Lenze DDS v2.3 - Manual (Page 220)

Drive PLC Developer Studio
Resources
8-30
l
DDS EN 2.3
Saving a start event
If an EVENT-type task is called again during processing, the new call is saved and restarted
immediately after the task cycle of the event-controlled task is complete. Only one call can be
buffered.
PRIORITY
Every user task (not the Cyclical task) has a defined priority between 0 and n
(0 = task never starts; 1 = highest priority ...n-1 = second-lowest priority ...n = lowest priority).
•
A priority cannot be assigned twice. It is thus not possible to have two tasks with the same
priority.
•
The priority of a task can not be changed while the automation system runs in online mode.
•
The priority of a task decides whether another currently processed task will be interrupted or
not. If a lower-priority task is started while a task is running, the lower-priority task will wait
and not interrupt the running task.
8.8.2 Data consistency
Multitasking operating systems or, in general, all software structures featuring “ concurrent”
processes running at different cycle times, have a problem with data inconsistency, which we will
briefly look into at this point.
When do data inconsistencies occur?
•
Variable A is used in different places within routine 1. At routine 1 runtime, variable A is
changed by a routine 2.
•
A data record consists of 4 variables that must be checked out from the memory for reading.
An Interrupt-Service-Routine 2 changes this data record during the read access.
8.8.3 Normal data processing/ IPO principle
Normal data processing for simple PLC systems, for example, always follows the IPO principle
(input/processing/output) where the data inconsistency problem does not occur!
The IPO principle deals with only one data processing branch:
•
All data required are read at the beginning of the loop.
•
In the middle of the loop, the data are processed and the output values calculated.
•
The data are output at the end of the loop (just before the ” loopback” ).
INPUT
PROCESS
OUTPUT
With this method, the input data are the same and constant during the entire processing time until
the new read at the beginning of the loop. Since there are no further processes that interrupt the IPO
process during the runtime of the loop and may access data of the running process, there is no
reason for data to become inconsistent !
Show/Hide Bookmarks