Adding to DW in S7

All good points.

And yes after you go through it the first couple of times, it is not much of an issue. AFTER............not during. During the first couple of times it is not pleasant.

In the textile world it has to be simple, in any manufacturing in the south east USA, it has to be simple.

Steel foundries, textiles, chemical plants.

The chemical plants were the most complex, LOTS of data movin around with the actual hardware manipulation.

Of course I am talking about the editor of the PLC,

If this were C++ or ANSI C or VB I would expect the type to be an issue.

Why would it be so unthinkable for the editor to take 10 as 10 dec? It is the default for humans.

How many times have you asked for 10 bolts at the hardware store and the guy gives you 16 or 8 because you left the base 10 out?

I would like 10.

Hex, decimal, or octal 10?

???

The default should be something intuitive for humans. 10+10.6 is intuitive for humans.

I understand the machine level difficulty. My background is in electronics. I get the arguement on a machine level.

The editor is for humans to translate from human into a compiled machine language.

As such I want a car that I do not ned training to drive and I think for 4000 USD an editor should be able to take 10+3.6+1200 and not choke because one is a DWORD real and one is a WORD INT and one is a DWORD INT.

I would favour a type system that uses "numbers" and "not numbers".

Find the decimal, assume base 10 unless told otherwise and perform the math, if the destination is an INT trunc and round up or down. (user choice)

16# is good..........L# is good..........10.0 ok thats a real...........10 WHOA now hang on a minute we have no idea what that is!!!!

You could use L10 nut the documentation says it is not needed to use any prefix.

"L L#2300"
"L 10"
"L 10.0"

Now when I assign symbols, I include the DT in the name.......OIL_LVL_SCALED_INT MW23 INT not word, got it.


Ok this is the second time I read "house keeping".

I seriously do not understand what you mean by that. No really.

I am no where near the level you guys are on, give an example of house keeping that is required if you turn off data type checking.

And now explain what is the benefit of turning off data type checking if it allows you to make more mistakes more easily.

UNless it removes the data type requirement I don't see the advantage.

And yes I like Siemens Step7 even with the pain.

Once the pain goes away , it should get even more fun.

Other PLC systems have had subroutines for a long time......FBs are really just subroutines and to some extint so are FCs.

I have yet to see any industrial software that just works.........some better than others, it generates a huge market for training though.

I can't believe there are only 2 books on Siemens. And one of them is just an expanded help file and the other is a collection of more words clouding the help file without actually improving the explaination.

The real pain is knowing I may never use any of this ever again.
 
I'm happy to agree with some of the criticisms dhanuguy offers, and yes, there are loads of things I wish you could do in STEP7 that you can't. But my perception (and it is just that, a perception) is that some of the issues he has still boil down to lack of familiarity and "it doesn't work like I expected".

I went through a similar problem many years ago - I decided to start trying to catalogue my CD collection. So I downloaded just about every software there was on this subject and tried them all, extensively. Some were easy. Too easy. Maybe they didn't have the reporting or filtering features I wanted. Some were OK. One I remember, a package called CATraxx, was monstrously impressive in what it could do but was way beyond my requirements and needed a lot of learning. So I settled on another one called Music Library and for 18-24 months used it happily, learning more about what I wanted and how I wanted to use it. After that period, all of a sudden new versions dried up, the support forum run by the author stalled, no support e-mails were answered (but he still responded to a couple of sales requests from my wife at her e-mail address!). I eventually went back on the market and this time I chose the one I'd previously discarded, CATraxx. My requirements had grown and I now appreciated this for what it is. I've been using it for over 5 years now and I'm a real evangelist for it. There is nothing you can't do with it. It is so customisable. I've even started learning the rudiments of XSL, XML, and SQL, not because I have to, but because this package permits me to get under the bonnet (hood?) and make it really mine.

I suspect my original view of CATraxx was similar to dahnuguy's current view of STEP7. I didn't know why I would ever want to use some of its features. And if I did, why would I expect to find them there, rather than here? But experience and a growing demand by me on the software has made it all clear now. I wouldn't move to anything else now.

