Power Management of Solar Cells in Space Microsatellite

Transcription

1 Power Management of Solar Cells in Space Microsatellite Power Controller Design Darren J. Wright 1 University of New South Wales at the Australian Defence Force Academy The aim of this project was to develop a power switching controller for use with the Microgravity Experiment Recoverable Satellite (MERS) Electrical Power System. The power switching controller forms part of the Electrical Power Sub-system of the microsatellite, its role is to control the transition or switching of the power source between the solar array supply and the micro-satellites storage battery during periods of solar eclipse. The design process required the development of a satellite power profile simulation tool which was used to determine the input power profile for the controller. The significant outputs derived from the simulation tool were the expected minimum and maximum power yield from the solar panels and the duration of the eclipse periods. Finally, the power controller was designed, constructed and tested based on the output from the simulation tool. A scaled working solution for the power management of the MERS micro-satellite was successful, the power controller maintained electrical power throughout eclipse periods for all of the required MERS orbits. The final power controller design as tested was successful and is proof of concept enabling the design to be scaled up to meet the specifications stated in the MERS concept documents. Contents I. Introduction 2 A. Aim 2 B. Methodology 3 C. Requirements 3 II. Characterisation of Solar Cells 3 A. The Photovoltaic Effect 3 B. Standardised Solar Spectrum 4 C. Solar Cell Characterisation 4 III. Power Profile Simulation Tool 6 A. Equations of Motion 6 B. Model Validation 7 C. Power Profile Output 7 IV. Power Controller 8 A. Design Decisions 8 B. Operation 9 C. Testing 9 V. Conclusions 1 VI. Recommendations 11 Acknowledgements 11 References 11 1 PLTOFF, School of Engineering & Information Technology. ZEIT451 1 Final Project Report 216, UNSW Canberra at ADFA

2 Nomenclature AM = Air Mass Zero (solar irradiance at Earth orbit) E g = Energy Band Gap [ev] EPS = Electric Power System I SC = Short circuit current [A] k = Boltzmann s Constant [J.K -1 ] MERS = Microgravity Experiment Recoverable Satellite q = Elementary Charge [C] R S = Series Parasitic Resistance [W] R SH = Parallel Parasitic Resistance [W] T = Temperature [K] = Open circuit voltage [V] V OC I. Introduction Since the beginning of the space race between the Soviet Union and the United States back in the early 195 s, billions of dollars have been spent on space exploration and research. From the humble beginnings of the launch on October 4, 1957 of the first artificial satellite Sputnik I to the much anticipated first successful manned lunar landing on July 16, 1969, by US astronauts Neil Armstrong, Edwin Aldrin. However, much of the emphasis in recent years has shifted to the development and testing of new technologies through research conducted via interplanetary space missions or earth orbiting satellites. This research has been the catalyst in the development of much of todays progressive technology and has been instrumental to the research capabilities across all fields of science, technology and engineering. The development of micro-satellites has vastly decreased the costs associated with conducting experiments in space with many schools and universities beginning to produce their own micro-satellites for this purpose. The University of New South Wales Canberra is currently developing a Microgravity Experiment Recoverable Satellite, a micro-satellite designed for Low Earth Orbit which is powered via a solar array with backup and eclipse power to be provided by an on board battery system. Currently little effort has been presented in the area of power management for the Electric Power System of the MERS project. This is an important part of the EPS sub-system as during some orbital conditions there will exist eclipse periods which will greatly impact on the effectiveness of the solar panels to supply the electrical power systems of the satellite. During these periods it is essential that the battery sub-system maintain power to the satellites payload modules. A. Aim The aim of this project was to develop a power switching controller for use with the Microgravity Experiment Recoverable Satellite (MERS) Electrical Power System. The power switching controller will from part of the Electrical Power Sub-system of the micro-satellite, its role is to control the transition or switching of the power source between the solar array supply and the micro-satellites storage battery during periods of solar eclipse. The project was broken down into three phases each completed in the following order. 1. Characterisation and modelling solar cells in space. 