Design of a Solar Powered Battery Charger

Ke Liu, member IEEE, and John Makaran, member, IEEE

Abstract -- A solar powered battery charger is presented, where a photovoltaic (PV) panel is used to convert solar power into electricity and a DC/DC converter is used to control the output power of the PV panel and the charging current for the battery. In the software, an optimal control algorithm is applied to obtain the maximum available power from the sunshine. The simulation and experimental results are presented and compared. The applications of this technique can be light electrical vehicles such as golf carts, scooters, airport utility vehicles, as well as other renewable power stations where batteries are used for energy storage. Index terms Battery chargers, DC-DC power conversion, Solar energy.

The paper contains the following sections: system design, optimal control algorithm, simulation of the DC/DC converter, experiment results, and conclusion.
Solar Energy

PV Panel

Battery (a) with diode

I. INTRODUCTION
Applications of solar energy have been a research topic for decades. In recent years, it has attracted even more interest due to the challenges on the environment, fuel source, and automotive industries. Using solar power to charge batteries is not a new idea. A simple way to accomplish this is to connect a photovoltaic (PV) panel directly to a battery. To prevent the reverse current from the battery, a diode is usually used as shown in Figure 1(a). The advantages of such a system are the simplicity and low cost. The disadvantages are: 1) the PV panel can only provide charging current when its output voltage is higher than the battery voltage; 2) the charging current will decrease when the voltage difference decreases due to a increasing battery voltage or a decreasing PV output voltage; 3) the system does not always work at the optimal to convert the available solar power into electricity. To improve the system performance, a controller can be added between the PV panel and the battery as shown in Figure 1(b). The controller is usually a DC/DC converter. It should be a buck/boost to take into account conditions where the PV panel voltage is higher or lower than the battery voltage. In such a system an optimal control algorithm can be implemented to convert the maximum available solar power into electricity and charge the battery. This paper presents the design of a solar powered battery charger with optimal controller. The goals of the proposed system are: 1) to convert the solar power into electricity as much as possible under the varying weather condition; 2) to charge the battery as fast as possible in accordance to the battery lifecycle condition. The application of the proposed system can be light electrical vehicles such as golf carts, scooters, airport utility vehicles, as well as other renewable power stations using batteries as energy storage.

Solar Energy

PV Panel

Optimal Controller (b) with optimal controller

Battery

Fig. 1. Solar Powered Battery Charger

II. SYSTEM DESIGN

As shown in Figure 1(b) the proposed solar powered battery charger contains the following parts: 1) a photovoltaic panel (PV), which converts the solar power into DC electricity; 2) an optimal controller, which controls the output power of the PV panel and the charging current for the battery; 3) a battery, which can be either the power batteries in an electrical vehicle (EV), or the energy storage batteries in a renewable power station. Table 1 provides the system specification. 1 2 3 4 5 6 7 8 9 10 11 Table 1 System Specification Power rating 200W PV voltage (input) 24V (18 ~ 32V) PV current (input) 8.3A Battery voltage (output) 36V (27 ~ 48V) Battery current (output) 5.6A Over-load protection 10A Over-current protection (cut-off) 15A Input over-voltage protection 32V Output over-voltage protection 48V Input under-voltage protection 18V Output under-voltage protection 27V

2009 IEEE Electrical Power & Energy Conference 978-1-4244-4509-7/09/$25.00 2009 IEEE

Figure 2 is a block-diagram of the optimal controller, which contains the following components: (a) DC/DC Converter: according to the system specification, a step-up (boost) converter is used, which consists of a power MOSFET, a power diode, an inductor, and a few input / output capacitors. The topology will be illustrated in the simulation model. (b) MOSFET Driver: an integrated circuit (IC) chip, which amplifies the PWM control signals from the microcontroller and control the switching of the power MOSFET. (c) Input Sensing Circuit: sensors measuring the input voltage (Vin) and current (Iin). Based on the measurement, the input power of PV panel can be calculated as Pin= Vin Iin. (d) Output Sensing Circuit: sensors measuring the output voltage (Vo) and current (Io). The output voltage (Vo) represents the charge level of the battery, and the output current (Io) is the charging current. (e) Microcontroller: a microchip 8-bit microcontroller, which includes some integrated hardware peripherals like A/D converter, PWM generator, and the optimal control algorithm in software.
PV Panel (a) E(PV) Iin Vin (b) MOSFET Driver DC/DC Converter Io Vo Ebt Battery

typical characteristics of a PV panel, which includes the following two curves: (a) V vs. I curve, (b) Power (P) vs. current (I) curve. The curve (b) indicates clearly that there is a maximum point for the output power. This maximum power is going to vary along with the environment condition. To track the maximum power during the real-time operation is the purpose of the control algorithm.

1.2 1 0.8 V, P 0.6 0.4 0.2 0 0 0.2 0.4 0.6 I

(a) V vs. I

Pmax

(b) P vs. I

0.8

1.2

Fig. 3. Typical output characteristics of a PV Panel Y axis: Voltage (V) and power (P) in P.U., X axis: current (I) in P.U.