Dahnuguy, if you feel PLCs are there to replace relays I suspect (without being judgemental) that your application must be relatively simple. Yes, STEP7 will do this, but it is also capable of oh-so-much-more. As are PLCs. My earlier remark about strongly-typed computer languages was deliberate. As far as I'm concerned a PLC is a computer. A highly evolved, very specialised computer agreed, but a computer nontheless. I agree with you that the inconsistencies with which STEP7 handles data-typing can be annoying, but let's be honest, they're not exactly show-stoppers. Overall STEP7 and S7 have served a large number of people very well indeed over a long period of time. They can't all be wrong.:)
 
My point is being narrowly missed.

Relay Ladder Logic issssssss based on replacing relays.

I suspected you would go with a similar analogy as you did with the CD software.

This misses my point entirely.

It is NOT the complexity of the task at hand, it is not my understanding of controls or logic or general practices that is the issue.

I will NEVER come to think as S7 as a good solid well documented easy to use software , even after it becomes easy for me to use.

Here is my point and this is my only point with S7.

If it is so tell me it is so and give a good solid explanation with an example.

If it is not so , tell me it is not so.

The documentation should be written by taking a new user and sitting down beside them and given NO instruction, have them just try it and observe their actions and missteps and write the instructions to address these issues.

This product is documented by the people who developed it as if they are reminding themselves of how it works.

It does not matter to me the little quirks and oddities ifffffffff they are discribed and laid out before hand.

I do like the process of discovery and I do not think a training class should be needed for an experienced user of other programs to use a similar program.

I never took a class on Windows, DOS, EXCEL, WORD, various HMIs etc. or a long list of other softwares.

I will write a small whats what for new users in a week or two, not as an arguement or to air out a bunch of greif , but I think it will bring some light to the issues new users face.

You experienced S7 users will probably not see the value in it. Which is another point of mine, it is difficult to see from the other side.

I used to teach a generic PLC class with motor controls.

I just made up the introduction as a preface to the official material and you could see the light bulbs comin on all over the class. It was all simple stuff but laid out for the new comer to grasp easily.
 
dahnuguy said:
All good points.

And yes after you go through it the first couple of times, it is not much of an issue. AFTER............not during. During the first couple of times it is not pleasant.

Well geez, what do you expect. Every new software I ever learned is full of gotcha's. I just wrote a GUI in wxPython, and in the beginning it made no sense at all, but I didn't blame it on the editor. But now I "get it", and can crank out the code. Whether I'm doing S7, S5, C++, SQL, Java, or whatever, it's the same story.
In the textile world it has to be simple, in any manufacturing in the south east USA, it has to be simple.

Really? Why in the southeast? Are you saying they are stupid or something? Hundreds of companies have relocated to the southeast over the years (BMW in SC, Daimler-Benz in AL, among many others). Their processes are not simple by any means, and they do just fine. I think you are making some bad assumptions.

Why would it be so unthinkable for the editor to take 10 as 10 dec? It is the default for humans.

Well, not if I happen to be doing masking. In that case, I would want to think in terms of hex. It would be obvious to the editor if I entered a FFF0 as a mask, but not 1010.

How many times have you asked for 10 bolts at the hardware store and the guy gives you 16 or 8 because you left the base 10 out?

Never. But that is a ridiculous comparison.

As such I want a car that I do not ned training to drive and I think for 4000 USD an editor should be able to take 10+3.6+1200 and not choke because one is a DWORD real and one is a WORD INT and one is a DWORD INT.

You can already combine INT, DINT, DW, and DD if you want. I don't know any editor that allows you to mix ints and floats though. Do you? It just exposes problems all over the place.

And now explain what is the benefit of turning off data type checking if it allows you to make more mistakes more easily.

Your code will be much more streamlined because you don't have to do trivial type conversions when they aren't necessary. STL take it even further to the next level. For instance, you can try this one yourself: Convert an analog input, scale it, multiply it by 34.67, and display it as an INT. Do it with type checking on, type checking off, and then in STL.
 
