Design and construction of a bidirectional DC/DC converter

(1)

UPTEC E 19009

Examensarbete 30 hp Juni 2019

Design and construction of a bidirectional DC/DC converter

Alexander Wallberg

(2)

Teknisk- naturvetenskaplig fakultet UTH-enheten

Besöksadress:

Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0

Postadress:

Box 536 751 21 Uppsala

Telefon:

018 – 471 30 03

Telefax:

018 – 471 30 00

Hemsida:

http://www.teknat.uu.se/student

Abstract

Design and construction of a bidirectional DC/DC converter

Alexander Wallberg

A four quadrant general single-phase bi-directional DC/DC converter was designed and constructed for high effect systems. The target application for the DC/DC converter was to be used to transfer energy between different energy storages, a miniature DC power grid and the high voltage AC power city grid. The converter is capable of step-up and step-down operations in both directions i.e. it is bi-directional at varying voltage levels.

Different DC/DC topologies were investigated, and thereafter simulations were performed in LTspice and Simulink to ensure its capabilities and functionalities. The result of the simulations was a two layered PI-regulator, controlling both the external DC-grid voltage and inductor current through the converter. Once a suitable topology and control strategy was found, a suitable power transistor investigated and a PCB driver card were developed with KiCad. The final converter is capable to seamlessly change between its four modes and controlling voltages up to 1200 V and currents up to 200 A.

Ämnesgranskare: Johan Abrahamsson Handledare: Martin Fregelius

(3)

Mother Martin Fregelius Johan Abrahamsson

Ping Wu

My future wife, Martina Pettersson

Thank you for everything

(4)

5.2.5 Practical test 4: A → B - Buck mode - Current control - Small steps 39 5.2.6 Practical test 5: A → B - Buck mode - Current control - Large steps 40 5.2.7 Practical test 6: A → B - Boost mode - Current control - Small steps 42 5.2.8 Practical test 7: A → B - Boost mode - Current control - Large steps 42 5.3 Component values . . . 43

6 Discussion 44 6.1 Power Transistors: MOSFET vs IGBT . . . 44

6.1.1 Dimensioning the inductance . . . 44

6.1.2 Topology . . . 45

6.1.3 Control Strategy and Code . . . 46

7 Recommendations for future work 47 8 References 48 9 Appendix 51 9.1 Bill of Material per PCB . . . 51

9.2 Protection circuitry - NAND subcircuit . . . 52

9.3 Protection circuitry - Signal input subcircuit . . . 54

9.4 TLP5214 - External connections . . . 55

9.5 Component value calculation . . . 56

9.6 MOSFET vs IGBT comparison . . . 57

9.7 KiCad Schematic circuit . . . 58

9.8 Simulink circuit . . . 62

9.9 Matlab code . . . 64

(6)

Abbreviations

BDC - Bidirectional DC/DC Converter BJT - Bipolar Junction Transistor CCM - Continuous Current Mode Back EMF - Back Electromotive Force

IGBT - Isolated-Gate Bipolar Transistor

MOSFET - Metal-Oxide-Semiconductor Field-Effect Transistor PCB - Printed Circuit Board

PWM - Pulse Width Modulation PSS - Periodic Steady State

SRA - Small Ripple Approximation UVLO - Under Voltage Lockout

(7)

List of Figures

1.1 Traditional topology . . . 2

1.2 New topology . . . 2

2.1 Periodic Steady State . . . 4

2.2 Bipolar junction transistor - Two doping types . . . 6

2.3 MOSFET - Two doping types . . . 7

2.4 IGBT . . . 8

2.5 Buck converter . . . 9

2.6 Buck converter - Operation states . . . 10

2.7 Boost converter . . . 11

2.8 Boost converter - Operation states . . . 12

2.9 Four quadrant bi-directional converter . . . 14

2.10 B to A - Buck . . . 15

2.11 B to A - Boost . . . 15

2.12 B to A - Buck . . . 15

2.13 B to A - Boost . . . 15

3.1 10 types of Single phase converters . . . 17

3.2 The Cascading buck-boost converter . . . 18

3.3 Block diagram of the control system . . . 19

3.4 Block diagram - Control signal . . . 20

3.5 Driver PCB . . . 21

4.1 CM200DX-24S - Internal connections . . . 23

5.1 Entire system - DC/DC bidirectional converter . . . 25

5.2 Driver Card PCB . . . 26

5.3 Test 1: The supercapacitor and DC-bus voltages . . . 27

5.4 Test 1: The inductor current and the desired current from the controller . . 28

5.5 Test 1: The four modes of the DC/DC converter . . . 29

5.12 Test 2: Control signal . . . 37

5.13 Test 3: Two control signals . . . 38

5.14 Test 3: Delayed gate signal . . . 39

5.15 Test 4: Buck Mode - Current control - Small steps . . . 40

5.16 Test 5: Buck Mode - Current control - Large steps . . . 41

5.17 Test 6: Boost Mode - Current control . . . 42

5.18 Test 7: Boost Mode - Current control . . . 43

6.1 Current control - Transient . . . 45

9.1 NAND subcircuit . . . 53

(8)

9.2 Signal conversion circuit . . . 54

9.3 TLP5214 - External connections . . . 55

9.4 TLP5214 - Internal functions . . . 56

9.5 KiCad circuit - Overview 1 . . . 58

9.8 Simulink circuit - Overview 1 . . . 62

9.9 Simulink circuit - Overview 2 . . . 63

List of Tables

2.1 Four MOSFET types . . . 7

2.2 Bi-directional converter - Modes . . . 14

3.1 Choice of mode . . . 19

5.1 Component Values . . . 43

9.1 Bill of Material . . . 51

9.2 Truth Table - Control signals & Error signals . . . 52

9.3 MOSFET vs IGBT comparison . . . 57

(9)

1 Introduction

1.1 Background

Since the industrial times, we have relied on fossil fuel, gas fired power stations and power plants to generate electricity. However, these kind of power generation options are no longer sustainable due to the negative impact they have on our planet. Therefore, new and renewable ways of generating electricity have been developed such as solar and wind generation.

These new options of generating electricity have significantly less negative impact on our planet. They do however impact our power grid in a negative way [12] [15].

Two things are absolutely critical when generating electricity: the generated energy must be equal to the energy that we consume and the power grids voltage and frequency must be constant. The previous ways of generating electricity have contributed with a sort of inertia to the power grid. This inertia has created a window of time for the power generation side to adapt to the changes in power consumption, making the power grid more robust and stable.

The issue with the renewable power generation alternatives is that they do not create this inertia, thus having a negative impact on the power grid.

Due to the negative impact that the older power generation plants and stations have on our planet, the aim is to replace them with the newer alternative renewable generation systems. Because of this change, new ways of creating synthetic inertia have to be invented to compensate for the loss of inertia. One way of creating the synthetic inertia to use energy storages to continuously exchange energy between the storages and the power grid. The idea is that the energy storages would counteract the changes in the power grid by either withdrawing energy from it or depositing energy to it.

At the Department of Engineering Sciences, Division of Electricity at Uppsala University, a PhD student is building a miniature DC power grid with different kinds of energy storages.

This proof of concept shows that these energy storages can be used to create synthetic inertia on the AC power grid via one DC-bus. Traditionally, each energy storage would be connected to the AC-grid directly via one power inverter each. However, this is not efficient since one power inverter is needed for each energy storage. Instead, the idea is to connect each energy storage to this miniature DC power grid and then to the AC power grid via just one inverter.

