What's new
What's new

Choosing a PLC vendor

Punkinhead

Cast Iron
Joined
Feb 13, 2010
Location
IN, USA
Our machine design group moved away from PLCs 15 or 20 years ago because with the nature of what we were doing it made more sense to use robot controllers and their native languages (such as Epson's SPEL). I'm designing a line now with a complicated enough transport system and peripheral devices like tray feeders that it makes more sense to use a PLC. Having been out of the PLC game for so long, I'm curious if there's a clear leader in the market or if they're just commodities with no real reason to choose one over the other. One thing to add, a distributor that we buy a lot of automation stuff from sells Mitsubishi PLCs and offers training and local support so that might weigh into the decision.

Also, back in the day it was all ladder logic. I'm reading up on structured text programming and think it fits better with my group's experience since we've been doing all control with C-like languages for so long. Is there any reason to choose ladder logic over structured text?
 
I would think Siemens. Type of code I'd go with what the group would be good with. I like ladder because it's very visual so debug is easy, but I've done it both ways. Text is maybe easier to walk up to and figure out what the program is doing. I've recently used an Automation Direct PLC for a simple project and it worked fine, but if you're not on a shoestring, I'd look at Siemens. They do baggage handling systems in airports so should have all the features you'd need.
 
Also, back in the day it was all ladder logic. I'm reading up on structured text programming and think it fits better with my group's experience since we've been doing all control with C-like languages for so long. Is there any reason to choose ladder logic over structured text?

Ladder logic is the preferred choice when the PLC programmer has no knowledge or experience in using a higher level language or understanding of structured programming. Ladder logic is suitable for "small" projects where there is no interest in writing reusable code. It is also the preferred choice by managers who have no experience writing code. It is a very time consuming and error prone method for larger projects and in my experience it requires a dedicated field service/software repairmen to keep the code running once the automation project is installed. These features are not considered disadvantages at many companies. A bug filled automation project can become a profit center under the right conditions.

The choice of a PLC vendor is determined by the location of your customer. If you are selling machines in Asia a Mitsubishi controller may be a requirement. In Europe it will be a Siemens controller. In the US it will be a Siemens or AB controller. The vendors were offering 30 to 50 percent discounts on list prices. So price can be a factor in the decision. Another factor to consider is the use of one of the device net systems for connecting remote I/O to the PLC The Siemens equipment may have more to offer.

It is also possible to install PLC simulation software that will run on a inexpensive office computer and provide the same capability as the PLC.

The cost of the PLC CPU may be small in comparison to the cost of the I/O devices. The decision to use a PC based system will be driven by factors other than cost. PLCs have very long product lifetimes. In most cases the automation project will become obsolete before replacement PLC components become unavailable.

Another consideration is the speed of the PLC CPU and its memory size. The CPU speed is important if there are a large number of sensor signals that need to be processed to determine the system state while running a "fast" process. A interlock signal is of little value if it cannot stop a failed motion command from damaging the machine or a in process part.

The memory becomes a concern if there are multiple process stations. For example if there are twelve process stations which are controlled by a single process subroutine that is called twelve times, there will need to be twelve data sets in memory. Each station data set contains timers, set points, sensor inputs, command outputs, and internal logic states.The memory needs to be allocated in advance when the software is being written even if the automation product will initially only be offered with six process stations.

If your automation project has multiple process stations the statement list type coding becomes essential. It will allow you to write subroutines with a structured data base. The subroutines can have multiple entry points for code initialization at startup, machine operation, machine diagnostics, machine repair, and machine shut down. The repair entry point may be needed when you need to remove damaged product from a conveyor type system but do not have physical access to the line. The unclogging of the conveyor in this instance needs to be done using the software.


PLCs do not do floating point arithmetic. This can be a problem if you need to control set points as a function of time or some other variable. The solution involves dividing the set point range and control variables into 256 increments and ramping the set point in small steps. This simple calculation will take a full page of code. The alternative is to do the ramping calculation in the user interface program and transfer the set point to the PLC.

