Rockwell versus Siemens Controllers

Hi everyone,

I'm doing some research and was wondering if any users could comment on whether they think Rockwell has an advantage over Siemens with its Logix controller family.

Just wondering if there are any opinions on whether Siemens and it's Simatic product can match Logix in terms of offering an "integrated architecture" as well as general performance.

What advantages/disadvantages do you see for either product.

Thanks for the help!!!!!

Do a search on this forum for advantage disadvantage and that should give you plenty.

Thanks for your reply.

However, I haven't found a forum discussing whether Siemens architecture is as "integrated" as Logix. Meaning Logix allows multiple control disciplines including motion and batch control. Does anyone have experience with this and do they think of Siemens versus Rockwell.

Thanks.

Al_k

What is 'integrated'?

Siemens claim their TIA concept (Totally Integrated Automation) gives you everything from a lamp, button or switch, through sensor/actuator networks like ASiBUS, to device networks like Profibus DP and PA, with Siemens' own drives, motors, instrumentation etc up to the controllers (PLCs, standalone loop controllers, positioning and motion controllers) with HMI from single line text displays to multi-client multi-server redundant systems using WinCC, running on Siemens' own industrial PCs, all coordinated using the software mothership of STEP7, with Ethernet connectivity throughout, on up to full DCS functionality with PCS7, and then on to MES / MIS systems for production management and scheduling control, running on Siemens servers.

So, that's their story and no doubt their engineers can back it all up. Equally I guess Rockwell's sales engineers will explain what a wonderfully 'integrated' thingummy they have and how no-one has ever put together quite as well as they have.

They're both in the business of selling. I don't know enough about both to answer your question of whether Siemens can match Rockwell. Why not ask your local Siemens engineer? "Here's what Rockwell can do for me. Now how would you do it with your kit?" Then comes the hard bit - making your decision.

Regards

Ken.

Gee... since I've been invited to take a leak, mind if I add a little pi$$ing into the pot?

Siemens bought out the TI PLC line including the TI-505 and the PLC system that has come to be known as S-7. Yes, the S-7 line originated at TI.

But I'm not here to talk about the S-7 line. I'm here to talk about the 505 line. But first...

The S-7 does have some capabilities that are not available on the 505 in a "function-form". However, one way or another, those capabilities are available on the 505 (including REAL-TIME Interrupts thus eliminating the need for signal-stretching. Also, in the latest version of TiSoft/SoftShop ladder-coded REAL Math, which is compiled directly into machine language - that portion of the program doesn't have to build a "stack" to examine the rung and then execute the stack - it is not stack-oriented like ladder and all other general programming methods - it is straight machine code... of course, you have to know what you are doing!).

There are at least a few TI-505 (Simatic 505) users on this site. I think that maybe some of them have also worked with various versions of AB PLC's and can relate to what I'm saying.

Over the past few years I have seen that programming an AB PLC is changing and that it is getting better.

Now, what I would like to hear from those TI users out there that have also followed the development of AB Software is...

...do you not also see that AB is slowly but surely moving to the TI-paradigm? (A Paradigm is a "model".)

The TI-Programming-Paradigm (TiSoft) has existed for longer than many of the young bucks have existed.

If you want to make a fair comparison between a Siemens product and an AB product, then compare the latest AB to the Siemens 505 (TI-505 and TiSoft or SoftShop, the Windows version).

You will find that the latest AB and its' software is almost as good as the 505 and its' software.

Keep going AB... you are slowly winning me BACK!

As soon as AB loses the Data-Files-by-Data-Type (that is, open up all memory without "zoning laws"), I'll be sold and ready to be a proponent of AB.

Ooooooop's..... I made a bad... I won't be sold until AB gets rid of the damned "zoning laws" AND removes the tight binding between the specific software versions and the specific CPU revision number! (It appears that this is becoming a more prevalent tactic to get your bucks! Thus leading to more aggrevation.)

From what I understand... I haven't had to upgrade my AB CPU... yet.
That last part of the previous means... if you upgrade your CPU, then you have to upgrage your SOFTWARE PACKAGE! That is, if, for whatever reason, someone has to get a replacement CPU, and that CPU is of a later revision, then you must also purchase the programming software to go with that specific CPU. Even if you are not planning on using any "advanced capabilities" that might exist on the new CPU, you can not even talk to the CPU unless you have the correct software for that specific version.

I'm quite sure, if this is not true, there will be plenty of guys to set the record straight. If I'm wrong... please explain how.

I'm sure Peter will set me straight, at least in terms of some issue from his particular (very unique) perspective...

At any rate, al_k, you would do better to compare the 505 to the AB.

Still waiting to hear from those TI users.

