FACULTY OF ENGINEERING ASSIGNMENT (1) (Multilevel DC-DC Buck Converter 42V-14V)

COURSE CODE : SUBJECT LECTURER DIVISON : : :

KKS 4113 POWER ELECTRONICS 1 AHMED KHAN ELECTRICAL ENGINEERING

Please fill the following particulars:

NAME/ MATRIX NO

(1) Mohamed Faisal Mohamed Ahmed

(408 2009 642)

DATE OF SUBMISSION : 14 October 2011

Introduction:

DC to DC converter is considered to be one of the great economical importances in Todays society. They are built in their millions every year, and are perhaps one of the few electronic disciplines that use in our daily life.

The DC-DC converter is an electrical circuit that transfers energy from a DC voltage source to a load. The energy is first transferred via electronic switches to energy storage devices and then subsequently switched from storage into the load. The switches are transistors and diodes; the storage devices are inductors and capacitors. This process of energy transfer results in an output voltage that is related to the input voltage by the duty ratios of the switches.

It is insightful and worthwhile to investigate why DC-DC converters are necessary before their detailed presentation and analysis. In addition to the constraints of size, weight, and cost, DC-DC converter technology also addresses the issues of efficiency and regulation.

In general, DC to DC converters are widely used at home and industries to produce the desired output power. Traditionally, different types of converters include buck, boost, buck-boost, cuk, voltage-source inverter, etc. With these converters, they can perform different types of applications: step-down of voltage, step-up, inversion of polarity, and conversion of dc to dc or vice versa. In recent years, with the proposed of a 42V in future automotive electrical system, which is three times higher than used in todays vehicle. It becomes inevitable that High Power Converter has gained popularity in automotive industry.

Objectives of this assignment: The objective of this assignment is to design and build a prototype Multi Level Converter capable of stepping down from 42V to 14V, DC-DC under high power and current condition. Comprising buck convertor, voltage regulator and driver embedded on to a single PCB board. In order to design the hardware, certain specifications need to be met.

Types of convertors: Dc/dc converters can be divided into two broad categories: a) Non-isolated dc/dc converters b) Isolated dc/dc converters

Non-Isolated DC/DC Converters The non-isolated converter usually employs an inductor, and there is no dc voltage isolation between the input and the output. The vast majority of applications do not require dc isolation between input and output voltages. The non-isolated dc-dc converter has a dc path between its input and output. Battery-based systems that dont use the ac power line represent a major application for non-isolated dc-dc converters.

Isolated DC/DC Converters For safety considerations, there must be isolation between an electronic systems ac input and dc output. Typical isolation voltages for ac-dc and dc-dc power supplies run from 1500 to4000V, depending on the application. Isolated dc-dc converters use a switching transformer whose secondary is either diode-or

synchronous-rectified to produce a dc output voltage using an inductor capacitor output filter. This configuration has the advantage of producing multiple output voltages by adding secondary transformer windings. For higher input voltages (48V) transformer isolated converters are more viable.

1. Convertor theory:

Multi-Level Converter 42V to 14V

It is a DC-DC converter capable of stepping down from 42V to 14V under high power condition. Advantage of 42V to 14V system Can meet increasing demand of load in the future. Cheaper to implement as no expensive and heavier wiring is required. No modification circuitry is needed for existing equipment. A universal standard for car manufacturers to follow. Easier to troubleshoot as the circuit is simple in the case of fault. Reduction in fuel consumption and CO2 emissions.

y y y y y y

Overview of 42V system 42 volts is one of the key enabling technologies in the automotive industry for the 21st Century. It will enable other advanced technologies that bring enhancement to all areas of the vehicle. Without it many would struggle or not be applied at all. The demand for electrical power in passenger cars has been growing steadily as vehicle performance increases and more features are added to improve performance, comfort and convenience, and safety. Power demand on small cars has doubled from 500W in 1990 to approach 1KW today, while medium cars now average 1.5KW and

large cars exceed 2KW. By 2005, demand will have reached 1.5, 2.8 and 3.5KW respectively. The practical limit for light vehicle electrical systems is about 200 amps, beyond which the wire diameter becomes too large and bulky to handle. 200 amps represents about 3KW for a 14V system, so it is clear that if future demand is to be met, then higher voltages will have to be used. Figure 1 display the electrical power consumption trend from 1980-2005.

