Labview for automation programming

[snarkysparky]

I need to write a program for machine with 8 servo motors. This program must be able to read and interpret a simpler
language ( home made ) that describes the machine operation because it will change frequently. My leader wants to use LabView for this but I am having some time getting dataflow programming in my head. At this point I am on a path to write a rats nest wires and boxes that will be hard to maintain. I can do it well in other languages.

Is this a learning curve thing, is it just as easy to write code with many states and interactions in in graphical language as say python once you get the hang of it.

I want to switch but I need some ammunition to back my point. Anybody know any public comparison for machine control.

Oh and none of the devices on the machine have any LabView blocks. It's all bit banging out instructions.

 

[itsmoked]

Ah man.. Just get a cheap PLC that handles servo control. Check the higher end Automation Direct PLCs. They just added another model too.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[FreddyNurk]

Its been years since I looked at Labview, but I would have thought that it isn't necessarily that different to some of the function block programming styles that are around for some PLCs.

Admittedly, Labview is more focussed on data acquisition rather than control, and to that end, it used to work quite well. Whether its suitable for use as a replacement for a different PLC platform is a different question.

Probably one of the more pertinent questions is what scope of version control is available, and how easy is it to do comparisons (or diffs) between revisions? Is there a better means in an alternative platform that would make it easier to track the changes?

EDMS Australia

 

[snarkysparky]

The business need for the machine is there needs to be flexibility for non controls people to specify servo
motor parameters such as speed, acceleration and position points without being exposed to the lower level details. That is why my managers prefer LabView. But I also have to code up logic to read a file of parameters and convert them to position instructions in LabView at run time not compile time. It seems to me to be difficult to write complicated logic in the data flow environment of LabView. I need a GUI interface with an environment that will easily let me code a basic instruction interpreter with lots of string manipulation. I could do it in C# in no time but the management is worried about finding support for C#. Any random thoughts appreciated

 

[controlsdude]

I remember seeing this type of situation at a company. I think they used visio as the interface between operators and the servo code. Maybe look into Microsoft Visio as a compromise instead of using Labview.

I know that visio worked in the interface that your asking for above.

 

[RyreInc]

I'm not really familiar with LabView, but if it were me I wouldn't think twice and go with a PLC + HMI. The HMI is the user-friendly interface you require, and can display pretty much whatever you want; and PLCs are designed for such tasks, rather than trying to re-purpose something. I would be worried about responsiveness with LabView - can you ensure the code will reliably scan through often enough? Or can some other program interrupt LabView and you can lose control?

 

[IRstuff]

The issue lies with standardization; something running on Labview with common interfaces is portable to another machine with Labview. Otherwise, everyone winds up picking their own favorite PLC+HMI and no re-use or commonality results and no one can work on anyone else's setup.

We had a contractor once who created their own unique digital video interface, which he then proceeded to "debug" for 3 months; he was let go, and the rats nest code he work had to be scrapped and re-written.

There seems to be examples of using Labview to control servos:

https://www.allaboutcircuits.com/projects/create-a-functional-pid-controller-on-the-ni-myrio-part2/

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[PNachtwey]

Why not get an 8 axis motion controller? An HMI can be attached using Ethernet. The PLC and LabView are not necessary but can be connected using Ethernet. The HMI can change positions, speeds, acceleration and deceleration.
The HMI can also change conditions required to go to the next step. This is standard stuff.

LabView will work but it isn't the best or simplest for motion control. We use LabView for our automated test machines. There are a lot of sequential steps but usually they are doing the same test over and over again. If there is another type of board to test we use a different program.

A cheap PLC is not so cheap when doing motion control. The hardware may be cheap but the cost of writing and supporting all the motion profiles and closed loop controls is prohibitive for a one of a kind machine.

There are many common motion controllers out there that are capable of doing basic motion and sequencing.




Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[snarkysparky]

Well I worked through it an got the basic stuff working. The way I did it was to code the top level logic in C# ( something I am familiar with) and then implement piece by piece in Labview. I know people say it is more productive to program in labview ( faster ) but not for me. There were many 4 letter words issued while i struggled with boxes an unfamiliar data flows.
My managers are Academic Engineering folks who put Labview assignments on students all the time and they consider that it must be easier to implement , and there are more hungry grad students who can support it.

I guarantee it is some ugly code. I just feel Labview is great for flow through programming with minimal branching and not so hot for complicated logic such as a machine control.