42V power system architecture development

M.A. Shrud, A.Bousbaine, A.S.Ashur,R. Thorn*, A. Kharaz

University of Derby, UK, University of Al-Fateh, Tripoli, Libya, *Victoria University, Australia.

Keywords: Dual-voltage electrical power system, automotive, power generation, dc-to-dc power modules, 42V power system architecture, and Matlab\Simulink. Abstract With the increasing demand for more fuel efficiency and environmentally friendly car coupled with the consumers drive for more comfort, safety and luxury car has led to the introduction of more electrical and electronic systems to the passenger car. This is further impacted by the current trend in automotive industry to replace mechanical and hydraulic system with their electrical counterparts. The handling capability of the current 14V DC system is getting very close to reaching the limits. To meet the new growing electrical power demands with minimum fuel consumption and minimum environmental effects, the automobile industry is looking into increasing the present voltage threefold, from 14V to 42V for future cars. A shift towards a 42V system will cut the current of the vehicle by a factor of three. With the lower current, the size and cost of power semiconductors can be reduced, allowing for their use in applications that could not use semiconductors before. In this paper, a detailed mathematical model for a 3-phase, 4kW and 42V Lundell alternator average electrical equivalent circuit will be presented along with the DC/DC converter based architectures for dual-voltage systems. The performance of the 42V Lundell alternator with the interleaved six-phase buck dc-to-dc converter system is modelled using Simulink software to assess the effectiveness of the model and its transient behaviour. 1 Introduction Today, one of the major trends in the automotive industry is the increasing amount of installed electrical and electronic system on the passenger car which results in a growing consumption of electrical energy. There are several reasons for the electrification of many automotive functions and the introduction of new features. Today's consumer wants more and more features in their cars to increase comfort, safety and luxury. The rain sensor or the electronic seat position control is an example of an electronic control system that improves the comfort, while navigation and entertainment accessories are examples of luxury. Another strong trend in the automotive industry is to replace mechanical and hydraulic powered components by introducing new electrically powered solution. In this way, they consume energy only when they are in use, resulting in lower fuel consumption and better

overall system efficiency. Examples of these, includes electric power steering, pump-motors for engine cooling fans and water pumps. Additional pressures are due to the increasing demand for environmentally friendly car with less pollution. However, looking back through the years, the increasing loads are not recent phenomena. In the near future, higher growth in the average power for vehicle loads is expected to rise to 3.5kW by 2015 as shown in Figure 1[27]. This further corroborated by the automobile industry which estimate that power demand will be in the range of 4 to 5kW by 2010. This tendency will push the electrical power demand beyond the handling capability of the today's standard 14V DC system which is around 1kW with peaks above 2kW [20]. In order to meet the growing electrical power demands with minimum fuel consumption and minimum environmental effects, the automobile industries have agreed to increase the present voltage to 42V, given the name 42V Power Net, which represents a three-fold increase in the system voltage. The chosen 42V is a compromise between the technical demand for increased voltage and personal safety, 60V [37]. The work on the new 42V supply systems for passenger cars is mainly carried out by two centres, namely the Consortium on Advanced Automotive Electrical/Electronic Components and Systems, established in Massachusetts Institute of Technology in America (MIT) and the Forum Bordnetz (or SICAN forum), which represents the European group of automobile manufacturers and suppliers. The two groups have outlined the existing and future voltage requirements of various vehicle components (Table 1). These specifications are now widely accepted in the automotive industry [12,35].
250 200 Current Wattage 3500 3000 2500 150 100 50 0 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2000 1500 1000 500 0

Current , A

Year

Figure 1: Typical vehicle average load current draw versus time.

Wattage ,w

Electrical system Battery voltage Nominal voltages Maximum voltages Maximum dynamic voltage Power Electronics rating

14V Car 12V 14V 24V -

28V Truck 24V 28V 34V -

42V PowerNet 36V 42V 50V ( ripple ) 58V (Load Dump)

60-40V

80-60V

100-75V Figure 2: Potential 42V Power-Net Alternator Architectures architecture, where the field current is used to regulate the 42V bus while the phase controlled rectifier is used to control the low voltage 14V bus. The drawbacks of dual-rectified alternator is the difficulty in independently and dependably regulate both the 14V and 42V buses under a range of loading scenarios and large output filters are required due to the very large voltage ripple created by switching of the power electronic devices, thyristors [7, 9]. An alternative approach is the introduction of dc-to-dc converter-based architecture [1, 36] (Figure 2c). The 42V alternator generates the required 42V to supply the heavyload and this is further processed by a dc-to-dc buck converter to supply power to the conventional automotive loads that expected to remain at 14V level. This architecture is considered to be, technically, a viable solution for automotive dual-voltage power system for passenger car in the near future. An interleaved dc/dc converter system with six-cells is chosen as it provides a fast dynamic response, a good power management and filtering processes [16, 5, 15, 21] The ability to model and simulate engineering design of the electric power system of a vehicle is essential before proceeding to the engineering experimental phase. Hence, DC-to-DC converter-based architecture, which is composed of a 42V Lundell claw-pole alternator and interleaved sixphase dc-to-dc buck converter, is modelled using simulink software. Initially, the system is subdivided into three main sub systems, namely power generation, conversion and load. Each subsystem is analysed separately, then combined together to form the whole system for real time analysis and evaluation using control-oriented simulators Matlab\Simulink. The simulation results add significant understanding to the behaviour of the system model under different load and transient conditions while at the same time obeying the automotive specifications.

Table 1: 14V, 28Vand 42V system specification A shift towards a 42V system involves the reduction of current for the same amount of power resulting in a decrease in wiring harnesses copper content, which consequently less weight and lower consumption, as well as the ability to operate high power loads more efficiently. Unfortunately, economic and technical considerations do not allow an easy transition to single 42V PowerNet and the abandonment of the 14V system to history. Therefore, in the medium to long term most car manufactures will provide cars with dual voltage systems (42V-14V), where the 42V power distribution systems will co-exist with the traditional 14V electrical system in the same road vehicle. In this way, the dual-voltage system provides a smooth transition period for loads to migrate to the single 42V system architecture. Various architectures for the dual-voltage electrical systems implementation have been proposed to meet 42V PowerNet system specification. The mains architectures under consideration in the automobile industry are a) dual-wound alternator architecture, b) dual-rectified alternator architecture and the dc-to-dc converter based architecture [26, 17, 31, 16] as shown in Figure 2. In the dual-wound alternator architecture (Figure 2a), a 42V alternator has two separate sets of stator windings, each supplying an output voltage via dedicated rectifiers. The 42V bus is supplied by one winding via a full-diode bridge rectifier, while the low voltage stator output is connected to phase-controlled rectifier to supply the 14V bus. The two outputs are controlled by a combination of field control and phase control. However, due to the fact that field control is common to both outputs, this poses serious difficulties in fully regulating both outputs, and in achieving good use of the alternator machine power capability under all operating conditions [16]. Figure 2b illustrates the structure of dual-rectified alternator [7, 32] where a single winding 42V alternator and dual-output rectifier used to supply both high and low voltage buses. The two outputs are controlled as in the dual-stator alternator