2. Development of the MERS micro-satellite power profile simulation tool. 3. Design, build and test the power controller. B. Methodology A review of current literature surrounding the MERS concept was initially conducted to determine the requirements for the design of the power controller. Critical requirements identified during this process were the input power profile determined by the satellite orbit and the satellite payload power system requirements. The design process began with a comprehensive theoretical analysis of the relevant theory of solar cells in order to understand the physics related to the production of electrical power via the photovoltaic effect. This knowledge permitted the characterisation and modelling of solar cells under space like conditions, a process which is key to understanding the possible input parameter to the power controller. The MERS micro-satellite is a multirole satellite designed to operate over a wide range of Low Earth Orbits, ranging from 4-5km altitude within a inclination and over a period of 2 months. Thus, for the 2 Final Project Report 216, UNSW Canberra at ADFA

3 controller to be designed fit-for-purpose, it was necessary to first determine the expected power profile provided to the controller of the EPS and determine the possible eclipse durations during all specified MERS orbits. This power profile was achieved through the development of a MATLAB simulation tool which accepts the orbital parameters of the micro-satellite and calculates both the solar cell output power and the eclipse duration of the orbit. Finally, the ultimate phase of the project was the design, build and testing of the power controller. This design process was greatly assisted by the data obtained from the power profile simulation tool and ensured the MERS micro-satellite design requirements were met. C. Requirements Identification of the MERS micro-satellite power subsystem requirements and design constraints. Altitude: 4 5 km. Orbit: Able to operate in a 47 and 98 Low Earth Orbit. Duration: 2 months LEO. Solar Panels:.25m 2 triple junction GaAs 6 panel solar array, with 28% cell efficiency and 8% packing factor. Maximum input power: 36 W. Subsystem power requirements: 33.5W (5V 4A (2W), 3.3V 4A (13.3W) busses). Electrical Power System: 38 W Clyde space flexible EPS, CSXUEPS2-42. Storage: 3 W-hour 8.2V Clyde Space CS-SBAT2-3 (Lithium Polymer). For the purposes of the design and construction of the physical power controller, these requirements were scaled down to operate at 5W max utilising a commercial silicon solar panel and a 12V 7Ah lead acid battery. II. Characterisation of Solar Cells One of the key technologies of interest for space satellite applications is the Photovoltaic solar cell, particularly its ability to extract unlimited clean renewable energy from the sun. When sunlight is incident on a solar cell its energy is converted into electricity without the need for any mechanical movement or the production wasteful by-products, making solar energy an appealing power source for satellite operations. It is important to understand the process behind the conversion of solar energy from the suns radiation into an electrical energy source for a micro-satellite. This will aid in the calculation of the theoretical maximum power that can be supplied to the micro-satellite electrical power system and the power controller. A. The Photovoltaic Effect The suns electromagnetic radiation is emitted in the form of photons which are converted into electrical energy in semiconductor materials via the photovoltaic effect. In order for a solar cell to generate power, a voltage and a current must be generated within the cell (Sze, 27). The generation of this voltage is due to what is known as the photovoltaic effect. A solar cell, typically a p-n junction, has a single bandgap energy level. When the cell is exposed to the solar spectrum, photons with energy greater than the bandgap energy will excite valence electrons into the conduction band thus contributing to the photovoltaic effect. Photons with energy less than the bandgap energy cannot excite an electron to the conduction band and are therefore absorbed as heat by the cell. The collection of these light-generated carriers within the p-n junction causes a build-up of electrons in the n-type material and holes in the p-type material of the junction. The short circuit current is the result of the motion of these carriers as they exit the cell as a light-generated current. The open-circuit voltage results from of the build-up of charge carriers and holes at the terminals of the p-n junction hence creating a potential difference. 3 Final Project Report 216, UNSW Canberra at ADFA

4 B. Standardised Solar Spectrum Since the output of a solar cell is dependent upon both the power and the spectrum of the incident light it is important to measure and compare solar cells at a standard metric. To facilitate an accurate comparison, the spectral distribution of sunlight has been defined by the radiant power per unit area perpendicular to the direction of the sun at the mean earth-sun distance (Green, 1982). Outside the earth s atmosphere, this intensity is referred to as the solar constant or AM radiation condition and the spectral distribution is shown in Fig. 1. There are many sources for representative measurements for the average intensity of solar radiation in free space, the power density at the average distance of the earth from the sun has been quoted in the range of 