With all energy storages connected together via the DC-bus the energy they hold could be balanced between each storage via the DC-bus. For example, if the supercapacitor was fully charged and the battery bank was nearly depleted, the energy from the supercapacitor could easily be directed to the battery bank via the DC-bus. This sort of balancing could not be achieved in the traditional way since the energy storages are isolated from each other.

(10)

Figure 1.1: Traditional topology Figure 1.2: New topology

However, these energy storages can’t be connected to the DC-bus since they most likely would operate at different voltage levels and the set voltage level of the DC-bus. Therefore, we need a device that can convert the voltages to suitable levels and control the amount of energy that flows between the DC-bus and each energy storage. The solution is to create an adaptable single phase bi-directional DC/DC converter that could be used for any energy storage with varying voltage levels and energy storage capabilities.

1.2 Purpose & Goals

The purpose of the project is to design, simulate and build a general purpose single-phase bi-directional DC/DC converter. In order to accomplish this, the following goals have to be achieved:

• Evaluate different bi-directional DC/DC converter topologies, determining which is most suitable for the desired system (See System Requirements 3.1)

• Simulate the chosen topology and investigate how the converter should be controlled and also how the peripheral components should be designed.

• Find a suitable power transistor type that fulfills the system requirements (See System Requirements 3.1)

• Design the driver circuit for the chosen power transistor type with suitable signal circuitry and protection circuitry

• Design a printed circuit board, PCB, for the power transistor and find suitable peripheral components

(11)

1.3 Report disposition

In chapter 2, Theory, different kind of power transistors and power converters are described and relevant equations are listed. The chapter gives the reader the necessary background and understanding of power transistors and power converters to fully understand the report. In chapter 3, Method, the project approach is described step by step, starting with the choice of topology and control strategy, PCB design and finally the construction of the entire system.

In chapter 4, Components & hardware, the chosen power transistor and driver module are described and also the Bill of Material. In chapter 5, Results, the simulation and practical tests are shown along with the PCB design and system design and parameters. In chapter 6, Discussion, the results are discussed along with improvement proposals.

(12)

2 Theory

2.1 Assumptions

When deriving the equations for the Buck and Boost converter several assumptions have to be made: All components are ideal, PSS, CCM, SRA and Volt-second balance [7].

2.1.1 PSS - Periodic steady state

Periodic steady state refers to the equilibrium condition that occurs after start-up when a circuit has stabilized and transients no longer influence the circuit. (See Figure 2.1).

Figure 2.1: Periodic Steady State

2.1.2 CCM - Continuous current mode

The continuous current mode assumption says that the current must flow through the inductor continuously and never become zero.

2.1.3 SRA - Small ripple approximation

The output voltage of an arbitrary non ideal DC/DC converter contains both a DC and an AC component. The AC component or ripple occurs due to the switching operation of the converter and is in most cases undesired. However, a converter that is properly designed will

(13)

have a very low AC component compared the DC component. Therefore, the ripple can be neglected when calculating the output voltage of a DC/DC converter.

v(t) = V_dc+ v_ripple(t) → |vripple(t)| << V_dc

→ v(t) ≈ V_dc (2.1)

2.1.4 Volt-second balance

In PSS, the average voltage over the inductor must be zero assuming that the inductor current changes linearly.

1 T

Z 0 T

V_Ldt = 0 (2.2)

2.1.5 Annotation

To differentiate between AC and DC components of voltages and currents, lower case letters are used to denote AC and upper case letter to denote DC components. For example: An AC voltage would be written as ’v’ and a DC current would be written as ’I’.

2.2 Transistors

2.2.1 Bipolar junction transistor - BJT

The bipolar junction transistor, BJT, is a current controlled semiconductor device that can be operated as a switch or as a current amplifier [5]. This transistor can be divided into two doping types: the n-type (NPN) and the p-type (PNP). Both of them have three terminals:

Collector (C), Base (B) and Emitter (E). They consist of both n-doped and p-doped sections but in different constellations, as can be seen in Figure 2.2. Both types can be used for the same purpose but they function in opposite ways. The NPN transistor conducts current from the Collector to the Emitter when a current is applied to the Base of the transistor.

The PNP transistor on the other hand conducts current from the Emitter to the Collector when no current is applied to the Base of the transistor.

(14)

Figure 2.2: Bipolar junction transistor - Two doping types

The amount of current allowed through the transistor is determined by its DC current gain, β, and the current applied to the Base. (See Equation 2.3).

DC current gain, β = I_C

I_B (2.3)

Due to the Base not being isolated from the rest of the transistor the Emitter current, I_E, is the sum of the Base current, I_B, and the Collector current, I_C. (See Equation 2.4)

I_E = I_B+ I_C (2.4)

The relationship between the three currents are describe by α.

α = I_C

I_E = I_C

I_B+ I_C (2.5)

Since the I_C is much larger than I_B, actually β times larger, α usually is very closed to 1.

The relationship between α and β describe the relationship between the transistors three currents.

β = α

1 − α or α = β

1 + β (2.6)

2.2.2 Metal-oxide-semiconductor field-effect transistor - MOSFET

The metal-oxide-semiconductor field-effect transistor, MOSFET, is a voltage controlled semiconductor device that can either be used as an amplifier or as a switch [5]. Similarly to the

(15)

BJT, the MOSFET has three terminals, however with different names: Drain (D), Gate (G) and Source (S). The MOSFET also have two doping types: NPN and PNP. The MOSFET can further be divided into two subtypes: the enhancement MOSFET and the depletion MOSFET. How these variations of MOSFETs behave when different Gate threshold voltages, V_th and V−th, are applied are shown in Table 2.1.

Type of MOSFET V_GS = V_th V_GS = 0 V_GS = V−th

NPN - Depletion ON ON OFF

NPN - Enhancement ON OFF OFF

PNP - Depletion OFF ON ON

PNP - Enhancement OFF OFF ON

Table 2.1: Four MOSFET types

The major difference between the MOSFET and the BJT is that the Gate terminal of the MOSFET is isolated from the two other terminals. The MOSFET is controlled by applying a voltage to the Gate terminal and charging it. The amount of charge on the Gate terminal determines the amount of current allowed through the transistor.

Figure 2.3: MOSFET - Two doping types

Since the Gate of the MOSFET is isolated no current flows from the Gate to the Source.

Thus, the Drain current is equal to the Source current (See Equation 2.7).

I_D = I_S (2.7)

(16)

2.2.3 Insulated-gate bipolar transistor - IGBT

The isolated-gate bipolar transistor, IGBT, [5] is a voltage controlled semiconductor device especially developed for fast switching and high power applications [5]. Similarly to the BJT and the MOSFET, the IGBT has three terminals: Collector (C), Gate (G) and Emitter (E).

The IGBT is a hybrid between the BJT and the MOSFET, inheriting the isolated Gate and high switching frequency from the MOSFET and the low saturation voltage from the BJT.

Like the MOSFET, the IGBT is controlled by applying a voltage to the Gate terminal and therefore charging it. However, the IGBT is unidirectional i.e. it can only conduct current in one direction. A power diode is used to conduct current in the opposite direction.