2 Power generation 2.1 Analytical model The system, for a non hybrid electric passenger car with a dual-voltage electrical architecture will be discussed with more emphasis on simulation. The proposed system is composed of a 4 kW power generation system buffered by a 36V energy storage battery and a 1kW interleaved six-phase buck converter The heart of any automotive electrical power system used in vehicles today is based on the claw pole alternator also known as the Lundell machine. This is a three-phase synchronous generator equipped with a field winding and brushes. It is modelled as a three-phase set of back emf voltages, with the leakage inductances, Ls, and winding resistance, Rs, connected in series as shown in figure 3. The output voltage of the alternator is controlled by regulating the field current, if and rectified using a six diode bridge rectifier. For each cycle of the input voltage there are six intervals of operation of the rectifier. During each of the six intervals, only two diodes conduct with the assumption that the amount of the short period of the overlap is zero. The mathematical model for operational characteristics of the circuit model of the synchronous generator connected to the full-bridge diode rectifier driving a constant-voltage load [8] is derived in appendix A and its equivalent circuit model is shown in Figure 4. Where Vg is the internally generated voltage and Zg the total synchronous impedance and given by equation (1), shown below.
3 3 3Ls 3 3R Vs cos = s + Is1 + 2Vd + Eo 6 2 2
Vg Zg ()

Figure 4: Synchronous generator model

id =

g (2V d + E o )

(2)

Where Vs is the peak phase voltage, Is1 is the magnitude of the fundamental component of the line current, and Vd is the diode voltage drop. 2.2 Alternators Electrical Behaviour Equations (1 and 2) are used to simulate the alternator averaged model using Matlab software, with the 14V and 42V Lundell alternator machines parameters shown in Table 2. In addition a simulink block diagram was created using the input parameters to produce the desired output simulation results, as shown in Figure 5. Figure 6 shows the performance curve for the present 14V automotive alternator which is capable of supplying approximately 93A of D.C. current at an alternator speed of 1800 rpm (ideal speed). As the alternator speed increases the maximum output dc current is about 124A at cruising speed (6000 rpm). For comparison the average alternator data for the Bosch NC 14 V 60-120A model is plotted which shows a good agreement between the developed model and the real characteristic of the machine. There are always losses when converting mechanical energy to electrical energy. A normally operated alternator has an average efficiency of 50% and decreases at higher speeds. The losses are related to iron losses (hysteresis and eddy currents), copper losses, friction and aerodynamic losses[6, 10,]. The relatively high losses are
k M a c h in e C o n s t a n t

(1)

The average current delivered to the constant-voltage load Eo by the averaged equivalent circuit model is given by equation (2).
Diode Rectifier
id D1 Rs Ls ia a ib b Rs V b Ls ic c D3 D5

Rotor field coil

Regulator

Rs V a

Ls

Slip rings

Eo

I F

F ile d

C u r r e n t O u t p u t P o w e r P o w e r

r p m

A lt e r n a t o r

S p e e d

V c

D4

D6

D2

E o

R e f e r e n c e

V o lt a g e

V d

D io d e

D r o p

R s

S t a t o r

R e s is t a n c e O u t p u t I n d u c t a n c e C u r r e n t C u r r e n t

Figure 3: Electrical model of a Lundell alternator system

L S

S t a t o r

N u m

b e r

o f

P o le s

Figure 5: Simulink functional model of the Alternator

Parameters Rs (Ohms) Ls( Henries) Vd (Volts) If( Ampers) k rpm p

Eo (Volts)

Description Stator winding resistance Stator leakage inductance

Diode drops Field current Machine constant Alternator speed Number of poles Alternator voltage

14V 33e-3 166e-6 1 3.6 4.2867e-3 700:1800 12

14

42V 133.65e-3 0.729e-3

alternator is illustrated in Figure7 which follows the same trend as for the standard 14V alternator. The power delivered at idle speed is approximately 3kW and 4kW at cruising speed. 3 Analysis and simulation of Lundell Alternator System

1 3.6 1.286e-2 700:1800 12

42

Table 2: 14V and 42V Lundell Alternator Parameters

a) Performance Curve for 1.7 kW / 14V Alternator
140 Output Current [A] 120 100 80 60 40 0 2000 Ideal Mode 4000 6000 8000 10000 12000 14000 16000 18000 Crusing Mode

To assess the behaviour of the output voltage, load current and load dump characteristic waveforms of the Lundell Alternator system a simulink model has been developed, figure 8, taking into account the parameters of the 42V alternator (Table 2). In addition the armature is modelled as a Y-connected set of sinusoidal three-phase back emf voltages, Va, Vb, Vc, stator windings inductances Ls and stator windings resistances Rs. The phase separation of the generated three-phase AC voltages are displaced by 2 / 3 radians. The generated three-phase AC voltage is rectified by a full-bridge diode rectifier to produce a DC output power required for the battery and the rest of electrical load system. Connecting the switch, SW 1 to a battery, permit the response of the system load dump to be activated.
Performance Curve for 4kW/42V Alternator Output Current [A] 100 80 60 40 20 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Ideal Mode Cruising Mode

b)
2500 Output Power [ W ] 2000 1500 1000 500 0 2000 4000

Output Power for 14V alternator

Cruising Mode Ideal Mode

Output Power for 42V Alternator Output Power [ W ]

6000 8000 10000 12000 14000 16000 Alternator Speed [ rpm ]

18000

5000 4000 3000 Cruising Mode 2000 1000 0 0 2000 4000 6000 8000 10000 12000 Alternator Speed [ rpm ] 14000 16000 18000 Ideal Mode

Figure 6: Performance curve and output power for 14V compensated by the lightweight, compact design and low investments costs. The demand for electric power in vehicles is increasing as more electrical loads are added in the cars. The economic feasibility of the present system continues to decrease. To meet the future electrical power demands, 42V automotive generators are becoming necessary. As illustrated in Figure 7, the 42V alternator is capable of supplying approximately 71A of D.C. current at ideal speed. As the alternator speed increases the maximum output d.c. current is about 15 A (cruising speed). Though the two alternators provide the same load current, the power capabilities are quite different. Figure 6, shows that the variations power curve for the standard Lundell automotive alternator with the speed. It can be observed that, up to the idle speed the power varies exponentially with the speed, linearly between the between the idle speed and 8000rpm and remains almost constant beyond this speed. The power delivered is about 1.35kW at idle speed, 1.7kW at cruising speed and a maximum power of 1.8 kW at speed beyond 1000rpm indicating a field weakening region. The effect of field weakening causes the torque to decrease as the speed increases while the electric power is kept constant. The power curve for the 42V Lundell

