Troubleshooting Intermittent PLC I/O Problem

[samuelgmaurer]

Attached please find the following: 

Ladder View Project 17MAR15_ATA2.PDF <-Problem in Stage 22 starts @ Rung 98.

PLC I-O Problem Description 13-Mar-15.PDF <-Description of what the PLC is doing.

Right Edge Trend 17Mar15.JPG <-Typical Trend View for several cycles but S22 not On, or Y31 On, or Y32 On....?

Stage 22 is where the intermittent problem seems to occur.  The Trend picture shows that the two triggers Y31 & Y32 should go on when the table is at the right edge of the object. 

Although there were nineteen successful edge find and returns, before a failure to return to center.  The Trend View shows S22, Y31 & Y32 do not go on during the entire all nineteen successful cycles or the last cycle (left edge, center, right edge, center)that does not work.

Suggestions on how to find the cause of this problem.  Thank you.. Sam

[Greg]

The first thing you must realize is that the Trend View is not going to show you the status accurately of everything if things are happening very fast. The data is not stored in the PLC but is gathered by DirectSOFT6 communications. So ever how many scans of the PLC happen between times that DirectSOFT6 polls the PLC for data, that is the time that the data will be missing on Trend View.

With that said, and considered, it seems to me that when the jump is made to Stage 21 and the Trend View seems to show that X20 and X21 come ON simultaneously, supposedly triggering a jump to Stage 22, that this is obviously not what is happening. Thus without studying it too deeply yet, it looks to me that it is possible for Rung 95 to become true and a jump be made to Stage 124 instead. In other words, though the Trend View seems to show X20 and X21 coming ON simultaneously, it is possible that X20 came on one (or so) scan(s) before X21 and thus the jump out of Stage 21 was not to Stage 22 (to SET Y31 & Y32 ON, but instead a jump was made to Stage 124.

[LWgreys]

The problem is that your trying to watch both table sensors on the same rung. Try separating the sensor logic for each table on their own rungs. A little more coding but you will get the results you need.