Design of a Highly Integrated, Multi-Functional Solar PanelPanel

example is the so-called "intelligent tile" dubbed CubePMT developed in the scope of the AraMiS project at Politecnico Torino [93]. This panel has no precision sun sensor, but includes a gyroscope, a magnetic field sensor, a magnetorquer, MPPT, and voltage regulators for local power supply to the electronics components. It is the panel with the highest integration density up to date, but still awaits IOD. Due to the use of six different connectors and some taller components like gyroscopes and buck-boost converters, the positive effect on overall volumetric utilization for the satellite are believed to be less than optimal. Notable about the CubePMT is the envisaged layout of the proposed AraMiS-C1 single unit CubeSat, where the satellite bus is made from four CubePMT tiles and two communication tiles, which leaves most of the internal volume available for payloads.

6.2 Design of a Highly Integrated, Multi-Functional Solar

different solar cells and antenna on the front of the panel. Less effort is required on the back, if the mechanical interface between actuator and panel is kept similar to the existing interface between BEESAT panels and structure.

The electrical interface needs to be implemented in a way that allows for robust connection.

6.2.1 Communication

The only component related to the communication system is the solar antenna (cf. chapter 3). It is located in one corner of the solar panel, and glued to matching contact areas using conductive adhesive (cf. page 31). DC current from the solar antenna is routed to the panel using two electrical contacts on the edges of the antenna (cf. figure 3.7). A straight miniature SMA connector is used to feed the RF signal on the back of the solar antenna. To install the antenna on the solar panel, a small opening needs to be provided on the solar panel.

To realize the three solar panel configurations, the solar antenna has to share copper pads with the alternative large solar cell. If used for the solar antenna, the pads must not be connected to electrical ground. If the second large solar cell is applied, the shared pads need to be connected to the other solar antenna pads. The allocation of copper pads to for the antenna and the solar cells is shown in figure 6.4. A 0 W bridge is provided on the back of the solar panel, that allows to connect the solar antenna pads with the solar cell pads in case a large solar cell is used (cf. figure 6.6(a)).

6.2.2 Power Generation and Distribution

Power is generated either by two large solar cells, or a combination of one large and four small solar cells together with the cell on top of the solar antenna.

As described in the previous section, the solar antenna and solar cell footprint pads on the panel allow for connecting the two different setups already. In order to have a footprint design, that in addition supports the use of large and small cells in the same place, a gridded design is applied to all solar cell footprint pads (cf. figure 6.4 and figure 6.6(b)).

Figure 6.4: Allocation of antenna and solar cell pads

Solar cell power is collected and combined using a set of bypass diodes on the back of the panel. From this point, electric power is guided through the MPPT hardware, and then further towards the single connector used for both power and data bus. MPPT uses a perturb and observe algorithm for finding the maximum power point, which depends on multiple factors like solar cell temperature and the connected load (cf. [125]).

Supply power is not generated locally on the solar panel. Instead, it is drawn from the power bus lines, which are supposed to be controlled by the superordinate PCDU node of the satellite. This node monitors the power drawn by the panel and latches it off in case of excessive power consumption.

Current-sense amplifiers and power-distribution switches located on the solar panel allow to locally monitor and latch subsystems like the sun sensor, the magnetic actuator, or the pFDA (cf. Grau, Tschoban, et al. in [65]). Therefore, malfunction of one of the local subsystems does not jeopardize the complete solar panel functionality or the satellite bus.

6.2.3 Attitude Determination

For attitude determination one TU Berlin sun sensor plus two MEMS magnetic field and two angular rate sensors are utilized on the panel. On the BEESAT satellites, the sun sensor is already located on the solar panels. It consists of a PSD, aperture, and amplifier circuitry [126]. Amplified signals are processed on the OBC board, and there combined with magnetic field and angular rate sensor data to estimate the attitude. In the present design, having all sensors and a microcontroller on the solar panel allows to perform attitude determination locally, and provide attitude estimates from each solar panel to the ADCS core, that is running on the OBC.

Implementation of attitude determination hardware and algorithms is designed based on BEESAT heritage. This reduces the likeliness of failure in case the multi-functional solar panel will be used on future CubeSat missions.

6.2.4 Attitude Actuators

Two types of attitude actuators are used on the solar panel: an optimized magnetic actuator for detumbling and coarse pointing (cf. section 4.4.3) together with a panel-mount pFDA for precise and agile maneuvers (cf.

section 5.4.6).

6.2.4.1 Magnetic Actuator Design

The magnetic coil is realized on eight internal layers of the PCB. The coil should produce a minimum magnetic dipole of 45 mA m², and consume a maximum of 200 mW of electric power at a supply voltage of 3.30 V. Track separation is set to 100 µm, and the values for all other technological parameters are set to the required minimum of the board manufacturer. Using the magnetic actuator optimization procedure (cf. section 4.3), a number of windings per layer of 19 and a track width of 550 µm is found. GNU Octave is used to display (cf. figure 6.5) and export the coil geometry to KiCAD, the free electronic design automation (EDA) software used to create all schematics and PCB layouts for the solar panel and the picosatellite fluid-dynamic actuator.

i i

-40 -30 -20 -10 0 10 20 30 40

-50 -40 -30 -20 -10 0 10 20 30 40 50

x [mm]

y[mm]

Figure 6.5:Magnetic coil layout

The magnetic coil driver is implemented using an H-bridge, which is controlled by the attitude control algorithms executed on the microcontroller on the solar panel.

6.2.4.2 Panel-Mount Pico Fluid-Dynamic Actuator

As precise and agile attitude control actuator, the pFDA shown in figure 5.10 on page 76 is integrated with the solar panel. Solar panel design work with respect to the actuator comes down to the mechanical and electrical interface, including power monitoring and latching.

Monitoring and latching circuitry is already discussed in section 6.2.2. As the pFDA implements an external I2C data interface, the physical electrical interface is realized using board-to-board connectors between the solar panel and pFDA PCBs. The interface provides logic and pump power to the actuator, and uses one of the two I2C buses that are available on the solar panel for communication.

6.2.5 Command and Data Handling

A 32-bit, low-power microcontroller is used as central processing unit on the solar panel. The controller connects to the redundant CAN bus using two transceiver chips. Algorithms for MPPT and attitude estimation are executed there, and the determined attitude is provided to the data bus. Commands for the attitude actuators are received and either translated and passed on to the pFDA, or directly executed for the coil driver. The implemented PCDU software monitors the subsystem power demands, and latches them in case of excess.