1, W/m 2 (Kopp, 211) (Sze, 27). However, the most accurate value for total solar irradiance representative of solar Air Mass Zero (AM) is ±.5 W m 2 (Kopp, 211). AM represents the solar spectrum outside the earth's atmosphere and hence is the relevant spectrum for satellite applications (Sze, 27). C. Solar Cell Characterisation In order to understand this process, a single monocrystalline silicon solar cell was modelled as a simple electrical circuit as shown in Fig. 2. The illumination current (I L ) provided by the absorbed photons within the semiconductor is represented as an idealised constant current source in parallel with a semiconductor diode. The model shown includes and identifies the series and shunt parasitic resistances of the cell. Figure 1. Spectral distribution of sunlight (Sze, 27). Circuit analysis of the equivalent circuit shown in Fig. 2 results in equation 1 for the current-voltage characteristic of a single solar cell. Figure 2. Electrical circuit model of a Solar Cell including series and shunt resistances.! =! # $ % &'() * +, +. +!/ # / # -! 2 (1) These parasitic resistances are generally the result of the manufacturing process of the solar cell. The shunt resistance is typically the result of defects in the structure of the cell while the series resistance is associated with the movement of current through the emitter and base terminals of the cell, the contact resistance between the metal contact and the silicon and the combined resistance of the top and rear metal contacts (Honsberg, 216). Due to the nature and location of the series and shunt resistances, to reduce overall power losses we require a high shunt resistance and very low series resistance. The model was tested under space like AM radiation condition through an experiment conducted using the solar simulator in the space laboratory at the University of New South Wales Canberra. The aim was to validate the model by measuring the I-V characteristics of a single cell under simulated AM conditions and use these results to determine the cells series and shunt resistances. The experiment was set up to expose a single silicon solar cell of area 8.5x1-4 m 2 to a solar irradiance of 1354W/m 2 in order to simulate the cell in space like power density conditions. The I-V characteristics of the cell was measured by varying the load resistance and recording the cell voltage. This method also enabled the calculation and characterisation of the cells power generation. The results are shown in Fig. 3 and Fig. 4. It is important to note that all measurements were recorded after the temperature of the cell had stabilised at approximately 55 C. This temperature falls within the expected temperature range of 2 C to 85 C of solar cells in Earth orbit (Landis, 1994), thus the experimental data was deemed to be acceptable. 4 Final Project Report 216, UNSW Canberra at ADFA

5 Current, A Voltage, V Voltage, V Figure 3: Monocrystalline silicon solar cell I-V curve, Irradiance 1354W/m2. Figure 4. Monocrystalline silicon solar cell P-V curve, Irradiance 1354W/m2. The shunt resistance can be measured directly by placing an Ohm meter across the terminals of the single cell since R SH >> R S. The measured shunt resistance was 1.6kW or 32 Ωcm 2, given the cell area was approximately 8.5cm 2. This value is significantly less than the expected theoretically predicted shunt resistance which is typically in the range of MΩcm 2 range for laboratory type solar cells, and 1kΩcm 2 for commercial solar cells (Honsberg, 216). The series resistance is much more difficult and can not be measured directly, one method to estimate the series resistance of a solar cell is to measure the slope of the I-V curve at the open-circuit voltage point (Honsberg, 216). Using this method it was estimated that the series resistance was approximately 1.57W. In addition to these physical structures modelling a cell there are three other parameters used to characterise solar cells; these are the short-circuit current, open-circuit voltage and the Fill Factor. The short-circuit current is due to the collection of the light-generated carriers within the p-n junction. Ideally, the short-circuit current and the light-generated current (I L ) are identical and hence the short-circuit current is the largest current which may be drawn from the solar cell (Honsberg, 216). The short-circuit current is therefore the current measured with the load resistance (R L =W). The maximum current produced by the single silicon solar cell as shown in Fig. 3 was 285mA. The theoretical short-circuit current for the cell under AM conditions can be determined by integrating the total area under the graph shown in Fig. 1 up to the maximum wavelength for which electronhole pairs can be generated for a specific semiconductor (Green, 1982). For silicon with a band gap energy E g of 1.1eV this corresponds to a wavelength of 1.13µm, which gives a maximum short-circuit current of 55mA/cm 2. Given the total cell area was 8.5cm 2 the estimated short-circuit current is 467mA. The discrepancies between the theoretical and measured values may be attributed to parasitic resistance losses stated above and defects in the semiconductor