Figure 2.4: IGBT

Since the Gate is isolated no current flows from it to the Emitter, therefore the Collector current is equal to the Emitter current (See Equation 2.8).

I_E = I_C (2.8)

(17)

2.3 Switched DC/DC converters

2.3.1 Buck converter

The Buck converter is a switch mode power device that outputs a DC voltage that is lower than the input DC voltage [2]. A simplified version of the Buck converter can be seen in Figure 2.5. The Buck converter requires a power switch, a power diode, a capacitor and an inductor.

Figure 2.5: Buck converter

During normal operation, the Buck converter can operate in two states: the on-state and the off-state (See Figure 2.6). Whilst in the on-state, current is supplied from the voltage source, charging the inductor and supplying a current to the load. The stored energy in the inductor also increases and the magnetic field around it expands. In this mode the diode is reverse biased, meaning that it doesn’t conduct current and is considered to be turned off.

(18)

Figure 2.6: Buck converter - Operation states

Whilst in the off-state, the switch is turned off and the magnetic field around the inductor starts to collapse, reinserting the stored energy to the circuit. The back e.m.f causes the voltage polarity of the inductor to become reversed. Due to the reversed voltage polarity, the diode becomes forward biased and it begins to conduct. During the off-state both the inductor and the capacitor supplies the load with current.

A desired output voltage can be produced by controlling the power switch of the Buck converter with pulse width modulation. The relationship between the voltages and the duty cycle, D, can be seen in equation 2.9.

D = V_out

V_in (2.9)

When designing a Buck converter, the following equations are useful [6]:

(19)

Inductor current ripple:

∆i_L= VinDD⁰Ts

2L [A] (2.10)

Output Ripple Voltage:

V_ptp = 2v_ripple = V_out(1 − D)T_s²

8LC [V ] (2.11)

Component calculation:

L = V_inDD⁰T_s

2∆i_L [H] (2.12)

C = V_out(1 − D)T_s²

16 ∗ v_ripple∗ L [F ] (2.13)

2.3.2 Boost converter

The Boost converter requires the same components as the Buck converter (Section 2.3.1) but in a different constellation, as can be seen in Figure 2.7 [3]. The Boost converter, as the name suggests, outputs a voltage that is higher than the input voltage.

Figure 2.7: Boost converter

As in Figure 2.8, the Boost converter has two states: the on-state and the off-state. During the on-state, current through the inductor increases its stored energy i.e. its magnetic field is expanding. In this mode the diode is reverse biased which separates the two sides of the converter, the inductor side and load side. Whilst the diode isn’t conducting, the energy stored in the capacitor is being supplied to the load.

(20)

Figure 2.8: Boost converter - Operation states

In the off-state the power switch is turned off. The magnetic field starts to collapse and due to the inductors back e.m.f, the diode becomes forward biased. The stored energy in the inductor results in a current flowing to the capacitor and the load. Unlike the Buck converter, the inductor of the Boost converter is always connected to the power supply. This means that its always being charged and that is how the converter can produce a voltage higher than the input voltage.

By controlling the power switch of the Boost converter with PWM, a desired output voltage can be produced. The relationship between the voltages and the duty cycle, D, can be seen

(21)

in equation 2.14.

Conversion Ratio:

V_out

V_in = 1

1 − D (2.14)

When designing a Boost converter, the following equations are useful [8]:

Inductor current ripple:

∆i_L= V_in

2LDT_s [A] (2.15)

Output Ripple Voltage:

V_ptp = 2V_ripple = V_outDT_s

RC [V ] (2.16)

Component calculation:

L = V_in

2∆i_LDTs [H] (2.17)

C = VoutDTs

2RV_ripple [F ] (2.18)

2.3.3 Four quadrant bi-directional converter

The single phase four quadrant bi-directional converter is a cascaded Buck-Boost converter [9]. It consists of four power switches, two capacitors and one inductor. Due to the symmetric design of the converter it can operate in four modes. It is capable of stepping up and down the voltage in either direction. All modes are shown in the figures below, and which power switches to activate in order to choose a mode is shown in Table 2.2. Since the converter behave in the same way as the Buck or Boost converter, depending on which mode is activated, the formulas presented in Sections 2.3.1 and 2.3.2 can be used for the bi-directional converter as well.

(22)

Figure 2.9: Four quadrant bi-directional converter

Direction Mode S₁ S₂ S₃ S₄

A to B Buck Switching OFF OFF OFF

A to B Boost ON OFF OFF Switching

B to A Buck OFF OFF Switching OFF

B to A Boost OFF Switching ON OFF

Table 2.2: Bi-directional converter - Modes

(23)

Figure 2.10: B to A - Buck Figure 2.11: B to A - Boost

Figure 2.12: B to A - Buck Figure 2.13: B to A - Boost

(24)

3 Method

3.1 Requirements

Listed below are the requirements that had to be met in order to make the system function as desired:

Topology requirements

1. The converter must be able to operate in all quadrants i.e. it must be capable of step-down and step-up operations in both directions.

2. The amount of components used should be as low as possible 3. The control strategy should have low complexity

System requirements

1. The main system requirement is that the converter has to maintain 100 V on DC-bus 2. The control strategy should not allow the inductor current to go above 200 A

3. The driver cards external communication logic should operate at 24 V 4. The converter should be rated for 1200 V and 200 A

5. The system has to be as small as possible

3.2 Converter topology

The first step of the project was to research different bidirectional DC/DC converter typolo- gies, determining which were best suited for the desired system. When designing a converter, three attributes have to be taken into account: simplicity, flexibility and efficiency. The bidirectional DC/DC converters can also either be isolated or non-isolated i.e. whether or not the converter has a transformer. The most important requirement was that the converter could step-up and step-down the voltage in both directions.

Due to the size requirements, isolated converters were dismissed early in the project since the transformers were to large. In A Review of Non-Isolated Bidirectional DC/DC Con- verters for Energy Storage Systems (2016) it is stated that the non-isolated bidirectional DC/DC converter can be divided into three subcategories: Single-phase, Single-phase with auxiliary ZVS/ZCS and Interleaved [10]. The two latter were dismissed due to their required complexity and component count.

By default, the single-phase bidirectional DC/DC converter was picked for this project. Once the type of converter was determined, an actual design was investigated. In A Review of Non-

(25)

Isolated Bidirectional DC/DC Converters for Energy Storage Systems (2016) 10 versions of a single-phase converters are presented (See figure 3.1) [10].

Figure 3.1: 10 types of Single phase converters

Out of the 10 converters in Figure 3.1 only the third topology (Figure 3.1 (c)) fulfills the first system requirement (See System requirements in Chapter 3.1).

However, the cascading buck-boost converter has some drawbacks due to its flexibility. Unlike other single phase converters, it requires four power switches and therefore a more complex control strategy is needed. Due to the higher count of power switch, the converter also has more switching losses than the other single phase converters.

(26)

Figure 3.2: The Cascading buck-boost converter

Since the cascading buck-boost converter is either in Buck or Boost mode, the same equations as for the Buck and Boost converters could be used to design it and dimension the peripheral components (See equations in Section 2.3.1 and 2.3.2). The allowed current ripple in the inductor, ∆i_L, and the voltage ripple of the output voltage, v_ripple, was arbitrarily chosen to 1 A and 1% (compared to the output DC voltage). The values for the peripheral components were calculated (See calculations and code in Appendix 9) and multiplied with 10 as a safety precaution. The final values for the components are shown in the Results (See Section 5.3).

