The first thing I noticed was that it does not look like you are releasing your queue when you are done with it.
You
have error clusters strung through but no error handling that I can
see. A good practice is to use shift registers on your error clusters
and place your dequeue in a case with the error cluster wired in it.
That way if an error occurs, you can run an error case to handle it
instead of dequeuing that loop.
You have a lot of
enqueuing and not a lot of dequeues. It looks like you are running
your stated with the preview queue and then dequeuing inside the subvi
to get the item out of the queue... I would suggest running the tates
from a dequeue and anything that needs top be scanned constantly, do
that in an idle case. this is done by setting a timeout on your queue
and if it times out, send an "idle" to the selector on the case
It
also looks like you are enqueing a lot at the oppisite end. This may
be necessary for you your doing, but that can result in your queue
building up (i.e. adding more to the queue than you are reading off,
espically since you are running your cases from the preview)
I can look at it in more depth later but those are the items that pop out at me.
You
may also want to change your top loop to use an event structure and
move all those sub vis that look like they are monitoring or scanning
into an idle case as I described earlier.
Try doing a
search for "queued state machine" and you will probably find quite a
bit of nice advice/archatechure/templates/etc for what you are trying
to do. Specifically it look like you are trying to use it as a
producer consumer dual loop structure.....
Hope this helps a little
Cheers
Greg
Gregory Osenbach