Really? Why in the southeast? (cuz thats where I am)Are you saying they are stupid or something? (nope)Hundreds of companies have relocated to the southeast over the years (BMW in SC, Daimler-Benz in AL, among many others and none of them started here). Their processes are not simple by any means, and they do just fine. (and I bet they brought their team of engineers with them to support said process)I think you are making some bad assumptions. (nope)

I am not making assumptions,bad or otherwise, except for the 6 years I spent in the USAF, I have lived in NC or SC. No assumptions, I am making a statement.

I installed textile machines for a couple years SLC500s, GE90-30 or is it 30-90?, TI545, old AB2 and AB5, toshiba T2, MItsubishi, Modicon.......and Continental tire used Beckoff the hands down worst and by far most complex control system ever , makes S7 look like playtime. Not in a good way either.

And at Conti, we had one guy in engineering on site at all times for nothing but maintenance troubleshooting support.........I was one of 3 , one for each shift. Like the fire department, 40 acres of machines under roof. over 150 maint persons. And one troubleshooter, because the logic was NOT simple.

100s of textile plants all over the SE USA and 5 to 10 maint guys in each plant and no trouble most of the time finding their own problems. When they did hit a snag, they called me.

If the code is written well for MOST processes a moderately trained person can get through it.

One programer I worked with wrote his ladder so complex it would stop working if changed or edited in any way. So many nested subroutines and so many latched coils and multiple coil assignments, it was a nightmare to debug his stuff. And it always needed help. If you loose power during a certain step it would hang , if the operator pushed several buttons at once it would hang, if 2 operators hit the load button from 2 differnet stations it would hang.

I never even tried to fix his code. His job security attitude landed him in many a dye house on Saturday at over 120 deg F and 100% humidity while I just waited .

He would always try to explain it to me like I was the one who didn't get it. Sorta like some of the people here.

I GET it , I don't AGREE with it.

I have always thought about what is the simpliest way to do XYZ.

Unless the process mandates complication, it is not needed.

High speed packaging machines at Phillip Morris for example, were very difficult to troubleshoot. Lots of data .....fast data......and zero allowable downtime.

Chemical plants have some complication but are a collection of fill and drain and pump on and off and analog levels and pressures and flows until you mix in the HMI and data logging it's not too bad except for the safety overkill which can't be neglected.

Never worked in an auto plant, I imagine the robotics and complex functions make them quite different.

Hydraulic presses are a place where complexity is required.

1200 tons regulated precisely at 45 strokes a minute can be complicated. As I recall scan time was an issue with that one. I did not write that one.

So yes some processes are complex, most are not. All can be MADE complex.

It is not because I am simple that I want to make things simple. I feel they should be simple. If it is tough for me to follow, it will be very difficult for the next guy down the skill ladder.

Why is it so difficult to accept that S7 is not well documented or as easy to learn as all other PLC programs?

It's just not.

It's not a big deal.

Some people actually like using complex systems to maintain a high ego boost. Never got that. I don't need a Casio programable graphing calculator to do Algebra or Trig. No one does. But there is that small group that will. Just because.

I imagine allot more people try to learn S7 than actualy do learn S7.

I can teach my 14 year old to use other PLC programs.

Maybe even my 9 year old.........they are rather bright though.

I think I will , maybe this weekend even...............family fun with PLCs.

And I bet people who defend S7 learned it first and "grew up" on Siemens.

Again, I like S7, it's fun, I like puzzles. If this project was a regular non emergency I would be havin a blast with S7.

WHy not just use C? ohyeah no 3000USD per software license. And it's not proprietary.

Should be one PLC prog software with plug ins for each manufacturer of PLC hardware. Now THAT would be fun.

you could do most PLC progs in BASIC or PASCAL. Oh yeah those aren't easy to troubleshoot with.

If any of you guys ever write a PLC prog in STL for a textile company your in for a surprise.

OK back to the "project".
 
