Control Systems Circuit Design Help

Thread Starter

ahmetkara4635

Joined May 5, 2024
12
First of all, hello. I want your help with an assignment. The assignment given to me for the control systems course is to design a lead compensator. For this, I first need to design the circuit of the plant given to me. My plant is 5/s*(s+9).
For this, I first took my simulink output. Then, I tried to implement this circuit with an integrator, a low pass filter and a differential receiver opamp. While my circuit works in LTspice it doesn't work in reality and Proteus. Can you tell me if I missed a point or something I should pay attention to while implementing my circuit?
Below are my outputs.
 

Attachments

MrAl

Joined Jun 17, 2014
11,734
First of all, hello. I want your help with an assignment. The assignment given to me for the control systems course is to design a lead compensator. For this, I first need to design the circuit of the plant given to me. My plant is 5/s*(s+9).
For this, I first took my simulink output. Then, I tried to implement this circuit with an integrator, a low pass filter and a differential receiver opamp. While my circuit works in LTspice it doesn't work in reality and Proteus. Can you tell me if I missed a point or something I should pay attention to while implementing my circuit?
Below are my outputs.
Hello there,

Just a few questions that will help me understand what is going on with the circuit(s). These are all important.

First, what exactly is not working right for your implementation of the circuit?

Also, since you have one true integrator in the circuit, did you check for a DC offset in the output that would be too high?

Another question, did you analyze the circuit you are using for the actual implementation and if so, did you get the right response?

Another question, when you created the low pass filter did you realize that's not a true integrator (or perfect integrator)?

If you are first concentrating on the forward part of the circuit then these would be questions for that, but if for the entire circuit then for that.
 

LvW

Joined Jun 13, 2013
1,778
First of all, hello. I want your help with an assignment. The assignment given to me for the control systems course is to design a lead compensator. For this, I first need to design the circuit of the plant given to me. My plant is 5/s*(s+9).
For this, I first took my simulink output. Then, I tried to implement this circuit with an integrator, a low pass filter and a differential receiver opamp. While my circuit works in LTspice it doesn't work in reality and Proteus. Can you tell me if I missed a point or something I should pay attention to while implementing my circuit?
Below are my outputs.
Your plant is a system with an I-T1 characteristic.
Hence, it consists of an integrator and a simple lowpass.
The most simple realzation would be to use the classical inverting Miller integrator in series with an inverting damped integrator (active 1st-order lowpass).
A possible DC offset of the integrator would create no problems as the plant will be part of a closed-loop control system.

* General comment to your simulations: As far as I understand, at first you are trying to realize the circuit for the plant (I-T1) only, correct?
However,all your simulations show closed-loop responses - why?

Question 1: Why do you want to test your circuits in several different simulation packages?
Question 2: You wrote"....doesn`t work in reality...". What does this mean?
Question 3: The step response shows a certain delay.....does your plant contain such a delay? This would be a very important property with respect to stability of the closed loop
 
Last edited:

MrAl

Joined Jun 17, 2014
11,734
First of all, hello. I want your help with an assignment. The assignment given to me for the control systems course is to design a lead compensator. For this, I first need to design the circuit of the plant given to me. My plant is 5/s*(s+9).
For this, I first took my simulink output. Then, I tried to implement this circuit with an integrator, a low pass filter and a differential receiver opamp. While my circuit works in LTspice it doesn't work in reality and Proteus. Can you tell me if I missed a point or something I should pay attention to while implementing my circuit?
Below are my outputs.
Hello again,

Here are a few hints.

If you look at the attached diagram you can see the feedback path has been broken and the associated input to that first op amp has been grounded. After doing that, analyze the output from point A to point B.

When you do this, you will find your coefficient of s^2 is not the same as in the forward section of your first diagram. To fix that, you need to change one or more values in your low pass filter consisting of R5, C1, and R6. Those three need to be looked at more carefully.
Once you change one or more of those values you will get the same response as the forward section in your first drawing when you consider all the components as ideal. The circuit should work better after that, provided there is no DC offset problem associated with U2. You might have to check for that also.

To calculate those values you will have to analyze the circuit as mentioned and solve for one (or more) of the values for one (or more) of the components. That can come from comparing what you need to what you analyzed.

I can see that it does look like you are on the right track with a reasonable grasp of what is required here.
 

Attachments

Last edited:
Top