Bill PLC

MrPLC Member
  • Content count

    1
  • Joined

  • Last visited

Community Reputation

3 Neutral

About Bill PLC

  • Rank
    Hi, I am New!

Profile Information

  • Country United States
  1. B & R Automation - What's the big secret?

    My company was looking for a platform solution that could handle a wide range of tasks (and size/complexity of machines), with a low hardware cost (Allen Bradley and Siemens were immediately disqualified), powerful, ease of use, etc. All of the typical requirements. We had positive experience with Beckhoff in the past (although in my personal opinion, the Beckhoff environment provides almost too much information in the hardware configuration side - so much so that it can be daunting to new users - the two different applications make the environment seem disjointed), but we did experience some issues with Beckhoff product availability (lead times) and especially local support in the U.S. in the Massachusetts area. The B&R platform appeared to offer the same benefits of Beckhoff with a bit better support in the US. We had struggled with Mitsubishi for about 2 years and, although the hardware is impressive, the software environment is disjointed and overly complex (up to 4 applications at once must be open to program a job), and they were very slow to make necessary improvements (even the US-based Mitsubishi team agreed that Japan was not that responsive and that in the Japanese culture, software ease of use has a much lower priority than in the US). The platform is decidedly old-school - you have to do a lot of manual accounting of memory addresses - something I look at as 1980s technology. As a result, we dumped the Mitsubishi platform. We've been learning and starting to use the B&R platform for the last 6 or 7 months. We attended training classes in Atlanta, which were pretty well done. They have a lot of formal training documentation which is quite nice. We've completed a few smaller automation jobs and encountered no negative issues, and we're now in the process of designing some full-scale machinery. I'll certainly comment on how these projects turn out in a few months. My opinions so far: 1) Automation Studio is a single application (not CoDeSys based - some would say a disadvantage, but B&R advertises this as an advantage because they control all of the code) and the only thing you need to code and debug your application. The only exception is the safety PLC editor, which was written by a third party, but it's fairly well integrated into the environment (still a separate application). Since we've started using the latest version of Automation Studio about 6 or 7 months ago, they've come out with 5 service packs, so for sure the support and attention to issues is really active. Still, the environment has some basic issues with window handling and simple things like occasional problems with cut & paste and find that, in my opinion, have existed for too long. The environment is very Visual Studio-like, so if you're familiar with that environment, you'll like it (although it's not as polished). Structured text is clearly their favorite language. The SFC editor and functionality is decent, but has some annoyances. The ladder editor pretty much sucks, so if you prefer ladder, I'd highly recommend that you check this out. Automation Studio version 4 (major version upgrade) is due out sometime soon and may address some of this. No way of knowing until it launches. 2) Hardware configuration is certainly a bit more manual than the Beckhoff system. Beckhoff software can automatically search the entire EtherCAT network and pull in all of the devices. The B&R system requires you to manually insert each of the POWERLINK (their real time Ethernet network) nodes, and only then can it automatically find all of the I/O modules. Apparently this was one of the major focuses of Automation Studio 4 which will supposedly do a lot to address the configuration issue (I've seen early previews and it does look a lot nicer). Still, hardware config is a one-time thing, and certainly not the bulk of the work in any project. It's certainly not that inconvenient, once you know the limitations and steps required to build the hardware tree. 3) The debugging environment is very responsive and quick. Debugging structured text in Beckhoff is superior in that it shows a split screen when viewing code, with all of the current variable values displayed right next to them. In Automation Studio, you have to hover over the variable or (more typically) open a watch window and manually add the variables you care to look at (the selection of variables is saved from one debug session to the next). Each has it's pluses and minuses - I wish Automation Studio did both. 4) The environment in general is MDI or tabbed based. I typically use the tabs. My complaint with this environment (and most others are similar in some way), is that the tabbed and/or MDI environments are very restrictive when it comes to PLC programming - you need to see a lot of code simultaneously. If you have two screens, it's difficult to make effective use of them. I'd rather each editor window was a separate floating window that could be placed anywhere on the desktop. Another thing that most environments lack is basically remembering the size and position, etc. of all windows, so that you can spend a lot less time moving and resizing things each time you open them. My suggestion was the ability to save editing "scenarios" - i.e. save my current window layout exactly as it is on screen now and later restore it. I'll get off my soapbox, but if you think how much non-value-added time you spend just manipulating windows.... 5) The built-in visualization editor is decent and powerful, but like any property-based system, can be complex to understand at first pass, and tedious once you really know what you want to do. Maybe once you build yourself a decent starting template it would be easier. For larger projects we've decided to continue to use Visual Studio .NET and either OPC, or preferably the more optimized and very powerful PVI interface to the PLC. This approach continues to prove itself from a power, cost, and flexibility standpoint. The PVI interface, to me, is certainly one of the significant benefits of the B&R system, and few other platforms (to my knowledge) offer this sort of thing. Of course, it's not open like OPC, but if you've standardized on B&R hardware, why not optimize the HMI to PC communications. 