Private Message

[Jay Anthony 118]

Guest_GaryH: Please register at MrPLC.com and join the MRPLC.com community Register at MrPLC.com Hi JayA, I hope you don't mind my e-mailing you but I have some questions. I don't have any real training in PLCs, just a few seminars in PLC basics, I find it difficult to explore the programming on work hours because of other responsibilities in the factory. The seminars are helpful but lack the complete picture. I have purchased a PLC ( C200HX-CPU85-ZE ) and a touchscreen ( NT631C-ST141B-EV1 ). I hope to find a small enclosure to mount them and start playing with it. ( most of the people at work don't like my term "playing" when it comes to the robots but I really enjoy programming and teaching. ) I like taking my time and doing a little extra effort in any project I do. Ok ok enough of that, my first question is what do you think about using steppers to control a whole process? I am working in a factory that uses steppers to control a line that has a loading jig, two transfer jigs, a p-q jig, three robots with weldheads, three robots for handlers and two robots for p-qing. Everytime someone puts the line in manual, it resets the steppers. So you have to empty the line out and reset everything ( scrapping all parts in the line ). I have been looking through the manuals for the PLCs and came across a instruction set of STEP and NEXT-STEP. this seems to be the better way to go but the engineer assigned to this line thinks that the steppers are the way to go. I would like to really know what you think? I'm not trying to pit your way against the way of this engineer, but it just seems waistful to keep scrapping parts everytime a new operator runs this line. I know that steppers are sequencers, just thought I would throw that in. they call them steppers at work and its hard to refer to them as sequencers. I really appreciate your time and information. GaryH

[Jay Anthony 118]

Please tell us what a stepper is. Is it a stepper motor? Or is it a step sequencer? If it is a step sequencer, what brand is it?

[Jay Anthony 118]

Ok, 007.. I'm putting two and two together, you've just been to a training in January and the instructor mentioned Step-StepNext sequencing. Try reading the following threads about Step Sequencing: http://forums.mrplc.com/index.php?showtopic...&st=0entry117 http://forums.mrplc.com/index.php?showtopic...st=0entry2325 My Webpagehttp://forums.mrplc.com/index.php?showtopic=713&st=0entry2542 Also try this beginning code for Step-SNXT: Template for Step-SNXT @ MrPLC.com

[flemming007 0]

Hi JayA, I am registered, Flemming007. Sorry, I thought I had explained it in the e-mail. The steppers are step-sequencers. I will attach a program I have to this reply. It should give you a better idea as to what I mean. Thanks again. Flemming007 LWR2_Sta_10_20.cxp

[flemming007 0]

Hi JayA, Sorry I wasn't thinking, I think if you look at the program I attached, it should show you the what I was talking about. I believe they are using shift registers to step thru a sequence of steps. I keep referring to this as steppers because the engineers at work call them steppers. The STEP - NXTS wasn't discussed at the seminar. I was looking through a old college manual of mine and found a section on Omron PLCs. The author referred to there being a advantage to using STEP programming to SECTION programming. I just thought it would make it easier to use STEP programming to reduce the down-time and scrap on the lines at work. It looks easier for troubleshooting the process as well. Thanks again. Flemming007 P.S. Do you need the comments for the program?

[Jay Anthony 118]

OK, a lonnnnnnng time ago.... I built a machine like this one using SFT as the sequence control. It's not a bad idea to use a shift register to define specific states of machine. It has the distinct advantage of defining where the machine was, where it is and where it should be going if it stops in cycle. This single fact should be used to include diagnostic code that at the very least generates error codes as to the state of the machine. The distinct disadvantage to using SFT as sequence control is that it is single directional; ie, once it is stepped forward, it cannot be stepped in reverse. My suggestion is that you substitute SFTR for your SFT instructions which then gives you direction control in case parts jam or there is a need to reverse the cycle of the machine. If the machine stops, plan for a semi-automatic mode which involves a three-position spring-return selector switch for forward and reverse. Be sure to differentiate (one-shot) the direction control in semi-automatic mode, so that you can move forward or reverse only one step at a time. As far as your inquiry on Step-SNXT instructions, they have distinct advantages over SFT instructions in that the Step-SNXT sequence has natural built in interlocking, has the ability to jump over and back to non-continguous steps and the ladder becomes extremely simple at each step. It also lends itself to diagnostics for the same reasons as stated above. As far as "restting the steppers" and "scrapping out all the parts on the line", this is the lazy man's answer to "I don't know what's wrong with the code, therefore I'm going to reset the machine." There are plenty of methods of clearing a jam and maintaining the state of the machine if removing a single station's parts is necessary. Save off the entire status of the words being used to hold the shift register, clear the jam, put a new part back in place, move the status of the shift register word back to the shift register, and continue on. I would like to hear some more suggestions from our other gurus, but I am sure that resetting the shift register is the poorest way of diagnosing the problem.