Figure 1: Electrical Power Consumption Trend Thus 42V will be essential to meet future electrical demands on passenger cars, but it will also bring major benefits, both in overall fuel consumption, and in the performance and efficiency of many vehicle systems. 42V alternators are significantly more efficient at all engine speeds than current 14V and will offer major savings in fuel consumption.

ii Bucker converters The most widely used dc-dc converter topology in power management and microprocessor voltage-regulator (VRM) applications those applications require fast load and line transient responses and high efficiency over a wide load current range. They can convert a voltage source into a lower regulated voltage. For example, within a computer system, voltage needs to be stepped down and a lower voltage needs to be maintained. For this purpose the Buck Converter can be used. Furthermore buck converters provide longer battery life for mobile systems that spend most of their time in stand-by. Buck regulators are often used as switch- mode power supplies for baseband digital core and the RF power amplifier.

Circuit diagram and key waveforms:

(a)

(b)

(C)

Figure 2 (a) Buck- converter (b) switch on for a time duration DT (c) switch off for time duration (1-D) T (d) key waveforms. Circuit description: The three basic dc-dc converters use a pair of switches, usually one controlled (eg. MOSFET) and one uncontrolled (ie. diode), to achieve unidirectional power flow from input to output. The converters also use one capacitor and one inductor to store and transfer energy from input to output. They also filter or smooth voltage and current. The dc-dc converters can have two distinct modes of operation: Continuous conduction mode (CCM) and discontinuous conduction mode (DCM). In practice, a converter may operate in both modes, which have significantly different characteristics. Therefore, a converter and its control should be designed based on

both modes of operation. However, for this course we only consider the dc-dc converters operated in CCM. Circuit Operation: When the switch is on for a time duration DT, the switch conducts the inductor current and the diode becomes reverse biased. This results in a positive voltage  = Vg  across the inductor. This voltage causes a linear increase in the inductor current . When the switch is turned off, because of the inductive energy storage . continues to flow. This current now flows through the diode, and  = - for a time duration (1-D) T until the switch is turned on again.

2. Design Consideration: In the field of High Voltage Power Conversion, the circuit designer is often confronted with difficulties in finding semiconductors capable of sustaining the desired voltage. In terms of the high power converter, the upper most important issues are power efficiency and the selection of components. This chapter will cover the design considerations of a DC-DC bucker converter, which are of particular concern to the selection of components. A DC-DC bucker converter comprises of N-channel mosfet, inductors and capacitors. The operation and characteristics of individual components are very critical as this will subject to poor efficiency and performance.

42/14 Volt standardization: A draft standardization all 42 Volt in vehicles was introduced so that this can as be a guide to all. The most important about this draft is about the maximum dynamic over-voltage, as it removes the current load dump pulse. The main objective of the standard is to establish permitted voltage levels and test procedures for the loads without specific technical solutions. 42V/14V standardization is also important to components such as capacitors as they are chosen on the maximum over-voltage that they can withstand.

Table below provides a 42V/14V standardization summary on multi-voltage in vehicle electrical system.

Specification for 42V/14V vehicle electrical system From the Table above, components selected must have their rated voltage greater than the maximum permissible voltage as mention above. This is one of the design considerations.

Operation of Power MOSFET: Before selection of Power MOSFET, it is better to have some understanding on the operation of Power MOSFET. A metal-oxide-semiconductor field effect transistor (MOSFETs) is a three-terminal device where the input, the gate controls the flow of current between the output terminals, the source, and drain. [10] In the DC-DC bucker converter application, the MOSFET is used as a switch to control the duty-cycle. By controlling the duty-cycle, the desired output voltage can be achieved.