6) For communication between B&R PLCs over standard ethernet, there is a library that is more difficult to configure than it should be, but works well once configured. It's certainly not like allen-bradley publish and subscribe - but I wish it was. Of course, you can also use MODBUS TCP (for which the master is built in and can be simply configured to talk to MODBUS slaves), or one of the other supported industry standard networks (Profibus/NET, EtherCAT, etc... they support most everything with an interface module). 7) The B&R POWERLINK real time ethernet interface (very similar to EtherCAT from a standpoint of how it is connected), seems to work just fine. Of course, the only problem is, there are very few 3rd party devices that have this protocol built in. Not a huge issue if you use all B&R hardware, but if you want to talk to an EtherCAT device (for example), you need an add-on interface card. POWERLINK is an open standard. Not sure why others have not adopted it yet - maybe because they figure it's too new and/or they already support enough networks. Certainly there is a critical mass point at which all 3rd party vendors jump on board - we haven't reached that yet with POWERLINK. 8) The selection of I/O modules (slice and IP65 blocks) is quite extensive (maybe not so extensive as Beckhoff), but they seem to have everything I need so far. 9) The large variety of PLCs (some with and some without an integrated HMI screen) is very extensive and seems to fit the bill from the very small system (a few air cylinders) to the very large (99+ servo axes). With the wide variety of things my company needs to do (lab systems to production OEM machines), this was quite attractive. 10) The selection of servo drives, stepper drives, and VFDs seems more than adequate to fit most needs. Be warned that lead times on motors (the only thing they don't build in-house) can be longer than desired (6-8 weeks). They have 3 levels of servo drives, ACOPOSmicro, ACOPOS, and ACOPOSmulti. The first two are stand-alone, and the third is rack-based with shared power supply. Features and capabilities go up. The micro drives are decent size (one and 2-axis are the same size). The second two are much more capable, but also physically quite large (you'll need a 12" deep enclosure). Network-based safety is built in to the ACOPOSmulti (not quite sure about the ACOPOS standard drive). The micro, for sure, only has enable lines and you need to use safety outputs to achieve STO. Not a show-stopper, but I wish network based safety was consistent across the entire line. 11) Motion is based PLC Open with supplementary function blocks that extend the capability to exploit specific B&R features. Built in servo tuning and commissioning is not quite as polished as Beckhoff, but you can display traces of motor velocity, position, following error, etc. Basically everything you need. I have not put this through it's paces much yet, so my opinion may change. 12) There are over 100 function block libraries provided with the software. Extensive capabilities, but sometimes you have to contact B&R support to ask if a function of a certain type exists because it's difficult to know with so many options. In some cases B&R has provided additional libraries that are not part of the official set, with the warning that they are not officially supported and use at your own risk - bla, bla. Nonetheless, they've worked just fine. There isn't too much 3rd party support out there for B&R yet, as you may have noticed if you've done any searching. You can create your own custom libraries of function blocks, but there is an acknowledged problem (to be fixed in the next version) with debugging function block instances - it's buggy if not impossible. 13) The B&R help system is very extensive, updated with almost every release (of course the download is huge and install takes a while). The help is really quite thorough and gives almost every detail that you need, as well as a large number of examples and full example code projects. It's some of the most well-written help that I've seen. The help search engine, however, pretty much sucks, because it returns way too many matches for any typical search. One thing I really like is that virtually EVERY error code returned by any library or device (e.g. servo errors) is in the documentation. Simply type the error number in search and up pops the documentation behind it (some helpful, some so-so, but at least it's there). They also provide files containing text descriptions of servo errors (at least in English and German, not sure about others) that can be used by your code for display to the operator. Not so (as far as I can tell) for many of the other libraries, unfortunately, so you're on your own to explain those to the operator. I could obviously go on forever! I'm sure I'll have more accurate opinions once we get through these larger projects. My general impression is that the system is very powerful, very capable, relatively easy to program and debug, but there are several areas that need improvement, especially related to the code editor and window management. All in all, so far so good - nobody seems to knock it out of the park in every area. The good news is that B&R seems to be aggressively making improvements. The B&R platform is certainly worth serious consideration. Be aware, however, if you're an Allen Bradley programmer, this type of system and approach to programming will feel VERY foreign. I'll freely admit that the Allen Bradley control logix platform and programming software is second to none. But if you look at the range of hardware and unique way that a modular system like B&R (and Beckhoff) can solve automation projects, I think you'll be impressed.