[flemming007 0]

Hi Again, I'm glad you agree that resetting the shift registers is the lazy way out. My company is owned by three mother companies - Honda, Takao and Kikuchi. Everytime there is a model change, we get one of the two smaller companies to engineer, purchase and install the whole line. Since I am not in the engineering department and do not know everything that is involved, I can't really say why the engineering department is writing its code this way. I'm just hoping that I can write or modify any logic better than it is now. I really like the touch-screens and feel that they are not being used to their fullest potenial either. If you haven't figured it out, my company makes automobile parts ( Honda - Hint - Hint ). And with the logic wrote the way it is, and putting another part in its place isn't possible because of inline process. I just thought that breaking the process down into steps, instead of one big process, and using the touch-screens to help the operators to start the line back up. The technology is there, I just don't think the company is utilizing it to its fullest. Thats were I need to learn to utilize it more. Thanks again Jay, You have given me quite alot of information and ideas. I look forward to more! Flemming007

[Jay Anthony 118]

If the logic is the problem, change it. These parts are expensive.

[Sleepy Wombat 27]

I agree with Jay. However in some instances clearing the production line and resetting the shift register might be related to the process, ie if a error occurs and the line stops, then if some parts are left in heat (the heat want stop immediaetly) or in an acid dip for too long then the quality of the product migth be reduced and in the long rung and can be cheaper to throw away these parts. Dealing with a supplier to the motor industry quality can be a very important factor. Once the manufactured parts go to the respective factories they are generally quality tested there also. if one fails, they don't just send back one but the whole batch...this is of course more expensive not just in replacement parts but the reputation of the company to supply products at a set standard..... Anyways..thats my two bits.... BTW here is a cool Avatar you could use...

[flemming007 0]

Hi Sleepy Wombat, You have hit the nail right on the head. The parts are lower dash panels, and the process adds about ten parts of various sizes and about thirteen threaded studs to the assembly. Once the part is damaged or misses a step in the process, it is usually thrown in the scrap. And everyone accepts this as the norm. The program I attached is for stations 10 thru 20, this line goes up to 100, in increments of 10 with as many as three pieces of equipment at each station. All of this is either spot or projection welded. I just thought that by breaking the process down by using the STEP and NXTS that I could - using the touchscreens - stop, start and remove parts from the line and reduce scrap and increase productivety. As long as they get their production numbers ( 480 in 8hrs ) no one seems to think it is worth their time. So theres not alot of support on my side. I do appreciate any input and thanks for the gif. Flemming007

[Guest Guest_Peter Nachtwey]

I definitely would use some sort of state machine. I like Moore state machines. I don't know what Omron uses or if you must roll your own but for this application a state machine must be used. The key two a state machine saving money is the restart routine. Many application can be restarted even if the PLC power is turned off. However, this requires extra programming and more information. A case in point is using absolute encoders. I am a big fan of absolute encoders because one can tell if the line is in the same state as when the machine was turned off. If so there state machine can be started from the current state and not need to be reset to state 0. The absolute encoder can also tell if the production line has been jogged forward or backward since the line was shut down. In these cases the machine can be jogged back to the last known position before restarting the state machine from the last state. This makes sure operations are not skipped or worse, done twice:( The price of an absolute encoder can easily be paid for by increased production.

[5150 1]

Jay/Wombat, I have had similar situations and am currently working on a prog (motoman loading two press bonders). If I understand the original post, when the machine drops to manual (fault or PB, etc.) the machine puts up a big ? mark as to status. I use a CYCLE STOP latch in ALL of my programs. Operator presses stop, stop bit latched. When this latched bit couples with cycle complete (a.ka. Home position) the machine latches a CYCLE STOPPED latch that drops the machine out of cycle. I, too, couple this with my faults.......latch the fault, but do not drop the machine out of auto until machine returns to Home position. A major fault, CRASH would fault out a Cycle Incomplete watch dog timer and halt the machine as is, but only in this fashion of fault: Major. Perhaps this is what he desires? --5150

[Guest Tom_Stevic]

Hello. Just my two cents on Jays advice to using a reversible shift register. Sometimes it's ok, but most times it is not. I have been using shift registers for 15 years ( wow)in my sequential machines. Most times, if not all, to get out of a jamb, the machine will need to use a separate recover sequence that is seldom the exact same as the sequence that got it into the jamb. So if it works, SFTR is fine, but be careful.

[Jay Anthony 118]

Guest_Tom_Stevic: Please register at MrPLC.com and join the MRPLC.com community Register at MrPLC.com Mr. Stevic...Welcome to MrPLC.com..a lot of water under both our bridges, eh?...Glad to have you join our community. Whether SFTR or Step-SNXT or task oriented programming, the main point is it idiotic to throw away expensive parts due to a lack of ability to recover from a jam condition.