The use of a single vendor can threaten the success of a automation project. Twenty years ago, Siemens sold its PLCs and PLC based motion control boards through one set of vendors. Their industrial automation motion control equipment was sold through a second set of vendors. The problem with this system was that the two sets of vendors were in competition. The PLC vendors at that time were selling motion control products that were considered defective by Siemens management. The vendor did not care as long as the product sold.

If you are primarily doing motion control it may be better to work with the industrial automation group and their specialized PLC rather than work with the PLC group and their specialized motion control.

For inexpensive motion control you will end up using another vendor's stepper motors and control boxes and interfacing with the PLC for start/stop and interlock signals.

One of the goals in automation is to avoid making a mess. Your control system will need to include sufficient sensors and interlock logic to insure that the process can be stopped quickly when a fault is detected. It should not rely on a operator tripping over a stack of bad parts that have fallen off the conveyor.

A well planned design will include a careful examination of the process by people from the mechanical. electrical, and software groups. This can be easily achieved if the planing is all being done by one person. It can sometimes be achieved if there is a competent manager directing the groups.

Planning includes detailed timing diagrams for all commanded motions. Once the timing diagrams are completed and the machine through put is known it will be possible to size the motors to meet the acceleration/ deceleration requirements for each motion. If this is a servo system, the mechanical design group will then make some quick guesses on mechanism settling time and maybe the first vibration frequency. This information is then used to adjust the acceleration profile of the motion command and to select the location of the command notch filter. Position sensors may be used either to detect a fault or to trigger a subsequent motion. The triggers are needed when the machine timing is variable and cannot be controlled by a system clock. Once the speed of each of the motions is known it will be possible to select sensors based on response time and eventually determine the minimum CPU cycle time needed to allow the machine to run.

There is also a need to determine machine reliability. If the machine has a required system reliability of 99 percent and the machine uses a sequence of twenty motions each with a reliability of 99.9 percent it will be best to abandon the project before anymore time or money is wasted. The system reliability in this case is 99.9% **20 = 98%

The planning needs to be completed before any software, mechanical, or electrical work is considered.

A project that relies on patches by one group to fix the omissions or mistakes of another group will be a project that is overdue, over budget, and does not meet the performance requirements of the customer.
 
Last edited:
Take a look at Beckhoff Automation.
I found a series of videos on Youtube showing structured text programming using Beckhoff's free software. It looks very promising. I've also downloaded and have been playing with Codesys.

My customers are all internal and are at plants around the world. The line I'm concepting now will go to Europe. I'm leaning towards a Siemens S7-1500 Compact. I'm also considering their Logo series but they're pretty limited on I/O and I'm not sure you can do structured text programming with them.
 
You also need to think about Safety - and whether you want to integrate motion in the mix. The S7-1500T by Siemens is becoming the new backbone of automation and motion control for Siemens including the new Sinumerik One control that will replace the 840D in the near future. Lots of I/O options, has the ability to do IRT for real time position capture . . . and it can be programmed by multiple users at the same time which if you have a large project with more than one person programming is a big bonus. And no problem with floating point math . . . and if you want to do a smaller less cost intensive project, the Logo is fine and also S7-1200 should be considered.

We are about to ship our first fairly complex S7-1500T with S210 drives . . . ProfiNet connection from the PLC to 8 servo drives in addition to 2 Fanuc Robots and a Keyence Laser Marker. A capable system with a lot of complexity and integrated safety programmed by 3 individuals simultaneously. We had lots of issues with the laser and some issues with the Fanuc Robots, but the Siemens gear has been rock solid.
 
the Logo is fine and also S7-1200 should be considered.

Have you used the Logo? I looked at it but it doesn't seem to be able to run ladder. It looked like a bunch if independent commands floating around flow chart style which didn't seem like you could do much but really basic stuff.
 
Last edited:
I don't know a heck of a lot about PLC's but I'd be interested to hear any comments on EZ Automation.
Hollywood style website, claim USA made, low cost, many happy users, etc. I've searched around the web looking for comments from satisfied customers and I don't find much. So either they're wishing and hoping, or they are the best kept secret in low cost USA made PLC's and related products. Which is it, anybody know?