OOOOOOOOO a puzzle ..........thanks, I'll post the 3 efforts and summary.

Your code will be much more streamlined because you don't have to do trivial type conversions when they aren't necessary. STL take it even further to the next level. For instance, you can try this one yourself: Convert an analog input, scale it, multiply it by 34.67, and display it as an INT. Do it with type checking on, type checking off, and then in STL.

This I will do ....................sounds like an educational exercise, thanks.

I don't always want the answer but a shove in the right direction is nice.
 
dahnuguy said:
Really? Why in the southeast? (cuz thats where I am)Are you saying they are stupid or something? (nope)Hundreds of companies have relocated to the southeast over the years (BMW in SC, Daimler-Benz in AL, among many others and none of them started here). Their processes are not simple by any means, and they do just fine. (and I bet they brought their team of engineers with them to support said process)I think you are making some bad assumptions. (nope)

No big deal. When you specified the southeast, I assumed you meant as opposed to the northeast, midwest, etc. I've worked with a lot of southeast firms, and know a lot of people in the southeast, and I couldn't figure out why you said it had to be simple for them.

He would always try to explain it to me like I was the one who didn't get it. Sorta like some of the people here.

I think the people here have been very helpful and patient. We know it isn't easy, and if I was thrown at an AB project, I would be asking the same questions. But based on your questions, you need to step back and understand a few things more thoroughly, such as in the last example you posted where you overlooked the initialization of the temp variable, and the difference between FBs and FCs.

Why is it so difficult to accept that S7 is not well documented or as easy to learn as all other PLC programs?

I woud have to disagree that it isn't well documented. The help files are very detailed, but from questions on this board (not from you), I can tell that people don't read them. And, you can download PDFs for free for just about every piece of software and hardware they have ever made.

I can't speak for whether it's as easy to learn as other programs. But I still would like to know if other PLCs allow something like you described above (comparing an int to a float and let it automatically detect the data type). That intrigues me.

Some people actually like using complex systems to maintain a high ego boost. Never got that. I don't need a Casio programable graphing calculator to do Algebra or Trig. No one does. But there is that small group that will. Just because.

I suppose. I don't know anyone like that, but I guess they are out there. No one that has been helping you out has even brought something like that up though.

If any of you guys ever write a PLC prog in STL for a textile company your in for a surprise.

I've worked on textile machines, and STL was pretty common on them, especially when it came to the motion control.
 
dahnuguy said:
Your code will be much more streamlined because you don't have to do trivial type conversions when they aren't necessary. STL take it even further to the next level. For instance, you can try this one yourself: Convert an analog input, scale it, multiply it by 34.67, and display it as an INT. Do it with type checking on, type checking off, and then in STL.

This I will do ....................sounds like an educational exercise, thanks.

I don't always want the answer but a shove in the right direction is nice.

Yeah, I don't have much time to crank out the finished code, but I usually tell people to give this a try to understand the major differences. The STL code will probably have about 10% of the instructions that the first method uses. A good compromise would be to use ladder for the boolean logic, and STL for the math. But any way will work fine.
 
I think the people here have been very helpful and patient. We know it isn't easy, and if I was thrown at an AB project, I would be asking the same questions. But based on your questions, you need to step back and understand a few things more thoroughly, such as in the last example you posted where you overlooked the initialization of the temp variable, and the difference between FBs and FCs.

Yes everyone here is most patient. Very nice place. And you would be surprised if you were thrown into an AB project at how few questions you would have. The communications is tricky esp one or two details with the drivers in RSLINX but in the ladder editor few issues if any.

And I am workin 12 and 16 hours a day for 3 weeks, I miss some dumb stuff........the next day I kicked myself on that temp thing.

I got the FB FC thing, it's a DB issue......mostly the same but not. It helps that I read DB as data base not data block..........looks just like SQL when you open one.



I woud have to disagree that it isn't well documented. The help files are very detailed, but from questions on this board (not from you), I can tell that people don't read them. And, you can download PDFs for free for just about every piece of software and hardware they have ever made.