Basic construction of N-channel Mosfet

N-channel mosfet in the off-stage Gate-to-source voltage (VGS) applied across the depletion region of the p-n- junction. Since the gate-to-source voltage is not sufficient high enough to attract the minority carrier to form a channel at the surface of the P-region, therefore no drain current is flowing.

N-channel mosfet in the on-stage Now that the positive gate-to-source voltage (VGS) applied is greater than threshold voltage and is sufficient to form a channel at the surface of the P-region, underneath the gate.

Heat sink During high speed switching, MOSFET tends to generate some power losses. This power loss is dissipated as heat and thus causes the temperature to rise. Often this temperature increases beyond boundary condition that the MOSFET can function properly. In order to tackle this problem, a heat sink is introduced. Heat sink can be used to cool down the heat dissipated by the MOSFET. The choice of the proper heat sinks depend on the allowable junction temperature (Tj,max), the maximum ambient temperature (Ta,max), the maximum operating voltage and maximum on-state resistance.

Principle of step-down buck converter: The principle of buck converter is illustrated by Figure below. When switch is closed for a period of t1, the input voltage Vs appears across the load. On the other hand, the voltage across the load is zero if the switch remains off for a time t2. The chopper switch can be implemented by using a (1) power BJT, (2) power MOSFET, (3) GTO, or (4) forced-commutated thyristor. These devices has a finite voltage drop ranging from 0.5 to 2V, and for the sake of simplicity the voltage drops of these power semiconductor devices are neglected.

Step-down chopper with resistive load

Copper track size In designing the hardware in a printed circuit board, every designer has to be confronted by the amount of copper track size needed. Traditional, copper track is used to carry current. Making the copper track size too small can cause the temperature to rise tremendously. This is due to the amount of current exceeded the current capacity of the copper track size. Resulting in severe damage to the copper track and the circuit board. Vice versa, making the copper track too big can cause the power losses to increase and thus reduce the power efficiency and the performance of the circuitry. The current capacity of the copper track is depended on:  Temperature rise above ambient  Cross-sectional area The internal copper track width can be calculated:

In the case of external layers, copper track sizes are determined by

Where I T A

= maximum current in Amps = temperature rise above ambient in C = cross-sectional area in square mil (A mil is 1/1000 of a inch)

3. Specification of the control circuits: Following the design considerations from previous second part of the assignment, this part will touch on an in-depth design process for the Multi-Level Converter Bucker converter, starting with analytical view from a block diagram. Multi-Level Converter consists of voltage regulator, pulse-width modulation controller, Mosfet driver and bucker converter. Due to time constrain and the complexity of pulse-width modulation controller, this design is not implemented.

Multi-Level Converters: Multi-Level Converters possess advantages over classical converters in high voltage applications. These advantages include greatly reduced switching losses, reduced common mode currents and avoidance of simultaneously switched series devices.

Multi-Level DC to DC Converter

Multi-level Bucker Converter:

Multi-Level Converters consists of 3 multiple levels which has two switching devices (upper and lower MOSFET) which is driven by a MOSFET driver. The upper MOSFET is to provide current to the load during on time, whereas the lower MOSFET is to pull the convertor to the off stage and increase the reduction in current flow.

Multi-Level Bucker Converter

Finding the power loss:

To find the power losses, first we must find the ripple current. Table below provide the require data needed to calculate the ripple current.

Data for Bucker Converter

Refer to Multi-Level Bucker Converters figure. Voltage across inductor, VL = Vin Vout => 42-14 = 28V Therefore, switching time for the MOSFET must be:Switching time, ts = 1/fs => 1/200 KHz = 5sec

Switching Time of MOSFET Figure above illustrates the turn on period of the MOSFET. It is known that from the Table (Data for Bucker Converter) that the duty-cycle is 33%. Therefore the turn on time of the MOSFET must be:Turn on time, = 33% x ts (switching time) = 0.33 x 5sec = 1.65sec Inductor:

Ripple Current (Inductor)

As we know that ripple current is caused by instantaneous switching of the MOSFET, ripple current can be found using:Ripple current,

= (28V / 10H) 1.65sec = 4.62A Where the inductance was chosen to be 10H, which has an impedance of 0.006. As mentioned before, this design is catered for 55A at the load. Since this is a 3 level converter, therefore each level must share a burden of 18A. This inductance is selected because it can take a maximum current of 20A.

