Arne

Thanks a lot. I did not know that this was possible.

Thanks for that info, but I don't know if I understand it completely. Can this be done without changing the program itself, and a new download? Will it not require a new autotuning after I change the input variable from slider to the analog input? The same variable can not be connected to the slider and to the analog input at the same time, or can it?

Edit: The problem with the autotuning appeared in a way smaller than I expected. You only have to do "a fake autotuning" one time (when you have a 4-20 mA source connected and not a real process). The the next time you turn on the PLC, it will of course pick up the old parameters and there will no need for a new autotuning. (And PID controller can be manually configured.) Still if anyone know how to make a program without an autotuning at all, it would be interesting to know.

I guess there is "to parts" of a stable closed loop control system. Half of it is the controller and the other half is the process itself, what is contains of time constants and dead time. If it is a some kind of heating fan with a temperature sensor, it might give a proses that easy get unstable. https://blog.opticontrols.com/archives/499 Maybe the problem is more related to the process than to the controller? A more detailed information about the process itself might make things a bit clearer.

I use Visilogic and Unilogic for teaching. To be able to demonstrate PID responses using a mA source on the input (simulated process value) and a mA meter on the output, I would (also) like to run the PLC as i PID controller with no process connected. I first made such a program using the Visilogic (V570) and it worked just fine to remove all auto-tuning and to have only manual configuration of P, I, D parameters. (And the output (CV) could be sat manually as well) Then I tried to do the same with Unilogic (USP-70-x10) and this seems to be more difficult. It looks like you first have to make an auto-tuning before the PID function will work. (The Status Indicator will show zero or some other value that will prevent the PID controller from running.) I have found out that if you take the simulated process value from the mA source up and down a few times, I can "fake" a process response and make it run that way, but it is not a really good solution. Is there any possible way of programming a PID control function without auto-tuning? If it is impossible it would also be a valid answer. Then I don't need to use more time to find it out, If it can not be done or not 🙂 (I think that the problem is that the PID_Config.Status bit is not so easy to manipulate without making an auto-tuning first, and that's the reason it is difficult to run it in manual mode, without making this auto-tuning first. Hopefully I'm wrong in some way, or there is some workaround to this "problem".)