Question: We have a GE series 5 PLC running a section of our plant. The other day, after power was cycled on the IC655CPU500, a handfull of forces (but not all forces) had changed state. For example I 102 is found Forced OFF when previously it was Forced ON. I have never seen this happen and power has been cycled many times before. In the manual, GFK-0023B pg. 4-10 at the top, it says "The ladder diagram logic cannot change overides; however, non relay functions can change the state of an overriden reference". Can you eleborate any on this? Do you know of any explanation for the states of some forces being changed?
The GE Fanuc Series 5 maintain 3 separate tables for its Boolean references.
- Status table – stores the states of the references
- Override table – has a bit for each Boolean reference that indicates whether or not the reference is overridden, Applies only to those reference types that supports overrides. Ladder logic cannot add or remove an override.
- Transition table – keeps track of Boolean references when they transition from 1 to 0 or 0 to 1. This information is used by one-shots and counters and is not accessible by the user.
In the case of overrides, no coil type (output, latch, one-shot, timer or counter) can change the state of a reference that is overridden. However, data moves, table functions, math function and other similar functions (AKA mnemonic functions) can change the state of an overridden reference. This is because these functions deal with bytes and words and any Boolean reference assigned to the output of one of these functions is being treated as part of a larger block of data and overrides become irrelevant. To check for this type of scenario, you must perform an implicit search for the reference. An explicit search will not find a bit in the middle of a block of data in a mnemonic function.
As far as the references that changed due to a power cycle, this problem could be due to a piece of logic that may not have been enabled or disabled on previous occasions when power was cycled. Another possibility is that there were power or ground transients that occurred during the power cycle. However, I would expect additional problems with data corruption in this particular case and that did not happen, even though it is possible that only certain data areas were affected. It is also possible that the memory chip and/or its supporting ICs may be having an issue.
Also, it is possible that an outside device communicating to the GE Series 5 via the serial ports on the CPU or a CCM may have written directly to the data table. An HMI/SCADA system typically cannot check if a bit is overridden. This would mean that the event occurred during/after a power cycle was purely a coincidence.