PID control

[TerraNova]

Hi everybody!

The thing is the "Transfer Function" from PID pack
does not have Limits, wich would make it worth modeling real-world systems (with the supplied voltage limits for instance).

So, can anywone help me to make "Limited Transfer Function".
May be someone has already done the thing alike with the "Transfer Function"?

Thanks in advance.

[Barp]

Sharonoff,

You can use the "In range and coerce?" to limit the signal after the Transfer Function below. Also, you could use the "Saturation.vi" that is in the Advanced >> Nonlinear.vi pallette.

Hope this helps...

Barp

[TerraNova]

Thanx, Barp!

Still, the thing is not just to limit the value :), but to limit the output value of the Transfer Function.vi ;).

The deal is to make, for instance, Integrator built with "Transfer Function" work THE SAME WAY as real-world Operational Amplifier Integrator, wich output value is limited, off course, by Suplied Voltage value. And this is the very simplified example I've shown.
Much more interesting thing is to make LabVIEW Transfer Function to work as real-world transfer function of complicated electric curcuit wich theoretical Transfer Function I can derive beforehand.

That's the question.
Thank you.

[Barp]

Sharonoff,

I think that you could mimic the behavior of the AmpOp Integrator with the PID functions, as shown below.

However, this solution requires you to have some conversion of numbers. Since you have the old PID Control Toolkit, you could use the Limited Integrator function in series with the Transfer Function that don't have the integrator, as shown below. That would get near of what you need. Maybe just the integrator with limit would model your Ampop.

Now, after showing all those options, do you still think that a built-in limitation of integrator in Transfer Function is necessary? That goes away from simple structure and get a more heavy weight implementation of the TF that is not needed if you are trying just model a normal Transfer Function. Would you prefer to have a special block that "mimics" the AmpOp with Saturation instead of TF with limits? I just trying to understand the root of this request.

BTW, thank you for your feedback!

Barp

Control Design and Simulation Group - National Instruments

[TerraNova]

Thanks again, Barp!

First of all, would you please attach those pictures in the post rather than insert - I just can't read them :(.

...And to clarify my problem...

You can see in the picture attached the difference in Transfer Function behavior and the behavior of a real Electric circuit: time intervals t2 and t3 are not equal (with the same input signal, off course).
And, the loss of time (t2-t3) is of importance!

The deal is to have both outputs identical.
That's when it is critical for me to have the truely Limited Transfer Function.

Appreciate any help!
Thanx in advance!