Capacitor:

Ripple Current (Capacitor)

= = 4.62 / 2 = 2.31A
From this, ripple current of the capacitor can be determined using:-

MOSFETs Current in 3 Level As shown in Figure above, this multi-level converter is divided into 3 stages. For simplicity, we just find power losses for 1 stage. Then we can get the total power losses for the 3 stage by multiplying by 3. Power losses for the MOSFET (Top and Bottom) can be determined:

The N-channel MOSFET each having 0.027 y Power Losses (Top) = 3 x Irms x R = 3 x 121 x 0.027 = 9.8W y Power Losses (Bottom) = 3 x Irms x R = 3 x 225 x 0.027 = 18.2W

Voltage regulator: The voltage regulator performs the critical task of supplying the desired voltage to the respective integrated chip. In this case, the MOSFET driver required 14.4V as supply. As mention before, the input voltage to the printed circuit board is 42V. Therefore, the voltage regulator needs to step down from 42V to 14.4V. An ideal voltage regulator is LM317HVT. It is a 3 terminal adjustable regulator capable of supplying exceeding 1.5A over a 1.2V to 57V at the output and input voltage can go up to 48V. Thus, this makes it an excellence choice for this application. In addition, better performance in line and load regulation can be achieved than any standard fixed regulators. LM317HVT offering full overload protection in IC can enhance the performance. The full circuit is shown in Figure below The operation of LM317HVT is simple. It makes use of voltage divider rule using two resistors. By setting R1 fixed at 240, R2 can be adjusted to get the desired output voltage.

Schematic of voltage regulator (LM317HVT)

Matching MOSFET Drivers to MOSFETs: Selecting suitable MOSFET drivers to match this converter design is critical. Before any decision is made, the basic functionality of the driver must be listed. Below are a few that must have in order to satisfy this converter design application. They are:o Capable of controlling high voltage of MOSFET for both high and low side. o Have external adjustable dead time feature. (Optional) o High driver current capability. o Provided with an internal patented circuitry. (Replaced external boot-trap diode) o Half-bridge driver functionality

Schematic diagram for L6384 driver

The full schematic diagram is illustrated in Figure below. It comprises of MOSFET driver, voltage regulator and multi-level bucker converter.

Schematic Diagram

4. Improvement of efficiency: The evaluation, selection, and design of converters to meet given requirements are considered very critical. One important performance such as desired output current, power and voltage are not met due to the fact that incorrect mosfet, inductors or capacitors were selected. Internal resistance in component plays a vital role in determining the efficiency of the bucker converter circuit. Selecting capacitors with very low resistance may affect the acceptable ripple current. In contrast, having capacitors with very high resistance will increase the power loss and thus making the bucker converter inefficient. There must be a tradeoff between these two that high efficiency and acceptable ripple current can be achieved.

We can find the power efficiency by:

Power Efficiency (%) =

  

=

= 95.77%

5. Simulation and results: In this part, detailed description on the results obtained from the experimental are presented. This is to verify the theory with the actual circuit constructed. With particular interest to ripple current, power efficiency and duty-cycle.

Ripple Current: See Analysis on the ripple current minimization (next topic).

