AB 540 E processor fault

I've been getting a weird major fault on one of our 540 E revision E processors.... S/8 watch dog timer fault... It has happened 2 days in a row. This processor has tons of I/O and memory left. Can anybody help?

Have any program changes been made recently? Watch dog timer faults activate when teh processor is unable to do everything you want in the time designated. A common cause is looping logic.

Greetings rszuszkiel ...

and welcome to the forum ... as my distinguished colleague Ken Moore has said, you should carefully check your program for excessive “looping” or “recursive” programming ...

you might also want to check the setting for the watchdog - as shown in the figure attached ...

the setting is normally entered as 500 for one-half second ... I do NOT (let me repeat that ... I do NOT) recommend just increasing the setting to “mask” the problem ... that’s sort of like just increasing the size of a fuse to keep it from blowing ...

the reason that I mention the setting is that sometimes it gets changed to a lower number by accident - or by inexperienced programmers who mistakenly think that a lower watchdog setting will somehow “speed up” their program ...

if you can’t spot the problem yourself, you’re welcome to ZIP and post your .RSP program file here ... we’ll be glad to take a look at it for you ...

(suggestion: before you post it, make sure that your employer, etc. doesn’t consider the program to be “proprietary” or contain “trade secret” information) ...

hope this helps ...

Ken& Ron ;
Thanks for the comments ... I'll be opening the program up directly....

Ken & Ron :
Guess what? ... The watchdog timer setpoint was 0 mS .. I've set it to 500 to solve this problem for good..... Thanks again..
Rob

sometimes we get lucky ...

not to rain on your parade, but now the big question becomes:

HOW did it get changed? ...

if you suspect a "fat finger" on the keyboard (either yours or someone else's) then everything might be OK from here on out ... or is something in your program possibly writing a 0 to that location once in awhile? ... double-check that - and be suspicious of any HMI/SCADA type devices connected to the system ... sometimes a programmer makes a typo error to an address - and ends up with another device writing into a PLC data location over a network cable ...

cause for concern:

It has happened 2 days in a row.

Click to expand...

now you’ve got us wondering how it was running at all - with that 0 setting ...

Ron:
I have no idea who or what changed the setting to 0 .... This is a stand alone processor with ten semi automated pushbutons on it;no HMI or SCADA.(It's a new processor& rack - we're just starting to automate ) However it is on our network.. I'll keep my eye on it- Once again thanx....
Rob