I never said undocumented.........it is not well documented. I have a stack of PDFs and 3 books.......I read the help and read the book and then ask the question after lookin at it and tryin a couple of times.

I can't speak for whether it's as easy to learn as other programs. But I still would like to know if other PLCs allow something like you described above (comparing an int to a float and let it automatically detect the data type). That intrigues me.

I may have misspoke on that count but I seem to remember not having ANY issues with data typing until S7.



I suppose. I don't know anyone like that, but I guess they are out there. No one that has been helping you out has even brought something like that up though. I have seen some in other posts, not from the main core here though.



I've worked on textile machines, and STL was pretty common on them, especially when it came to the motion control.[/QUOTE]


I never saw motion on any textile machines..........dye house and yarn, weaving, drying, warping, carding, transport, spinning ummm prolly more, but I am sure I didn't see it all.
 
I've gone the other way from the heaven of S7 to AB (n)


Can't go on-line with a PLC where the code is different, can't compare between off and on-line to see what the differences are, having to upload to be able to get on-line with a client who's a 5 hour flight away, then taking 1.5 hours to upload via the internet...

One of the many things I don't like about AB...

Horses for courses, AB isn't bad at the end of the day, I prefer S7 though.
 
dahnuguy said:
I never saw motion on any textile machines..........dye house and yarn, weaving, drying, warping, carding, transport, spinning ummm prolly more, but I am sure I didn't see it all.

That blows my mind. Textiles, due to the intricate weaving patterns, lends itsellf perfectly to motion control. The ones I worked on were replacements of cam-driven machines that were used between the 40's through '95. The new ones had servo controls and vision feedback to keep the patterns perfectly consistent.
 
PeterW said:
I've gone the other way from the heaven of S7 to AB (n)


Can't go on-line with a PLC where the code is different, can't compare between off and on-line to see what the differences are, having to upload to be able to get on-line with a client who's a 5 hour flight away, then taking 1.5 hours to upload via the internet...

One of the many things I don't like about AB...

Horses for courses, AB isn't bad at the end of the day, I prefer S7 though.

HEHE..........this is fun...........

Why would you WANT to go online with a code mis match???????

The other PLCs wont let you do it either, have to upload and merge symbol file or download.........option A is best so what your looking at is actualy what is happening. I think one of them would allow it though don't member which one maybe a AB5? but it was screwy and not good. Might have even put the CPU in STOP I dont recall.

You CAN compare to see the diffs, I did it many times........but its easier to just upload and merge and get on with it.

They could send you a copy through email or post it on a file share server with DSL or highspeed. Of course that would negate their shiney new MODEM though and would "GASP" put their "secret" code on the evil web.

I have heard the arguement recently.

1.5 hours while you surf the web or read up on the last version is still better than 10 hours flying plus expenses and time in the plant.

I never defend any of them, never have, and I am no expert, but I might be able to help with some AB issues. I would have to go home and dust off the software though. Dont have it all in my head anymore.
 
S7Guy said:
That blows my mind. Textiles, due to the intricate weaving patterns, lends itsellf perfectly to motion control. The ones I worked on were replacements of cam-driven machines that were used between the 40's through '95. The new ones had servo controls and vision feedback to keep the patterns perfectly consistent.


I left textiles in 93 or 94.............never saw anything high tech that worked.

The weaving was all mechanical never did much with weaving.

Dye machines are just little chemical plants or huge washing machines, old ones were controled by a big cam and pnuematic valves..........no electrons.

I upgraded several of those to PLC. But think about it, if you can do it with a cam and a switch,,,,,,,,,,,how complex can it be?

It was all pretty easy stuff if you let it be.

Of course the fresh out of college guys haaaaaaad to add every little thing and test it out on a production machine 400 miles from home ..........on Friday afternoon............on a long weekend............when you had plans.........and they were alll gone..........and when they come back and read the report they would all act amazed that it didn't work!

One guy I offered to take and show and explain the machine.........as he had NEVER actually seen one!!!!!!!!! He said all he needed was a text discription someone gave him in an email.

