INTRODUCTION
Motivation
A healthy transportation system and a national economy promote and support each other An efficient transportation system of products and people – both public and personal – has become an important factor in the economic empowerment of a nation Conversely, as the economy grows it translates into increased traffic of goods and people.
According to statistics by 2022 from the Vietnam Association of Motorcycle Manufacturers (VAMM), sales of the whole market have reached more than 3 million vehicles, an average of more than 8,000 units sold per day It can be seen that, with the characteristics of economy, culture and infrastructure in Vietnam, motorbikes are still the most optimal and most popular means of transportation On the other hand, all signals indicate that the trend of vehicle production in the world in general and Vietnam, in particular, is towards environmentally friendly vehicles, using energy sources that can be regenerated One of them is an electric vehicle.
Combining humanistic, economic, social and market backgrounds, we have every reason to believe that electric motorbikes will soon become a dynamic future for Vietnamese people.
However, the issue that many consumers are concerned about is: Are electric scooters safe to use and charge? How long does an electric motorcycle take to fully charge? Is there a way to make this process faster? And many more questions To be able to replace traditional motorcycles, electric motorcycles need to ensure convenience and quickness - which are the strengths of traditional motorbikes To catch up with this changing trend, applying knowledge of automotive technology, especially automotive electronics, we decided to choose the topic of Research and design charging systems for electric motorcycles using pulse transformers as our graduation project.
Research status
The article "Nghiên cứu ảnh hưởng của chế độ làm việc dòng điện liên tục và gián đoạn máy biến áp đối với bộ nguồn chuyển mạch cao tần" by Cao Xuan Tuyen and Nguyen Anh Tuan's team was published in the journal Science and Technology [1] The authors have calculated, tested, and simulated the operation of transformers for high frequency switching power supplies in two working modes: continuous and intermittent current working mode In this study, the Flyback converter was used to control the transformer and applied formulas to calculate the required parameters at three different output power levels with the same input power With the same input power source Vin = 9-15 (VDC), the team surveyed with output = +5 (VDC), = 1.1 (A); = +15 (VDC), = 0.6 (A); = -15(VDC), = 0.6 (A) Simulation of the power supply is performed by the author's team using PSIM software Based on the simulation results in two working modes, both working modes of the transformer are stable Besides, the power supply working in intermittent current mode has better voltage quality than the other mode However, the circuit size working in intermittent current mode is smaller, requires more turns of wire, and in continuous current mode is larger in size with a smaller number of turns.
The research project “120-V, 200-W, 90% Efficiency, Interleaved Flyback for Battery Charging Applications Reference Design” [2] was conducted by researchers from Texas Instrument Corporation Monitor the charger for the battery pack consisting of 5 Lithium-ion batteries This power circuit uses input power Vin = 100-12- (VAC) with frequency 50-60 (Hz) and power at output is = 21 (VDC), = 9.5 (A) The charger adopts a Flyback pulse source type combined with a pulse transformer and produces an output power of up to 200W.
The research topic "Buck-Boost/Flyback Hybrid Converter for Solar Power System Applications" [3] conducted by Sheng-Yu Tseng and Jun-Hao Fan has researched and applied two types of Buck-Boost and Flyback pulse sources into the process Lithium-ion battery charging and discharging process This power supply uses solar energy to provide input power to the Buck-Boost charging circuit with Vin = 17.5-20.6 (VDC), = 8-12
(A) to charge the Lithium battery- ions during the day Then, these batteries will discharge at night and through the Flyback power circuit to generate output power= 10 (VDC), = 2(A) to supply LEDs at night.
Objectives of the research
Research objective: Design charging circuit for electric vehicle using pulse transformer.
- Learn the theory and operation of Flyback pulse sources, electronic components, and simulation software.
- Calculation and design of charging circuit for electric motorcycle.
- Experiment, evaluate the stability of the circuit.
Research methods
Method of document review: Scientific articles from the most reputable sources
(ScienceDirect, Springer, IEEE, Elsevier) were downloaded, read, and analyzed by the research authors to find out the problems of scientists learning ahead has not solved.
Simulation and control: use Proteus simulation software to simulate electrical circuits using STM32CubeIDE software to program microcontrollers.
Experiment and evaluate results: conduct circuit tests with different conditions, based on calculated parameters and experimental results to evaluate the effectiveness and safety of the circuit.
Object and scope of the research
- Pulse Transformer and Pulse Source Flyback.
- MOSFET 11N90 and MOSFET Driver IR2103.
The topic focuses on research on the structure, basic operating principles, and how to apply objects to the design of charging circuits for motorcycles and will not delve into algorithms or in-depth knowledge of the basic electronics field.
Expected result
Design the charging system that can use to charge the 72-battery-connectedLithium-ion The system can adjust the duty cycle base on the feedback signal.
LITERATURE REVIEW
Pulse power supply
2.1.1 Elements used in pulse power supply a The Law of Inductance
The instantaneous voltage drop across an inductor is directly proportional to the rate of change of the current passing through the inductor The mathematical relationship is given by: v = L × ( )
With: L is inductance of the inductor in Henries (H)
( ) instantaneous rate of current change in amperes per second
Important characteristics of the inductor according to the law of inductance:
The voltage across the inductor is valid if and only if the current varies with time.
If a constant DC current is applied to the inductor, there will be no voltage across it (except in the case of a small voltage drop across the copper coil).
In an inductor, the voltage can change instantaneously but not the amperage (this only happens when the voltage value is infinite) The greater the voltage drop across the inductor the faster the current changes.
Below is an illustration of the voltage characteristics of the inductor:
Figure 2 1 Inductor Voltage/Current Relationship [4]
According to the phenomenon of self-inductance, the current through the inductor does not change instantaneously but increases and decreases with time.
There are two modes of inductor current operation:
Continuous operating mode (CO): the amperage in inductor increase from the value to and then decreases to again when the switch is closed, continuously increasing, and decreasing but will never decrease to zero.