as this cell was from an old cheap commercial product. The open-circuit voltage is obtained by setting I equal to zero in equation 1.! "# = %& ' () * + *, + 1 (2) The value of the open-circuit voltage is determined by the properties of the semiconductor, namely by its dependence on I S, the diode saturation current. Minimising the saturation current results in a maximum opencircuit voltage which is important for both efficiency and maximum power. An estimate for the maximum value for the open-circuit voltage for silicon is approximately 7mV (Green,1982). The experimental results indicate an open-circuit voltage of 548mV. Again this lower than expected value can be a result of the parasitic resistance losses and defects in the cell. The short-circuit current and the open-circuit voltage represent the limitations on the maximum output power from the solar cell. However as shown in Fig. 4 the cell output power is zero at these points, due to zero current flow at open circuit condition and zero potential difference at short circuit condition. The Fill Factor is the measure which determines the maximum power that can be drawn from a solar cell. The fill factor is defined as the ratio of the maximum power from the solar cell to the product of the open-circuit voltage and short-circuit current (Honsberg, 216). A common empirical expression for estimating the fill factor is shown in equation 3. The calculated fill factor for the single silicon solar cell using equation 3 is.8. This value is within the expected range of for a commercial silicon cell (Honsberg, 216), however this value seems relatively (3) 5 Final Project Report 216, UNSW Canberra at ADFA

6 high as practical measurements of fill factor tend to be on the low end of the range due to the effects of the parasitic resistances. The above results indicate the experimental solar cell characterisation is an accurate representation to the theoretical models of solar cells. This data will aid in the understanding of how practical solar cells provide electrical power for satellite operation and the development of the power profile for use in the remainder of the project. III. Power Profile Simulation Tool A fundamental problem to be resolved in satellite power sub-system design is associated with the power management of the supply and storage systems and the associated effects of the charge discharge cycling of the batteries, clearly the satellites orbit has a major role in determining how significant this task is (O Sullivan, 1993). The MERS concept is a low earth orbit satellite that will provide microgravity conditions for experimental research, however to maximise the wide applicability of experiments that can be conducted, the MERS microsatellite has been designed to operate in a wide range of orbits (Tuttle, et al., 215). For this reason, it becomes a difficult task to design the EPS power distribution system without some means to predict the power that can be supplied by the on-board PV panels. The power profile simulation tool was developed as a means to predict the input profile to the micro-satellite EPS and hence the controller input. This analysis tool permits an accurate calculation of both the duration of the eclipse period and the expected input power provided by the PV panels. In addition, the simulation tool calculates several other orbital parameters such as the orbit period and hence the number of orbits expected to be completed during the two-month orbit duration, this information can be used to determine the number of charge discharge cycles of the battery system. A. Equations of Motion There are two methods used to determine the orbital mechanics of a satellite in low earth orbit, the first is Newtonian mechanics which uses the forces and masses of the objects to determine the orbit and the second is Keplerian celestial mechanics in which the orbits are determined by geometry and the force of gravity. The Keplerian celestial technique was be selected for use with this modelling tool as this method does not require masses of each component to be known. As there will be some difficulty determining an accurate mass of the MERS satellite given the wide applicability of payload modules that may be fitted. The simulation is a combination two two-object systems, the earth-satellite system and the earth-sun system. The limitations imposed on the use of Kepler s equations of motion are; the central body must be spherically symmetric, the central body s mass must be much greater than that of the satellite and the central body and the satellite must be the only two objects in the system. Since the approximate masses of the satellite, Earth and Sun are approximately 5kg, 5.972e 24 kg and 1.9e 3 kg respectively, both of these systems can therefore be modelled with Kepler s equations of motion. The equations of motion were modelled in MATLAB in order to calculate the instantaneous position of the satellite orbit around the earth and the earth orbit around the sun. Once the satellites position is known it can be used to provide a distance vector between the satellite and the sun. Thus, the mean solar radiation intensity H (W/m 2 ) incident on the satellites solar panels can be determined using equation 4.! " = ( $ %&' (! %&' (4) ) %&'-%+,-../,- where, H sun is the power density at the sun's surface (W/m 2 ), R sun is the radius of the sun (m), and D sun-satellite is the distance between the sun and the satellite (m). The model first determines the irradiance incident on the micro-satellite solar panels based on the distance from the sun, then the power supplied by the solar panel is calculated based on the incident irradiance power, the solar cell area and the panel efficiency. The model includes the ability to adjust input orbital parameters such as the orbit altitude, inclination angle and specific solar panel parameters including solar panel area and panel efficiency. 6 Final Project Report 216, UNSW Canberra at ADFA