Figure 7: Performance curve and output power for a 42V alternator

Figure 8: Simulink model for the internal structure of the 42V

3.1 Simulation results

45

Rectified output voltage Output Voltage [V]

3.1.1 Rectified output voltage and load current. The simulations were performed at cruising speed, 6000rpm and at constant full field excitation current, if = 3.6 A . The simulated output behaviour of the alternator connected to a 4kW load and the battery is displayed in Figure 9. The first sub plot shows the systems rectified output voltage while the second sub plot shows the systems output load current. It can be observed that at the beginning of the cycle the output voltage rises from around 28V, with an overshoot, to 44V, over a 0.0015ms time span and settles to steady state of 42V behaviour within 0.01ms. The transient voltage deviations Vmin, Vnom, Vmax observed in this simulation are within the allowable required range of 42V PowerNet specifications and satisfy the 42V alternator system, table 1. The load current ramps up from 63A to 100A and then return to its steady state within 0.003ms. 3.1.2 Response to a step change in the load

40 35 30 25 Overshoot=44V

42V Setting time=0.005

0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 Load current at full load

0.01

120 Load Current [A] 100 95A 80 60 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 Simulation time [s] 0.01

Figure 9: The simulated output voltage and load current.

Step changes in load 100 Load Current [A] 75% 80 50% 60 40 20 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 full load-95A

Since the electrical load varies for various driving conditions such day or night, summer or winter; and city or country side driving. The simulation of load change is therefore a very important parameter for circuit behaviour. In order to study the 42V power generation dynamic performance under load variations, step change in loads have been investigated. Figure 10 shows simulation results of alternator rectified voltage to step changes in the load. The operation of 42V alternator system when load activated from 2kW to full-load, 4kW, is satisfactory with very little variation in the output voltage observed. These variations have only small influence on the output voltage and still respect the specifications of the 42V automotive standard. 3.1.3 Load dump control

Rectified output voltage response to step changes in load 60 Output Voltage [V] 50 40 30 20 10 0 0.05 0.1 0.15 0.2 0.25 0.3 Simulation time [s] 0.35 0.4 0.45 0.5 45V 43.6V 42V

Figure10: Dynamic response of system to loads variation

Rectified output voltage without load dump control

The fault condition which occurs when the battery is suddenly disconnected while the alternator is charging is called load dump and is a rather rare event. The standardised specification requirements impose transient voltage of less than 58V. Past research studies have been conducted to characterise and analyse the 14V automotive transient events [4, 13]. However, for the 42V alternators[22, 23, 25, 29], they have mainly concentrated on understanding the effect of different transient suppression devices on the bus voltage during a load dump, or have concentrated on active centralised load dump suppression using various high-current semi-conductor devices. The load dump simulation for the 42V electrical system has been conducted using the developed Simulink circuit model shown in Figure 8. With the battery switch closed, the DC bus voltage rise is maintained at the required value. When the system is suddenly interrupted by opening switch SW1 at 0.1s, the DC bus voltage rise reaches 74.84V, figure 11. This is greater than the standardised

80 Output Voltage [V] Peak voltage = 74.8 V 70 60 Overshoot = 44 V 50 40 Switch open 30 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 42 V

Load current without load dump control 200 Load Current [A] 170 A 150 100 A 100 95 A

50

0.02

0.04

0.06 0.08 Simulation time [s]

0.1

0.12

0.14

Figure 11: Simulated alternator voltage and load current during load dump

specification requirement limit of 58V. To meet the 58V load dump transient voltage of the 42V standard, a voltage limiting circuit has been designed, figure 8, where the damping resistor is activated using SW1 whenever the DC bus voltage exceeds the specification requirements. Figure 12 represents a transient event that occurs as result of a load dump limiting circuit. This is achieved by feeding back the output voltage that has been attenuated by a factor of 0.1 and this satisfies the specification range limits [4V-5V]. This output is used to generate a positive pulse to activate the switch SW. 4 DC-to-DC converter power module A switching converter is an electronic power system which transforms an input voltage level into another for a given load by the switching action of semiconductor devices. A highly power efficient dc-dc converters are strongly desired and have found widespread applications including aerospace [12], sea and undersea vehicles [34], electric vehicles (EV) and Hybrid Electric Vehicle (HEV) [19], and microprocessor voltage regulation [33]. A single step-down converter system typically involves switching circuits composed of semiconductor switches such as, MOSFETs and diodes, along with passive elements such as inductors, capacitors, and resistors (Figure13(a)). The gating signal, q (t) [0, 1] is the control variable that models the MOSFET switch. When the value of the control variable q(t) =1, the MOSFET is ON, and zero when control MOSFET is OFF. The main switching waveforms of the inductor voltage and current under steady-state conditions are sketched in Figure13(b) for Critical Discontinuous Conduction Mode (CDCM) and in Figure 13 (c) for Discontinuous Conduction Mode (DCM) of operation. Assuming that the converter is operating at Critical Discontinuous Conduction Mode (CDCM) of operation as shown in Figure13 (b), the operation of the circuit can be explained as follows; During the turn-on period of the high-side switch, the input voltage is connected to the inductor and the diode is OFF. According to Lenz's law, the voltage across any coil is equal to the instantaneous change in current multiplied by the self inductance of the coil. Therefore, the mathematical equations for this interval maybe written as;

average value of inductor current, is described by the following relationship;

I avg = Vi * D 2 (Vi Vo ) / 2 L * fs * Vo

(5)

At the end of turn-on period, as soon as the switch is turned OFF the diode turn ON to keep inductors current flowing. The rate of fall in the inductor current during the freewheeling period is
i / t = Vo / L
Rectified output voltage with load dump control
Output Voltage [V]

(6)

50 40 30 0 42 V 0.02

peak voltage = 45 V

0.04 0.06 0.08 0.1 Voltage attenuation factor [ 0.1 ]

0.12

0.14

Load Current [A]

6 4 2 1 0.5 0 150 95 A 100 50 101 A

0.02

0.04 0.06 0.08 0.1 The generated postive pulse

0.12

0.14

Output Voltage [V]

0.02

0.04 0.06 0.08 0.1 Load current with load dump control

0.12

0.14

Load Current [A]

0.02

0.04

0.06 0.08 Simulation time [s]

0.1

0.12

0.14

Figure12: The circuit feedback waveforms for load dump control

VL = L * I / t
The maximum inductor current is given by;

(3 )

Im ax = ((Vi Vo) / fs * L) * D

(4)

Where Vi and V0 are the input and output voltages respectively, fs = 1 / Ts is the converter switching frequency and D is switching duty cycle for MOSFET switch, defined as ton / Ts . Upon inspection of figure 3 (b), Im ax is twice the phase averaged current, Im ax = 2 * Iavg . Therefore, the

Figure13: (a) Single-buck converter topology, (b) typical CDCM excitation and (c) DCM excitation.

