FFL Issue

[BM80]

I have a project that is loading pallets into 10 test bays and then unloading them when the test is complete. I am using an array to store which test bays have completed testing and using an FFL command to store them in the array, so they are unloaded in a FIFO order. I am having an issue with the FFL consistently loading a completed test bay's number. Out of 100 times 6 or so may not load into the array. There is a prox sensor in each test bay that senses the pallet and when the test is complete the operator presses a push button at the test bay. If the button is pressed and the prox sensor senses a pallet then a test complete tag is latched for that test bay. An OSR of the this latched tag is used to MOV a test bay number into a tag which is then used as the source for the FFL command to load into the array. The test complete tag is getting latched, so I know that part is working correctly. I can't seem to figure out why this is randomly not working. Attached is the routine if anyone cares to look at it. The part of the routine starts at rung 10.

 

[drbitboy]

It seems extremely unlikely, but if two or more TESTBAYn_PB1's are pressed on the same scan cycle while their corresponding TESTn_PROX's are 1, then the last bay number value written to TESTBAY_COMPLETE_FIFO_SOURCE would overwrite the previous values, the previous values would be lost, and you would get the symptom you are describing.

Also a suggestion: the PB+PROX-driven logic (rung pairs 10+11; 14+15, etc.) could be done more clearly with the canonical Start/Stop Circuit pattern i.e.

• Put a branch with a XIC TESTBAY1_COMPLETE instruction around/in parallel with the XIC TEST1_PB1 instruction on Rung 10
• Change the output of Rung 10 to be a coil OTE TESTBAY1_COMPLETE instead of a latch (OTL).
• Remove Rung 11

 

[drbitboy]

It seems extremely unlikely, but if two or more TESTBAYn_PB1's are pressed on the same scan cycle while their corresponding TESTn_PROX's are 1, then the last bay number value written to TESTBAY_COMPLETE_FIFO_SOURCE would overwrite the previous values, the previous values would be lost, and you would get the symptom you are describing.

As unlikely as it is, you could replace Rungs 10-52 with 10 rungs like the one below, replacing the n's with 1 through 10, and those rungs would not have that problem i.e. you would be able to have all 10 bays complete on the same scan cycle and all 10 would get loaded into the FIFO.

Another question: where are the FFU instructions?

​

 

[BM80]

I don't think it two or more push buttons are being pressed during the same scan cycle. They are pushed by one operator and are a couple feet apart. Thanks the logic suggestion I will implement that to see if it helps at all. The FFU is called in a different sub routine. Is there anyway a FFU you could be taking place at the same time as the FFL and somehow prevent the FFL from running?