EZAutomation.net - Exceptionaly Innovative, Made in America, Automation Control Products Factory Direct at Great Prices
 
We are about to ship our first fairly complex S7-1500T with S210 drives . . . ProfiNet connection from the PLC to 8 servo drives in addition to 2 Fanuc Robots and a Keyence Laser Marker. A capable system with a lot of complexity and integrated safety programmed by 3 individuals simultaneously. We had lots of issues with the laser and some issues with the Fanuc Robots, but the Siemens gear has been rock solid.

This is OT to the original question, but, if you don't mind me asking, what was the driver for mixing Siemens PLC and Fanuc robots as opposed to using a Siemens-based robot like a Kuka?
 
You also need to think about Safety - and whether you want to integrate motion in the mix. The S7-1500T by Siemens
I'm leaning towards one of the S7-1500 models based on my research. I'm also curious what you use and see others use for the language. As I mentioned in my original post, we have a couple grey beards with vague memories of ladder logic but all of my guys and gals are proficient in high level languages so structured text seems like an easy transition. Each of our factories also have a few support engineers with decent high level language skills. Do you see structured text being used a lot or are one of the other IEC 61131-2 languages more prevalent?
 
I was disappointed when they sold as Kuka has a good reputation and played well with Siemens. My sense is that Robot companies that want to participate in the US aerospace market will likely need to keep some distance from Chinese owned factory automation suppliers . . . look to ABB and Fanuc to become standards and perhaps Comau when a "run my robot" app using G-Code with a Siemens 840D is needed.
 
I'm an Automationdirect fan. Their prices are good, the parts are good, and the support is great! Plus they'll have almost everything you need.
 
Consider Omron. They are very well respected, fair priced, high quality and probably the best tech support out of anyone else.

Also you'd be surprised how good AutomationDirect tech support is.
 
We're really big on the Siemens platform where I work and the support, quality and options on the Siemens is outstanding.

Those I know who have worked on other platforms all tell me the Siemens platform is superior.

The biggest problem I see with them however in our area however (Connecticut) is most of the other places are all using Allen Bradley. It seems to make hiring a lot more difficult for us as apparently there is a steep learning curve for guys to go from AB to our Siemens systems. Siemens programmers in Connecticut are either all happily employed and not looking for work or just dont exist as in my five years with this company I haven't seen anyone new come in with Siemens experience and we have pretty high turnover.

Most of the new guys we get in tend to struggle for a long time before they come up to speed on the TIA portal. I don't know if that is due to the complexity of what we do and how our standard code is written or if it is the challenge of learning a new environment. I am told coming over from the AB world is a different way of thinking than Siemens. Our most senior Siemens programmer (20-30yrs experience) always used to tell me AB was almost put some one at a disadvantage.

Maybe a good idea before picking a platform would be to survey a site like indeed.com and look at the job postings to see what platforms are big in your area before picking the one that is really hard to find skilled staff for if you have any plans for growth.

Maybe motion guru can provide more insight on whether he thinks local talent pool should factor into the decision or not. It definitely seems to be the only disadvantage I see with Siemens.
 
I've been warned away from Allen Bradley due to their software pricing practices. Perhaps that's a thing of the past since the engineers who warned me haven't worked with them in awhile. Still waiting to here if my group will get the project that triggered my original question, but if we get it the team is leaning towards Siemens.
 
Do not rule out Allen Bradley over the cost of the software. We design many of our custom manufacturing machines with Allen Bradley Micro850 PLCs. The software is called Connected Components Workbench, and it is free for the Micro850 line of PLCs. One of the best features is that you can call Allen Bradley's Tech support without a support contract to ask questions on the Micro850 line. How many axis (motors) will you be controlling? How many IO? Are you programming an HMI. Also safety is easy to implement.

Micro850 Programmable Logic Controller Systems

you can direct message me with any questions.
 








 
Back
Top