(c) Input Sensing

(d) Output Sensing

A/D Converter

PWM Generator

Optimal Control Algorithm (SW) (e) Microcontroller

Fig. 2. The optimal controller

III. OPTIMAL CONTROL ALGORITHM

The main goal of the control algorithm is to obtain the maximum available solar power from the PV panel. The input voltage (Vin) and current (Iin) of the system are the output voltage and current of the PV panel. The relationship of Vin vs. Iin is the key feature of a PV panel. Figure 3 presents the
Current set point: I* +

The converted solar power is used to charge the battery, therefore the charged level of the battery, usually the battery voltage, is checked to see if the battery is capable of taking the maximum available solar power. When a fully charged battery voltage is detected, a power limit will be given, then the charging power will be much less than the maximum available solar power. In most of the time during operation, the battery is not fully charged, and the maximum power is applied in the system, allowing charging the battery as fast as possible. Figure 4 provides a block diagram of the control algorithm. Based on the measured input voltage and current, the input power is calculated. The optimal power control algorithm keeps trying new current set-point (I*) until the maximum output power is achieved. The measured battery voltage indicates if the battery is fully charged, hence determines whether a limit on the current set point is needed. The measured input current is compared with the set point, the error is fed into a controller, which can be either a PI or a Bang-Bang (on-off) controller, which outputs a PWM signal to drive the DC/DC converter.

Output PI or B-B control Current feedback: I PWM DC / DC Converter Input

Vo

Optimal Power control algorithm

P = VI

Sensing V, I

Limit
Fig. 4. Block diagram of the control algorithm

Start Sampling V, I P(k) = VI Y Y I* + + P (k) > P (k-1)? N P (k) > P (k-1)? N I* + +

I* (k) > I* (k-1)? N I* -Y

Y I* limit

Vo>= Vlmt?

Fig. 6. Simulation Model of DC/DC converter

I* output

Fig. 5. Flow-chart of the control algorithm

The optimal power control algorithm is implemented by the software in the micro-controller. Figure 5 shows the flow chart of the algorithm. The input variables are the output voltage (V) and current (I) of the PV panel, and the output is the current set-point (I*) for the power converter. Within every sampling period, the output voltage (V) and current (I) of the PV panel are measured, and the power P(k) is calculated. Comparing with the power in the previous period P(k-1), the algorithm decides how to change the current set-point (I*). If the power is increasing, the current set-point is changed in the same way as in previous period. Otherwise, the changing direction of the current set-point will be reversed. The algorithm keeps trying the new current set-point to ensure the maximal available solar power is achieved. The output voltage (Vo), usually the battery voltage, is checked against a limit value to see if the battery is fully charged. If the answer is Yes, a limitation value is applied to the current set-point (I*).

Fig. 7. Simulation result: Start-up process First graph, PV current Is (blue) in A, Battery current Ib (red) in A Second graph, PV voltage Vs (blue) in V, Battery voltage Vb (red) in V X-axis: time in ms

IV. SIMULATION OF THE DC/DC CONVERTER

Figure 6 shows the simulation model for the DC/DC converter. According to the PV source voltage 24V, and battery voltage 36V, a boost converter is used, which consists of a power MOSFET (FET), a diode (D), an inductor (L), and capacitors for both input (Cin) and output (Co). To simulate the PV panel, a voltage source (Vpv) plus a series resistorinductor element (RL1) are used. The battery is simulated by a resistor-inductor-capacitor bank. The MOSFET is controlled by a PWM signal with a certain duty cycle, which will control the output voltage, current, and power. Sensors are used to measure the input voltage (Vs), input current (Is), inductor current (IL), MOSFET voltage (Vq), and output voltage (Vb) and current (Ib).

Figure 7 provides the simulation result for start-up process. At the first moment after the DC/DC converter is connected into the system, both the input capacitor (Cin) and the output capacitor (Co) are charged up by the PV panel and the battery respectively. Both the input current (Is) and output current (Ib) increase significantly at the beginning, and decrease while the capacitors are charged up. The battery current (Ib) is negative since it flows from the battery to the output capacitor (Co). After the capacitors are charged up, the DC/DC converter starts working, boosts the PV voltage (around 25V) up to the battery voltage level (around 38V) and provides a current to charge the battery. Both input and output current become stabilized. The simulation model can be used to analyze the effect of changing values of certain components, and provide a reference for components selection.

V. EXPERIMET RESULTS
A prototype has been built following the design described above. Figure 8 shows a picture of the prototype.

