Problems working with the Integrator of the PID Controller
Verfasst: Freitag 14. April 2017, 09:40
Dear All,
I truly apologize for asking help on what would be easy things for most of you here, however I have some great difficulties to get the PID to work correctly.
In the plotter the trace colors represent
Red = FB signal
Blue = Set point(X+)
Orange = Controller output,
I have set the limit properties of the Controller components as shown in the attached screen print.(in a separate post as the forum allows only 3 attachments per post)
In my understanding of reading the plot is
(1)As the error goes bigger Error = (Set-point - FB signal) the controller output increases ,correctly so as the output in my understanding is trying to correct the error.
(2)As the error is not reduced the integrator is increasing the controller output until the max limit is reached ,which it is doing as the output reached the 117 max limit in the plot.
(the error was deliberately not reduced to show this and indicate my problem) .
(3) With the control output at maximum and the error is staring to correct (going to zero),the controller output stay at maximum until the error reached the zero (minimum setting ) then the output fell from maximum to zero.
This is my issue once saturation is reached the control output does not reduce gradually back to minimum value.it hangs on maximum until the error is zero(0) then control output drop to zero.
In my lack of knowledge it seems to me the integrator does not stop when maximum output is reached and when the error is deceasing the high integrator value keeps the output high.
Any advise on how to correct this issue would be be greatly appreciated .
I have tried the following by resetting the integrator by pulling the RST pin low when the output has reached the maximum.
This did not give good results as the controller output drop to a lower value, then the RST pin goes , integration is started , output goes to max again and RST is pulled low again.
This process just keeps going in this way .
As shown in the 3rd screen print.
In the red marked area is a reset function when the error is small , I stop the integration.
In blue a function that stop the integration when max is reached and was the result of trying to solve the issue I described above.
I truly apologize for asking help on what would be easy things for most of you here, however I have some great difficulties to get the PID to work correctly.
In the plotter the trace colors represent
Red = FB signal
Blue = Set point(X+)
Orange = Controller output,
I have set the limit properties of the Controller components as shown in the attached screen print.(in a separate post as the forum allows only 3 attachments per post)
In my understanding of reading the plot is
(1)As the error goes bigger Error = (Set-point - FB signal) the controller output increases ,correctly so as the output in my understanding is trying to correct the error.
(2)As the error is not reduced the integrator is increasing the controller output until the max limit is reached ,which it is doing as the output reached the 117 max limit in the plot.
(the error was deliberately not reduced to show this and indicate my problem) .
(3) With the control output at maximum and the error is staring to correct (going to zero),the controller output stay at maximum until the error reached the zero (minimum setting ) then the output fell from maximum to zero.
This is my issue once saturation is reached the control output does not reduce gradually back to minimum value.it hangs on maximum until the error is zero(0) then control output drop to zero.
In my lack of knowledge it seems to me the integrator does not stop when maximum output is reached and when the error is deceasing the high integrator value keeps the output high.
Any advise on how to correct this issue would be be greatly appreciated .
I have tried the following by resetting the integrator by pulling the RST pin low when the output has reached the maximum.
This did not give good results as the controller output drop to a lower value, then the RST pin goes , integration is started , output goes to max again and RST is pulled low again.
This process just keeps going in this way .
As shown in the 3rd screen print.
In the red marked area is a reset function when the error is small , I stop the integration.
In blue a function that stop the integration when max is reached and was the result of trying to solve the issue I described above.