The electronics is contained on two boards. The power control board (on left) contains power converters and other support functions. The CPU board (on right) contains the CPU with associated program and data buffer memory.
The electronics is contained on two boards. The power control board (on left) contains power converters and other support functions. The CPU board (on right) contains the CPU with associated program and data buffer memory.
The battery pack consists of two rechargeable lithium polymer cells contained in the base of the box. A single cell is shown here along side the box. The power control board mounts on top of the battery box. The CPU board stacks on top of the power control board. The space inside the box above the CPU board is available for the transmitter and command receiver.
The mass breakdown of individual components is shown in the following table.
| Electronics Total | 385 g |
| CPU Board | 77 g |
| Power Control Board | 88 g |
| Transmitter (estimate) | 60 g |
| Receiver (estimate) | 20 g |
| 2 Li-polymer cells | 140 g |
| Shield Box total | 489 g |
| base (battery box) | 172 g |
| sides | 180 g |
| cover | 146 g |
The battery pack may be larger than needed. Early mass budget estimates had allocated a large percentage to the battery pack. Given the high energy density of the Li-polymer pack, a very conservative pack can be used and still take less mass than originally planned. This pack is rated at 22W-hr which may be a factor of 3 or more oversized. Roughly 90 grams could be saved depending on eclipse requirements.
A major constraint on the design is the requirement for radiation tolerance. The prototype is built with commercial grade parts but all devices used in the design have radiation rated parts available with similar electrical specifications. In many cases the packages used for radiation rated parts are different from the commercial grade parts. The board layouts would need to be re-done to accommodate flight parts. An attempt was made to allow extra spacing so that the overall mass and volume of the boards would be similar to flight boards.
A few modifications were done to the board to reduce its power consumption. Some support parts for the unused PC104 buss were removed. The crystal was removed and wiring for an external clock was added.
This photo shows the functions provided by the power control board and the relative board area required by each function.
The controller provides monitors of voltage and current to the CPU but battery charging is independent of CPU control. The CPU can choose to change power consumption but the charge circuit will always attempt to recharge the pack with whatever excess power is available from the panels.
The charge controller also includes a low ohm FET switch to allow disconnecting the battery from the load via an external control signal. This function is not needed during flight but likely would be needed before launch.
| 0 | spare |
| 1 | spare |
| 2 | spare |
| 3 | solar panel temperature |
| 4 | battery charge current |
| 5 | load current |
| 6 | battery voltage |
| 7 | electronics temperature |
A 12Mhz frequency was picked to minimize the power required by the oscillator while providing adequate CPU speed. The 80188 divides the external clock by 2 so the actual CPU clock rate is 6Mhz. The 12MHz is divided down to provide timing for the DC-DC converters, solar aspect sensor and data sampling. A total of 28 counter stages are required to generate a 22.37 second period interrupt for data sampling.
The major power consumption in active mode comes from the magnetometer. A CPU controlled dummy load of 100mA active and 5mA idle was connected between +5V and -5V to simulate the magnetometer power requirements.
The efficiency of power transfer from the solar panels is about 96%. It is somewhat dependent on the battery charge state. Some detailed measurements of battery charge-discharge curves were made with an external supply set up to simulate solar panel characteristics. See battery charging measurements. The rough estimate is that a panel area of 30 square cm per facet would be minimal and an area of 60 square cm per facet should provide an adequate safety margin.
| Average Power from Panels | 750 mW |
| Average Load (8.3V buss) | 723 mW |
| Total Active Current | 230 mA |
| Magnetometer | 160 mA |
| CPU board | 35 mA |
| ADC | 5 mA |
| TXCO etc. | 30 mA |
| Total Idle Current | 46 mA |
| Magnetometer | 8 mA |
| CPU board | 8 mA |
| TXCO | 10 mA |
| Charge Controller | 5 mA |
| DC-DC Converter quiescent | 15 mA |
The transmitter and receiver power are not included in this measurement because although the transmitter power demand is very high, the RF subsystem is only active over small percentage of the orbit and doesn't contribute significantly to the orbit average. It is assumed that the RF system will be entirely shutdown for most of the orbit. The battery pack is capable of supplying the maximum power required when the transmitter is on and there is plenty of time during the orbit to recharge.
For example an S-band transmitter with 500mW RF power would require about 5W from an 8V buss. That would be 625mA which is well within the 3000mA battery rating. A transmitter-on time of 10 minutes requires 0.83 W-hr. For a 1 day orbit this increases the orbit average power requirement by less than 5%.