View Full Version : Grafcet
May 24th, 2002, 07:36 PM
I looked at the program today. It seems to be a very indirect route of getting something done. It's a little hard to follow through the steps. They way it is written appears as if someone was writing a sequencer.
I guess the part that was really throwing me off was the print out that I was describing earlier. The program itself is in ladder logic but what was given to us for a hard copy was the flow chart and boxes.
They claim that is easier to follow through when trouble shooting. The flow chart is ok but it doesn't allow one to a person to track everything back to everything that's required to get things started.
If your given just the rungs of a program that run a motor for example. It is pretty useless if there are other bits that need to become true or false for the motor to run and you don't have the rungs that show what sets the state of those bits.
I hope this is as clear as mud. I think I might have even confused myself.
May 24th, 2002, 08:15 PM
You are correct on several counts.
It looks like someone was writing a sequencer, because someone WAS writing a sequencer.
That's principly what SFC are used for (the S stands for Sequential).
And the way AB implemented it, it is very hard to watch what's going on. If something gets hung up, it isn't so bad, because then you have time to go to the transition step that isn't transitioning. But if you've got lots of things going on - well, I haven't used them in years for precisely that reason.
One trick that might make things easier: I think you'll agree that the easiest way to monitor it would be to have two windows open at the same time - one looking at the SFC (so you know what step it's on), the other looking at the relevant ladder. But RSLogix only lets you have one open Ladder window at a time. But you can make that one Window a split screen. If you've done it in Excel or Word, you know how it works: Postion your mouse to just above the scroll bar. The cursor should change to a horizontal line. Drag that line down, and you've have two independent windows (The new one will only have LAD 2 open). Click in there, and you can open up a new ladder file (like the SFC). You can now work back and forth, and perhaps better follow what is going on.
To track multiple bits (even if you can't see the logic behind them), try creating some CDMs.
Good luck You'll need it
May 24th, 2002, 09:18 PM
Personally from both posts surfman has not shown that he can or has the capabilities to monitor the plc online.
The program itself is in ladder logic but what was given to us for a hard copy was the flow chart and boxes.
If its written in ladder it should look like ladder and not be that difficult to understand(barring bad spaghetti code). I think the main problem is that a (my description is funky) printout/pi&D/flowchart was given out NOW the people involved are confused.
To L with all the other stuff..have them allow you to SEE the program online and when in action...spend an hr (or a few hrs) just looking at the system.
Learn the Inputs an Outputs then find them in the program...back step to learn the conditions ...inputs, timers etc that fire the outputs.
If its ladder...eventually it will make sense, if you study it.
Technically if its SFC, IL, Statement List etc...if you look at it long enough you can understand it, if you take the time to learn the basics of any of those techniques.
Note: I said understand, not that you can program...that requires alot more study, practice, and time.
The deal is, learn to be efficient(if not good/great) in any one of them...then you can learn the others fairly easily. Hopefully.
I am not experienced with programming plc's but have dealt with many brands in many ways and have learned different ways to diagnose/troubleshoot using the plc LED's and software.
As an example today the electricians I work with had a machine problem. The machine is packager/sealer. A product is inserted into plastic film ( to be wrapped ) then a button is pushed and the machine cycles...ram extends and heat wire cuts/seals package. The natural state is for the "jaws" to be closed...when product is inserted then button pushed the "jaws" open...product falls..jaws close and cut/seal. The problem was the machine was suppose to close to cut but after several cycles the "jaws" that closed would stay open. The electricians spent 4 hrs checking/changing parts...after lunch I kinda strolled over to where they were to see how they were doing. THe found a wire to a switch bad and replaced it then cycled the machine a few times..it did the same thing..jaws stayed open. OK from observing the sequence I noticed that the cycle PB(I:0/2) was only on when pushed...but when cycled didnt complete(jaws open) that cycle PB was still energized..this told me the PB (I:0/2) had a problem...being the subliminal kinda guy I try to be..I pointed the electricians towards it and left. They figured out the PB was bad and replaced it.
The point of this example is I dealt with JUST the LED's. Why..because I studied what they did. It helps on larger machines to have a printout or ability to monitor online the program.....but isnt always necessary when doing diagnostic...darn it helps alot.
NOTE TO PROGRAMMERS: DOCUMENTATION IS A BIG ASSET