7 B. Model Validation Following the development and implementation of the model and prior to the use of the model output for the controller design process it is important to ensure the model is providing accurate data. Therefore, several of the key parameters of the model were validated against available data. These included the orbital period, the initial irradiance and predicted solar cell power at the start of the simulation and the absolute maximum possible values for satellite power. The models orbital period over a full year was compared using two methods, first against Kepler s 3 rd Law for determining orbital periods and secondly against the orbital period of the moon. The model calculated the orbital period to within -.25 days of that using Kepler s 3 rd Law, this difference is due to the fact that Kepler s equations are adjusted to account for leap years while the MATLAB model did not. This will have little effect on the accuracy of the power profile as the duration for the MERS orbit is 2 months which can be input into the model in days eliminating the need to account for leap years. The model was accurate to within 3.7% when determining the orbital period of the moon around the earth, a figure which is respectable considering the simplicity of the tool compared to those used to determine the reference data. The model also calculates the initial and maximum solar cell output power to within 1.3% of the value calculated using equation 4 and the mean sun earth distance data available from NASA (Spaceplace.nasa.gov, 216). C. Power Profile Output The use of the power profile simulation tool enabled the determination of the minimum and maximum eclipse durations and hence the best and worst case scenarios during the designed satellite orbit for which the controller is required to maintain satellite power from the storage batteries Time, s Figure 5. PV power profile for a single orbit indicating minimum eclipse duration. Altitude - 5km, Inclination angle - 91, Eclipse duration 5:54 minutes Time, s Figure 6. PV power profile for a single orbit indicating maximum eclipse duration. Altitude - 4km, Inclination angle - 47, Eclipse duration 35:7 minutes. The minimum eclipse duration within the MERS operational orbits was 5:54 minutes and occurred while tracking a 5km altitude orbit with an inclination of 91, the power profile for this orbit is shown in Fig. 5. The maximum eclipse duration was 35:7 mins for the 4km orbit with an inclination angle of 47 as shown in Fig. 6. Any orbit with inclination angle greater than 93 (4km altitude) and 91 (5km altitude) does not have an eclipse period. These are the critical orbits used for the EPS system and controller design, they indicate the total expected PV supply power during the non-eclipse part of the orbit along with the required battery supply duration or eclipse period. Important design information may also be determined from the power profile tool such as the ability to size the capacity of the battery sub system which may be optimised using the maximum eclipse duration data. The power profile tool was used to develop the power controller test cases, the input to the scaled constructed controller design was produced from power profile plots of the critical orbits required for the MERS micro-satellite. 7 Final Project Report 216, UNSW Canberra at ADFA

8 IV. Power Controller The electrical power system of a satellite is one of the most important subsystems as its role is to supply and manage the power available to all other satellite systems. The EPS serves as the interface between the two sources of satellite power, the solar panels and storage batteries, at the core of this management process is the switching configuration used for the power controller. Figure 7. MERS EPS circuit block diagram (Clyde Space CSXUEPS-2-42), Switch configuration is the Power Controller (Clyde Space LTD, 21). The MERS micro-satellite electrical power system features the Clyde Space flexible electronic power system: CS- XUEPS2-42. The flexible design of this COTS EPS system enables it to be integrated within the overall design requirements of the MERS micro-satellite. However, as a result of this flexibility in design it is left to the user to define the switching configuration as is shown in the EPS system block diagram in Fig. 7. The solution designed for this switching configuration is detailed in the block diagram for the power controller