Discontinuous Operation mode (DO): when the switch is closed, the amperage in the inductor increases from 0 to and the amperage returns to 0 when the switch is open. c Transformer in pulse power supply
The main element used in this project is the pulse transformer, unlike the common transformer, the coil in the pulse transformer also acts as an inductor, and each transformer usually has two or more pairs of magnetic coils Below is a schematic diagram of the transformer:
For a transformer, the input and output voltages depend on the number of turns of the winding The coil with the higher number of turns will have higher voltage but lower current and vice versa The dot on the transformer determines its polarity relative to the other winding, reversing the dot leads to reversing polarity and output current. d Pulse Width Modulation (PMW)
Pulse width modulation is a method of adjusting the output voltage The output voltage depends on the width of the square pulse sequence in one cycle And this voltage is controlled directly by turning on/off a switch in the converter Output DC voltage is calculated as shown in the following figure:
Figure 2 3 Basic Principle of PWM [4]
A DC output voltage that is equal to the peak pulse amplitude multiplied, , times the duty cycle (the duty cycle is defined as the switch ON time divided by the total period ).
The working principle of Flyback circuit:
The pulse transformer circuit uses a Flyback circuit that acts as an inductor or a device that accumulate magnetic energy.
Figure 2 4 Basic Flyback Converter Schematic [5]
While 1 is closed and directs current to the primary coil, which stores energy in the pulse transformer 1 acts like a pure inductor and primary current builds up linearly in it to a peak On the other hand, the secondary coil diode 1 will not conduct and so there will be no current in the secondary coil During this time, the output load current will be supplied only by the output filter capacitor 1
While 1 is open, all winding voltages reverse under flyback action, bringing the output diodes into conduction and the primary stored energy 1 2 2 is delivered to the output to supply load current and replenish the charge on the output capacitors (the charge that they lost when 1 was closed) This current not only supplies the load but also charges the capacitor 1 , as a way to recharge the capacitor to be able to supply the load again in the closed 1 phase.
Figure 2 5 Current and Voltage in primary & secondary coil [5]
One thing to keep in mind when applying the Flyback circuit is that the voltage in the pulse transformer will exist between the primary and secondary windings even though they do not conduct at the same time When the 1 is OFF, the input voltage combines with the reverse voltage due to the secondary coil acting back Therefore, the 1 is required to withstand the magnitude of these two voltages The voltage is calculated according to the following formula:
And the maximum voltage on Q1 is defined as:
Flyback converters operate in either continuous mode, CCM, (where the secondary current is always larger than 0) or discontinuous mode, DCM, (where the secondary current falls to zero on each cycle) The circuit is discontinuous if the secondary current has decayed to zero before the start of the next turn “ON” period of 1
Another difference between DCM and CCM is that in DCM, the current in the secondary coil will drop to 0 (point ) before the current in the primary begins the next conduction cycle (point ) This period is called the dead time The average output current at the secondary is the average of the triangle multiplied by the interrupt ratio of the 1 lock,
The DCM will give the output diode 1 in the secondary better switching because the current through the diode will return to 0 before its reverse bias, and the energy stored in the transformer is also less which helps to reduce the size of the pressurized machine. its output current will be lower than that of the CCM Therefore, the 8 discontinuous current mode is often applied in circuits with high voltage and low current output.
The change of current in the DCM is shown as shown below:
Figure 2 8 Waveforms of a discontinuous-mode flyback [5]
The current of the CCM will be shown in the following figure:
Figure 2 9 Waveforms of a continuous-mode flyback [5]
It can be seen that CCM and DCM have relatively similar current forms Then operating mode is determined by the magnetized inductance and output load current The main difference between the current form of the two types is the current storage in the two windings because there is no dead time In CCM, the current in the primary coil has a step, and its slope increases from that point During the 1 off time, the current in the secondary coil takes the form of a combination of a triangle plus a stepping of magnitude from point to And similarly, when the key 1 starts the closing phase, the current in the primary coil also has a buffer current with magnitude from point to
Although it is assumed that the output currents in the two modes are equal, the current in DCM gives a higher peak current than in CCM Therefore, a circuit operating in DCM mode requires a large filter to remove this current ripple The continuous current mode, CCM, is applied in circuits that require low voltage and high current output.
Lithium-Ion Battery
Lithium-ion batteries are devices that generate DC electricity through chemical reactions When the battery discharges or charges, lithium ions move back and forth between the plates (anode and cathode) Usually, the cathode plate is made of metal oxide material with the base metal being Cobalt, Nickel, or Manganese The anode plate is made of graphite Both the positive and negative plates are structured as thin layers. Lithium ions are in those layers During charging, lithium ions move from the cathode to the anode During discharge, lithium ions move from the anode to the cathode The electrolyte is an important part of a Lithium battery, providing a conductive medium for Li+ ions between the plates but also being a non-conductive solvent [7]
2.2.1 Structure of Lithium-ion Battery 18650
Figure 2 10 Structure of Lithium-ion battery [8]
The structure of a lithium-ion battery consists of 4 main parts: cathode (positive electrode), anode (negative electrode), separator and electrolyte.
The negative electrode of a typical lithium-ion battery is typically comprised of carbon-based graphite A metal oxide is frequently used as the positive electrode A lithium salt in an organic solvent serves as the electrolyte A separator prevents the anode (negative
10 electrode) and cathode (positive electrode) from shorting A metal component known as a current collector isolates the anode and cathode from external electronics Depending on how current flows through the cell, the anode and cathode electrodes have different electrochemical roles.