Power Efficiency: Power efficiency are particular important in every aspect. Poor efficiency equipment will be costly to maintain. Thus, power efficiency is always the first priority in circuit design. The power efficiency for the constructed converter is calculated as below: y Experimental Results: Given that load resistor, RL = 3 and load current, IL = 5.18A Vin = 42V and IIN = 2245mA Output Power can be calculated using:-

Input Power can be calculated using:-

As mentioned in previous parts of this report, the theoretical power efficiency is 95.77%. From the actual result, the power efficiency is 85.37%. This is still acceptable as this is tested without a proper heat sink. The heat sink will cool down the MOSFET so that it will reduce power losses in switching due to heat dissipated.

Operation Of The Converter:

Analytical View on the Operation

Figure above shows the block diagram on the actual operation of the circuit. Voltage regulator constantly supply 14V to the half-bridge driver. This is to ensure that the driver chip can operate under optimal condition. Frequency generator with TTL signal is input to the driver at the same time. Control the duty-cycle of the frequency generator to 33% so that the half-bridge driver can drive the gate of the

upper MOSFET to turn on for 33% duty-cycle. When the upper MOSFET is off, the lower MOSFET is turn on for 66% to pull the bucker to 0V. This cycle will continue for all the 3 set of bucker converter. Having a duty-cycle of 33%, this can ensure that the output voltage is step down to 14V from the initial at the input. This is the overall summarize on the operation of the multi-level converter. The waveform for the output voltage with respect to the duty-cycle of the frequency generator is shown in next figure.

Output Voltage At 14V Versus TTL Signal

The diagram above provides the resultant waveform of 14V at output after adjusting the duty-cycle of about 33%. This design has proven the theory behind the bucker converter and thus it is successful.

Current Limiting:

Without a heat sink, it is observed that the power supplys current limiting is activated after a short while. This can be easily explained. The power supply is set to a minimum permissible current initially. Due to fast switching in a continuous conduction mode, the temperature in the MOSFET increased tremendously without a heat sink. Thus, this causes the power loss in the MOSFET to increase since power loss is directly proportional to the temperature.

Since,

6. Analysis on the ripple current minimization: In continues conduction mode of a bucker converter, the ripple current is created. It is due to the difference between the instantaneous values of the waveform and its fundamental-frequency component. Figure below provides detail of the input ripple current flowing in a bucker converter circuit. Since the input is a DC source, therefore the input is a square wave.

Input ripple current in a bucker converter

Input ripple current waveform Selection of components in a bucker converter is very critical as this may determine the success or failure of the whole operation. Thus, it is important to know how much ripple current is flowing into the capacitors and whether the capacitors are suitable for this amount of ripple current.

When designing a bucker converter, several considerations need to be taken. One of them is ripple current flowing in the capacitor as shown in Figures below Figure below shows the direction of output ripple current in a bucker converter. The output ripple current in a repetitive waveform are shown in Figure 2-7

Output ripple current in a bucker converter

Output ripple current waveform

Ripple current minimization: Capacitors are needed to reduce the actual ripple current flow. The input capacitors are used near the source is to filter away the unwanted noise. Whereas the output capacitors near the load are used to get a purer waveform. The capacitors rated ripple current must be sufficient to take in the actual ripple current flow.

Insufficient capacitors will cause some noise in the input. Thus, this will cause the waveform at the output to be impurity and generate more noise. Figure below is the waveforms comparing between insufficient and sufficient capacitors.

Ripple Current across Inductor As illustrated in Figure above, theoretically lesser capacitors will deduce larger ripple current. This can be proved by the experimental result shown below.

(a) With 3 output capacitors

(b) With 5 output capacitors Voltage across the inductor

From the constructed circuit board, we can see how the capacitors affect the value of the ripple current. Using oscilloscope, voltage across the inductor can be measured using the differences between the upper and lower voltage level of the inductor. According to ripple current calculation below, voltage is directly proportional to the current.

Where VL is the voltage across inductor  = Inductor current

= the change in current over the change in time

Thus, the characteristics of the ripple current can be observed. As illustrated before, voltage across the inductor is higher with 3 output capacitors as compared to using 5 capacitors. This supports the theory as mention above, which state that more capacitors will deduce smaller ripple current.

