BusCoupler Addressing - Newbie Problem - Read, Laugh ( and assist ?).

ME :
I have absolutely no plc experience other than the free tutorial on this site.

SITUATION BACKGROUND :
I have 'inherited' a machine - It has Siemens plc (with Beckhoff Profibus and MX2 CNC ).

SITUATION :
The machine has gotten itself 'stuck', leaving a significant tool weight, plus the additional weight of the now hydraulically relaxed machine part that was meant to pick the significant weight up in the first place - sat, immovable on a now supported, enclosed, supporting table ( it's worse than it sounds too !! )

EVENT :

1. The machine did not complete a hydraulic/mecanical sequence within a predefined time interval (25s).

2. Unknown if the hydraulic/mecahnical sequence was completed or not.

3. Actions will not advance until this hyd/mech sequence is completed - This hyd/mech sequence would not reverse either, so unable to attempt a back-tracking of machine actions.

4. People been messin' with it electro/hydraulically to 'make it work again'.

5. Still don't work.

6. Machine now way out of prescribed operational sequence.


So to try and get some idea of what to do.............

QUESTIONS :

1. What, normally, is best practice for sequenced heavy operations exceeding timed delays ? if there is any.

2. If I drive everything back to home position manually, switch all power off and then restart, will the plc see everything in home position and be as per normal operation ?

3. Might the plc itself be stuck somewhere on the ladder, waiting for an input before advancing ?




A ways' back replaced BK3000 BusCoupler may have a different address set on it compared to the original, though its' rail still shows correct I/P O/P states.

so, can you remark on :-

4. The address set ( turning dials in casing )on the BK3000 must be specifically referred to by the plc for the BusCoupler to be seen and commands sent to it ?

5. If the address on the BusCoupler is not contained/referred to anywhere within the plc, the Buscoupler may as well not even be fitted?


Which leads me to this :-

6. Under what circumstance would a BusCoupler with a different set address ( otherwise a direct replacement ) be seen and instructed by the plc ?

If you have been good enough to read this far I have nigh on 40 tonnes that could really do with being moved into the right palce !!

Yes I am aware of diagnostics and 'first aid kits', and of good folk like some ( or all - na, don't want to crawl too much on my first real post ) of you who do this while asleep - er... I mean who could do this in their sleep. It's just a new situation that I am not yet equipped to deal with properly ( make that at all !! ) yet.

Any replies, please describe in laymans terms first, then go technical if you want.

If something of this nature has been discussed elsewhere, then my apologies for duplicating.

Well, Mr Uberfault, you've got a real mess on your hands. The first piece of advice I can give is not to get anyone killed (as a fluid power engineer, I have great respect for hydraulics).

To answer some of your questions:

1. What, normally, is best practice for sequenced heavy operations exceeding timed delays ? if there is any.

A: It depends. As a rule, I return everything to a safe home position and then stop the machine if it cannot finish its cycle properly. Of course, without seeing your machine, it's hard to say in your case. But having 40 tons of steel stuck in place is definitely not good.

2. If I drive everything back to home position manually, switch all power off and then restart, will the plc see everything in home position and be as per normal operation ?

A: It depends on the programmer and the machine. At a minimum, the machine should "wake up" and do nothing at all until an operator tells it to do something (such as find a home position). Or, it's possible that it was programmed to remember what it was doing before (I do that often), but I still force the operator to reinitiate the sequence.

3. Might the plc itself be stuck somewhere on the ladder, waiting for an input before advancing ?

A: It's very possible, but it's poor programming practice. I always fault out the machine in an orderly manner if a sequence cannot be executed, but I have seen conditions like you describe. It's kind of dangerous, because a machine could sit still for an hour, when suddenly a certain prox switch is made and the machine takes off again.

Of course, none of this is going to help you right now, so let's move on to the addressing issues.

Remote I/O must have the correct addresses, or things will go haywire. Hopefully, if someone gets them wrong, they are wrong in a way that causes an external fault on the PLC. But I have seen cases where someone got the addresses of two nodes reversed, and the PLC wrote to the wrong physical outputs (could be dangerous). That said, if it looks like the correct outputs are coming on when they should, then I'm guessing that the addresses are correct.

Uberfault,
It is difficult to trouble-shoot remotely, but I have some experience at doing just that, so here goes.

I agree with S7Guy's excellent comments completly. Your addresses are probably okay. I think that a sensor or input switch has failed, and has given the PLC a false indication, resulting in allowing the operation to continue when it should have been halted, hence the lock-up. Probably it did have a fault detection routine which started a timer and then waited for some signal input to switch and it did not see a switch within the allowed 25 seconds, so the PLC just stopped the machine, not knowing what else to do. Your stated that (quote):

"The machine did not complete a hydraulic/mecanical sequence within a predefined time interval (25s)"

This makes me think that the machine knew something is not right, but it does not have the logic to determine completely the problem, so it just quit.

Unless you have a very extensive fault-detection routine built into the program, it is always difficult to EXTERNALLY determine what is making the PLC "quit". I assume that the program does not have such a routine, otherwise you would not need a bunch of people messing around with the machine. Ideally you want the machine to detect its own faults and display a clear message on a text screen.

1. But right now, the next step is to get someone with a portable computer and a copy of the program, and have him go INTERNAL, an get into the program on-line, and find the input that caused the machine to mal-function. It is usually quick and easy to find the "stopping point", IF you have a copy of the PLC software, and a copy of the PLC ladder logic program, and a portable computer, and the proper connection cable to access the Siemens PLC.

