current sensor is varying throughout the experiments as the Hall effect sensor is sensitive to external ferromagnetic field and temperature. Therefore,
33
a proper method of moving average and offsetting was applied on current data in PLX- DAQ to obtain a smooth pattern and more exact current pattern.
Relationship between Propeller Thrust and Rotor Speed
Figure 4. 3: Graph of propeller thrust against rotor speed
Figure 4.3 shows that the thrust generated is concaved upward as the speed increases and the collaboration between these two parameters is literally related to the formula for lift approximation where the in-plane upward force produced by the airfoil is proportional to the square of velocity in which the velocity can occasionally be represented by the frequency of a movement. Besides, Figure 4.3 also points out the performance of this motor-propeller set up literally stalls a maximum work with the thrust of 1.699 kg at the rotor speed of 9394 rev/min.
34 Overall Performance Parameters of Motor-Propeller
Figure 4. 4: Overall performance parameters of BLDC motor with KN1160 propeller Figure 4.4 illustrates that both the speed and its generated thrust rise when the current increases and the performance parameters of the motor-propeller stalls where the thrust generated remains constant at around 1.683 kg when the rising current exceeds 23.62 A. The motor stalling means that designed load of the motor is already overrated and can no longer supply enough torque to accelerate the spinning even if the current level keeps increasing. In addition, Figure 4.4 also proves that the flight of rotor-propeller has a nonlinear thrust-current dynamic which is crucial to be focused on when designing a flight controller, not to mention that there might be other three or more rotors-propellers set up for a multirotor that may hold other trends of performance with slightly deviated from the pattern shown. The experiment is practical to declare the applicable range of operation performance of the rotor-propeller set up when setting the signal output limit for the flight controller such as in PID control system.
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
0 500 1000 1500 2000 2500
0.00 0.47 1.14 2.58 3.66 4.83 4.92 6.83 8.89 11.46 15.28 19.93 23.34 24.73 23.89 22.94 23.17 24.22 25.05 26.13 25.69 25.74 25.33 Speed (rpm)
Thrust (g)
Current (A)
Thrust & Speed vs Current
Thrust Speed