Yeah, yeah, yeah,... so I'm biased... tell me something new.

(52) That's for you, George!

That last part of the previous means... if you upgrade your CPU, then you have to upgrage your SOFTWARE PACKAGE! That is, if, for whatever reason, someone has to get a replacement CPU, and that CPU is of a later revision, then you must also purchase the programming software to go with that specific CPU. Even if you are not planning on using any "advanced capabilities" that might exist on the new CPU, you can not even talk to the CPU unless you have the correct software for that specific version.

Terry,

It is not necessary to upgrade your software to use a newer rev processor in your system. Usually.....

When you receive a new/replacement processor there is essentially no specific rev on the unit. You flash it to the firmware revision you want to use and thus the software version you want to use.

There is an exception of course and I think you would agree that it makes sense, if a new piece of hardware comes out then it is possible that an old rev of software may not recognize the newer hardware component. This might require you to upgrade your software.

I don't think this is as big an issue as it used to be. On my computer I have software installed to support rev 8, 10,11,12, and 13. I launch one package and the software automatically switches from one revision to the next as needed. The install CD includes previous versions and the web site has them as well. Prior to rev 10 it was only possible to have one version installed at a time. So you ended up uninstalling and reinstalling a lot. This hasn't been the case for quite some time now.

Also, hopefully you get a chance to try out the ControlLogix/Logix5000 and witness the end of the data file structure. Grab a demo disk of the software if you get a chance.

OG

WHAT???

Terry Woods said:

Siemens bought out the TI PLC line including the TI-505 and the PLC system that has come to be known as S-7. Yes, the S-7 line originated at TI.

Click to expand...

You are almost right about the S7-200. You are definitely wrong about the S7-300 and S7-400. The S7-200 was developed using a lot of the code from the old TI505, but after it was bought by Siemens.

The S7-300 and S7-400 are definitely German. You can tell by how difficult it is to program. German's don't seem to put a priority on keeping things simple.

Terry Woods said:

Over the past few years I have seen that programming an AB PLC is changing and that it is getting better.

Click to expand...

Agreed. RS5000 is pretty good and much better than RS500.

Originally posted by Terry Woods
Now, what I would like to hear from those TI users out there that have also followed the development of AB Software is...

Click to expand...

Terry, the TI505 is slowly dying. Along with the PLC/5 and the SLC/5.

Originally posted by Terry Woods
...do you not also see that AB is slowly but surely moving to the TI-paradigm? (A Paradigm is a "model".)

Click to expand...

I don't think so and I hope it never will. The Control Logix is far superior to the TI505. We are talking about 15 years difference in technology.

Originally posted by Terry Woods
The TI-Programming-Paradigm (TiSoft) has existed for longer than many of the young bucks have existed.

Click to expand...

Terry is right about that. Until the Control Logix came around the TI505 had one of the most sophisticated back planes. It has store and forward capability long before Rockwell had it. It also has pseudo DMA that allows special function cards to read and write to PLC memory. That is a big advantage over almst all PLCs that exist today.

Originally posted by Terry Woods
If you want to make a fair comparison between a Siemens product and an AB product, then compare the latest AB to the Siemens 505 (TI-505 and TiSoft or SoftShop, the Windows version).

Click to expand...

As neat as the TI505 is, I would still go with the Control Logix because of the Ethernet/IP and the consistant way Rockwell PLCs can do communications.

Originally posted by Terry Woods
You will find that the latest AB and its' software is almost as good as the 505 and its' software.

Click to expand...

I have the RS5000 V13 and I we have the next version too. The latest TISoft I have is 6.3. I like RS5000 for a couple of reasons.

1. Trends.
2. Simpler instruction set. Especialy when it comes to indexed and indirect address. I don't like Rockwells paranoia about overflows halting the process though.

Originally posted by Terry Woods
Keep going AB... you are slowly winning me BACK!

As soon as AB loses the Data-Files-by-Data-Type (that is, open up all memory without "zoning laws"), I'll be sold and ready to be a proponent of AB.

Click to expand...

Try RS5000 with the Tags, UDTs and array indexing.

Originally posted by Terry Woods
Ooooooop's..... I made a bad... I won't be sold until AB gets rid of the damned "zoning laws" AND removes the tight binding between the specific software versions and the specific CPU revision number! (It appears that this is becoming a more prevalent tactic to get your bucks! Thus leading to more aggrevation.)

Click to expand...

It was a problem. I haven't seen it lately. Terry, you know I know.

Originally posted by Terry Woods
I'm sure Peter will set me straight, at least in terms of some issue from his particular (very unique) perspective...

Click to expand...

Some day I will explain the very unique perspective. There are a few of you that know already. I could write a small novel.