3.3 Control Strategy

Once the topology was chosen, a proper control strategy had to be derived. As stated in the system requirements (Section 3.1), the bidirectional DC/DC converter’s goal is to maintain 100 V on the DC-bus and allow a maximum current of 200 A. In order to ensure that these requirements were met, a system with two PI regulators was created: one inner PI regulator controlling the current going through the BDCs inductor and one outer PI regulator controlling the voltage on the DC-bus (See Figure 3.3) [17].

(27)

Figure 3.3: Block diagram of the control system

The entire system was simulated and tested in Simulink. The results of these tests are shown in Results (See Section 5.1) and the code is shown in the Appendix (See Section 9.8). The mode of the converter is chosen by the relationship between the DC-bus voltage and its set value of 100 V and also by the relationship between the DC-bus voltage and the energy storage voltage. (See Table 3.1). Table 3.1 the DC-bus is referred to as ”A” and the energy storage is referred to as ”B”.

Voltage DC-bus [V] Voltage DC-bus [V] Mode

>100 V >Voltage Energy storage Buck - A to B

>100 V <Voltage Energy storage Boost - A to B

<100 V >Voltage Energy storage Boost - B to A

<100 V <Voltage Energy storage Buck - B to A

= 100 - None

Table 3.1: Choice of mode

3.4 Protection circuitry

When controlling a converter with four power switches, it is extremely important that under no circumstance create a shoot through. As can be seen in figure 3.2 the power switches S₁ and S₂ (and also S₃ and S₄) are connected in such a way that if both are ON at the same moment, current would be allowed to rush through the two switches, creating a short

(28)

circuit. Such a short circuit would most likely destroy vital parts of the circuit. Therefore some protection circuits had to be developed and included. The entire input step is shown as a block diagram in Figure 3.4 and each protection feature is described below.

Figure 3.4: Block diagram - Control signal

3.4.1 Control signal - Filtering and conversion

The Single Board RIO used to control the converter uses 24 V logic signals and all ICs on the driver PCB are designed for 5 V. Therefore, the received control signals had to be converted to the desired voltage level of 5 V. A BJT (See Section 2.2) and some peripheral components were used for this conversion. The entire circuitry for the filtering and conversion is shown in the Appendix (See Section 9). Note that the technique used inverts the signal but is later inverted again in the NAND subcircuit.(See Section 3.4.3). The conversion circuit also includes a pull down resistor, a low pass filter and a baker diode. See the entire conversion circuit in the Appendix in Figure 9.2.

3.4.2 Delay circuitry

If ideal switches were used, the protection circuitry mentioned in Section 3.4.3 would be sufficient in order not to create a short circuit between the power switches. However, in non-ideal switches it takes a short moment to either turn on or off. Therefore, a delay had to be added between the control signal for the HIGH side and the LOW side. Such a subcircuit was designed with the software LT-spice and later on implemented on the IGBT driver PCB.

This delays the feedback from the HIGH side to the LOW side, would allow the switch on the HIGH side to turn off fully before the switch on the LOW side could activated. This means that both signals act as enablers for each other, ensuring the safety of the circuit. By adding such an analog delay circuit, no delay functionality has to be added in the control software.

(29)

3.4.3 Error circuity

To ensure that the driver only operates if no errors are registered from the drivers, a subcircuit of 2-input Schmitt NAND Gates (See Component 4.2) was created. The software Logisim was used to create the necessary circuit based on a truth table (See Figure 9.2 in the Appendix).

Note that the error signals from the TLP5214 are active LOW, meaning that if no error occurs, both error signals are pulled HIGH. Figure 9.1 shows the NAND subcircuit that was created from Table 9.2. If no error occurs and only one input signal is active, the rest of the circuit would operate as normal.

3.5 Driver PCB design

Once the control strategy and the protection circuitry was designed and deemed good enough, the entire driver circuit was created with the software KiCADs Schematic layout editor.

The entire schematic layout can be seen in the Appendix (See Section 9.7). Thereafter, each component was assigned a suitable PCB-footprint and then finally the PCB board was designed. The complete PCB is shown in the Results (See Section 5) in figure 5.2. All PCB layers alongside with the complete PCB can be found in the Appendix (See Section 9).

Figure 3.5: Driver PCB

3.6 Converter construction

The final step of the project was to build the actual system and test its functionality. Two CM200DX-24S IGBT modules (containing two IGBTs each) with drivers, an inductance, two capacitors, a resistive load and measurement equipment were placed on a test bench. The final system is shown in Figure 5.1 in the Results (See Section 5). A myRIO-1900 was used to control the entire system. The Simulink control code was implemented on the myRIO with

(30)

LabVIEW. In the construction part of the project there were no supercapacitors available, therefore a low resistance load was used for testing instead. Tests were performed to show that the converters Buck and Boost functionality worked as desired. These tests, along with the simulation tests from Simulink, are shown in the Results (See Section 5.2).

(31)

4 Components & hardware

4.1 IGBT - CM200DX-24S

The CM200DX-24S is a half bridge high power switching IGBT module rated for collector current of 200 A and collector to emitter voltage of 1200 V. Common applications for the IGBT are: AC motor control, power supply and servo control. Two IGBT modules could be used to create a H-bridge or a bi-directional DC/DC converter. The IGBT is also equipped with a NTC thermistor for internal temperature measurements.

Figure 4.1: CM200DX-24S - Internal connections

4.2 Driver - TLP5214

The TLP5214 is commonly used for driving an IGBT or a Power MOSFET used in power inverter applications. The TLP5214A is a general-purpose isolated IGBT/Power MOSFET gate driver which can output a maximum current of 4.0 A to the IGBT/Power MOSFETs gate. The driver has built-in functions, such as desaturation detection, isolated fault status feedback, soft turn-off, active miller clamping and UVLO. Equipped with two infrared light- emitting diodes and two high-speed ICs, it is suitable for high-speed switching and controlling of power devices. The driver also has a very high isolation voltage of 5000 V_rms. The internal and external connections are shown in Figure 9.3 and 9.4 in the Appendix (See Section 9.4).

(32)

4.3 Components

The rest of the components are listed in Appendix in Section 9.1.

(33)

5 Results

A suitable bi-directional DC/DC converter topology was found for the desired system that fulfills the Topology requirements in Section 3.1. A two layered PI controller system that fulfills the System requirements in Section 3.1 was implemented, simulated and tested. Fi- nally, a system was built that fulfills the the System requirements in Section 3.1. The final system is shown Figure 5.1 and the code is presented in the Appendix (See Section 9).In Figure 5.2 the final version of the driver card PCB is presented.

Figure 5.1: Entire system - DC/DC bidirectional converter

(34)

Figure 5.2: Driver Card PCB

5.1 Simulations

In the simulation tests, the DC-bus is referred to as the ”A-side” and the supercapacitor is referred to as the ”B-side”. Also all step responses occur after one second if nothing else is stated.

5.1.1 Simulation test 1: A → B - Buck mode

The resulting voltages and currents of a 10 V step response are shown in figure 5.3 and 5.4. In this test the supercapacitor voltage is initially far lower than the voltage on the DC-bus.