2. The second-best method is to get a print-out of the ladder logic, and try to determine by looking at the PLC outputs and inputs, which ladder rung was being executed when the malfunction happened. Then look for a failed electrical device, a limit switch, a pressure switch, frayed or broken wires, LOOSE or DISCONNECTED wires. Vibration can cause things to get loose after many hours of operation.

3. The third thing is to attempt to manually move everything to the "home" position, turn off the power, and try to reset the machine. BUT DO NOT DO THIS UNTIL YOU CONSIDER 1 & 2 ABOVE. When you move the machine from its stalled position, you will destroy any hope of finding the problem that made it fail, making trouble-shooting much more difficult, and the machine may repeat the same lock-up all over again the very next cycle.

Uberfault,

It can be a VERY difficult task to sort out another persons program.
Especially if it has to be in a hurry.

Is this a "standard" machine, manufactured by an OEM company ?
Or is it a "special" machine, manufactured for your company by an external company ?
If so, get in contact with that company !
Try to establish the exact conditions before the stoppage occurred.
Maybe there are some diagnostics or some procedure to follow in your case, so that it isnt required to go online with the PLC.

If you cannot get in contact with the original supplier/programmer then someone has to dig into the manuals + diagrams + program etc.
But you say: "I have absolutely no plc experience other than the free tutorial on this site."
With all respect, to get into PLC programming can be easy, but to find your way into an existing and probably complex sequence program might be too difficult for you.
You could try to get in contact with an experienced S7 programmer.

And yes, there SHOULD have been a proper fault identification on the HMI. And there SHOULD be a procedure to get the machine back to its initial position.

Edit: About the Bus coupler:
The Profibus node address MUST be properly set for the node to be recognised by the CPU.
Depending on how the S7 program is set up, Profibus errors could stop the CPU, be handled in a proper way, or not be recognised at all.
But as you say that it was exchanged some time ago, and the problems first started now, then I dont think its related to your current problem.

Thanks for the info - someone coming to look 'inside' hopefully on Monday.
Having read a load on plc's and the various valves used in the hydraulics, the problem could originate from either a physical perspective (electro/fluid/mecanical) or, as you guys have said, sommit not 100% foolproof somewhere in the programming.

< Have just found the required Buscoupler addresses - Question at the end of this post on that >

One thing I'd not appreciated was the storage of 'current position'. With the current position being 'Upper Tool Clamping' ( hence username )this actually won't be dangerous should they activate at any point, the clamps themselves ( to clamp the tooling before the lifting sequence ) are internal of the 20t machine head.

Now, so I don't look a complete idiot on Monday, hypothetically :-

Questions ( 1+2 of 3 ) :
i. If the machine were driven to home manually, is there a way to either reset (clear? ) the potential 'current position' stored data to achieve home position ?......

ii ....or might it be possible to drive the ladder on, offline from the machine, through the current sequence to get to home position ( having 'homed' the machine manually )?.

A power down-power up would then have everything 'seen' correctly ?.



Regarding the Buscoupler addressing - I have this info :-

"AAT By Slave
Slave @=2 'Pendant'
Slave @=3 'Junction Box 1'
Slave @=...etc..........."

On the BK3xxx series, there are two rotating plastic 'disks' by which to set the address. Both disks are situated on the left side of the unit, one at the bottom edge, and the other above it. Each has 0 to 9 inlusive, settable by rotating the disk with a terninal screwdriver.

Question ( 3 of 3 ):
iii So set the address to 2 do I dial in 2,0 or 0,2 ?.
(ie is 2,0=20? or is 2,0=02 ) May sound a silly question, but I
really don't know, and certainly don't want to guess. I have the
current units to look at, but I'd appreciste confirmation
on this.



I do appreciate that either of the possibilities i or ii will not resolve the problem, but at the moment I want to establish if it's possible to relive the sagging weight of the machine head, plus the tooling, on the support table, by getting it all home and locked.
( Curent weight not really an issue, just be nice to get some weight off and achieve some machine movement ).
After checking valve function and electrical connections, I
can then remove the tooling, and try clamping again with a reduction of 5tonnes in weight. At least if i or ii is do-able, the machine and plc can be sync'd when all back home again !!

Oh, S7Guy - You may be interested to know that the machine operates at 4200bar - yes, that's FourThousandTwoHundredbar - with steel and bulletproof polycarbonate guards.

Anticipating more (very helpful) replies,

Cheers, UF (or should that be oooooff!)

Oh, and wish me luck !

Uberfault...

Your program sounds very inadequate.

It sounds like the program was written using "minimalist engineering techniques". That is, the programmer believed that both his code and the process were infallible - foolish.

It also sounds like the program was written to provide only a specific sequence... nothing more, nothing less. Again, foolish.

On a machine like this the program should provide "homing routines" or at least, allow you to manually position your tooling by hand-crank or program manual functions to whatever positions are necessary.

It should also be the case that when you power-cycle the PLC it should "look to see" the current physical state of the machine and then give an indication of what it (the PLC) thinks the situation is.

For example, on a power-cycle the PLC should expect to "see" that the machine is in a proper start-state. If it doesn't see that condition then the program should automatically allow you to "call" for "homing routines". Of course, these routines have to be carefully designed so that they do not cause more trouble than existed.

Generally, the amount of code necessary to drive the automatic portion of a process is relative small compared to the amount of code necessary to handle the problem situations (i.e. "exceptions").

One must certainly design code for the most efficient execution of the process. However, one must also include code for the "exceptions". Neglecting to address "exceptions" can be devastating to what should otherwise be a simple process.