Originally posted by Terry Woods
At any rate, al_k, you would do better to compare the 505 to the AB.

Click to expand...

The TI505 has seen its day. The Control Logix is the new king.
I agree that the S7 was not a step forward. From what I have heard from the Siemens guys it seems that the S5 is superior to the S7 when it comes to the back plane. The S7 has an advantage over the S5 when it comes to processor speed and better programming software.

Originally posted by Terry Woods
Still waiting to hear from those TI users.

Click to expand...

Ok, you have heard from someone(me) that has known the TI for 16 years. I have a TI525, a TI535, a couple TI545s, a TI555 and a TI565. Is that enough TI. I also have a S7-315 DP and S7-316DP. The S7's big advantage is STL. The learning curve is steep but once one knows the quirks, it is much easier to write re-useable code and do sophisicated math that on a Rockwell computer. What can't be forgiven is the unituitive programming software and unnecessay quirks of the S7. The back plane is also slow compared to the Control Logix.

For doing PLC things I would go for a Control Logix. For doing communications I would go with a Control Logix. For doing complex algolrithms and reusable code I would go with a S7. The S7 seems better for those that can invest a lot of time in software. OEMs fit in this category. Rockwell computers are better for getting simple systems done quickly. System integrators like that.

Terry, what do you have to say about that?

I will chip in as I use both AB and S7. And as I am neither US or German, maybe I am can be completely neutral.
Notice: I use SLC/PLC not ControlLogix, but I follow it very closely from the sidelines in order to evaluate when/if to make the transition. We have CLX in-house and I have the chance to try it out.

AB CLX is possibly more ambitious than S7. But CLX is also much younger (CLX started approx 99, S7 started approx 92).
There have been a LOT of issues with the CLX, but possibly they are slowly petering out. There are new versions of the CompactLogix that will help AB to cover the lower range, but I hear that the issues had a comeback when they were shipped.
When S7+STEP7 came out there were also a lot of issues, but today S7 is rocksolid.
I use Siemens Windows based HMI's and they have been faultfree from day one. AB's new Windows based HMI's have had a very troubled start so far.

Advantages S7:
Hardware spans the range from bottom to top very well.
Hardware is markedly cheaper (15-35%).
All CPUs have an acceptable online port (MPI).
Fieldbus is the most universally accepted (Profibus).
Has the largest market share worldwide.
You are not tied in to a support aggreement in order to get bug-fixes.

Advantages CLX:
Networking strategy.
More user friendly programming software.
AB Ethernet/IP is more "finished" than Siemens ProfiNet.

MAIN GRIEF WITH SIEMENS: The documentation is extensive but does not convey the information very well. This causes the learning curve to be vey steep (Germans simply do not know how to write manuals). I think that its the main reason why many people dislike S7.

MAIN GRIEF WITH AB: AB is really in a transitional stage from old to new. It means that there is an immense amount of hardware and software that you must master and support at the same time.

Lastly: S7 carries 5% of its genes from TI and 95% from S5.

I've used TI and S7 and I agree with Peter : the S7-300 and -400 are most definitely products of Germany. The S7-200 was developed by the guys at the old TI HQ in Johnson City, Tennessee and although they presumably had a completely different design brief to the 545/555, nevertheless you could see many common concepts. The S7-200 is really an excellent little PLC with easy programming: it probably doesn't get the credit it deserves because it's branded as "Siemens S7" and to many people that means "difficult". In truth it has next to nothing in common with the S7-300 and S7-400. You don't use the same software and you don't use the same functions or structure.

With respect to Terry and his praise for the TI505, would he really want to continue these days writing 'toilet-roll' code with no structure? This is really the fundamental shortcoming of any PLC of that vintage nowadays. TI were great in maintaining continuity between the 500-series and the 505-series, but it does mean that today's 2MB 555 is constrained by memory arrangements created when a 520 with 4K of memory was a big deal. If you want to go all starry-eyed about TI, I would give my vote to the APT software. That was a real advance in process control technology for PLCs and even now, 15 years or so after it was launched, I've never seen anything so radical or effective. It was the work of a genius.

S7 nowadays is a bit of a mixed bag. I've no idea why Siemens chose to maintain all these residual S5 features in S7. S5 Timers - yeeuugh! And silly little things like STL - instead of making it IEC61131-3 compliant they chose to stick with 'A' and 'O' as in S5 instead of following the IEC mandate of 'AND' and 'OR'. On the other hand there are a bunch of features in S7 which are really well done. I actually like the formality of the program structures, especially the use of all the system controlled OBs for handling the stuff outside the main scan cycle - cyclic interrupts, time-of-day interrupts, hardware interrupts, error interrupts etc. The transportability of code from one CPU to another, regardless of whether you stick with S7 PLCs or use Siemens WinAC PC-based controller is excellent. And as you might expect, the integration of Profibus in the software and hardware is second-to-none.

