Thanks for the reply @tomas_hora although not very useful. As you can see from the graph in my original post, the Xp and Ti parameters are correctly set. Stepping up from 2 to 3, the PI controller works fine. Stepping down from 3 to 2, the behavior is strange and unexpected. Neither changing the parameters nor switching to PID control solves the problem.

Also your reply about the B34 and B35 blocks does not bring any useful information. I can see the heating/cooling description in the title of the block. I'm trying to control pressure. Which one should I use then? The documentation is the same for both blocks.

Please explain. Thanks.

Here are the parameters. In the meantime, I have switched to B34 as I was trying to find the difference between the two PI controllers provided by Mervis. Can you please explain when to use B34 and B35?

Thx.

Hi @r3sist,
any progress with your project? Very interesting.

I was trying to use the built-in PI controller only to find out that it is behaving unexpectedly. When stepping up, it works as expected - Proportional term kicks in and afterwards the Integrating term makes it to the desired value. But when stepping down, no action of the P term is observed and a big undershoot happens.

Is this normal and is this an intended behaviour? If so, what is the purpose? And how to make the PI controller behave in a symmetrical way?

This older post asks moreorless the same question, but there is no reply: P term cutoff in PID controller :-(

Hi guys,
I know this is a Mervis section, but if PID control is what you are after, consider switching to REXYGEN. Works like charm.

Codesys clearly states that the version for Raspberry Pi can be used for testing and training only. Thus by using it in your automation projects you would be breaking the terms of use. This is not the case of Mervis or REX. So I guess you are comparing apples and oranges…