anti tie-down circuit

[tlvaun]

I use the following code on all of my programs for a two PB cycle start on a machine. BTW, machines where I can legally go through a plc.

Just curious Tim, I was under the impression that now you have to have a safety relay or safety plc for an anti-tie down. Is there certain machinery where it is not required?

 

[Rick Densing]

In my case, I usually have a light curtain or other safety device between the operator and the machine. The two hand setup is just extra and helps stop accidental starting of the machine. With the safety devices, no one should get hurt, but you could break the machine.

 

[Allen Nelson]

ganutenator

All I was doing was simplifying your code, removing double references to the pushbutons (the PB one-shot along with the PB on your timer rung), etc.

The code can be simplified further. First off, take the last rung (written correctly):

                  BUTTON A             BUTTON B     TOGETHER

     BUTTON_A     TIMER     BUTTON_B    TIMER       OK TO START

---+----| |--------|/|--------| |--------|/|----+----( )

   |                                            |

   | TOGETHER                                   |

   | OK TO START  BUTTON_A  BUTTON B            |

   +----| |--------| |--------| |---------------+

.
Since -| |- BUTTON_A and -| |- BUTTON_B appear on both branches, the rung can be reduced to:

                             BUTTON A   BUTTON B    TOGETHER

     BUTTON_A   BUTTON_B     TIMER      TIMER       OK TO START

-------| |---------| |----+---|/|--------|/|----+----( )

                          |                     |

                          | TOGETHER            |

                          | OK TO START         |

                          +----| |--------------+

.

And then you ask yourself "Why do I need two timers?", and so you remove them, and you are left with this:

[attachment]

which, gee, is the logic that Rick posted (only using the NOT /DN bit instead of the /TT bit).

But you wrote it, not me. I just eliminated the one-shots, since the timers were only going to run while the PB was pushed anyway, and removed the logic that was making the OK TO START bit a one-shot.

I did cheat and add the seal branch, since when I was done, I could see that the OK TO START bit was only going to be ON for the duration of the shortest timer, making it neither a one-shot nor a sustained bit (but plenty good enough to initate a start cycle in the way that you intended it to be used).

ganutenator (Mike is it?), you said:

I kept going back and adding logic to try and fix the code that I origionally started instead of just re-doing.

One rule of thumb in debugging I use is that I know I'm doing it right when I'm removing code. Adding code is usually just a bandage on top of a wound. You have to remove existing bandages, and clean out the wound for it to heal right. That isn't the same as re-writing it

 

[Greg Gauper]

Something else to consider....

I have worked on a number of machines in which the required number of operators could vary from 1 to 4 people, depending on how big the part was that was being made by the machine (bigger parts = more people). This could change from day to day depending on the production run requirements.

These machines all had operator T-stands with dual palm buttons, connected to 'Joy' plugs for quick set up. On each T-stand all of the palm buttons had a N.O. contact and a N.C. contact. On each pair of palm buttons the N.C. contact were wired in series to a single input thru the 'Joy' plug connector and all of these inputs were programmed in series to trigger a 'time delay off' timer for the anti-tie down circuit. So once any single pushbutton had been pressed, the series circuit lost power and started the timer. Each individual pushbutton had its N.O. contact wired to a discrete input. All of the active pushbuttons on each T-stand had to be pressed within the anti-tie down time period as sensed by all of the N.O. contacts.

The thing that made this flexible was that each T-stand 'Joy' plug could be replaced with a 'Dummy' plug that had the inputs jumpered out for any T-stand not being used. The anti-tie down logic didn't have to be reprogrammed when they changed the number of operators. And no, the machine was not permitted to run with all 4 dummy plugs installed. The machine also had hard wired light curtains and safety mats for primary safety, the anti-tie down circuit was for redundancy.

 

[rogerhollingsworth]

Tim

Just another flavor
HOLEY SMOKE RICK, things sure are easy once you know, huh?
By the way I used your ladder, just rearranged things in PAINT.