7. Applications in automotive electrical system: The availability of 42V power will result in the enablement of many new systems throughout the vehicle, and in the demand of many components. Almost all the accessories driven mechanically from the engine could be operated much more efficiently by electricity if sufficient power were available. In terms of comfort and convenience, 42V power will be important as a source of heat. In terms of safety, electronically controlled electric powered systems will offer benefits in steering, braking and handling. 1- Electric Power Steering Current 14V systems limit the application of electric power steering to small and medium size cars because of the large current needed for slow speed man oeuvres, 42V will allow the use of electric power steering on all sizes of cars and light commercial vehicles. 2- Engine Water And Oil Pumps Electrical loads such as speed variation and electronic control needed for the water pump motor is best provided by a 42V system, as this will save fuel while improving engine efficiency. 3- Electromechanical Brakes 42V are essential to handle the power required for an emergency stop, and to minimize the system weight. 4- Electrically Powered Suspension Control 42V system can cover suspension, steering and brakes.

5- Electromagnetic Valve Operation (EMV) Allows engine valves to operate completely independently, giving more flexibility than any crankshaft driven variable valve system. EMV will not be possible until 42V systems are developed.

8. Future improvements and conclusions: An overview on the results obtained and the future development of higher power in vehicle are in the concluding statement. This part concludes this assignment by suggesting some possible future recommendations to the enhancement of this Multi Level Converter 42V to 14V. Future Improvements: Input Voltage Warning device Another future feature is adding an external warning circuitry to monitor the input Voltage. If the input voltage drops below a minimum value, this converter will have difficulties in converting the desired 14V at the output. Thus, this implementation can serve as a tool for a better multi-level converter with the minimum interruption to the load.

Integrated With PWM Controller In order to make it more users friendly, we can include the PWM controller into

the printed circuit board as well. This future enhancement can run the multi-level converter without using external frequency generator to input logic signal TTL to the driver for controlling the duty-cycle of the switching devices. Thus, making it more compact and multi-tasking.

Conclusions: The objectives of this assignment were to demonstrate the successful implementation of the Multi Level DC to DC Converter 42V to 14V. After a brief overview of the background information, design considerations of the operation of the MOSFET and standardization of 42V were discussed. This assignment particular made to the calculation on the ripple current. It is found to be 4.62A. Thus, the capacitors rated ripple must be greater than this actual ripple current. Theoretically, if the actual ripple current is more than the capacitors can take. There would be unwanted noise and the output waveform will not be pure. The current passing through the inductor will be steeper. This theory is proven by the experimental results obtained. Simply adding capacitors, we can see a significant drop in the ripple current. This proven the theory that having more capacitors will reduce the ripple current. Calculations on the power losses in capacitors, MOSFET and inductors have been made. It is found that theoretically this Multi Level Converter has an efficiency of 95.77%, which is within the tolerance level. Meanwhile, the power efficiency from the experimental one is found to be 85.37%. This is still acceptable as this circuit board was tested without a proper heat sink. Therefore, the MOSFETs temperature will rise after a short time. Thus, this causes the power losses to increase and reduce the power efficiency. Power efficiency will improve once there is a good heat sink.

References: [1] Mohan, Undeland and Robbins, Converters, applications and design 2nd edition, Power Electronics, John Wiley and Sons, pp.161-195, pp.737-739, 1995. [2] R.W. Erickson, Converter Circuits, Fundamentals of Power Electronics, Kluwer Academic Publishers, 135-176, 1999. [3] T.A Maynard, Multi-level Conversion: High voltage choppers and voltage-source inverters, IEEE Industrial Electronics Society, 1992. [4] Hans-Dieter Hartmann, SICAN GmbH, Hannover, Standardization of the 42V Power Net History Current Status, Future Action. [5] Quarterly Technology And Applications Magazine, On Display, Vol 1, No 3, 2000. [6] NATIONAL SEMICONDUCTOR Datasheet, LM117HV/LM317HV 3-Terminal Adjustable Regulator, April 2000.