However, the above procedure could be repeated easily during the OFF interval, (1 D) * Ts to determine the same state variable equation of the average value of inductor current. Furthermore, as it can be seen from figure 3(c), the value of the inductance needed to ensure that the converter remains in Discontinuous Conduction Mode (DCM) of operation (where the inductor current is zero during part of the switching period and both semiconductor devices are off during some part of each cycle) must be less than L critical , and can be determined as follow;

L critical = (Vi * D 2 ) /( 2 * fs * Im ax ) * ((Vi Vo) / Vo ( 7 )

Where L < L critical However, the standard dc/dc converter with single structure is not feasible due to the low voltage, high current, and high operating temperature characteristics of the converter. Therefore, most power stage of the converter would have to be built in parallel for practical implementation. A common practical approach is to use the interleaved multi-phase technique instead of a single larger converter [24, 38, 39, 40].
4.1 Multiphase switching of dc-to-dc converter

The basic building block of the multi-cell interleaved converter is shown in figure 14. This represents six-cell interleaved buck converters which are connected in parallel to a common output capacitor and shares a common load with the associated control system. The low-voltage side is connected to the 14V automotive electrical loads while the high-voltage side to the on-board power generator (alternator) with nominal input voltage of 42V, and a range between 30V to 50V during normal operation. In this interleaved six-cell dc/dc converter architecture, the cells are switched with the same duty ratio, but with a relative phase shift or time interleaved of 60 introduced between each cell in order to reduce the magnitude of the output ripple at the output port of the converter. The overall output current is achieved by summation of the output current of the cells. With the phase of 60 the output of the converter is found to be continuous. Ripple reduction helps to reduce the filtration requirements needed to contain any EMI the converter produces and thereby decrease the constraints on the electronics components connected to the low-voltage bus. Furthermore, due to the equal sharing of the load current between cells, the stress in the semiconductor switches is reduced and thereby reliability is improved. Another advantage is the ability to operate the converter when a failure occurs in one cell as well as the possibility to add new cell to the converter with minimum effort. The ideal design is that the power management system should be smart enough to manage the key-off loads from depleting the high voltage battery to the point that the car cannot be started [20, 5, 30, 11].

Figure 14: Simulink implementation of the interleaved six cell buck converter circuit with PID controller Using dc/dc power modules two system structures are possible, distributed and centralized power conversion as shown in Figure 15. For automotive applications where volume, weight, and cost are particularly important. The preferred choice is the 42V/14V DC dual-voltage supply system centralised architecture with single battery and is based on the principle that the power processing is achieved by only one dc/dc power module to supply the existing 14V electrical loads. Also it lend itself to low cost, low weight, reduced packaging problems created by the second battery and reduction of the electromagnetic interference (EMI) generated by power switching devices in the system. Furthermore, the removal of the 12V battery does not alter the dynamic operation of the power converter, but the power of the converter must cover the power requirement of all the 14V power loads (approximately1000W) under the worstcase scenario. In addition, the non-isolated dc/dc converter topology is the most appropriate architecture because isolation between 42V and 14V buses is not required in the automotive power net and has the advantage over the transformer-isolation types in terms of an easy to design, low volume, weight, and cost. In this topology power flows in one direction, but bidirectional power flow can be achieved by replacing the lowside switch diode with a MOSFET. However, using two MOSFETs may result in lower losses but will require another gate driver and additional complexity in the control of the converter.

Parameter

Description

Value

Vi Vo Po T Vo Io

Input voltage range Output voltage range Power Rating Temperature range Output ripple voltage Output ripple current

30V < 42V<50V 11V < 14V <16V 1kW -40 C < T< 90C 300mV 1A

Table 3: Design specifications for a power converter in a dual-voltage automotive electrical system To design this six-phase interleaving buck converter system, the following automotive specifications for dual-voltage automotive electrical systems [9, 14, 18, 28] must be fulfilled as shown in Table 3. The specifications of the converter should meet the demand of the 14V electrical loads of 71A at low output voltage approaching 14V and to meet the operating temperature range, (105-125C) for thermal design requirements. For low voltage/high current power converter, the usage of MOSFETs switching devices with low onresistance is required for more efficient and practical power conversion. The inductance value L that guarantees the converter cells should run in the DCM over the entire operating range and can be calculated using Equation (7). Since this is a six-phase interleaving converter, the power stage inductance of each phase is therefore equal to 2.4H. The output capacitor is another important element, which may reduce the system cost in multi-phase converter system and is needed to keep the output voltage ripple VO within allowable output voltage range to meet the constraints of the design specifications. The necessary output capacitor has no severe effect on the value of the inductor current and the switching frequency of the converter. Also, the capacitor value does not necessarily have to be very large to smooth the output voltage. Table 4 shows the capacitor variation from 100F to 400F along with the value of voltage/current ripple. Figure 16 shows the plot of output ripple voltage versus capacitor value from the simulation analysis obtained. To meet the constraint of the design requirements concerning the voltage ripple of the converter system, a capacitor value of 300F is chosen. Figure15: Dc/Dc converter implementations: a) distributed and b) centralized

Capacitor value 100F 150F 200F 250F 300F 350F 400F

Voltage ripple 6mV 4mV 3mV 2.5mV 2mV 1.5mV 1mV

Current ripple 30mA 20mA 16mA 12mA 10mA 9mA 7mA

Table 4: capacitor value versus voltage /current ripple

Ripple voltage versus capacitor variation 0.02 V o lt a g e r ip p le [ m V ]

0.01

-0.01 100

150

200

250 300 Capacitor [ F]

350

400

Figure 16: Ripple voltage versus capacitor value

4.2 Control design strategies

The control strategy of the proposed converter is based on voltage-mode-controlled Pulse Width Modulation (PWM) with a proportional-integral-derivative (PID) which takes its control signal from the output voltage of the switching converter instead of current-mode (or current-injected) PWM, which utilises both the output voltage information and the current information from the inductor to determine the desired duty cycle. Simulink model for the internal structure of the PID used to control the converter is shown in Figure 17. The aim is to regulate the output voltage of the converter Vo across the load resistance RL to mach a precise stable reference voltage Vref. This is achieved by subtracting the desired reference voltage Vref from the sensed output voltage VO of the converter. The voltage-error thus obtained is passed through PID to obtain the desired signal. The individual effects of P, I, and D tuning on the closed-loop response are summarized in Table 5[3]. The desired output generated signal of the PID enters the PWM unit, where it is compared with the constant frequency saw tooth voltage Vpwm. The frequency of saw tooth voltage is the switching frequency fs of the converter which is100 kHz. The output of the PWM is the switching control signal, a sequence of square pulses that drives the semiconductor switch, as seen in Figure 18. The proposed converter necessitates a phase-shift of 60 between the cells to generate the six-switching control signal which are used to drive the six active MOSFET switching devices of the converter system. Figures 19 and 20 show the implementation of the six-phase interleaving circuit in Simulink and the six phase control signal waveforms respectively. Another approach is given in [2].