So I let him debug all his in person.

2 weeks in Alexander City Alabama will change things. 400 miles away from the office near CHarlotte.

I have seen the whole process from cotton to shirt...........it aint complicated. But I suppose it has changed since 93 94, if there is any left. I worked for 4 or 5 companies that went under, was like leap frog.

Then contract work.............Seen allot of stuff...........not sure it's a good thing.

And the weaving thing is just open , send shuttle, close, open send shuttle back , close...........repeat. Think about it people did it for a long time by hand.

Circular knitt is just a series of needles that go up and down and make a tube of cloth. All one color. usualy white Greige ( grey) goods then dyed and dried and cut and .............

The patterns I saw were screen printed and it looks quite real.........most furniture is actualy , or was, screen print and all curtains, some sheets and bedspreads.

COmplicated is making tires...............oh boy........and cigarettes.

I could have stayed at Continental for 10 years and not learned it all and of course they keep adding new stuff.

Textiles never did anything until they HAD to.

I saw ONE spinning machine that had a custom gee whiz PC based motion control to make odd textured yarn for carpet. ONE and it was way custom, the owners son in law did it "on the side" and would never tell anybody "the secret"

LOL anybody could just look at it and tell what it was, but I acted all impressed. Place is in Pageland SC.

OK I'm all done today............
 
dahnuguy said:
HEHE..........this is fun...........

Why would you WANT to go online with a code mis match???????

To fault find of course! With Siemens you can go on-line and see what is in the PLC, which is good.

You have to update the block off-line before making changes though, but at least you can see what is running without having to do an upload.

dahnuguy said:
The other PLCs wont let you do it either, have to upload and merge symbol file or download.........

Shame on them then


dahnuguy said:
option A is best so what your looking at is actualy what is happening. I think one of them would allow it though don't member which one maybe a AB5? but it was screwy and not good. Might have even put the CPU in STOP I dont recall.

You CAN compare to see the diffs, I did it many times........but its easier to just upload and merge and get on with it.

Not with Controllogix you cannot.. you have to upload what is in the PLC and do the compare off-line, where of course you have to wait for it to convert both files to text and then do a compare. Siemens S7 not only will compare off-line to on-line, it also allows you to compare small sections of the code.

You can also upload a block at a time, so of your change is only in one or two blocks you do not need to spend ages uploading the whole program.


dahnuguy said:
They could send you a copy through email or post it on a file share server with DSL or highspeed. Of course that would negate their shiney new MODEM though and would "GASP" put their "secret" code on the evil web.

They can only do such things if they have people and equipment on-site to do this.

dahnuguy said:
I have heard the arguement recently.

1.5 hours while you surf the web or read up on the last version is still better than 10 hours flying plus expenses and time in the plant.

With Siemens in this time I could have compared the blocks I wanted to work on and uploaded if different, solved the problem, and downloaded the changes. Job done

dahnuguy said:
I never defend any of them, never have, and I am no expert, but I might be able to help with some AB issues. I would have to go home and dust off the software though. Dont have it all in my head anymore.
 
Last edited:

Similar Topics

Hello all! So I have one project with a S7-1214 and a MTP1500 "Project1" and one project with another S7-1214 "Project2". Both of these PLC:s need...
Replies
6
Views
102
Hi Everyone, Currently we have three plants running with Controllogix PLCs (L72, L73, L74). In each of these plants we have 2 FTView SE...
Replies
0
Views
80
Hello, Im building project with 1756-L82ES controller and 1756-IB16S card but i cant find it when trying to add the card to IO configuration...
Replies
3
Views
155
Hello, I have a pair of redundant 1756-L71 controllers rev 24. These controllers currently have produced and consumed tag interfaces to 3 other...
Replies
2
Views
179
I'm adding an IAI Gateway with 2 axes connected to it. To an ethernet network on my PLC. So, I think the math is correct on the Input and Output...
Replies
0
Views
166
Back
Top Bottom