Table 2 1 Main specifications of 18650 Lithium-ion batteries
Capacity Rated capacity: 2600 mAh (0,52A discharge; 2,75V)
Ambient temperature Charging: 0 - 45 o C, Discharging: -20 - 60 o C a Cathode
Typically, cathode materials are made of LiCoO2 or LiMn2O4 A pseudo-tetrahedral structure that supports two-dimensional lithium-ion diffusion emerges in the cobalt-based material Because of their high theoretical specific (per-mass) charge capacity, high volumetric capacity, low self-discharge, high discharge voltage, and strong cycle performance, cobalt-based cathodes are the best choice The material's high price and poor thermal stability are limitations The cubic crystal lattice system used by the manganese- based materials enables three-dimensional lithium-ion diffusion Because manganese is less expensive than other metals, manganese cathodes are appealing If its drawbacks could be resolved, they might be used to create a battery that is more effective and lasts longer Some restrictions include the cathode's poor cycling stability caused by manganese's propensity to dissolve into the electrolyte during cycling Although cobalt-based cathodes are the most popular, additional materials are being investigated in an effort to reduce prices and prolong cell life [7] b Anode
Although newer silicon-based materials are increasingly being employed, negative electrode materials are typically made of graphite and other carbon-based materials 89% of lithium-ion batteries in 2016 contained graphite (43% synthetic and 46% natural), 7% amorphous carbon (either soft carbon or hard carbon), 2% lithium titanate (LTO), and 2% materials made of silicon or tin-based These substances are utilized due to their availability, electrical conductivity, and ability to intercalate lithium ions to retain electrical charge with only a 10% volume expansion Due to its superior performance and low intercalation voltage, graphite is the predominant material There have been several suggested alternate materials with higher capacities, but they typically have higher voltages, which lowers energy density Anodes must operate at low voltage to maximize energy density; otherwise, the excess capacity is useless [7] c Separator
In order to prevent battery self-discharge and short circuit issues between the two poles, a microporous film consisting of polypropylene (PP), polyethylene (PE), and other plastics is sandwiched between the positive and negative plates Lithium ions can pass through the dense micropores in the separator, enabling the battery to create a full charge and discharge circuit [9] d Electrolyte
The common primary elements of lithium-ion battery electrolytes, such as ethylene carbonate (EC), dimethyl carbonate (DMC), etc., play a crucial role in the performance of lithium-ion batteries as a medium for the transfer of lithium ions between the positive and negative electrodes You can start by improving the electrolyte formulation and electrolyte additives if you want to increase battery cycle life, safety, and lithium-ion transmission characteristics The performance of lithium-ion batteries can be improved with the right electrolyte [9]
2.2.2 Working principle of Lithium-ion battery
The materials in the two electrodes are the reactants of the electrochemical reaction in the Lithium-ion battery; Lithium ions move between two electrodes in an electrolyte solution During the reaction, lithium ions move in the two electrodes. Existing electrode materials can allow lithium ions to enter or leave the lattice, with little or no disturbance of the remaining atoms in the lattice.
During charging, Lithium ions are released from the positive electrode At this time, the Lithium ion uses the electrolyte as the transmission medium, penetrates through the separator and attaches to the anode plate.
Figure 2 11 Lithium-ion battery charging process [9]
Figure 2 12 Lithium-ion battery discharging process [9]
During discharge, Lithium ions are released from the cathode plate Lithium ions use the electrolyte as a conducting medium, penetrating through the separator and returning to the cathode plate.
The following equations exemplify chemistry.
The positive electrode (cathode) half-reaction in the lithium-doped cobalt oxide substrate is:
CoO 2 + Li + + e - ↔ LiCoO 2 (left to right: discharging, right to left: charging) The negative electrode (anode) half-reaction for the graphite is:
LiC 6 ↔ C 6 + Li + + e - (left to right: discharging, right to left: charging)
The full reaction for charging and discharging is:
LiC 6 + CoO 2 ↔ C 6 + LiCoO 2 (left to right: discharging, right to left: charging)
The overall reaction has its limits During discharge, C 6 1- (Cathode) is oxidized to
C 6 0 , Co 4+ is reduced to Co 3+ , and vice versa during charging If a Lithium-Ion battery is over-discharged, a saturated Lithium Cobalt Oxide will convert to Lithium Oxide, in one direction of the following reaction:
If the LCO battery is overcharged with a voltage above 5.2 V, it will convert to Cobalt IV Oxide, in one direction of the following reaction:
2.2.3 Charging and discharging characteristics of lithium-ion battery
A Lithium-ion battery cell has a 2-stage charging process:
Stage 1: Constant current (CC) charging
During constant current (CC) charging, the current is kept constant, usually equal to C/2-C (where C is the capacity [Ah] of the battery) The larger the charging current, the shorter the stable charging process, but the longer the voltage stabilizer charging process; However, the total charging time for both phases is usually no more than 3 hours At the same time, a large current will increase the temperature of the battery. During the charging process, it is necessary to monitor the temperature closely because too high a temperature can cause the battery to ignite or explode.
Normally, the temperature should not exceed 45ºC Some Li-ion batteries use Lithium-Ferro-Phosphate (LiFePO4) technology that can push the charging temperature up to 60ºC Using a fast charger (quick charge) only pumps a steady current into the battery (constant current charging), so a larger temperature limit means a larger charging current or a faster charging time.
Figure 2 13 Charging characteristic of lithium-ion battery [10]
Stage 2: Constant voltage (CV) charging
During constant voltage (CV) charging, the battery charging voltage is usually kept at 4.2V/cell constant As the capacity is nearly restored, its electromotive force increases, reducing the current When the current drops below 3% C, the voltage stabilization phase ends At this point, the battery capacity has been restored to about 99% [10]
Li-ion cannot absorb overcharge The charge current needs to stop right after the battery is fully charged A constant trickle charge would endanger safety and plate the metallic lithium Keep the lithium-ion battery's peak cut-off as brief as feasible to reduce stress.
Some chargers offer a brief topping charge to lithium-ion batteries when they need to be left in the charger for operational readiness in order to make up for the minimal self-discharge the battery and its protection circuit incur When the open circuit voltage reaches 4.05V/cell, the charger may start working and then stop at 4.20V/cell The battery voltage is frequently allowed to drop to 4.00V/cell on chargers designed for operational readiness or standby mode, and they recharge to only 4.05V/cell as opposed to the full 4.20V/cell This lessens stress caused by voltage and increases battery life [11]
Electronic components used in the project
2.3.1 MOSFET a Structure and working principle of MOSFET
MOSFET (Metal-Oxide Semiconductor Field-Effect Transistor) is a semiconductor metal oxide field-effect transistor MOSFETs work on the principle of magnetic field effects on metal oxide and semiconductor junctions MOSFETs can switch quickly around a few MHz, so they are often used in pulse source circuits and high voltage control circuits.