(35)

Figure 5.3: Test 1: The supercapacitor and DC-bus voltages

The current starts to flow from the DC-bus to the supercapacitor until the voltage on the DC-bus is stabilized around 100 V. The slow increase of the current, relative to the desired current, is caused by the systems large inductance and the simple nature of the controller (See more about this in the Discussion 6.1.1).

(36)

Figure 5.4: Test 1: The inductor current and the desired current from the controller

Due to the increased voltage on the DC-bus, and since the supercapacitor voltage is lower than the DC-bus voltage, the system is operating in Buck mode (A → B) as can be seen in Figure 5.5.

(37)

Figure 5.5: Test 1: The four modes of the DC/DC converter

When the initial voltage of the DC-bus and the supercapacitor was set to 100 V and 80 V, it took 5.056 seconds for the system to adjust to the 10V step response whilst in Buck mode.

5.1.2 Simulation test 2: A → B - Boost mode

The resulting voltage and currents of a 10 V step response is shown in figure 5.6 and 5.7.

However unlike test 1, the supercapacitor voltage is initially far higher than the voltage on the DC-bus.

(38)

Just like in Test 1, the current starts to flow from the DC-bus to the supercapacitor until the voltage on the DC-bus is stabilized around 100 V. Whilst in Boost mode, the inductor current follows the desired current much better than in Buck mode, however with much more ripple.

(39)

Due to the increased voltage on the DC-bus, and the fact that the supercapacitor voltage is higher than the initial DC-bus voltage, the system is operating in Boost mode (A → B) as can be seen in Figure 5.8.

(40)

When the initial voltage of the DC-bus and the supercapacitor was set to 100 V and 110 V, it took 4.755 seconds for the system to adjust to the 10V step response whilst in Boost mode.

5.1.3 Simulation test 3: A → B - Buck and Boost mode

Test 3 shows how the systems operates when the supercapacitor voltage is just slightly lower than the DC-bus voltage and a 10 V step response is introduced to the DC-bus. The voltages and currents are shown in Figure 5.9 and 5.10. Due to the small initial voltage difference between the two sides, and since the converter is in Buck mode, the current is limited by the inductance and can’t be as high as desired. This is a limitation and design flaw of the controller, and possible improvements are presented in the Discussion (See Section 6).

(41)

Once the step response is introduced, the converter operates in Buck mode because the supercapacitor voltage is lower than the DC-bus voltage. After 3.457 seconds, the supercapacitor voltage exceeds the DC-bus voltage and the converter switches to Boost mode (See Figure 5.11)

(42)

Like in Test 2, the initial voltage difference between the two sides is very low and therefore the converter can’t produce a sufficiently high current in Buck mode. This is shown in Figure 5.10 between 1 and 4.457 seconds. Thereafter, the supercapacitor voltage exceeds the DC-bus voltage and the converter switches to Boost mode (A → B). However, the stored energy and the nature of the inductance prevents fast changes in current flow and therefore the current can’t follow the desired current. This is also shown in Figure 5.10 after 4.457 seconds. The converter modes in this test are shown in Figure 5.11.

(43)

When the initial voltage of the DC-bus and the supercapacitor was set to 100 V and 95 V, it took 5.363 seconds for the system to adjust to the 10 V step response.

5.2 Measurements and tests

5.2.1 Practical test restrictions

Due to restrictions when it comes to component types, shipping delays and available components, the practical tests had to be adjusted. The driver cards are designed to be used with the Single Board RIO (24 V logic voltage level for communication) used in the PhD students miniature DC power grid mentioned in the Introduction (See Section 1.1). Instead a myRIO-1900 was used to control the converter, which operates at a logic level of 3.3 V.

Therefore, several conversion modules had to be used for the communication to function properly. Another test limitation was that no suitable energy storage was available when

(44)

testing the system. Instead different kind of resistive loads were used to demonstrate the converter functionality and capabilities. Additionally, the miniature DC power grid or DC-bus wasn’t available either and therefore two DC-cubes were used instead.

These limitations restricted how much current that could be produced by the DC-cubes and sent through the converter. It also meant that the converters bi-directionality couldn’t be tested simultaneously, but instead one direction at a time. Since the real DC-grid wasn’t available only the inner PI controller, responsible for the current through the inductor, was tested.

In the practical tests, the two DC-cubes (acting as DC-bus) are referred to as the ”A-side”

and the resistive load is referred to as the ”B-side”.

5.2.2 Practical test 1: Both directions - Buck & Boost mode

In this test, the systems capabilities of acting as a Buck and Boost converter in both directions were tested by measuring the output voltage. When in Buck mode, a input voltage of 40 V and a resistive load of 1 Ω was used. However, in Boost mode, a resistive load of 25.6 Ω had to be used because of the Boost converters critical resistance. Whilst in Buck mode (in either direction), the output voltage could be anywhere between 0 and 37.8 V and in Boost mode (in either direction), the output voltage ranged between 40 and 98.9 V.

5.2.3 Practical test 2: Logic signal → Gate signal

To verify that all driver card functionalities (Mentioned in 3.4) were fully operational, each switch was activated one by one with pulse width modulation. The 3.3 V signal sent from the myRIO-1900 and the ± 15 V at the IGBT gate is shown in Figure 5.12. This confirms that the protection circuitry functions properly. The blue signal is the control signal (3.3 V) from the myRIO-1900 and the yellow signal is the associated gate voltage (± 15V).

(45)

Figure 5.12: Test 2: Control signal

5.2.4 Practical test 3: Signal conversion - Shoot through protection

In this test, the driver cards protection circuitry is tested and verified. Two signals are sent to the driver card, as shown in Figure 5.13 and the resulting gate voltage is shown in Figure 5.14.

(46)

Figure 5.13: Test 3: Two control signals

(47)

Figure 5.14: Test 3: Delayed gate signal

Even though the control signals are each other’s inverse, the resulting gate signals are sep- arated by a 6 us time window. Thus protecting the driver card from shoot though when switching between the two IGBTs.

5.2.5 Practical test 4: A → B - Buck mode - Current control - Small steps In this test the systems current control was tested, within the scope explained in Section 5.2.1.

The input voltage was set to 40 V and a resistive load of 1 Ω was used. The desired inductor current was increased with steps of 5 A at the time. The desired current is represented in orange and the measured current in the inductor is represented in blue in Figure 5.15. To avoid high transient currents when switching the system, the duty cycle of the switch was not at 0% in the beginning of the test, just very close to zero resulting in a small current as can be seen in Figure 5.15. The reason for these transient and how to remove them is presented in the Discussion (See Section 6.1.1).

(48)

Figure 5.15: Test 4: Buck Mode - Current control - Small steps

5.2.6 Practical test 5: A → B - Buck mode - Current control - Large steps Test 5 uses the same voltages and loads as Test 4 but the desired current increase from 1 A to 20 A instantly.

(49)

Figure 5.16: Test 5: Buck Mode - Current control - Large steps

(50)

5.2.7 Practical test 6: A → B - Boost mode - Current control - Small steps

Figure 5.17: Test 6: Boost Mode - Current control

5.2.8 Practical test 7: A → B - Boost mode - Current control - Large steps Test 7 is similar to Test 4, however in Boost Mode. The input voltage was 40 V but a 25.6Ω resistance was used as the load.

(51)

Figure 5.18: Test 7: Boost Mode - Current control

5.3 Component values

Table 5.1 show the calculated component values for the converter used in the simulation and the component values used for the practical tests.