Fig. 8. Prototype of the designed solar powered battery charger

Experiments have been conducted on the prototype. Figure 9 shows a measured start-up process, where Chanel 1 is the output voltage to the battery, and Channel 2 is the input voltage from the PV panel. It can be seen from the waveform, that the output voltage is controlled at a constant value after the start-up process.

battery charging current. The controller is increasing the input current gradually, the input voltage from the PV panel decreases according to its I-V curve. The battery charging current increases accordingly until the maximum power is achieved. Then the battery charging current will decrease due to the decreasing power. The controller will shut down the PWM signal when the PV voltage drops below the under voltage protection limit (18V), and starts again after the PV voltage resumes. Since the battery voltage dose not change very fast, the maximal power is achieved when the battery current reaches its maximal value as shown in Figure 10. The goal of the optimal control algorithm is to keep the system working at the max power area. Figure 11 shows the waveforms with the optimal control algorithm. Chanel A is the PV panel voltage and Chanel B the battery charging current. The controller is successfully tracking the optimal power from the sunshine and keeping the PV panel and charger system working at that point, at the moment of measurement, PV voltage = 24V, battery charging current = 0.53A.

Fig. 9. Measured start-up process, Channel 1: battery voltage, 10V/div, Channel 2: PV voltage, 10V/div, Time base: 50ms/div

Fig. 11. Waveforms with optimal control algorithm Channel A: PV panel voltage, 10V/div, Channel B: battery charging current, 0.5A/div Time base: 5ms/div

VI. CONCLUSION
This paper presents the design of a solar powered battery charger with optimal controller. The optimal controller consists of a DC/DC converter and a microcontroller, where the optimal power control is implemented in the software of the micro-controller. Simulation is conducted on the DC/DC converter, and the results can provide guidelines for components selection. A prototype is built following the design, and the experiment results indicate that the prototype is working properly. Further experiments are planned to verify the dynamic response of the optimal control algorithm.

Max Power

Fig. 10. Output waveform without optimal control Channel A: voltage of PV panel, 10V/div, Channel B: battery current, 1A/div Time base: 50ms/div

Figure 10 shows the output characteristics without optimal control. Channel A is the PV panel voltage, and Channel B the

VII. REFERENCES
[1] D.K. Bose, P.M. Szczesny, and R.L. Steigerwald, Microcontroller [2] [3]
control of a residential photovoltaic power conditioning system IEEE Trans. Ind. Appl.vol. IA-21. pp. 1182-1191, Sept. / Oct. 1985 J.H.R. Enslin, D.B. Snyman, Combined Low-Cost , High-Efficient Inverter, Peak Power Tracker and Regulator for PV Applications IEEE Trans. PowerElectronics,.vol. 6, pp. 73-82, Jan. 1991 B.M.T. Ho, H.S. Chung, An Integrated Inverter With Maximum Power Tracking for Grid-Connected PV Systems IEEE Trans. Power

VIII. BIOGRAPHIES
Ke Liu received his Bachelor, MSc. and PhD. degrees in 1986, 1989 and 1993 respectively, from Tsinghua University, Beijing, China. After teaching at Tsinghua University for a year, he worked as a postdoctoral research fellow at Technical University of Berlin, Germany from 1994 to 1999, and at University of Waterloo, Canada from 1999 to 2000. Then he moved into industries in Canada, mainly working on power electronics and motor controls. He is currently a professor at Fanshawe College, London, Ontario, Canada. John E. Makaran was born in London, Canada in 1964. He received his B.E.Sc. in electrical and M.E.Sc. in mechanical engineering, and Ph.D. in electrical engineering from the University of Western Ontario. From 1990 to 1992 he held the position of Advanced Electrical Engineer in Engineering Research at 3M Canada Ltd. in London, Ontario. From 1992 to 1993, he held the position of Senior Electronics Engineer in machine controls at Husky Injection Molding Systems in Bolton, Ontario. From 1993 to 2006, he worked at Siemens VDO Automotive Inc. in London, Canada. Presently, he is the Chair of Applied Science and Technology at Fanshawe College in London, Ontario. He has authored or co-authored 25 patents in the area of brush motor and brushless motor controls. He was responsible for the electronics design on the worlds first brushless motor used for automotive engine cooling applications for which he received an innovation award from Siemens Canada Ltd. in 1997. His research interests include motor drives, automotive electronics, EMC, mechatronic systems simulation, and industrial systems design.

Electronics, vol.20, pp. 953-962, July 2005

[4] S. Jain, V. Agarval, A Single-Stage Grid Connected Inverter

Topology for Solar PV Systems With Maximum Power Point Tracking , IEEE Trans. Power Electronics, vol.22, pp. 1928-1940,

Sept. 2007

[5] Q. Li, P. Wolfs, A Review of the Single Phase Photovoltaic Module

Integrated Converter Topologies With Three Different DC Link Configurations IEEE Trans. Power Electronics, vol.23, pp. 1320-

1333, May2008

[6] K.K. Tse, M.T. Ho, H.S. Chung, and S.Y. Hui, A Novel Maximum
Nov.2002

Power Point Tracker for PV Panels Using Switching Frequency Modulation IEEE Trans. Power Electronics, vol.17, pp. 980-989,