Integral for PI controller? Why the value has been magnified?

[spwm]

Hi,

I want to build a PI controller using the original form, which consists the Proportional and Integral function. I chose the function "Time Domain Math". I  selected the Integral (Sum[Xdt]) and continuous calculation. When I did the testing, I found that the output value has been magnified. In my program I used a sine function the test the Integral function. The magnitude of input is 1 and the output become 320. Could anyone help me? Thanks!

Note: For testing purpose, in my program I use different input signals for P and I functions. For my understanding, I can have the PI controller, if I make the same input for P and I function and the sum the result. 

[BillMe]

It's no wonder since all you are doing is integrating the sine wave's amplitude over time and adding some constant (P side) to it.

What you have is not a PI controller. Where's your feedback, error and setpoint terms? In control theory, for a closed loop system, the P and I gains are applied to the error (difference between the feedback and your setpoint) and then summed to give a new control output which (hopefully) reduces the error.

There are a lot of PID tutorials on the web. My suggestion is to take advantage of them.

[spwm]

Hi Bill@NGC,

I agree what you said. But There are two points I need to mention. Firstly, It is the structure of the PI controller. It just has two parts the Proportional and Integral function if I don't consider others. Secondly, the program I attached is part of the control scheme. I test my program and found errors, so I break the whole program into piece. Then I found this magnified value. So I make this simple test to integral a sine signal. For ordinal condition, this sine signal should become cosine signal after integral function with same amplitude. But the result has been magnified, and I have no clue.  

Best regards, 

[spwm]

Hi,

I thank I should to ask this question in a different way. I actually want to know the procedure of the integration. For example the integral of sine signal between 0 to 90 degree, is -cos(pi/2)+cos(0), which is 2. But the signal generated from my program become far larger than this value. I don’t know what cause this and I am confused. So please anybody can give me a suggestion? Thanks!