Contrologix Alarming

It has been awhile since I posted a question... Thanks again for the help in the past.

I have the need to set a bit true if any of 100 INT'S values are greater then 0, basically like a global alarm. How would I go about doing this?
I really don't want to have to do a Greater then for each INT so I was hoping someone has a better Idea.

Tag Properties....... DISCRETE_ALM and the data type is INT[100]

Thanks

jgunnels said:

It has been awhile since I posted a question... Thanks again for the help in the past.

I have the need to set a bit true if any of 100 INT'S values are greater then 0, basically like a global alarm. How would I go about doing this?
I really don't want to have to do a Greater then for each INT so I was hoping someone has a better Idea.

Tag Properties....... DISCRETE_ALM and the data type is INT[100]

Thanks

Click to expand...

I'm super stoned right now, but.

pointer:= 0
While pointer < 98 Do
If DISCRETE_ALM[pointer] <> 0 Then
GlobalAlarm:= True;
Break;
Else
Pointer:= pointer + 1;
End_While;

Ganutenator,
Thanks for the quick reply. Hope this doesn't harsh your buzz but I have to do this in ladder logic no other way at this point.
Have a good one.

Check out the File Search Compare instruction (FSC).

Also, hello from Vancouver!

jgunnels said:

Ganutenator,
Thanks for the quick reply. Hope this doesn't harsh your buzz but I have to do this in ladder logic no other way at this point.
Have a good one.

Click to expand...

so going to try to do this sober.

did the structured text version even work? I'm drunk now as well.

That's the best I've got. All of the complexity should compile out.

Why not add up all of the individual integers into a total integer. If total integer >0 then alarm.

Depending on the application, YMMV...

The loop does not increment once the alarm happens, so you'll know which INT raised the alarm:

There's a possible solution in the RA forum, using ladder and FBC:

http://forums.rockwellautomation.com/posts/64cd74a521

sym_man said:

That's the best I've got. All of the complexity should compile out.

Click to expand...

You actually don't need the test for equality. If any array element is not zero, XOR will result in a value greater than zero, which evaluates the expression as true.

That said, if you have any multiple of two INTs of identical, simultaneous coolness, the expression evaluates as false. No alarm!

That said, its execution time appears to be 83us versus my 53us (worst case on an L33ER).

The FSC is really the best solution as dmroeder mentioned. It is built to do exactly what you are asking. The FAL solution above can certainly work too, but it involves copying the data and then looking at the copy. The FSC is looking at the data without having to copy it to a temporary location. The FSC can work in ALL mode where the FAL really has to be in INCREMENTAL to work for this. The FAL has to execute over a period of 100 scans to completely test for all alarms. Technically, 200 scans the way it is written as it must see a FALSE-TRUE transition in order to transition to the next value. Although I can't see how the subroutine is being called. ALL mode does all 100 in one scan.

In my example it stops at the first location that is greater than 0. You can see I highlighted that element #10 has a value of "1". All other tags had a "0". Rungs 4 & 5 are just to let me know I had an alarm and where.

OG

Thanks dmroeder ( Hello from Vancouver to you too ) ,opereaghost and everyone else for the help.
The FSC was the ticket...!