next up previous contents
Next: Science Data Header Up: Science Data Format Previous: Data Structure

Pointer Header

Each 1 Mbit block of CUBIC data dumped to the ground will begin with a 42 byte Pointer Header, which is constructed by the ICP during the data dump. The first 4 bytes of the Pointer Header act as a unique 32 bit CUBIC \ ID/synch word at the beginning of a CUBIC data dump. The remainder of the Pointer Header specifies the first and last addresses that need to be dumped, plus a copy of the serial command that produced the dump. The format of this header is described in Table 30.

  
Table 30: Pointer Header Format

The first address to be dumped is given by the Read Pointer, defined by a page address and an offset. The last address to be dumped is defined by the Write Pointer, also defined by a page address and an offset, which specifies the next memory location to be written into. Three copies of each pointer are transmitted to the ground to guard against transmission errors (these pointers are not written into Science RAM, so they do not have ECC bytes associated with them). The Read and Write Pointers ``wrap around'' the memory. In the event that the memory dump does not empty the memory, the Read Pointer is updated following the dump to allow the remaining data to be dumped on the next pass. In this case, the next data dump will begin with a Pointer Header, followed by the Science Data Header for the orbit just finished, followed by the remaining data from the previous orbit. The data must be patched together with the previous data set during ground processing. If the Write Pointer overlaps the Read Pointer, data collection stops until data is dumped from the memory to make room for new data.



next up previous contents
Next: Science Data Header Up: Science Data Format Previous: Data Structure



David N. Burrows
Thu Oct 24 10:59:06 EDT 1996