Component Type Values for simulation tests Values for practical tests

Capacitor, C 50 µF 330 uF

Inductor, L 1.25 mH 0.40 mH

Table 5.1: Component Values

(52)

6 Discussion

6.1 Power Transistors: MOSFET vs IGBT

In the early stages of this project, the idea was to use the SKM350MB120SCH15 Silicon Carbide (SiC) power MOSFETs instead of the CM200DX-24S IGBTs. Historically, the traditional MOSFET has been able to switch a lot faster than the IGBT, but at much lower effects. However, these new Silicon Carbide power MOSFETs are designed for similar, if not higher, voltage and current levels than IGBTs and they offer a lot higher switching frequency.

As can be seen in Table 9.3, the SKM350MB120SCH15 is far superior to the CM200DX-24T and the most important difference is the significantly lower gate charge, Q_G, that is required.

The lower gate charge makes the increased switching frequency possible and would put less stress on the gate driver IC. Unfortunately, the SiC power MOSFETs were not available for purchase and therefore the CM200DC-24T modules were used instead. But due to the flexibility of the driver IC and the constructed driver PCB, the design PCB could still be used to control a SiC power MOSFET as long as the pin configuration is the same.

6.1.1 Dimensioning the inductance

Generally when designing a DC/DC converter, the goal is to have a switching frequency that is as high as possible, and by doing so reducing the inductor size. However, the inductors size serves another purpose in high effect DC/DC converters: it determines how fast the current flow can change. When the voltage difference between the two sides of the converter is high and it starts to switch, the initial current can be extremely high as can be seen in Figure 6.1. If the current isn’t properly limited by the inductance, these transients could have a devastating effect on the power transistor and driver IC. It is very important that the measurement device is fast enough to detect the current transients, otherwise the current controller won’t be able to safely control the system. Therefore, it is very important to find a balance between the switching frequency and inductor size and not use the smallest inductor allowed by the switching frequency.

(53)

Figure 6.1: Current control - Transient

6.1.2 Topology

Out of the ten non-isolated single-phase converters mentioned in A Review of Non-Isolated Bidirectional DC/DC Converters for Energy Storage Systems only the cascading buck-boost converter was suitable for this project [10].

However, if a low complexity control system and low amount of components weren’t required, other more efficient topologies could be used, such as the Interleaved converter or a converter that uses soft-switching. An interleaved converter would have several power stages that are evenly distributed over a phase shift of 360^◦. Such a converter structure would offer higher efficiency, current ripple cancellation and better thermal performance. The current through each switch would also be lower with each interleaved section of the converter. Soft- switching, such as ZVS and ZCS, could also offer a higher overall efficiency by only switching the power transistor either when current through transistor or the voltage over the transistor is zero. This would eliminate or at least reduce the switching losses, improving the converter

(54)

efficiency.

6.1.3 Control Strategy and Code

The converter mode is determined not only by the relationship between the voltage on the DC-bus and its set voltage value of 100 V, but also by the relationship between the DC-bus voltage and the energy storage voltage as can be seen in Table 3.1. However, as shown in Figure 5.10 in Test 3, the converter isn’t able to produce a sufficiently high current whilst in Buck mode when the voltage difference, between the converters to sides, is to low. To optimize the system, the controller would have to be able to switch from Buck mode to Boost mode at an earlier stage than in Test 3. One way of doing this would be to also use the derivative of the inductor current to determine the converter mode. If the converter is in Buck mode and the current slope is to low or decreasing, the converter could change to Boost mode and therefore draw a higher current. This would improve the converters overall performance and efficiency.

(55)

7 Recommendations for future work

For future work, I recommend using the Silicon Carbide (SiC) power MOSFETs instead of the IGBTs due to them being far superior. I also recommend using either the interleaved converter or a converter that uses soft-switching since they both offer higher efficiency than the single phase converter. Finally, I recommend using a more sophisticated control strategy, such as the one described in the Discussion.

(56)

8 References

[1] Analog Devices, Linear Technology. (1999). LTspice. https://www.analog.com/en/

design-center/design-tools-and-calculators/ltspice-simulator.html#

[2] Coates, E., (2018). Buck Converters. http://www.learnabout-electronics.org/PSU/

psu31.php

[3] Coates, E., (2018). Boost Converters. http://www.learnabout-electronics.org/PSU/

psu31.php

[4] Gabrysch, M., Universitetslektor, Docent vid Institutionen f¨or teknikvetenskaper, Elek- tricitetsl¨ara, Uppsala Universitet., (2017). Equations Sheet – Power Electronics I, HT2017.pdf. [online].https://studentportalen.uu.se/portal/

[5] Gabrysch, M., Universitetslektor, Docent vid Institutionen f¨or teknikvetenskaper, Elek- tricitetsl¨ara, Uppsala Universitet. (2017). Power Transistors, Lecture 7.

[6] Gabrysch, M.,; Universitetslektor, Docent vid Institutionen f¨or teknikvetenskaper, Elek- tricitetsl¨ara, Uppsala Universitet. (2017). DC-DC Switch-Mode Converters Step-Down / Buck Converter, Lecture 8.

[7] Gabrysch, M., Universitetslektor, Docent vid Institutionen f¨or teknikvetenskaper, Elek- tricitetsl¨ara, Uppsala Universitet. (2017). Conditions for Analysis, Lecture 9.

[8] Gabrysch, M., Universitetslektor, Docent vid Institutionen f¨or teknikvetenskaper, Elek- tricitetsl¨ara, Uppsala Universitet. (2017). DC- Switch-Mode Converters Step-Up / Boost Converter, Lecture 10.

[9] Hedlund, M., (2010). Design and construction of a bidirectional DCDC converter for an EV application. Mastere thesis., Uppsala University.

[10] Kostiantyn, T., Oleksandr, H., Oleksandr, V., Roman, Y. (2016).A Review of Non-Isolated Bidirectional DC-DC Converters for Energy Storage Systems. DOI:

10.1109/YSF.2016.7753752. Ukraine: Biomedical Radio-electronic Apparatus and Sys- tems Department, Chernihiv National University of Technology (CNUT).

[11] KiCad Developers Team. (1992). KiCad EDA. http://www.kicad-pcb.org/

[12] Kirkwood, I., (2018). Renewables making grid ‘more unstable’. The Her- ald. 20 March 2018. https://www.theherald.com.au/story/5292937/

renewables-making-grid-more-unstable-says-government-agency-report/

[13] MathWorks. (2018). Simulink 9.2 (part of R2018b). https://se.mathworks.com/

products/simulink.html

[14] Molin, A., (2013). 8. MOS-transistorn. Analog elektronik. 2nd Edition. Sweden, Malm¨o:

Studentlitteratur AB, Lund, 193 - 211.

(57)

[15] Nohrstedt, L., (2018). När kärnkraftverk avvecklas: Hopp om tröghet fr˚an vindkraft. Nyteknik. 4 April 2018. www.nyteknik.se/energi/

nar-karnkraftverk-avvecklas-hopp-om-troghet-fran-vindkraft-6908556.

[16] Svenska kraftn¨at. (2019). Kontrollrummet. Malin Stridh. https://www.svk.se/

kontrollrummet.