Currently, MOSFETs are divided into two main types: N-MOSFETs and P- MOSFETs And each is divided into enhancement (enhancement) and decline (depletion) MOSFET has 3 terminals G (Gate), D (Drain), S (Source) In which, the G-S pole is mounted in the control circuit, and the D-S pole is mounted in the power circuit.
Depletion Type – To turn the device "OFF," the transistor needs the Gate-Source voltage (V GS ) The depletion mode MOSFET is equivalent to a “Normally Closed” switch Enhancement Type –To switch the device “ON”, the transistor needs a Gate- Source voltage, (V GS ) The enhancement mode MOSFET is equivalent to a “Normally Open” switch [13]
The following diagram illustrates the symbols and fundamental construction for both MOSFET designs.
In this project, the Enhancement-mode N-Channel MOSFETs are used in the control circuit and the feedback circuit. b Enhancement-mode N-Channel MOSFETs
Enhancement-mode MOSFET or eMOSFET, is the reverse of the depletion-mode type Here, the conducting channel is either undoped or only mildly doped, rendering it nonconductive As a result, when the gate bias voltage V GS is equal to zero, the device is generally "OFF" (not conducting) An enhancement MOS transistor's circuit symbol employs a broken channel line to represent a normally open, non-conducting channel.
The n-channel enhancement MOS transistor is a transconductance device because a drain current will only flow through it when a gate voltage (V GS ) is applied to the gate terminal that is larger than the voltage level at which conductance occurs ( ℎ ).
The channel resistance will decrease even further when the positive gate voltage is raised, increasing the drain current or ID, through the channel To put it another way, a zero or - V GS switches an n-channel enhancement mode MOSFET "OFF," whereas a +V GS turns it "ON." A "normally open" switch is what the enhancement-mode MOSFET is comparable to [13]
Figure 2 17 V-I characteristic of enhancement-mode MOSFET [13]
The V-I characteristic of enhancement-mode MOSFET is divided into 3 main regions:
1 Cut-off Region – with V GS < ℎ the gate-source voltage is much lower than the transistors threshold voltage, so the MOSFET transistor is switched “fully-OFF” thus, I D = 0, with the transistor acting like an open switch regardless of the value of V DS
2 Linear (Ohmic) Region – with V GS > ℎ and V DS < V GS the transistor is in its constant resistance region behaving as a voltage-controlled resistance whose resistive value is determined by the gate voltage, V GS level.
3 Saturation Region – with V GS > ℎ and V DS > V GS the transistor is in its constant current region and is therefore “fully- ON” The Drain current I D = Maximum with the transistor acting as a closed switch [13]
IC (Integrated Circuit) is an integrated circuit consisting of many semiconductors and passive components They are combined to perform a predetermined task.
IR2103 is a high voltage and fast processing MOSFET or IGBT driver IC with dependent low and high side output channels The logic input corresponds to CMOS orLSTTL standard output, the logic voltage level is 3.3V The output drivers have a high side pulse current buffer stage designed to minimize driver cross-conduction The output channel can be used to drive an N-channel power MOSFET or IGBT in high-side configuration operating up to 600 volts.
HIN Logic input for high side gate driver output (HO), in phase ̅̅̅̅̅ Logic input for low side gate driver output (LO), in phase
HO High side gate drive output
High side floating supply return Low side and logic fixed supply
LO Low side gate drive output
- V CC : Low side and logic fixed supply voltage: 10V - 20V
- V S : High side floating supply offset voltage: -5V - 600V
- V B : High side floating supply absolute voltage: V S +10V - V S +20V
- V HO : High side floating output voltage: V S - V B
- V LO : Low side output voltage: 0 - V CC
- V IN : Logic input voltage (HIN & LIN): 0 - V CC
Figure 2 19 IR2103 input and output control pulse [14]
At the HO pin, when the voltage at the HIN and LIN pins is HIGH, the voltage at the HO pin is HIGH, when the voltage at the HIN and LIN pins is LOW or different, the voltage at the HO pin is LOW.
At the LO pin, when the voltage at the HIN and LIN pins is LOW, the voltage at the LO pin is HIGH, when the voltage at the HIN and LIN pins is HIGH or different, the voltage at the LO pin is LOW.
In addition, IR2103 has an internal deadtime to prevent cross transmission; this function is applied to control MOSFET or IGBT in Full-Bridge circuits.
In this project, the input signal at the HIN pin of the IR2103 is the pulse output from the A0 pin of the STM32 microcontroller In order to prevent the microcontroller and the circuit from the reverse current, an opto HCPL2631 is applied between STM32 and IR2103 to be the intermediary for transferring the pulse and isolate the microcontroller from the reverse current if any.
Optocoupler (opto) is a device used to convert electrical signals to light and then transmit The main advantage of opto is the way voltage is between input and output circuits The only contact between the input and output at the opto is a beam of light The isolation resistance between the two circuits is up to thousands of MΩ Used in high voltage circuits and the voltages of the two circuits can be several times different.Optocoupler is also called Opto-isolator, photo coupler or optical isolator. a Structure
The optocoupler is made of a light emitter and a light detector.
Light emitter: This device receives electrical signals from the input side and converts them into light signals A light emitting diode is commonly used as the light emitter (LED).
Light detector: In an optocoupler, the light detector is used to transform the incoming light signals from the light emitter into the original electrical signals The light detector could be a phototransistor, photodarlington, photodiode, etc. b Working
STM32CubeIDE
STM32CubeIDE is an all-in-one multi-OS development tool, which is part of the STM32Cube software ecosystem.
STM32CubeIDE integrates STM32 configuration and project creation functionalities from STM32CubeMX to offer an all-in-one tool experience and save installation and development time After the selection of an empty STM32 MCU or MPU, or preconfigured microcontroller or microprocessor from the selection of a board or the selection of an example, the project is created, and initialization code generated At any time during the development, the user can return to the initialization and configuration of the peripherals or middleware and regenerate the initialization code with no impact on the user code.
STM32CubeIDE includes build and stack analyzers that provide the user with useful information about project status and memory requirements STM32CubeIDE also includes standard and advanced debugging features including views of CPU core registers, memories, and peripheral registers, as well as a live variable watch, Serial Wire Viewer interface, or fault analyzer.
