Abstract logic when High+HIGH give you Low output

[darkesha]

I am running out of ideas for this problem I have on one of our machines.
In advance I want to say there are no MCR's or doubled outputs in some other subroutines.
I will also upload the trend screenshot and logic screenshot.
Trend was set for max possible resolution, 1ms.
The white line is degrees (coming from encoder) - machine makes full revolution - 0-360 degrees, and home position is around 68.
Problem only happens when encoder makes jump from 359 to 0 degrees. It doesn't happen always.

What I don't understand is the actual logic behavior. You can see that my output 010.6 is a green line and it is being energized when 014.5 goes high (cyan) and then it will latch itself together with CAM2 (yellow line). What I don't understand is why it drops out ?
I can see that yellow is staying high (and this is basically position of the encoder - everything thats not in between 60-75 degrees). and since it's latching itself in, it should not go LOW.
Red output is just an upper rung and it goes high when green goes low. The pink one is the actual start/stop of the VFD that runs our transfer table.
I makes me go crazy, since I have no other ideas of what's goin on here.

Thanks to everyone,

 

[darkesha]

I had some problems with u/l. There are screenshots.

 

[bernie_carlton]

I would bet that this is a 10 line BCD encoder as opposed to a 9 wire Gray Code necoder.

 

[darkesha]

We installed 2 transfer tables at the same time with same hardware and to the same ctrl lgx card. One behaves this way, the other never gave us grief.
Not sure about encoders, since they are in production now.

 

[milldrone]

Problem only happens when encoder makes jump from 359 to 0 degrees. It doesn't happen always.

I would bet that this is a 10 line BCD encoder as opposed to a 9 wire Gray Code necoder.

See Bernie's post (#4) over at MrPLC for an explanation.
http://forums.mrplc.com/index.php?showtopic=18645

 

[kamenges]

Humor us

What are the conditions driving BUCKER_TT.CAM2?

I'm guessing Bernie and milldrone are on the right track, given the way your encoder feedback looks when you have the problem. Did you notice that the feedback goes negative? That will set a whole lot of bits you wouldn't normally expect to be set.

Keith

 

[darkesha]

It doesn't go negative, after 359 it goes to 0 (this is acording to trend of course).
And it's ( AQB)quadrature encoder, not BCD.

And for .CAM2 it set to be high whenever NOT in range of 60-75 degrees of feedback. So we have a NotCAM2 which goes high when in this range, and in next rung we have xio of this output that drives CAM2.
Can not recall direct reason for this, I think it was something releated to this whole revolution loop. But it shouldn't matter as long as after 359 I always get 0.
Also worth of mentioning is that this is the only instance of CAM2 used in program.

 

[bernie_carlton]

Is it possible for mechanical vibration to give 1 complete count backward 360 - 0 - 360 - 0 ? Would that cause a problem?

 

[kamenges]

Originally posted by darkesha:

It doesn't go negative, after 359 it goes to 0 (this is acording to trend of course).

The trend plot you posted is a little misleading. The actual position scaling is from -10 to 370. The last position cycle in the plot looks like it swings through that whole range. The three or four cycles before that look like they cycle through the correct 0-359 range.

But even given that I can't see how the count would get into the 70-75 degree range. It just seemed odd that the positon plot looks different on the cycle where the problem seems to appear.

Keith

 

[darkesha]

"The trend plot you posted is a little misleading. The actual position scaling is from -10 to 370."

When I trend, I always try to put the whole range + a little more into graph. That's why I put digital values from -1 to 2 or in this case ti +10 so I have flexibility in moving differen bits up and down.


I noticed it's misleading after I took the screenshot. The reason why that white line seems it's going in minus is because i clicked on that position to read out values. Otherwise if I positioned cursor anywhere else the bottom tip would be in the same position as other ones.
Other then that, how can I open the TBS trend log files. It looks like there is no default app installed on my computer, and short search through rockwell page didn't give me an answer.