Hey guys. I have tried to manually control this control system by taking out I and D gains and incrementing P gain till the system oscillates. But in this case my systen never reaches oscillation. Is there any other way?

 

Image is a thumbnail. Remember that tuning requires a step change and tat step change MUST be greater and likely much greater than 1/PB

 

What do you mean 1/PB. I just use a step input of 1 and increase gain

 

Sorry. 1/Gain = Proportional Band (PB)

A small step will do noting in helping you tune, Your diagram which I really can;t see, shows "step Change"

Higher gains give better control with the chance of oscillation.

The integral time tends to make the measured value and setpoint agree

The derivative time tends to reduce overshoot.

Implemented software PID need to include code to prevent "restet windup". There is no point for the I term to be over +-100%; i.e. -200% or +240% It can be limited further than 100% Another way is to limit the output of error ad the third way is to limit the I term.

But, in order to tune, you need a large setpoint change.

 

Yes that is exactly what I am trying to tell you.
https://ibb.co/f24NSS



Maybe you can see better from this Image.
I have increased the Kp gain gain to a 100 , 1000 , 10000 and monitored to scope. And my system is never reaching constant oscillation. It is always stable! And in order to manually tune your system you to achieve constant oscillation

Moderators note: insterted essential part of the image