Thanks

 

[Rick Densing]

Roger- if you make it a .jpg, it will display

 

[KEN_KACEL]

From experience I would not let a PLC control the anti-tie down circuit. Whether it is OSHA approved or not. I once ran a supervisory
bit through the PLC to maintain fluid level and then sound the alarm.

The MITSU puked and hence no level control, thing overtemped, bled all of its refrigerant into the air(10 55gal drum$)$$$$$$, and did not sound the PLC controled alarm. Embedded intelligence only works if its scanning.(I love the scare crow in the Wizard-of-Oz, if I only had a brain)

Another story: The DH+ networking all MMIs puked lost all control, no stop/starts, no VFD% control (already running), machine was on its own with no way to input any commands. This stuff happens. !!ESTOP!! ...then crunch...then Alot of lost time. (This is the reason why cars have airbags and houses have smoke detectors, and Murphies law thrives on imperfection).

Please design for failure. (Watchdogs, Hardwired safetys, un defeatable lockouts,ect.)

"Its hard to idiot proof when idiots are so ingenious"

 

[Tim]

tlvaun,

We have safety lights on all our machines. I'm wondering if this is why this code came that way from Japan. You have the saftey lights, but I would still rather start using the proper anti-tiedown like Rick D.
I don't know Japans standards, but I've learned most of my programming from the Japanese Engineers.

You'll have to look up the safety requirements on your particular machine. I'm not talking about a press, its just what Rick is doing adding a ___________ feature, "I'm hesatent to say ..safety, because the plc's I'm going through are no way, control reliable". To my knowledge you can program an anti-tiedown, but this cannot be considered in no way apart of your control reliable circuit,except on some plcs as other have pointed out.

Have a good Day,
Tim

 

[ganutenator]

Allen, once again I am impressed. Yes, it is Mike

Have been experimenting around with the ole Boolean Algebra thingy.

lets let

Button_A = A
Button A Timer = B
Button_B = C
Button B Timer = D
and
OK to start = E

THEN

                  BUTTON A             BUTTON B     TOGETHER

     BUTTON_A     TIMER     BUTTON_B    TIMER       OK TO START

---+----| |--------|/|--------| |--------|/|----+----( )

   |                                            |

   | TOGETHER                                   |

   | OK TO START  BUTTON_A  BUTTON B            |

   +----| |--------| |--------| |---------------+

could be written as

(A*B'*C*D') + (E*A*C)

could be reduced to

AC(B'D')+AC(E)

and further reduced to

AC(B'D'+E)

which is the same as

                             BUTTON A   BUTTON B    TOGETHER

     BUTTON_A   BUTTON_B     TIMER      TIMER       OK TO START

-------| |---------| |----+---|/|--------|/|----+----( )

                          |                     |

                          | TOGETHER            |

                          | OK TO START         |

                          +----| |--------------+

or at least I think.

Been messing with this stuff just a little. Has anyone used Boolean algebra to simplify code? How do you deal with oneshots, timers, scan, etc.?


Short painful story:
Had this boss once that would draw these Boolean Logic Diagrams (at least that is what he called them) which were drawn up as logic gates. He would hand them to the PLC programmers and they would go to town programming the PLC code. Nice thought, but it would make for some disaster code because the code wouldn't appear to follow any logical order. Everything would be dumped in one program file, no organization. A coil that would be used to fire a sequence in rung 4 might not appear until rung 400. I reduced a page of Logic gates down to only one N.O. contact once using Boolean Algebra. He was red. Contributed to me finally getting laid off. hence the www.prousucks.com domain that I purchased for $11.00. Now, I am being paid by his ex customers to fix the logic. Can't seem to escape it.

So I defintely need some pointers in reducing code more efficiently. I am going to study the above posts some more before I turn in. I wish I could reduce code that quick.

 

[drbitboy]

Necro-thread alert

Caveat: necro-thread alert!

As if drbitboy, O.C.D. could leave this alone when there was a whole XIC instruction to be saved!

​