HP 1650B Programming Reference Manual page 87

HP 1650B/HP 16518
Programming Reference
In the State acquisition mode with tags, data is
information Bytes.
obtained from the target system with each clock and checked with the
trace specification. If the state does not match the trace specification, it is
checked against the prestore qualifier. If it matches the prestore qualifier,
then it is placed in the prestore buffer. If the state does not match either
the sequencer qualifier or the prestore qualifier, it is discarded.
The type of information in the bytes labeled Data depends on the Prestore
vs. Tags bit. When the Data bytes are used for prestore information, the
following Count bytes (in the same column) should be ignored. When the
Data bytes are used for tags, the Count bytes are formatted as
floating-point numbers in the following fashion:
12 bits U&KU& 1
EEEEE MMMMMMMMMMM
The five most-significant bits (EEEEE) store the exponent, and the eleven
least-significant bits (MMMMMMMMMMM) store the mantissa. The
actual value for Count is given by the equation:
Count = (204S + mantissa) X 2exponent - 2048
Since the counts are relative counts from one state to the one previous, the
count for the first state in the data structure is invalid.
If time tagging is on, the count value represents the number of 40
nanosecond ticks that have elapsed between the two stored states. In the
case of state tagging, the count represents the number of qualified states
that were encountered between the stored states.
If a state matches the sequencer qualifiers, the prestore buffer is checked.
If there are any states in the prestore buffer at this time, these prestore
states are first placed in memory, along with a dummy count row. After
this check, the qualified state is placed in memory, followed by the count
row which specified how many states (or 40 ns ticks) have elapsed since
the last stored state. If this is the first stored state in memory, then the
count information that is stored should be discarded.
DATA
System Commands
6-13