[17] Sheehan, R., (2007). Understanding and Applying Current-Mode Control Theory. Liter- ature Number: SNVA555. Dallas, Power Electronics Technology Exhibition and Confer- ence.

[18] Toshiba. (2018). Smart Gate Driver poupler TLP5214A/TLP5214 Application Note -Advanced edition-.

(58)

(59)

9 Appendix

9.1 Bill of Material per PCB

Nr Description Part References Value Quantity

1 2STR1160 BJT1 U13 BC848BLT1G 2

2 Unpol. Cap. C C1 C2 C30 0.1uF 3

3 Unpol. Cap. C C31 0.33uF 1

4 Unpol. Cap. C C9 C10 100pF 2

5 Unpol. Cap. C C18 C19 10uF 2

6 Unpol. Cap. C C15 C16 C17 1nF 3

7 Unpol. Cap. C C3-C8 C22-C25 1uF 10

8 Unpol. Cap. C C13 C14 220pF 2

9 Unpol. Cap. C C11 C12 300p 2

10 Pol. Cap. CP C20 C21 C26-C29 47uF 6

11 Schottky diode D Schottky D20 D21 D22 BAT46JFILM 3

12 Schottky diode D Schottky D5 D6 D9 D10 D13 D14 D DESAT 6

13 Schottky diode D Schottky D16 D17 D Retur 2

14 Schottky diode D Schottky D18 D19 D23 D Schottky 3

15 TVS diode D TVS D4 D7 D8 D11 D12 D15 D TVS 6

16 Zener diode D Zener D2 D3 D Zener 2

17 Light diode LED D1 LED 1

18 Inductor L L1 L2 L3 L4 10uH 4

19 Resistor R R1 0 1

20 Resistor R R8-R13 0 6

21 Resistor R R16 R17 100 2

22 Resistor R R26 10M 1

23 Resistor R R4 R5 R22 R23 10k 4

24 Resistor R R3 200 1

25 Resistor R R6 R7 220 2

26 Resistor R R24 R25 22k 2

27 Resistor R R18-R21 3,3k 4

28 Resistor R R14 R15 30k 2

29 Resistor R R2 47k 1

30 IC 2STR1160 U2 2N7002 1

31 IC 74HC132D U14 U15 U16 74HC132D 3

32 IC D-SUB U1 D-SUB 1

33 IC H2415SH U17 U18 H2415SH 2

34 IC L7805 U19 L7805 1

35 IC TLP5214A U3 U4 TLP5214A 2

Table 9.1: Bill of Material

(60)

9.2 Protection circuitry - NAND subcircuit

Schematic layout of driver circuit

Input HIGH Input LOW Error HIGH Error LOW Output HIGH Output LOW

0 0 0 0 0 0

0 0 0 1 0 0

0 0 1 0 0 0

0 0 1 1 0 0

0 1 0 0 0 0

0 1 0 1 0 0

0 1 1 0 0 0

0 1 1 1 0 1

1 0 0 0 0 0

1 0 0 1 0 0

1 0 1 0 0 0

1 0 1 1 1 0

1 1 0 0 0 0

1 1 0 1 0 0

1 1 1 0 0 0

1 1 1 1 0 0

Table 9.2: Truth Table - Control signals & Error signals

Input HIGH = H Error HIGH = EH Output HIGH = HO

Input LOW = L Error LOW = E_L Output LOW = L_O

(61)

Figure 9.1: NAND subcircuit

(62)

9.3 Protection circuitry - Signal input subcircuit

Figure 9.2: Signal conversion circuit

(63)

9.4 TLP5214 - External connections

Figure 9.3: TLP5214 - External connections

(64)

Figure 9.4: TLP5214 - Internal functions

9.5 Component value calculation

Parameters:

∆i_L 1A

V_ripple 1V f_sw 10kHz

Ts 100µs D 0 − 100%

V_out 100V Inductor current ripple:

∆i_L = V_in∗ D ∗ D⁰∗ T_s

2 ∗ L (9.1)

(65)

L_min = V_in∗ D ∗ D⁰∗ T_s 2 ∗ ∆iL

=MAX(D ∗ D⁰) = 0.25 = 1.25mH. (9.2) Output Ripple Voltage in CCM:

V_{ptp ripple}= 2 ∗ V_ripple = V_out(1 − D)T_s²

8 ∗ L ∗ C (9.3)

C_min = V_out(1 − D)T_s²

16 ∗ Vripple∗ L =MAX(1 − D) = 1 = 50µF (9.4)

9.6 MOSFET vs IGBT comparison

Name SKM350MB120SCH15 CM200DX-24T Comparison

Type MOSFET IGBT

V_DSS/V_CES[V ] 1200 1200 Same

I_D/I_C[A] 523 200 MOSFET +1

V_isol[V ] 4000 2500 MOSFET +1

R_CC⁰_+EE⁰[mΩ] 0.55 0.71 MOSFET +1

QG[µC] 1.512 4.5 MOSFET +1

R_G_int/r_g[Ω] 0.6 0.67 MOSFET +1

f_sw 1MHz 10kHz MOSFET +1

V_GS/V_GES[V ] −6...22 ±20 -

V_GS(th)/V_GE(th)[V ] 5.4...6.0...6.6 1.6...4.0 -

Table 9.3: MOSFET vs IGBT comparison

(66)

9.7 KiCad Schematic circuit

Figure 9.5: KiCad circuit - Overview 1

(67)

(68)

(69)

(70)

9.8 Simulink circuit

(71)

(72)

9.9 Matlab code

1 c l c; c l o s e a l l ; c l e a r a l l ;

2 %% V a r i a b l e s

3 f r e q s i m = 1 e6 ; % S i m u l a t i o n f r e q u e n c y

4 freq pwm = 1 e4 ; % P u l s e width m o d u l a t i o n f r e q u e n c y

5

6 t s i m = 1 0 ; % S i m u l a t i o n t i m e

7

8 P I i n n e r P = 0 . 0 5 ; % P c o n s t a n t f o r c u r r e n t c o n t r o l l e r

9 P I i n n e r I = 0 . 0 0 8 ; % I c o n s t a n t f o r c u r r e n t c o n t r o l l e r

10 P I o u t e r P = 1 5 ; % P c o n s t a n t f o r v o l t a g e c o n t r o l l e r

11 P I o u t e r I = 5 ; % I c o n s t a n t f o r v o l t a g e c o n t r o l l e r

12

13 %%%%%%%%%%%%%%%%%%%% FOR BUCK %%%%%%%%%%%%%%%%%%%%%

14 d e l t a I L = 1 ; % Allow i n d u c t o r c u r r e n t r i p p l e

15 Vout = 1 0 0 ; Vin = 1 0 0 ; % I n i t i a l i n p u t and o u t p u t v o l t a g e

16 V r i p p l e = 0 . 0 1 ; % 1 % v o l t a g e r i p p l e

17 D = 0 . 5 ; % Duty c y c l e us ed f o r component c a l c u l a t i o n

18

19 M u l t i p l i e r L = 1 0 0 ; % M u l t i p l i c a t i o n f a c t o r f o r i n d u c t o r

20 M u l t i p l i e r C = 1 0 0 ; % M u l t i p l i c a t i o n f a c t o r f o r c a p a c i t o r

21 L Buck = M u l t i p l i e r L ∗ Vin ∗ D ∗ (1 − D) ∗ (1/ freq pwm ) / (2 ∗ d e l t a I L ) ; % I n d u c t o r v a l u e