Regards

Ken

I have worked with Siemens , TI and Allen-
Bradley products. 99% of the time with
no formal training and just learning from the manuals. Allen-Bradley or TI would
be my choice. Like most things manufactured
or developed in Germany S7300/400 is overblown , overhyped, make the most simplist task difficult , user-ugly , and the manuals suck.....

Thanks, you guys are awesome!!!! Great feedback...

Please keep it coming if there are additional comments.

From Ken M:

And as you might expect, the integration of Profibus in the software and hardware is second-to-none.

Click to expand...

Blech. We use S7-300/400's and AB/Rockwell here, with a smatterring of other brands (GE, Omron, AD, etc).

I cringe when I get handed an S7 job. Especially if it has drives on Profibus. Profi I/O is fine, but that's about it. Having to manually use send and receive blocks for cyclic comms is just plain stupid.

AB <SLC, PLC, CLX> + SST Profibus Scanner is a much nicer combo as far as I'm concerned.

Another thing is that CLX will run absolute rings around S7 (even the 'New' processors) speed wise. CLX may be somewhat more expensive, but having highly intelligent I/O modules and ultra-fast CPU's is a huge plus. Now, if could only have Siemens Style DB's in CLX, I'd be a happy camper

Siemens does do a great job of being able to "Totally Integrate" PLC's, Drives, HMI's, and other Field Devices into Step-7 Manager, but it is really such a cumbersome interface I hate working with it. And archiving a project and seeing 1500+ files related to even a simple project is scary.

We've done systems with both the S7-300 and the A-B SLC platforms. We haven't used the ControlLogix yet, since our applictions don't need the horsepower.

Of the two I definitely prefer A-B. The Siemens hardware is good, but the software is awkward. There are a lot of restrictions on use, a lot of non-intuitive aspects to the interface, and the documentation is terrible, horrible, and very bad. Tech support from Siemens is very difficult to access, but in fariness when you do finally find the right guy they are knowledgeable and helpful.

Siemens is an interesting contrast in philosophy. Their PLC lines and programming software are clearly optimized for discrete control, and analog/process control are not well implemented. However, they are also clearly based on using function block and statement list programming, with "pure" ladder almost impossible to implement. That's fine if you started out in that background, but if you cut your teeth on ladder the learning curve is steep indeed, and the difficulty is compounded by the crappy documentation.

A-B clearly shows that ladder logic as the preferred programming structure, but they are quite good at handling anaolg data and process control. The data structure is different than many PLCs, but it is still simpler and more flexible than the Siemens. Siemens is also a memory hog - to execute the same kind of programming we needed a CPU with two or three times the memory of an A-B PLC.

A-B has a lot of quirks in their communications, particularly in the hardware. Nonetheless, I find it easier to set up and use than Siemens. One thing that blows my mind is that with Siemens, even on Ethernet, you need message blocks in both the sending AND the receiving PLCs to execute data exchange!

Tom Jenkins said:

Siemens is an interesting contrast in philosophy. Their PLC lines and programming software are clearly optimized for discrete control, and analog/process control are not well implemented. However, they are also clearly based on using function block and statement list programming, with "pure" ladder almost impossible to implement. That's fine if you started out in that background, but if you cut your teeth on ladder the learning curve is steep indeed, and the difficulty is compounded by the crappy documentation.

Click to expand...

After 8 years of GE & Mitsubishi I'm 2 years into working almost exclusively Siemens and I'm still suffering from this exact problem. As seen in some of the threads I've started I'm trying to get a handle on STL because Ladder just isn't properly supported.

One thing that blows my mind is that with Siemens, even on Ethernet, you need message blocks in both the sending AND the receiving PLCs to execute data exchange!

Click to expand...

I have done one job using the CP-343-1 and found the Native TPC to be one of the few shining stars for Siemens. We setup a project where I interfaces with a custom C++ interface that controlled several separate parts of a complex project. I sent 40 BYTES of status info every 20ms to the C++ app and they sent 4 BYTES of control info as needed. I used Native UDP with a send function to send then the data and they used the Passive write function that doesn't even require any code, just a configuration setup. The best part is on their end Windows COM objects and Sockets connections were all that were needed. No other drivers!

I would have to disagree with the previous posts that say RSL5000 doesn't handle resuable code as well as Step7. With UDT's, automatic documentation and SIMPLE cut/copy/paste. I think AB is just as good if not better at reusable code. AND I didn't have to take a two week course to figure out how, like I did with Step7. I wouldn't say RSL5000 is perfect by any means but Step7 sure makes it look even better than it is.