ClosedLoop Response
Increasing Kp Increasing KI Increasing KD

Rise Time
Decrease Small Decrease Small Decrease

Over shoot
Increase Increase Decrease

Settling Time
Small Increase Increase Decrease

SteadyState Error
Decrease Large Decrease Minor Decrease

Table 5: Effect of independent P, I, and D tuning

PID signal, PWM voltage

Comparsion of PID singal and Vpwm 1.5 Vpwm 1 0.5 0 2.62 PID signal

2.64

2.66

2.68

2.7

2.72

2.74

2.76

2.78

2.8 x 10
-3

Control singal features a constant duty cycle

Switch control signal

1.5 duty cycle 1 0.5 0 2.62

2.64

2.66

2.68

2.7 2.72 2.74 Simulation time [s]

2.76

2.78

2.8 x 10
-3

Figure18: Implementation of Pulse Width Modulation Simulink model.

Figure17: Implementation of PID controller in Simulink

Figure19: Six phases of interleaving in Simulink

Six-phase control signals

Cell current [A]

Current in each cell

2 1 0 2 1 0 2 1 0 2 1 0 2 1 0

40 20 0 1.3 1.35 1.4 1.45 1.5 x 10

-4

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9 x 10

7
-4

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9 x 10

7
-4

7 1 [ A ] - Total output current Load Current [A]

80 60 40 20 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9 x 10

7
-4

14 [ V ] - Output Voltage

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9 x 10

7
-4

Output Voltage [V]

20 15 10 5 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

6 2 1 0 6

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9 x 10

7
-4

Simulation time [s]

6.1

6.2

6.3

6.4 6.5 6.6 Simulation time [s]

6.7

6.8

6.9 x 10

7
-4

Figure 22: Top trace is the individual cell currents then total output current and bottom trace is the output voltage of the converter system
Parameter Name Input voltage Symbol Value Units

Figure20: Six-phase control signal waveforms

Total output current ripple C u rre n t rip p le [A ] 71.44

Vin Vo N L C RL fs

42 14 6 2.4 300 .196 100

volts volts H F kHz

Output voltage Number of phases Inductor value Capacitor value Load resistance Switching Frequency

71.42

71.4

2.02

2.03

2.04

2.05

2.06

2.07

2.08

2.09

2.1 x 10
-3

Output voltage ripple 14.005 V o lt a g e rip p le [V ]

Table 6: Parameters of the simulated converter Number of phases

Voltage ripple
2.02 2.03 2.04 2.05 2.06 2.07 Simulation time [s] 2.08 2.09 2.1 x 10
-3

14

10

13.995

8.7mV 45mA

2 mV 9mA

1.1mV 6mA

0.5mV 3.2mA

Current ripple

Figure 21: Current and voltage output ripples

5 MatLab/Simulink simulations and results

Table 7: Output voltage and current ripples versus the number of cells.

5.1 Ripple cancellation

The complete model implementation of the internal structure of the interleaved six-phase buck converter system shown in Figure 14 is implemented in Simulink software to obtain the necessary waveforms that describe the converter system operation for steady-state and transient conditions, using the parameters tabulated in Table 6.

The first step in the analysis of the multi-phase interleaved converter system is to investigate the effectiveness of ripplecancellation related to the variation of current and voltage as a function of the number of cells using the same settings for the control system. The results obtained are summarised in Table 7. From the results it can be observed that the converter achieves a very good current and voltage ripples cancellation for four-cells and above. However, though, eight or ten-cells cancellations are better than the four-cell, the implementation cost outweigh the gains in accuracy.

The result also point out that the EMI filter is not needed to reduce the peak to peak voltage ripple on the 14V terminals. This may lead to the elimination or redesign of the protection circuitry connected to the 14V bus. It can be seen from figure 21 that the ripple of the output voltage and the total inductor current of the power converter system are better than the desired specified limits indicated in Table 3. Figure 22 shows the steady-state waveforms of the individual cell currents, the total output inductor current and the output voltage. The simulated results show that the curves of the individual cell currents are balanced and the time interleaved of the cells is apparent from the relative time delay of each cell's inductor current. The inductor current in each cell rises to 30A during each switching period and goes through an interval in the discontinuous conduction mode. The sum of the individual cell currents result in a total current of 71A with a ripple current of around 9mA which is less than the individual cell ripple current. The simulated results indicate that, the operation of the power converter system is stable and accurate. The converter is able to respond and produce the desired stable output voltage and deliver the required total output current to the load with very low ripple. As a result, no negative effect on the connected loads, such as small motors, lights and accessories.
5.2 Transient simulation for load variation

voltage from 33V to 50V is applied to the model. Figure 24 shows a transient response of output voltage behaviour waveform due to the sudden changes in the input voltage of the power converter system. At the beginning of the cycle, at time t=0.004s, the input voltage suddenly rises from nominal system voltage of 42V, to 50V. The maximum output voltage (bottom trace) transient is 15.09V, but after a short period this error is leveled out in approximately 200ms with a maximum overshot of 1.09V. At time instants t=0.01s, when the input voltage suddenly changed from 50V down to 33V, the output maximum transient is11.704V. The settling time to return to 14V is approximately 0.4ms with maximum overshot of 2.296V. Finally, at time t=0.016s, when the input voltage suddenly jumps from, 33V to the nominal system voltage, the maximum output transient is 15.85V. The settling time is approximately 0.4ms with maximum overshot of 1.85V. The simulation results illustrate that the converter system has a strong immunity against line voltage disturbances even with the 12V energy storage battery being abscent.
5.4 Load variations and supply-voltage variations

The converter is used to supply power to various loads such as;

Small motors (2 to 8A @12V). Very small motors ( less than 2A @12V ) Lighting system: internal and external lights. ECU and Key-off loads

The combinations of both the supply-voltage and load variations that occur in the converter system have been simulated and the outputs are presented in Figure 25. It can be observed that the designed system has a low-sensitivity to the load and supply-voltage variations. These variations have only small influence on the output voltage and load current, they still respect the specifications of the automotive standard. It can be concluded that from the results obtained the proposed converter can maintain designed output voltage independently of load and supply-voltage variations.
Step load change from 50% to 75% to 100% and from 100% to 75% to 50%

Load current [A]