22 C Buck = M u l t i p l i e r C ∗ (1/ freq pwm ) ˆ2 / (16 ∗ L Buck ∗ V ri pple ) ;

% C a p a c i t o r v a l u e

23 %%%%%%%%%%%%%%%%% SUPER CAPACITOR %%%%%%%%%%%%%%%%%%%

24 R SC = 13 e −3; % Super c a p a c i t o r r e s i s i t a n c e

25 C SC = 7 0 ; % Super c a p a c i t o r c a p a c i t a n c e

26 %%%%%%%%%%%%%%%%%%% COMPONENTS %%%%%%%%%%%%%%%%%%%%%%

27 DC BUS R = 1 e −4;

28 DC BUS C = 7 0 ;

29 V D C B U S i n i t i a l = 1 0 0 . 0 1 ;

30 V S C i n i t i a l = 9 3 ;

31

32 %% S t e p r e p s o n s e 1

33 S t e p t i m e 1 = 1 ; % At which t i m e t h e s t e p r e p s o n s e s h o u l d o c c u r

34 S t e p v a l u e 1 = 1 0 ; % S t e p r e s p o n s e a m p l i t u d e

35

36 %% S t e p r e p s o n s e 1

37 S t e p t i m e 2 = 4 ; % At which t i m e t h e s t e p r e p s o n s e s h o u l d o c c u r

38 S t e p v a l u e 2 = 0 ; % S t e p r e s p o n s e a m p l i t u d e

(73)

39 40

41 t o f f s e t = 0 . 1 ;

42 %% S i m u l a t i o n

43 sim ( ' BDCDC simulink . s l x ') ; % A c c e s s i n g t h e S i m u l i n k f i l e

44

45 data M1 = yout . g e t E l e m e n t ( ' Mosfet 1 ') ;

46 t M1 = data M1 . V a l u e s . Time ;

47 M1 = data M1 . V a l u e s . Data ;

48 %

49 % data M2 = yout . g e t E l e m e n t (' Mosfet 2 ') ;

50 % t M2 = data M2 . V a l u e s . Time ;

51 % M2 = data M2 . V a l u e s . Data ;

52 %

56 %

60

61 data M5 = yout . g e t E l e m e n t ( 'Mode ') ;

62 t MODE = data M5 . V a l u e s . Time ;

63 MODE = data M5 . V a l u e s . Data ;

64

65 data VSC = yout . g e t E l e m e n t ('V SC ') ;

66 t VSC = data VSC . V a l u e s . Time ;

67 V SC = data VSC . V a l u e s . Data ;

68

69 d a t a I L = yout . g e t E l e m e n t ( ' I L ') ;

70 t I L = d a t a I L . V a l u e s . Time ;

71 I L = d a t a I L . V a l u e s . Data ;

72

73 data VDC = yout . g e t E l e m e n t ('V DC BUS ') ;

74 t VDC = data VDC . V a l u e s . Time ;

75 V DC BUS = data VDC . V a l u e s . Data ;

76

77 data DC ERROR = yout . g e t E l e m e n t ( 'DC BUS ERROR ') ;

78 t DC ERROR = data DC ERROR . V a l u e s . Time ;

79 DC ERROR = data DC ERROR . V a l u e s . Data ;

80

(74)

81 data PID OUTPUT OUTER = yout . g e t E l e m e n t ('PID OUTPUT OUTER') ;

82 t PID OUTPUT OUTER = data PID OUTPUT OUTER . V a l u e s . Time ;

83 PID OUTPUT OUTER = data PID OUTPUT OUTER . V a l u e s . Data ;

84

85 % data IL AV = yout . g e t E l e m e n t (' IL AV ' ) ;

86 % t IL AV = data IL AV . V a l u e s . Time ;

87 % IL AV = data IL AV . V a l u e s . Data ;

88 % IL AV = s q u e e z e ( IL AV ( 1 , 1 , : ) ) ;

89

90 data DUTY CYCLE = yout . g e t E l e m e n t ('DUTY CYCLE ') ;

91 t DUTY CYCLE = data DUTY CYCLE . V a l u e s . Time ;

92 DUTY CYCLE = data DUTY CYCLE . V a l u e s . Data ;

93

94 data DUTY CYCLE FILTER = yout . g e t E l e m e n t ( 'DUTY CYCLE FILTER ') ;

95 t DUTY CYCLE FILTER = data DUTY CYCLE FILTER . V a l u e s . Time ;

96 DUTY CYCLE FILTER = data DUTY CYCLE FILTER . V a l u e s . Data ;

97

98 data ERROR CURRENT = yout . g e t E l e m e n t ( 'ERROR CURRENT') ;

99 t ERROR CURRENT = data ERROR CURRENT . V a l u e s . Time ;

100 ERROR CURRENT = data ERROR CURRENT . V a l u e s . Data ;

101

102 % data m = yout . g e t E l e m e n t ('MODES? ') ;

103 % t m = data m . V a l u e s . Time ;

104 % m = data m . V a l u e s . Data ;

105 % m = s q u e e z e (m( 1 , 1 , : ) ) ;

106

107 %% P l o t s

108 % s u b p l o t ( 3 , 2 , 3 ) ;

109 % p l o t ( t M1 ,M1)

110 % t i t l e ('MOSFET 1 ') ; x l a b e l ( ' Time ' ) ; y l a b e l ( ' Amplitude ' ) ;

111 % a x i s ( [ 0 t MODE( end ) −0.1 1 . 1 ] ) ;

112 %

113 % s u b p l o t ( 3 , 2 , 4 ) ;

114 % p l o t ( t M2 ,M2)

116 % a x i s ( [ 0 t MODE( end ) −0.1 1 . 1 ] ) ;

117 %

118 % s u b p l o t ( 3 , 2 , 5 ) ;

119 % p l o t ( t M3 ,M3)

121 % a x i s ( [ 0 t MODE( end ) −0.1 1 . 1 ] ) ;

122 %

Design and construction of a bidirectional DC/DC converter

Examensarbete 30 hp Juni 2019

Design and construction of a bidirectional DC/DC converter

Alexander Wallberg

Abstract

Design and construction of a bidirectional DC/DC converter

Contents

Abbreviations

List of Figures

List of Tables

1 Introduction

1.1 Background

1.2 Purpose & Goals

1.3 Report disposition

2 Theory

2.1 Assumptions

2.2 Transistors

2.3 Switched DC/DC converters

3 Method

3.1 Requirements

3.2 Converter topology

3.3 Control Strategy

3.4 Protection circuitry

3.5 Driver PCB design

3.6 Converter construction

4 Components & hardware

4.1 IGBT - CM200DX-24S

4.2 Driver - TLP5214

4.3 Components

5 Results

5.1 Simulations

5.2 Measurements and tests

5.3 Component values

6 Discussion

6.1 Power Transistors: MOSFET vs IGBT

7 Recommendations for future work

8 References

9 Appendix

9.1 Bill of Material per PCB

9.2 Protection circuitry - NAND subcircuit

9.3 Protection circuitry - Signal input subcircuit

9.4 TLP5214 - External connections

9.5 Component value calculation

9.6 MOSFET vs IGBT comparison

9.7 KiCad Schematic circuit

9.8 Simulink circuit

9.9 Matlab code