shown in Fig. 8. The purpose of the power controller is to maintain supply power for satellite operation and the payload modules from either the solar panels or the battery system throughout all of the required MERS orbits. The controller achieves this by continuously monitoring both the solar panel and battery output voltages which are fed to the ADC of the microcontroller and based on predetermined logic supply control signals to the battery and PV supply relays. Two DC-DC Buck converters are used for power processing; the first stage is used to reduce the solar panel output voltage down to 14.5V to provide a charging voltage for the battery during the non-eclipse period. The second and output stage is required to further reduce the output voltage and boost the output current to a suitable level for bus regulation. Figure 8. Block diagram of the Scaled Power Controller design. A. Design Decisions Design decisions for the controller were primarily based on the key constraint used for the selection of the MERS EPS system. The power controller design utilises only COTS components centred around a simple 16 MHz, 8-bit microcontroller. The Arduino Mega 256 was selected for the controller design as it is a reliable microcontroller with an operating temperature range (-4 C to 85 C) consistent with the selected electronic power system and within the wide temperature variations expected for a LEO micro-satellite. The two DC-DC conversion stages utilise LM2596 adjustable 15kHz step-down switching voltage regulators, these provide the 8 Final Project Report 216, UNSW Canberra at ADFA

9 capacity to output both a 14.5V battery charging voltage and 6.5V controller output voltage. Power switching was provided using solid state components (BJT transistors), the objective of which was to improve the reliability of the controller design due to the limited ability to perform maintenance or repairs during operating life. Although the controller design was limited to 5W, all of the design decision were intended to produce scalable design solution that can be modified to suit the MERS operating requirements. B. Operation During non-eclipse periods the micro-satellite power is supplied entirely by the solar cells through the PV supply relay. The PV and battery output voltages are continuously monitored and when the microcontroller senses the PV voltage drop below the battery supply voltage (during eclipse) the controller switches the supply to the output DC-DC converter stage to battery via the battery supply relay. Initially after the satellite exits eclipse, bus power is immediately supplied by the PV array and will also be supplied through the battery charge relay in order to replenish the battery. Once the battery reaches maximum SOC the battery charge relay will open and switch the supply directly to payload module. C. Testing The operation of the controller was verified using four separate test cases. These test cases were developed using the output from the power profile simulation tool and represent the power profiles of the critical orbits for the MERS micro-satellite. The orbital parameters and duration of the eclipse period for each of the test cases is shown in Table 1. Table 1. Power Controller Test Cases. Test Case Orbit altitude (km) Inclination angle ( ) Eclipse duration (mins:sec) : : : :37 These orbits were selected as test cases as they produced the minimum and maximum eclipse durations possible and characterise the complete altitude orbit range designed for the MERS micro-satellite. The time scale for the duration of each test was scaled down by a factor of 1 to reduce the time taken for a complete orbit from 92:56 minutes (5576 seconds) down to 56 seconds. This was elected in favor of completeing each test case multiple times rather than testing the controller over the full minute orbital period for each test case. Figure 9. demonstrates the operation of the controller during test case 2. The output power of the controller is maintained throughout a complete orbit despite having an eclipse period of appriximately one third of the orbit duration. Power is initially supplied from the battery during the initial eclipse period before switching to PV supply during solar irradiance and again repeated over the second half of the orbit. Voltage, V Current, ma PV Supply Battery Supply Figure 9. Power controller output, Altitude - 4km, Inclination angle Current, ma Voltage, V Figure 1. 5V bus output, Altitude - 4km, Inclination angle Final Project Report 216, UNSW Canberra at ADFA

