well, Robert, this is one for the books - but at this point I’m not sure exactly what book to put it in ...
first let me say that I have very little doubt that you saw exactly what you say you saw ... but ... based on everything that I know about the system, it couldn’t happen that way ...
let’s start with what the official Allen-Bradley book has to say on the subject ...
http://literature.rockwellautomation.com/idc/groups/literature/documents/um/1747-um011_-en-p.pdf
here are some pertinent excerpts taken from page 70:
LOSS OF POWER SOURCE
The chassis power supplies are designed to withstand brief power losses without affecting the operation of the system. The time the system is operational during power loss is called program scan hold-up time after loss of power. The duration of the power supply hold-up time depends on the number, type, and state of the I/O modules, but is typically 20 ms ... 3 seconds.
INPUT STATES ON POWER DOWN
The power supply hold-up time as described above is generally longer than the turn-on and turn-off times of the input modules. Because of this, the input state change from On to Off that occurs when power is removed may be recorded by the processor before the power supply shuts down the system. Understanding this concept is important. Write the user program to take this effect into account.
the way that the quiz question NORMALLY works is that the IMMEDIATE loss of 120VAC power through the NORMALLY-CLOSED contacts of the input device in the field APPEAR to be an OFF signal on the PLC’s Input Image Table ... specifically, IMMEDIATELY upon the loss of field power, the bit/box on the Input Image Table changes from a ONE status to a ZERO status ... the tricky part is that the PLC CONTINUES TO SCAN for a certain amount of time before it shuts down due to the loss of field power ... during that period of “hold-up” time, the XIO instruction is evaluated as a TRUE condition - and that causes the OTU to write a ZERO status to the output bit/box ... then later when the power is subsequently restored, the output device in the field will NOT turn back ON as intended ...
as I said, that’s the way it’s SUPPOSED to work, and I personally have NEVER seen an example of any other type of operation ...
now then, ever since you’ve reported a different type of operation in your experiment, I’ve been wracking my one remaining brain cell trying to figure out HOW that could be happening ...
here are some of the things that I’ve come up with which COULD cause the results that you’ve reported ...
(1) the output might have been FORCED ON ... but ... the program that you provided shows no sign of a force being applied ... and you clearly said that you had “TOGGLED” the bit ON - rather than using a force ...
(2) the input might have been FORCED ON ... but ... once again, the program that you provided shows no sign of a force being applied ...
(3) the I/O Slot Enable bit for the output module might have been written to a ZERO status ... but ... the program that you provided shows that the I/O Slot Enable bits were all in their normal ONE status instead ...
(4) the input module might have been a DC type ... but ... the program that you provided shows a 1746-IA16 which of course is a 120VAC type module ...
(5) the “unlatch” rung might not have been scanned ... but ... the program that you provided indicates that the rung WAS set up for a normal scan ...
(6) some other instruction (for example a Latch) in the program might have turned the output bit/box back to a status of ONE ... but ... the program that you provided indicates that there were NO other rungs in the program ...
(7) an HMI device might have been continuously writing a ONE status to the output’s bit/box ... but ... surely you’d have noticed and reported anything like that ...
as of right now, I’m out of ideas ... if I come up with anything in the future I’ll let you know ...
in the meantime, here’s something that might prove interesting in case you get a chance to do another experiment on the same system ...
the idea is that the value in F8:0 should increment whenever the system is powered down in the manner which we’ve been discussing ... if the value DOES increment, then it would be an indication that the “TRUE APPEARANCE” of the XIO is indeed working as the quiz expects it to ... (every time I’ve tried the experiment this way, the value shows at least 100 scans or so - each and every one of which SHOULD unlatch the output’s bit/box) ... so if F8:0 does indeed increment, the question would then become: why the heck doesn’t the bit/box get written to a ZERO status? ...
finally (at least for now) ... finding “weirdness” like this is one of the most entertaining parts of my job ... if there’s any chance that you could make this system available for me to play (correction: WORK) with for a few minutes, I’d certainly enjoy that opportunity ... I’d pay considerably more than pocket change for a chance to see what’s causing this system to act the way you’ve reported ...
thanks for the puzzle ... never a dull moment around here ...