After starting the software, the initial interface of the software is as follows:
Figure 2 35 Initial startup window of STM32CubeIDE
At this startup window, depending on the purpose of use, the user can choose the corresponding shortcut In this topic, the microcontroller used is STM32F103C8, we will start with selecting the shortcut Start new STM32 project.
Figure 2 36 MCU/MPU Selector window
Enter the microcontroller name in the Commercial Part Number box, select
STM32F103C8T6 and click Next to jumping to naming step After naming the Project, click Finish to complete the STM32 board selection process and start the programming process.
Figure 2 37 STM32CubeIDE naming window
The pins are selected as follows:
- PD0 and PD1: External crystal/ceramic resonator
- PA13 and PA14: Used when loading the program without pressing the reset button to run the program
We choose PWM Generation for pin PA0:
Figure 2 39 PA0 Mode and Configuration
Select the parameters Prescaler (PSC), and Counter Period (ARR) - the maximum value of the counter to obtain the desired PWM frequency In this project, the PWM frequency is 100KHz and the frequency is calculated following the below formula:
With is timer clock, we choose the value 72
Then, we choose PSC = 0 and ARR = 719
Figure 2 41 PWM generation mode at PA0
STM32F411 supports 2 PWM modes as follows:
- Mode 1: If using the count-up mode, the output will be at logic 1 when CNT <
CCR and vice versa, at 0 if CNT > CCR.
- Mode 2: If counting up mode is used, the output will be at logic 0 when CNT
CCR.
After selecting the mode for the control pins, we configure the processing time and frequency of the microcontroller.
Finish setting up the pins and related parameters, click Project > Generate Code, or use Alt + K shortcut to open the programming workspace, built into STM32CubeIDE
After clicking on Generate Code, a window with the name main.c will pop up, we will enter our code in this window.
Figure 2 43 Coding window of STM32CubeIDE
The interface of the code window consists of 3 main areas:
- Zone 1: Users will enter the code here
- Zone 2: is the location of the project window including imported projects, here users can choose main.c to program the code in Zone 1 or select the ioc file to reset the pins Users can also manage libraries, subroutines, and files in the project.
- Zone 3: This is the display area of the program compilation process In addition, this area also shows us the number of errors and warning in Problems tab of the project.
The microcontroller will receive the signal from pin PA4 If the voltage at PA4 > 0, will output a PWM pulse at pin PA0 with a Duty cycle of 20% This signal is the input of the IR2103 that drives the MOSFET to operate at 100 If voltage at PA4 = 0, Duty cycle will be 10%.
CALCULATION AND DESIGN
System description
On electric motorcycles, the vehicle's power source is from a Lithium-ion battery pack that generates 288V DC power This battery pack is made up of 72 batteries connected in series Therefore, to charge this battery pack needs to meet two requirements for voltage and amperage In addition, each battery has a maximum voltage of 4.2V with 72 batteries in series, so the voltage from the charger must be greater than 302 and the amperage is in the range of 0.32A Below is a Lithium-ion battery charger circuit using a pulse transformer.
Figure 3 1 Lithium-ion battery charger circuit using pulse transformer
Lithium-ion battery charger circuit has an input power source of 220V -50Hz AC and is rectified to 311V DC Then the voltage 311V is passed through the pulse transformer, controlled by the control circuit, creating a current of 320V - 0.32A to charge the Lithium-ion battery pack.
This is a Flyback type isolating load circuit The circuit is isolated by a pulse transformer, so in the event of a power failure, the battery pack will not be affected.
Operating principle of the system
The input power of the charging circuit is AC 220V -50Hz , so this power will be fed into the full-wave rectifier bridge circuit to rectify it into direct current However, the output DC voltage has an undulating waveform, which is not stable for the system.
Therefore, a capacitor will be connected at the output of the rectifier bridge to smooth this voltage.
A snubber circuit is a device that protects an electrical circuit from residual voltage spikes and oscillation effects When the MOSFET disconnects continuously, the voltage at the Drain terminal increases, which can cause the MOSFET to burn [22] And the snubber circuit reduces the voltage spike without changing the frequency of the main circuit.
The circuit used in this study, which is a Flyback type is Resistor Capacitor Diode (RCD) snubber circuit including a resistor, diode, and capacitor connected to the circuit as shown below:
Figure 3 5 Drain pole voltage without snubber circuit [24]
Figure 3 6 Drain pole voltage with snubber circuit [24]
The switching speed between the two on and off states of the MOSFET can be controlled by the on and off trigger resistors The switching speed and closing and interrupting times can be obtained by the values of the resistors of the two control gates, and , respectively Below is the structure of an N-channel MOSFET.
In the structure of the MOSFET, there are parasitic capacitances that affect the closing and interrupting process of the MOSFET Therefore, we must consider these capacitances.
Figure 3 7 Parasitic capacitances in MOSFET [25]
The MOSFET activation process consists of three main phases:
Figure 3 8 VGS as a function of gate charge [25]
The process from 0 to 1 : at time 0 , pole begins to be powered and the voltage starts to increase from 0 At this time, most of the current through the terminal is charged to the capacitor And there is also a small amount of charge current through the capacitor, but this capacitor has a smaller capacitance value than the capacitor, so this can be considered the charging period for the CGS capacitor This stage is also known as , because both the current and voltage through the source remain unchanged and the MOSFET remains in the off state.
The process from 1 to 2 : this is the MOSFET stage that is almost fully conducted.
At this point, the voltage increases very slowly or not even at all, and the voltage increases rapidly.
2 to 3 : MOSFET completes the excitation cycle at this stage Capacitors and are loaded and increases to the final point.
The MOSFET interrupting process is the reverse of the triggering process. a ON resistance
As shown above, the current through the terminal G and the voltage does not have an equation, so it is not possible to calculate the resistance accurately In fact, there are other ways to calculate these values, but IR Rectifier manufacturer has come up with a simple but highly effective method.
Let be the average excitation current, is the switching time from the beginning of the excitation process to when the MOSFET is completely closed, corresponding to the time interval from 1 to 3 in the analysis above.
With is the average voltage for the period 2 to 3 which was provided by the manufacturer in the datasheet.