10 Current, ma Voltage, V Figure V bus output, Altitude - 4km, Inclination angle The most important performace parameter for the power controller is the ability to supply the satellite systems and payload module with a constant and stable electrical supply during all orbit profiles. The success of the controller design is shown in Fig 1. and Fig 11., these figures demonstrate the uninterupted 3.3V and 5V output busses during the orbit with the greatest eclipse duration. Each of the four controller test cases shown in Table 1 provided continuous bus outputs for the duration of the orbit. Close inspection of Fig. 1 and Fig. 11 reveals a small variation in the output voltage and current supplied to each bus during solar panel and battery operation. The maximum change in the output voltage for the 3.3V and 5V busses across all test conditions was.3% and.6% respectively. While the bus currents showed a variation of 1.8% and 1.1% respectively. This is primarily due to the mismatch between the maximum power that can be supplied by the 5W commercial solar panel and the 12V 7Ah lead acid battery used in the experiment. The power efficiency of the controller was measured with the controller operating in both the battery supply mode and PV supply mode. Whilst operating in battery supply mode the controller efficiency was 68%, however when the controller was being supplied by the PV panel the efficiency increased to 7%. V. Conclusions During the initial research phase of this project it was concluded that in order to design a purposeful power controller for the MERS EPS system their needed to be some way of predicting the expected power input to the controller. This resulted in the requirement to provide the power profile simulation tool, this tool was exceptionally useful in the design process as well as providing the structure of the input to the controller during the testing phase. Without this tool the design and operation of the final controller would have proven difficult to test. To extend the duration of the MERS missions, orbits with inclination angles greater than 93 for a 4km altitude orbit or greater than 91 for a 5km altitude orbit may be selected, ensuring the solar panels alone can provide continuous electrical power to the EPS sub-system. This will eliminate the possible battery degradation caused by the continuous charge discharge cycles. The controller was designed with robustness in mind, choosing to use solid state components (BJT transistors) for switching rather than conventional mechanical components in order to make the controller less susceptible to component failure. However, as a result of the voltage drops across the multiple P-N junctions this decision had a considerable impact on the efficiency of the resulting circuit. The decision to use solid state components, although it did result in a significant decrease in efficiency of the overall system it was shown from the predicted power profiles in section IV that there is sufficient power available exceeding the requirements for the MERS micro-satellite. Therefore, the sacrifices in efficiency were deemed acceptable in favour of the increased robustness of the design, since a failure within the electrical power system of the micro-satellite will result in a complete failure of the MERS mission. A scaled working solution for the power management of the MERS micro-satellite was successful, the power controller designed maintained electrical power throughout eclipse periods for all of the required MERS orbits. The final power controller design as tested was successful and is proof of concept enabling the design to be scaled up to meet the specifications stated in the MERS concept documents. 1 Final Project Report 216, UNSW Canberra at ADFA

11 VI. Recommendations Future work might be directed towards the testing of the operation of the controller under more realistic space temperature conditions. This would include the use of a heating/ cooling chamber to analyse the effect and limitations of the electronic components under the high stresses caused by rapid changes in temperature conditions. Also further testing of the controller over complete real-time mission lengths is essential to verify controller performance. Acknowledgements I would first like to thank my supervisors Dr Haroldo Hattori and Associate Professor Hemanshu Pota for the guidance and advice throughout the course of my project. A special thankyou to all my family and friends for their continued support throughout my studies. References Clyde Space LTD, (21). User Manual: FleXible Electronic Power System: CS-XUEPS2-41/-42. Glasgow. Green, M. (1982). Solar cells. Englewood Cliffs, NJ: Prentice-Hall. Honsberg, C. and Bowden, S. (216). PVEducation. [online] Pveducation.org. Available at: [Accessed 15 Mar. 216]. Kopp, G. and Lean, J. (211). A new, lower value of total solar irradiance: Evidence and climate significance. Geophys. Res. Lett., 38(1). Landis, G. (1994). Review of solar cell temperature coefficients for space. In: Space Photovoltaic Research and Technology Conference. pp O Sullivan, D. (1993). Space Power Electronics, Design Drivers (I). EPE Journal, 3(2), pp Spaceplace.nasa.gov. (216). NASA Space Place: Home: NASA Space Place. [online] Available at: [Accessed 9 Jun. 216]. Sze, S. (27). Physics of semiconductor devices. 3rd ed. John Wiley & Sons. Tuttle, S. (213). Mission and System Requirements Document: MERS- Micro Gravity Experiment Recovery Satellite, Canberra: The University of New South Wales. Tuttle, S., Johnson, S., Davies, K., Woodward, M., Neely, A. (215). A Concept for a Microgravity Experiment Recoverable Satellite, Canberra: The University of New South Wales. Wertz, J., Everett, D. and Puschell, J. (211). Space mission engineering. Hawthorne, CA: Microcosm Press. 11 Final Project Report 216, UNSW Canberra at ADFA