Campbell Scientific CR7 Measurement and Control System - Manual (Page 29)

1-1
SECTION 1. FUNCTIONAL MODES
1.1 PROGRAM TABLES - *1, *2, AND *3
MODES
Data acquisition and processing functions are
controlled by instructions contained in program
tables. Programming can be separated into two
tables, each having its own programmable
execution interval. A third table is available for
programming subroutines which may be called
by instructions in Tables 1 or 2 or by a special
interrupt. The *1 and *2 Modes are used to
access Tables 1 and 2. The *3 Mode is used to
access Subroutine Table 3.
When a program table is first entered, the
display shows the table number in the ID Field
and 00 in the Data Field. Press A and the CR7
will advance to the execution interval. If there is
an existing program in the table, enter an
instruction location number prior to A and the
CR7 will advance directly to the instruction (e.g.,
5 will advance to the fifth instruction in the
table).
1.1.1 EXECUTION INTERVAL
The execution interval is entered in units of
seconds as follows:
0.0125 .... 0.1 seconds, in multiples of 0.0125
0.1 .....6553 seconds, in multiples of 0.1 second
Intervals less than 0.1 second are allowed in
Table 1 only. Execution of the table is repeated
at the rate determined by this entry. The table
will not be executed if 0 is entered. Values less
than 0.1 are rounded to the nearest even
multiple of 0.0125. If the Interval is 0.1 or
greater, the CR7 will not allow entry of digits
beyond 0.1.
The rate at which the CR7 can execute a given
table must not be confused with the sample
rates for the measurements contained within
the table. When a table is executed and a
measurement is made, the Control Module
instructs the I/O Module which measurement to
make and how many times to repeat it on
successive channels. The I/O module then
repeats the measurement as fast as possible
and stores the data until the Control Module is
ready for it. The Control Module takes the raw
data and scales it as required by the instruction
initiating the measurement. The next instruction
in the table is not executed until the scaling is
completed. The maximum sample rate for a
measurement is the rate at which the I/O
Module can make a number of measurements
specified by a single input instruction. Because
the sample rate does not include the processing
time required to scale the measurements into
engineer units, the execution time of an
instruction will be greater than the sample rate
for the measurement specified by the
instruction. The execution times for the
instructions are given in Section 3.9.
The throughput rate is the rate at which a
measurement can be made and the resulting
value stored in Final Storage. The maximum
throughput rate for fast single ended
measurements is approximately 310
measurements per second.
If the specified execution interval for a table is
less than the time required to process that
table, the CR7 overruns the execution interval,
finishes processing the table and waits for the
next occurrence of the execution interval before
again initiating the table (i.e., when the
execution interval is up and the table is still
executing, that execution is skipped). Since no
advantage is gained in the rate of execution
with this situation, it should be avoided by
specifying an execution interval adequate for
the table processing time.
NOTE: Whenever an overrun occurs,
decimal points are displayed on both sides
of the sixth digit of the CR7 display (e.g., L
O.G. in the *0 Mode).
When the Output Flag is set high, extra time is
consumed by final output processing. It may
be acceptable if the execution interval is
exceeded at this time. For example, suppose it
is desired to measure every 0.1 seconds and
output processed data every ten minutes. The
table requires less than 0.1 seconds to process
except when output occurs (every 10 minutes).
With final output processing the time required is
one second. With the execution interval set at
0.1 seconds, and a one second lag between
samples once every 10 minutes, 10
measurements out of 6000 (.17%) are missed:
an acceptable statistical error for most
populations.