One note is that is large or small depending on the excitation current, the smaller the , the faster the switching time and the lower the loss on the component Therefore, is usually chosen according to design criteria and suitable for carrier frequency The optimal switching time when triggered by the driver IC and is usually chosen is:
(3.5)Thus, the value of ON resistance has been defined. b OFF resistance
With driver ICs, they are provided with separate trigger and interrupt pins Then the trigger resistor is selected with a smaller value than the interrupt resistor because the triggering of the interrupt occurs faster, helping to reduce Dead Time.
But we need < ℎ of the MOSFET, and then we have:
To implement an excitation circuit, the selection of the on and off resistor values needs to be calculated carefully because they affect the performance of the circuit and avoid damage to other components.
When using a MOSFET to control a circuit, we will usually have two basic ways as shown below:
These two types of MOSFET excitation are distinguished from each other in the position before or after the load With the high-side excitation circuit, the MOSFET will be connected to the high-voltage source and the consumption load to the ground.
Figure 3 11 N-MOSFET driver circuit high (left) and low side (right)
In contrast, with the low-side excitation circuit, the MOSFET will be connected between the load and ground.
In the high-side driver circuit, for closing the MOSFET, must be higher than (with a 20N60 MOSFET, = 5V) When the MOSFET is closed, = 0, or the whole voltage will fall on the load, which means ≈ = 310V In addition, = − , and has a value of 7.5V < 310V of , so the MOSFET will not open.
In the low-side trigger circuit, when the MOSFET is closed, ≈ 0V Pole S is grounded, so ≈ = 0 And the entire voltage = 310 V will fall on the load and not affect Therefore, as long as > , we can control the MOSFET even though is much larger.
Therefore, controlling the high-side excitation circuit is much more complicated than the low-side triggering circuit With the input source of the pulse transformer being310V, we need a voltage greater than 310V to be able to trigger the MOSFET, for example, to create an isolation voltage source or a Bootstrap circuit Therefore, our team will choose to control the low-side MOSFET to simplify the control process.
Calculation and design circuit elements
In this MOSFET trigger circuit, the trigger resistor in two cases needs a small power of 0.25W, but to ensure the safety of the circuit, this semiconductor trigger resistor will be selected as a 2 resistor.
High side floating supply absolute
High side floating supply offset - - 600 voltage
High side floating output voltage -
Low side and logic fixed supply 10 - 20 voltage
Low side output voltage 0 - V CC ̅̅̅̅̅ 0 - V CC
Logic input voltage (HIN & LIN)
Gate - Source voltage enough to (min) 5 - - open the gate
Drain – Source diode forward - - 1.5 voltage
Peak diode recovery voltage slope - 20 -
We choose the gate resistor = 1 Ω)
In Vietnam, residential electricity is single-phase alternating current with a value in the range of 85-220V with a frequency of 50 Therefore, we will have = 85V and = 220V.
We will use the rectified DC current by the diode bridge, so VDC will be calculated according to the following formula:
The output required to charge the Lithium-ion battery is 320V - 0.3A, so the output power will be:
Assuming the efficiency of the circuit is = 70%, we will determine the input power:
Besides, the voltage of the alternating current will also be flattened by the effect of the capacitor, with two factors and ℎ Where:
: The value of the input capacitor per Wattage of input power With the AC voltage range defined above, = 2 – 3 We will choose = 3
ℎ: charge ratio of input capacitor This ratio is referred to as in the following figure:
Figure 3 12 Waveform graph of DC voltage after input filter capacitor [27]
Table 3 3 Input factors and output goals
Because the output requires high voltage and low current, we will choose the discontinuous current mode for the calculation of this pulse transformer.
First, we will choose the voltage ratio as well as the turn ratio of the pulse transformer.
As we have selected above, the 20N60 MOSFET has a Power-Source voltage rating of 600V To meet the working requirements and be able to withstand spikes or voltage noise during working, choose the maximum stress on the transistor in the “off” state (excluding the leakage inductance spike):
Where is the voltage drop of the output rectifier diode In the circuit using general- purpose rectifier diode type 1N4007 with = 1.1V.
With as 440 V Then even with a 25% or 110V leakage spike, this leaves a 50V margin to the maximum voltage rating We have:
Thus, = 126 and we can choose the maximum “on” time by the formula:
Next, we will calculate the primary inductance :
Besides that, we can determine the value of peak current :
The primary RMS current is calculated by:
And the secondary RMS current is:
Next, determine the number of turns of the primary winding of the pulse transformer to prevent core saturation, will be calculated by the formula:
: maximum input DC voltage through the primary wire (V)
: cross-sectional area of the pulse transformer core ( 2 ) : saturation flux density (T)
For = 13.51 turns, choose larger than this to ensure the pulse transformer is working properly and leave the rest to wind the secondary winding We choose = 15 rounds.
Then, the number of turns of the secondary winding is:
The ferrite core of the pulse transformer must have a clearance to prevent premature core saturation. The clearance length is calculated according to the following formula:
: is the inductance of the core According to the manufacturer's specifications, for EE42 type ferrite cores, the value will be equal to 1029
: is the inductance of the primary coil ( )
: is the number of turns of the primary coil (turns)
: cross-sectional area of ferrite core ( 2 )
Obtaining the core clearance G length is:
According to the manufacturer's specifications, the G clearance parameter of the pulse transformer EE42 is 0.25 mm > 0.0755 mm So, this pulse transformer meets the requirements of the circuit.
Calculate the output filter capacitor 1 of the circuit according to the circuit's highest output current of 0.5A and a voltage drop across the capacitor of 0.05V Capacitance of capacitor 1 is calculated according to the formula:
Select output filter capacitor 1 is 100 - 450V capacitor.
There is a table of statistics as follows:
The voltage in the primary coil is caused by the input voltage and reverse voltage on the secondary coil And this voltage is calculated by the formula:
The leakage inductance on the primary coil is: = 0.1 × = 6.26 ( )
Peak current in primary coil: = 7.99 ( )
, clamp voltage is the safe voltage for the component to operate The lower the clamping voltage during operation, the more protected the component is The is calculated from the 20N60 MOSFET's VDS with a safety margin of 90%.