80 60 40 20 0 0
50%

full load 75%

The electrical loads demand varies and depend upon the weather and the driving conditions. A full load condition is rarely present for a prolonged period of time and most of the devices run at light loads (stand-by-mode) for most of the time. To study the effect of the load variation on the dynamic behaviour of the converter system, the load at the output of the converter system is suddenly changed from 50% to 75% and to 100% and than back from 100% to 75% and 50% of the full load at time t=0.002, 0.004 and 0.006s respectively. The simulated results are shown in Figure 23. It can be seen that the output voltage undershoot varies from 13.12V to 13.28V while the overshoot from 14.735V to 14.89V. When the load at the output of the converter system was suddenly changed from 50% to 75% and to full-load (1kW). The results show that the performance of the system is stable and well behaved under load variations (disturbances) and the output voltage remains within the desired specified limits presented in Table3.
5.3 Input voltage variation

0.001 0.002

0.003 0.004

0.005 0.006 0.007

0.008 0.009

0.01

Output voltage due to load change Output voltage [V] 20 15 10 5 0 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 Simulation time [s] 0.008 0.009 0.01

Figure 23: Transient response of the output voltage due to step changes in load.
6 Complete system model

In real conditions, the alternator output voltage ranges from 30V to 50V during normal operation, with nominal voltage at 42V. To study this line of variation, a step change in the input

Figure 26 shows the complete model of the Simulink implementation of the internal structure of the of 42/14V DC dual-voltage supply centralized architecture system with a single battery. The proposed complete system is composed of the Simulink model of the 42V voltage alternator (Figure 8), and the Simulink implementation of the interleaved six-phase buck conveter (Figure 14), load dump circuit and the PID

controller. Using the parameters tabulated in Tables 2 and 6, the Simulink model above is used to determine the steadystate and load dump transient characteristics.
6.1 Simulation Results

Figure 27 shows the simulated results of the two buses at different level of voltages and powers. The high voltage deliver 3kW at an average output current of 71A, while the voltage, 14V bus, delivers a current of 71A at 1kW. The simulation results illustrate that, the operation of the complete system is stable and accurate. The system is able to respond and produce the desired output voltages and required level of currents.
Input voltage variation 55 In p u tV o lta g e [V ] 50 45 40 35 30

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

20 O u tp u tV o lta g e [V ] 15 10 5 0

Output response of input voltage variation

Figure 26: Simulink model of the complete dc/dc converterbased system architecture with single battery.
0 0.002 0.004 0.006 0.008 0.01 0.012 Simulation time [s] 0.014 0.016 0.018 0.02

42[ V ] -42V bus voltage Ou tp u t Vo lta g e [V]

Figure 24: Output voltage due to step line voltage disturbance

Output current due to line and load variation 100 Load current [A]

40 20 0 80 60 40 20 0 20 15 10 5 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 71[ A ] -14V bus current 0.02 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
14 [ V ] - 14V bus Voltage

0.002 0.004 0.006 0.008

0.01 0.012 0.014 0.016 0.018

0.02

71 [ A ] - 42V bus current

50

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

Output voltage due to line and load variation Output voltage [V] 20 15 10 5 0 0 0.001 0.002 0.003 0.004 0.005 0.006 Simulation time [s] 0.007 0.008 0.009 0.01

L o a d C u rre n t [A]

Figure 25: Load current and output voltage due to under line voltage and load change disturbance The transient response of the 42V bus at 6000 rpm is shown in Figure 28. The transient is induced by disconnecting the 42V battery at time t=0.01s. The peak transient voltage on the 42V bus is approximately 54.46V which is less than the 74.8V peak voltage observed with a single voltage system (Figure 11). The reason for the relative stability is the remaining loads. The 14V bus voltage undershoot is 13.69V and the overshoot is 14.28V. The results show that the performance of the system is stable and well behaved under load dump and the output voltage remains within the desired specified limits presented in Table 3.

Ou tp u t Volta g e [V]

L o a d C u rre n t [A]

80 60 40 20 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

Figure 27: Voltage and current of the two buses The voltage can be further suppressed by the use of the voltage limiting circuit (Figure 8) and these shown are in Figure 29. It can be seen that the voltage rises only to 45V on the 42V bus and less overshot/undershoot on the 14V bus.

Load dump-dual voltage system

60
Output Voltage [V]

50 40 30 20 10

42V

54.46V
Step load changes from 50% to 75% to 100% and from 100% to 75% to 50% L o a d C u rre n t [ A ] 80 60 40 20 50 40 30 20 20 15 10 5 80 60 40 20 0 0.01 0.02 0.03 0.04 0.05 0.06 Simulation time [s] 0.07 0.08 0.09 0.1 0 0.01 0.02 0.03 0.04 0.05 0.06 Output current at 14V bus 0.07 0.08 0.09 0.1 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 14V bus voltage due to load change 0.08 0.09 0.1 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 42V bus voltage due to load variation 0.08 0.09 0.1

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

Load dump-dual voltage system

20
Output Voltage [V]

14.28V 14V 15 13.69V 10

0.002

0.004

0.006 0.008 0.01 0.012 Simulation time [s]

0.014

0.016

0.018

Figure 28: Dual architecture load dump of 42V battery

42V bus load dump using load dump limiting circuit 60
Output Voltage [V]

Figure 30: Dynamic response of the system to step change in load on the 42V voltage bus
Step load changes from 50% to 75% to 100% and from 100% to 75% to 50%

50 40 30 20 10 0 0.002

42V

45V
Load Current [A] 80 60 40 20 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

L o a d C u rre n t [ A ]

O u t p u t V o lt a g e [ V ] O u t p u t V o lt a g e [ V ]

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

Output Voltage [V]

14V bus voltage due to load variation 20 15 10 5 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

14V bus voltage 20 Output Voltage [V] 14.2V 15 13.8V 14V

Output Voltage [V]

42V bus due to load change 50 40 30 20 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

10

Load Current [A]

0.002

0.004

0.006 0.008 0.01 0.012 Simulation time [s]

0.014

0.016

0.018

Output current at 42V bus 80 60 40 20 0 0.01 0.02 0.03 0.04 0.05 0.06 Simulation time [s] 0.07 0.08 0.09 0.1

Figure 29: Dual architecture load dump of 42V battery with voltage limit circuit Figure 30 shows the simulation of the dynamic response of the system to a step change in load on the 42V bus at 6000 rpm (cruising speed). The transient is induced by sudden change of the 42V bus load from 50% to 75% and 100% and then back from 100% to 75% and 50% .The loading in the 14V bus remains at its nominal value of 71A. The voltage on the 42V bus deviates from its nominal by approximately 3V. This load step change has little effects on the 14V bus, this is due to efficient output regulation. The transient voltage deviations observed in on the 42V bus in this simulation are within the preliminary voltage limits specifications that were given in Table 3. As depicted in Figure 31, step change in load on the 14V bus again from 50% to 75% and 100% and than back from 100% to 75% and 50% the undershot is 13.3V while the overshoot is approximately14.7V. The voltage on the 42V bus with 3kW load also deviates from its nominal by approximately 1V. This transient output voltage remain within the allowable specification reported in Table 3.

