Slow Communication between 1769-L33ER and Ignition Edge

Hello,
i have project developed in Ignition edge vision module and i have deployed it in project site with 2 vision client running on two separate PC. in order to reduce device loading i have to set system overhead slice in PLC to 70%. so is it okay ? after setting that to 70% device loading is now 0% and actual sampling interval is 1000 ms (1 second). but still i feel it is not instantaneous. if i press start button on HMI it starts motor and changes status in 2-3 seconds and in PLC it takes long to visualize this execution when online. is there anything i can do or check to improve this performance ? thanks

There are many things you can do, mostly involving structuring the PLC code and data for best performance. In some cases, my 3rd-party alternate driver is needed. Start reading this topic and following all the links:

Sounds like something else might be going on unless this claim isn't true.

@Mitesh_Maniar maybe let support take a look at your system?

2 Likes

Thanks Kevin. i had it it looked by support and they said it is okay. but one thing still i am worried is "System Overhead Slice" Setting to 70% can cause any impacts on logic execution ? i saw some post in user forum where user mentioned to increase but Rockwell suggests to keep it 20% which doesn't work for me. any thoughts ? My project has 9000 Tags in total. please let me know if any question.
Thanks for your help.

thanks Phil. changing continuous task in PLC to periodic requires any other modification. i have my all logic written under continuous task and only PID is under periodic task with 250 ms. while my project is running scan time max goes to 80 ms. thanks for your help.

That is an astonishingly busy PLC task. Your PLC code is possibly too complex for that model, or perhaps has loops or other non-PLC program structures blowing it up. (Fairly common problem when IT-ish types try to write code for PLCs.)

I agree but it was not that high until i changed system overhead slice to 70%.

Then don't do that. Put it back to 10% or so.

Figure out how fast your PLC task really needs to run. (Based on the requirements of the connected process.) If mixed, consider breaking the PLC code in multiple programs at separate speeds. Then your PLC code will all be in scheduled tasks, where the task pace is appropriate for what needs to be physically controlled.

Then, you will get the maximum messaging performance possible without breaking your physical process.

If that level of messaging performance is still not sufficient, start implementing all of the other PLC tag organization recommendations in the linked topics.

If that is impractical, or only partially practical, and you still need more messaging performance, test my alternate driver.

(Or skip to my alternate driver to see how much of the other stuff you actually need to do.)

There is NO Magic Wand!

2 Likes

Thanks Phil. Appreciate your detail explanation.

Are all 9000 tags at 1sec scan class? Do they need to be?