Choosing the maximum switching frequency: = 100000 ( )
In the Snubber circuit, the capacitor is calculated according to the following formula:
Then, we choose = 22 and the rated voltage = 630
And the diode needed for the spike-suppressing circuit will be a fast recovery diode. Therefore, diode CBB22 will be used.
In order to optimize the performance of the charging circuit, the feedback signal is one of the most important signals, allowing the microcontroller to adjust the pulse width by increasing or decreasing the duty cycle to give the output voltage and current suitable for operating conditions of the charging circuit.
The principle of the feedback circuit is that, at first, there will be a voltage of 5V put on pin 4 of the optocoupler and is connected in series with pin PA4 of the microcontroller, when the output voltage is enough to pass through the Zener diode, it will activate the MOSFET and then allow the feedback signal to turn on the optocoupler.
At this time, the voltage of 5V put on pin 4 of the optocoupler will flow through pin 3 to ground, which causes a voltage drop on pin PA4 from 5V to 0V This voltage drop is considered as the feedback signal for the microcontroller to reduce the duty cycle.
The 1N4740A Zener diode combined with a voltage divider bridge is used to facilitate the feedback of the circuit, returning the signal to the microcontroller.
With a breakdown voltage of 10V, we adjust the input voltage to the diode with a voltage divider circuit.
To achieve the breakdown voltage of 1N4740A Zener diode, must be equal to or higher than 10V, much more lower than the expected output voltage of the charging circuit is ( ) = = 311 , hence the 1 and 2 are chosen to reduce the output voltage of the voltage divider to meet the requirements as below:
Due to the voltage drop on Zener diode and the circuit, the calculated voltage must be higher than the breakdown voltage, ensuring that the feedback signal can turn on the MOSFET, which needs a voltage at around 5V-12V for operating.
After achieving the breakdown voltage of Zener diode, the signal then activates the MOSFET in the feedback circuit, which allows the feedback current to flow through pin 1 and 2 of optocoupler to ground At this time, the optocoupler will be activated at conduction state As mentioned above, the voltage of 5V put on pin 4 of the optocoupler will flow through pin 3 to ground and result in a voltage drop on pin PA4 of the microcontroller Once the voltage put on pin PA4 equal to 0V, the duty cycle will immediately be decreased to the initial set level by programming.
EXPERIMENT RESULTS AND DISCUSSION
After calculating the components of the circuit, our group soldered the components according to the circuit diagram of Figure 3.1.
For operating the circuit, there must be 3 separate power sources as follows:
- 220V AC as the power supply for the charging circuit.
- 12V DC generated to 5V DC as the power supply for the microcontroller and put on pin 4 of optocoupler in the feedback block.
- 12V DC as the power supply for the MOSFET driver IR2103 and generated to 5V for supplying power for optocoupler HCPL2631.
The purpose is to isolate the power supply of the microcontroller and other components Preventing them from shorting or negative effects may occur.
When tested for safety, the pulse width will be increased from 0 to the value at which the output voltage is targeted and can be met by the system.
In this charging circuit, the maximum duty cycle must be less than 20% to ensure stable operation, avoid damage due to overheating or over-capacity.
Figure 4 2 Output PWM on pin A0
The control of the charging circuit using pulse transformer is a chain of signal rotation, from the output of the microcontroller, through the opto and MOSFET driver to trigger the MOSFET switch Requiring high continuity and accuracy during operation, just a small mistake can cause huge damage to components in the circuit.
Figure 4 3 Output signal on pin 8 of optocoupler HCPL2631
Figure 4 4 Output signal on pin 5 of IR2103 to trigger MOSFET
In this project, the experiment will be conducted in 2 phases:
- Phase 1: Tested with 220V AC At this stage, our group will examine the capacity of the charging circuit A 220V-60W incandescent bulb is connected with the output as load.
- Phase 2: Tested for a long operating time After surveying the capacity at phase
1, the charging circuit is tested for a longer time to evaluate the stability of the whole circuit Within this phase, the battery packs are connected alternately to collect the charging capability of the charging circuit Currently, there are 2 battery packs available at the laboratory, the first one consists of 60 lithium-ion
18650 cells, the other consists of 72 lithium-ion 18650 cells.
4.2 Tested with 220V AC, incandescent bulb as load
Under no-load test, the output of the transformer gives a voltage of 320V.
Figure 4 5 Output voltage with no load
Monitoring the temperature of the MOSFET is also a very necessary task in the process of optimizing the performance of the charging circuit, by installing an aluminum heatsink for the MOSFET, the operating temperature is maintained below 50°C, make sure this component is not damaged during operation.
Figure 4 6 Temperature of MOSFET with no load on the circuit
By connecting an incandescent bulb as load, the output voltage and current of the charging circuit are collected as below:
Figure 4 7 Test charging circuit with 220V-60W with filament bulb
4.3 Tested with 220V AC, battery pack as load
By connecting a battery pack of 60 cells, the initial voltage of the battery pack is212V and the output voltage and current of the charging circuit after connected to the battery pack are collected as below:
Figure 4 8 Initial voltage output of the 60-cell battery
The output voltage of the charging circuit dropped to 220V, nearly the same with the voltage of the battery pack, and the charging circuit fluctuated around 1.1-1.2A
Figure 4 9 Output voltage and current of charging circuit when connect to the battery pack
At this stage of testing, to check the capability of the charging circuit, the test was extended continuously for 3 hours After 3 hours charging, the battery pack was fully charged by the charging circuit and gained the output voltage of 254V.
Figure 4 10 Output voltage of the battery pack after charging
Table 4 1 60-cell battery pack charging result
Charging time Charging voltage Charging current Battery voltage (V)
Charging voltage (V) Battery voltage (V) Charging current (A)
Figure 4 11 Graph of charging result by time
Monitoring the MOSFET while charging, the temperature remained around 45- 46°C during the charging period This temperature maintenance is suitable for operation of the charging circuit, MOSFET would not be damaged due to high temperature.
Figure 4 12 Temperature of MOSFET during charging
CONCLUSION AND RECOMMENDATION
During the period of implementing the project of designing charging circuits for electric vehicles using pulse transformers, our group has implemented and completed the following contents:
Learning about the specifications of Lithium-ion batteries and battery packs available at the lab From there, determine the objectives, research objects, and delineate the research scope and development direction of the topic.
