HSLink Introduction
•
17
Note: The minimum polling cycle time is 90.3µsec, even the maximum
address is less than 3.
1.4.7
Communication error handling
Though the HSLink communication protocol is dedicatedly designed to
avoid any error, there is inevitably still some chance that communication
error may occur, eg: light striking, suddenly off-line… etc. If this is the case,
never be too worried, the master will know it. In HSLink, the master holds an
accumulated slave-no-response count
for every individual slave I/O
module. The c ount value will be updated for every slave module during every
polling cycle.
l
If communication to certain slave I/O module is successful, the
no-response count value for this slave will be set to ‘0’.
l
If the communication failed, the no-response count value increases by
‘1’.
l
When the count is larger then or equal to ‘3’, a binary flag indexing
communication error will be set to true.
l
The maximum value of no-response count is ‘7’. It retained at ‘7’ even
if error continue to occur.
The
no-response count value
and the
communication error flag status
can
be obtained by software function call, or every time when user want to set or
get I/O values, these error handling data will also be returned.
In addition to no-response count, the HSLink supports a self-diagnosis
function to detect off-line or out-of-communication of any slave module. A
software communication error-handling driver does this. In a
programmable period of time (default 20 ms), the master send IRQ to
trigger this driver to check every slave’s
no-response count value.
If the count
value is ‘7’, the drive will inform system, by event, that some slave I/O is in
mal-communication situation.