Figure 31: Dynamic response of the system to step change in load on the 14V voltage bus
7 Conclusion

It can be concluded from the results obtained that the dc-dc based architecture system is a potential solution for a more efficient and stable automotive electrical power system. The system complies with the demanding requirements of the automotive industry in terms of current and voltage surges. It is anticipated that the system proposed here will be of value in future dual-voltage automotive electrical systems.

8 References

K.K. Afridi. A Methodology for the Design and Evaluation of Advanced Automotive Electrical Power Systems, Ph.D. thesis, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, February, (1998). [2] J.Abu-Qahouq, H. Mao and I. Batarseh. Multiphase Voltage-Mode Hysteretic Controlled DC-DC Converter with Novel Current Sharing, IEEE Transactions on Power Electronics, volume19, No.6, pp.13971407.Novmber, (2004). [3] K. H. Ang, G.Chong, and Y. Li. PID Control System Analysis, Design, and Technology, IEEE Transaction on Control Systems Technology, volume.13, No. 4, JULY, (2005). [4] J.Aikalay,H,Kendall,,M.Laskowski,A.Lee,and D.Noderer. Survey o Conducted Transients in the Electrical System of a Passenger Automobile, IEEE International Symposium on Electromagnetic Compatibility, Denver Co, (1989). [5] J.M.Altes ,E.G. Dolcet , and B.P. Solorzano. Analysis of the Most Appropriate Electrical Architecture and Communication Bus for the New Dual Voltage 14/42v System, IECON-2002: Proceedings of the 2002 28th Annual Conference of the IEEE Industrial Electronics Society, volumes 1-4 ,pp. 1687-1692, (2002). [6] K.Bolenz and W.Schleuter. Design Modifications of Alternator, Starter and Battery to Reduce Fuel Consumption., Proceedings of the International Congress on Transportation Electronics; pp.355-362, Dearborn, MI, USA, October, (1994). [7] J. C. Byrum. Comparative evaluation of dual-voltage automotive alternators, Laboratory for Electromagnetic and Electronic Systems (LEES), Cambridge, Tech. Rep.TR-00-012, November (2000) [8] V.Caliskan, D.J. Perreault, T.M. Jahns, and J.G. Kassakian. Analysis of Three- Phase Rectifiers With Constant-Voltage Loads, IEEE Transaction on Circuits and System-I: Fundamental Theory and Applications, volume50, No.9, pp.12201-226, September (2003). [9] V.Caliskan. Design and Analysis of Dual/High Voltage Automotive Electrical Power System Architectures, Ph.D. thesis, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, (200). [10] E. Ceuca, R. Joldes, E. Olteanu, Simulation of automotive alternator - solution for increasing electrical power, Proceeding of 2006 IEEE - The International Conference on Automation, Quality and Testing, Robotics, Cluj-Napoca, May 2006, vol 1, pp 292-297. [11] A.Emadi, S.S.Williamson, and A. Khaligh. Power Electronics Intensive Solutions for Advanced Electric, Hybrid Electric, and Fuel Cell Vehicular Power Systems, IEEE Transactions on Power Electronics, volume.21. No.3, May, (2006). [12] A. Emadi, M Ehsani, and J.M.Miller. Vehicular Electric Power Systems: Land, Sea, Air, and Space Vehicles, New York: Marcel Dekker, (2003).

[1]

[13] E.fland, M.Manternach, A. Marshall and J.Mings. The Load Dump ,Proceedings of IEEE Workshop on Electronic Applications in Transportations, Dearborn, MI, USA ,pp.73-78, October, (1990). [14] O.Garca, P. Zumel, A. de Castro, and J. Cobos. Automotive DCDC Bidirectional Converter Made with Many Interleaved Buck Stages, IEEE Transactions on Power Electronics, volume21, No.3, pp.578-586, May, (2006). [15] J.V.Hellmann and R.J. Sandel. Dual/high voltage vehicle electrical systems, SAE Paper 911652, SAE Future Transportation Technology Conference and Exposition, Portland, OR, August, (1991). [16] G. Hassan, D. J. Perreault, , and T. A. Keim. Design of Dual-Output Alternators With Switched-Mode Rectification, IEEE Transaction on Power Electrons., volume 20, No. 1, pp. 164172, January, (2005). [17] P. Hartnett, P. Miller, M. 42 V Powernet Enabling Technologies: Overview, Passenger Car Electrical Architecture (Ref. No. 2000/088) IEE Seminar, pp. 1/11/4, June, (2000). [18] L.Jourdan, J. L. Schanen, J.Roudet, M. Bensaeid, and K.Segueni. Design methodology for non insulated DCDC converter: application to 42 V14 V Power-net, in Proc. IEEE Power Electron. Spec.Conf. (PESC02), volume 4, pp.16791684, (2002). [19] L.A.Khan. DC-to-DC Converters for Electric and Hybrid Vehicles, Proceedings of IEEE Workshop on Power Electronics in Transportations, pp.113-122, (1994). [20] J.G.Kassakian, J.M.Miller. Automotive Electronics Power Up, IEEE Spectrum, pp.34-39, (2000). [21] J.G. Kassakian. The Role of Power Electronics in Future 42 V Automotive Electrical Systems, in Proc. 10th Int. Power Electronics Motion Control Conf. (EPEPEMC"02), (2002). [22] L.M.Lorilla. Enhanced Next Generation Alternator, Ph.D. thesis, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, September, (2005). [23] P. Le Bars and A. Regini. 42 V Load Dump Transient and Centralised Active Suppression, Passenger Car Electrical Architecture (Ref No. 2non~n88) IEE Seminar, pp.4/1 - 4/3, J u n e (2000). [24] B. A. Miwa, D.M. Otten, M.E. Schlecht.High efficiency power factor correction using interleaving techniques, in Proc. IEEE Appl. Power Electron. Conf. Expo (APEC92), pp. 55756, (1992). [25] C.P.Mudannayake and M.F.Rahman.A MATLAB/Simulink Model for a Prototype Integrated Starter Alternator for Automobiles", Power electronics and Motion Control Conference, IPEMC,The 4th International, volume 3, pp.1679-1684, (2004). [26] S. Muller and X. Pfab. Considerations implementing a dual voltage power network, Proceeding of IEEE SAE International. Conference on Transportation Electronics (Convergence), Dearborn, MI, (1998).