Searching the documentation and learning the influence and parameters of electronic components such as MOSFET and MOSFET Driver IR2103 Based on those documents to conduct circuit creation and testing.
Comparing the difference between power and operation of pulse sources and make plans based on Flyback pulse sources.
Learning how to calculate the required material parameters and how to run the pulse transformer.
Gaining knowledge on programming and compilation environment of STM32CubeIDE software to program input signals and pulse width and PWM frequency for pulse transformer control.
Experimenting with wiring and electronic components while making actual circuits and correcting errors after failures to achieve results that approximate the original purpose.
In the process of implementing the project, although our group has output the voltage and amperage suitable for charging the battery pack, there are still many shortcomings that need to be overcome Although the initial goal has been set, the actual results of the circuit have large errors and are not stable Therefore, the subject should be improved performance In terms of operation, the charging circuit has been able to operate for a long time, specifically, the circuit was able to charge the battery pack continuously for more than 3 hours, but because of shortcomings in accurately calculating the capacity, the output of the charging circuit is not in the best state Besides that, the battery pack available at the laboratory does not achieve the best operating conditions, some battery cells may be damaged during operation, resulting in the charging circuit not being able to be tested for output power in the most complete way, so our group has not been able to accurately test the capacity of the charging circuit.
In addition, the circuit needs to be tested more through charging the battery pack to be able to assess the capacity of the designed charging and feedback circuit most accurately The next direction is to calculate more accurately and optimize the stability of the output
70 power, laying the foundation for designing a balanced battery charger circuit in the near future.
[1] C X Tuyển and N A Tuấn, "Nghiên cứu ảnh hưởng của chế độ làm việc dòng điện liên tục và gián đoạn máy biến áp đối với bộ nguồn chuyển mạch cao tần," Tạp chí Khoa học & Công nghệ, vol 74, no 12, p 115, 2010.
[2] T Instrument, "120-V, 200-W, 90% Efficiency, Interleaved Flyback for Battery Charging Applications Reference Design.," Texas Instruments Incorporated, Texas, 2016.
[3] S.-Y T a J.-H Fan, "Buck-Boost/Flyback Hybrid Converter for Solar Power System Applications," Electronics, vol 10, no 4, p 414, Feb 2021.
[4] T Instrument, "Switching Regulator Fundamentals," Texas Instruments
[5] A I Pressman, K Billings and T Morey, Switching Power Supply Design Third Edition, McGraw-Hill Education, 2009.
[7] W contributors, "Lithium-ion battery," Wikipedia, The Free Encyclopedia., 25 June
2023 [Online] Available: https://en.wikipedia.org/w/index.php?title=Lithium- ion_battery&oldid61868886 [Accessed 28 June 2023].
[8] P Industry, "Lithium Ion Batteries," Panasonic Industry, 2007.
[9] H F Chemicals, "The 4 major components of the lithium-ion battery," 2021. [Online] Available: https://www.hopaxfc.com/en/blog/the-4-major-components-of- the-lithium-ion-battery.
[10] J Bilansky, M Lacko, M Pastor, A Marcinek and F Durovsky, "Improved Digital Twin of Li-Ion Battery Based on Generic MATLAB Model," Energies, vol 16, p.
[11] B University, "BU-409: Charging Lithium-ion," Battery University, 2021 [Online]. Available: https://batteryuniversity.com/article/bu-409-charging-lithium-ion? fbclid=IwAR06OOdj_vR86agpS6O1M8CLtRrEDN-CF8C0d- ikdql_YL5WdffGsYWs88s.
[12] P Battery, "NCR-18650B Datasheet," Panasonic Battery.
[13] E Tutorials, "The MOSFET," Electronics Tutorials, [Online] Available: https://www.electronics-tutorials.ws/transistor/tran_6.html.
[14] I Rectifier, "IR2103PBF Datasheet," International Rectifier, 2013.
[15] Basar, "Optocoupler or Optoisolator – Construction, Working Principle and How it works?," Know Electronic, 2022 [Online] Available: https://www.knowelectronic.com/optocoupler-or-optoisolator/.
[16] F S Corporation, "Single-Channel: 6N137, HCPL2601, HCPL2611Dual-Channel: HCPL2630, HCPL2631High Speed 10MBit/s Logic Gate Optocouplers," Fairchild Semiconductor Corporation, 2005.
[17] S Corporation, "PC817 Datasheet," Sharp Corporation.
[18] N L K Duy, "TÍNH TOÁN, THIẾT KẾ BỘ TẠO TẢI ĐO CÔNG SUẤT ĐỘNG
CƠ ĐIỆN DÙNG TRÊN XE ĐIỆN," 2023.
[19] W contributors, "Diode," Wikipedia, The Free Encyclopedia., 28 May 2023. [Online] Available: https://en.wikipedia.org/w/index.php? title=Diode&oldid57378806.
[20]J Grover, "Diodes: Types, Symbol, Characteristics and Applications," Collegedunia, [Online] Available: https://collegedunia.com/exams/diodes-types-symbol- characteristics-and-applications-physics-articleid-751.
[21] F S Corporation, "LM7805 - Datasheet," Fairchild Semiconductor Corporation, 2001.
[22] L ROHM Co., "Methods of Designing PWM Flyback Converter," ROHM Co.,Ltd, 2016.
[23] R Ridley, "Flyback converter snubber design," Switching Power Magazine, vol 12, 2005.
[24] L Đ V Bình and N T N Vũ, "Nghiên cứu, thiết kế hệ thống cung cấp điện áp thấp cho các tải điện trên xe máy điện," 2020.
[25] R McArthur, "Making Use of Gate Charge Information in MOSFET and IGBT Data Sheets," Application Note APT0103 Advanced Power Technology, 2001. [26]STMicroelectronics, "STF20NM60D - STP20NM60FD," STMicroelectronics, 2006.
[27] F S Corporation, "Design Guideline for Primary Side Regulated (PSR) FlybackConverter Using FAN103 and FSEZ13X7," Fairchild Semiconductor Corporation,2009.