Technical data

Response time The response time to a time interrupt request corresponds to the
processing time of a block or a STEP 5 operation (depending on the
selected preset). If, however, process interrupts are still in the queue
when cyclic program execution is interrupted, the time-driven
program is only processed after all pending process interrupts have
been completely processed.
The maximum response time between the occurrence and processing
of a time interrupt is then increased by the processing time of the
process interrupts. If you want to exclude as far as possible the chance
of a collision for a particular time interrupt OB xy, remember the
following rules:
A + B + C< D where A = the sum of the processing times of
all higher priority program
processing levels (process,
controller, time interrupt OBs)
B = processing time of the time
interrupt OB xy
C = runtime of the longest block of all
lower priority processing levels
D = time base of the time interrupt
OB xy
Note
If you run your program not only cyclically but also time and
interrupt-driven, you run the risk of overwriting flags.
This can occur if you use flags as intermediate flags both in the
cyclic and in the inserted time-driven or interrupt-driven
programs and the cyclic program is interrupted by a time or
interrupt-driven program.
For this reason, save the signal states of the flags in a data block
at the beginning of time or interrupt-driven program execution
and rewrite them into the (doubly assigned) flags at the end of the
interrupt.
Four special organization blocks are available for this purpose:
OB 190 and OB 192 "transfer flags to data block" and OB 191
and 193 "transfer data fields to flag area" (refer to the relevant
section).
To avoid double assignment of flags, you can also use the S-flags
for most applications. Special "saving procedures" for flags are
then no longer necessary (there are enough S flags available).
RUN Mode
CPU 928B Programming Guide
4 - 44 C79000-B8576-C898-01