M. Miller. Multiple voltage electrical power distribution system for automotive applications, 31st Intersociety Energy Conversion Engineering Conference (IECEC), volume 3, pp. 19301937, (1996). [28] T.C.Neugebauer and D.J.Perreault. Computer-aided optimization of dc/dc converters for automotive applications, IEEE Transactions on Power Electronics, volume18, No.3, pp.775-783, May,( 2003). [29] C.S.Namuduri, B.V.Murty,and M. G.Reynolds. Load Dump Transient Control of a 42V Automotive Generator, 35th Annual IEEE Power Electronics Specialists Conference, Aachen, Germany, pp.389-394, (2004). [30] P. Nicastri and H. Huang. 42V PowerNet: providing the vehicle electrical power for the 21st century, SAE Future Transportation Technology Conf. Expo., Costa Mesa, CA, (2000). [31] D.J.Perreault and V.Caliskan. Automotive Power Generation and Control, IEEE Transaction on Power Electrons., volume 19, No. 3, pp. 618630, May (2004). [32] C. Patterson, J. O.Dwyer ,and T. Reibe. Dual voltage alternator", IEE Colloquium on Machines for Automotive Applications, London, England, November, pp. 4/14/5 (1996). [33] A. V. Peterchev. Digital PulseWidth Modulation Control in Power Electronic Circuits: Theory and Applications, Doctor of Philosophy in EngineeringElectrical Engineering and Computer Sciences, University of California, Berkeley, spring, (2005). [34] C.H. Rivetta, A.Emadi, G. A. Williamson, R. Jayabalan, and B. Fahimi. Analysis and Control of a buck DC-DC Converter Operating With Constant Power Load in Sea and Undersea Vehicles, IEEE Transactions On Industry Applications,volume.42.,No.2,March/April, (2006). [35] Z.J.Shen , F.Y. Robb, S.P.Robb, and D.Briggs. Reducing Voltage Ratting and Cost of Vehicle Power Systems With a New Transient Voltage Suppression Technology, IEEE Transaction on Vehicular Technology, volume 52, NO.6,pp. 16521661,November, (2003). [36] J.G.W. West. Powering up - a higher system voltage for cars, IEE Review, pp. 29-32, January (1989). [37] International Electro technical Commission. Impact of current passing through the human body: Part I, Tech. Rep., International Electro technical Commission, Report IEC479-1, (1984). [38] High Efficiency High Density Polyphase Converters for High Current Applications, Application note 77, Linear Technology Inc., Sep. (1999). [39] Poly phase, high efficiency, synchronous step-down switching regulators, Tech. Rep. LTC1629, Linear Technology Datasheet, (1999). [40] High-frequency multiphase controller, Tech. Rep. TPS40090, Texas Instrument Datasheet, (2003).

[27]

Appendix A

Using KVL equation around a closed loop with phases

Figure 1 shows the automotive electrical power system which includes the alternator, rectifier, battery and loads. The combination of the energy storage, battery, and the associated electrical loads is referred to as a constant voltage load voltage.

a and c , diodes D1 and D2 and the voltage source E o

gives; Figure3: Three phase diode bridge rectifier equivalent circuit diagram during the interval when diode D1 and D2 conducting

V ll = V a ( ) V c ( ) di a = Ls + R s ia dt di c Ls R s ic dt + 2V d + E o

(1 )

Substituting expressions of

va ,vc , ia , and ic

yield the following expression:

Figure 1: automotive electrical power system: the alternator, battery and load. Figure 2 show the six conduction intervals within each operating period for the circuit shown in Figure 1. During each of the six intervals, two of the six diodes conduct. It is sufficient to consider one of these intervals for the averaging process.

V ll=V ssin (t ) V ssin (t + 2 / 3) =L s d (I s1 sin (t )) dt + R s ( I s1 sin (t )) L s

d (I s1 sin (t + 2 / 3)) dt R s ( I s1 sin (t + 2 / 3))

+ 2Vd + E o
3 Is1 id

Vll = 3 Vs sin(t / 6) = I s1[ Ls cos(t ) + Rs sin(t )]

t

2 3

+ Rs sin(t + 2 / 3)] + 2Vd + E 0

I s1[ Ls cos(t + 2 / 3)

Where: a = L s ,

b = Rs

Figure 2: A sketch of output current waveform for threephase diode circuit

id

Va

Ls

D1 Rs

3Vs sin(t / 6) = I s1[a cos(t ) + b sin(t )] I s1[a cos(t + 2 / 3) + b sin(t + 2 / 3)] + 2Vd + E o (2)

Vc

Ls

Rs

Eo

Applying the Trigonometric Identity :

a cos + b sin = A cos( )
Where: A = a 2 + b 2 ,

D2

(3)
= tan 1 (b / a )

Figure 3 shows the circuit diagram during the interval when diodes D1 and D2 are conducting.

on the right hand side of equation (2)

Vll = 3 Vs sin (t / 6) = I s1 [ A cos(t )] + 2Vd + E o

+ 2 / 3 A I s1[ sin ( t )] + / 3

I s1[ A cos(t + 2 3 )]

Vll = A I s1 cos(t ) A I s1 cos(t + 2 3 ) + 2Vd + E o ( 4)

+ 2 / 3 2 ) A I s1 sin ( t + 3 + / 3
3

+ 2Vd + E o =

First, averaging the left hand side of the equation (4) over

2 ) sin ( ) A I s1 sin ( 3 3 3 4 A I s1 sin ( ) sin ( ) 3

3 + 2Vd + E o (6)

a conduction angle of

. 3

3 3

+2 / 3 Vs sin (t / 6) d (t) + / 3

1 A I s1 sin + ( 3 cos + sin ) 2 + 2Vd + E o 3 3 3 cos A I s1 sin + 2 2 + 2Vd + E o 3

Vg = 3 3

+2 / 3 = V s [ cos (t / 6)] + / 3 3 3 = Vs cos( + ) cos( + ) 6 2

3 3 = = 3 3 3 1 V s cos + sin 2 2 3 3 V scos( ) 6 from (3) (5)

(7)

V s cos ( 3 L s 2

) ,

Zg =

3Rs + 2

Refer to (figure 2), I d can be calculated as follows

Now, averaging the right hand side of equation (4).

Id =

3 + 2 / 3 [A I s1[cos(t ) + / 3 2 cos(t + ) 3 + 2Vd + E o ]]d (t ) =

1 2 3 f ( ) d T 3 3 2 3 = I s1 sin d 3 3 2 3 = I s1 cos 3 = 3 I s1 cos cos 2 = I s1 3 3 3

+ 2 / 3 cos (t ) d (t ) + / 3 + 2 / 3 3 2 A I s1 cos(t + ) d (t ) + / 3 3
3 A I s1 + 2Vd + E o

The average current delivered to the load by the averaged circuit model is:

Id =

Vg (2Vd + E0 ) Zg

(8)