Power Budget (4/10/01)

 Item                 +5V            -5V
                  idle  active    idle  active   
CPU UT80CRH196KD   6mA   66mA
PROM UT28F256      2mA   70mA
SRAM UT7Q512       1mA   60mA
8 chan ADC         1mA   25mA
TXCO 12MHZ         4mA    4mA
Housekeeping       3mA    3mA
Solar sensors      1mA    1mA
Magnetometer       5mA  100mA      5mA   100mA
Dosimeter         20mA   72mA     20mA    72mA
                 -----  -----    -----   -----
                  43mA  401mA     25mA   172mA

Total power from +-5V supplies:

   340mW idle
  2865mW active

Duty cycle of 5 sec active, 15 sec idle:

   971mW average
The numbers are based on estimates of worst case power using specs from rad-hard chips. The estimate for TXCO isn't based on an actual rad-hard part. It is 2mA for a 2.5ppm 1V p-p osc plus 2mA for a 12MHz buffer to bring it up to 5V logic level.

The Dosimeter estimate is based on the assumption that the 100mW for detector bias must be maintained during idle but the 260mW for analog signal electronics can be duty cycled. Counters and housekeeping functions have been moved to the CPU eliminating 250mW. The Dosimeter turns out to be a significant portion of the total but this estimate makes it under the 1W goal even with it included.

The duty cycle can be reduced to further reduce power. The trade-off is how many magnetometer spins that will be averaged into one data sample. The 5 sec active time is somewhat arbitrarily based on averaging 20 spins at 4Hz spin rate. Dropping back to 2sec of active time gives 8 spins for 600mW.

A low power mode for eclipse can be implemented by slowing the base sample period to 80sec. This drops the average to 500mW but allows full data sampling to continue (at the slower rate).

This example doesn't show transmitter power. A 5W transmitter on for 10mins doesn't contribute much to the orbit average. It is assumed that the RF system can be completely powered down during most of the orbit.