function transfert of a first order system (without any toolkit)

[scls19fr]

Hello,

I'm using a quite old LabVIEW version (6.0.1)
I'd like to simulate a system and its control.
Unfortunately I haven't enough money to buy a toolkit for this.
I'd like to do it without any toolkit

The function tranfert of my system is
H(s) = OUT(s) / IN(s) = K/(1+tau*s)

So I can write OUT*(1+tau*s) = K*IN
and so OUT = K*IN - tau*s*OUT

Inverse Laplace transform is

out(t) = K*in(t) - tau * d out(t) / dt

so the schematic should be

     K--|
          x----------(K*in)
in------|            |
                       +
                         -------------------------out
     tau---         -                  |
             | -------(tau*s*O)   |
            x                            |
            |-----<---d/dt-----<--

         
Input comes from a slider
and output goes to a scope


Is it possible to do it with LabVIEW ?

I have the following problems :
 - I don't see how to calculate the derivative function of OUT
 - I can't compile because LabVIEW seems not to like cycle in my diagram

Regards

[cosmin]

Hi,

you can make a simulation of your system with backward-difference discretization.

cosmin

[scls19fr]

Thanks for your .vi

But there is some problems...

First I don't understand how it works (sorry I'm a beginner)

I don't understand for example what is the difference between T and tau
(I have never use the "i" and the "+1" icon)

On your scope, I see that the initial condition is 0...
but the graph start at 0,5...

I don't understand why it is not possible to do a cycle in a shematic...

Regards

[scls19fr]

That's ok
The problem was with shift register
...

But if you can explain me a little I will be very pleased...

The second part is now to do a sub-vi for my first order model
to use it in other simulations

The third part will be to make a second order model
(beause a second diff equation can be write as
a system with 2 first order equations)

[scls19fr]

There is still a problem with initial value

[cosmin]

Hi,
regarding the initial value showing in the chart, wire the waveform chart directly on the left shift register (as I wired the chart, the initial value is not shown).
look for backward discretization on digital control system books.
cosmin

[scls19fr]

Ok, I see what you mean (sorry for my poor english... I'm french)
You replace the derivative of a function by a finite difference

But what I don't understant here is that T should be very inferior (<<) from tau
and the 100 ms constant is same as T

Do you think that this sim can become modular to be in a sub-vi ?
(and used for example in a feedback)

Thanks

[cosmin]

Hi,

looking at your system, the pole is -10, and settling time 0.4, so 0.1 as T yelds good enough results. Descrese T as you wish

cosmin

[scls19fr]

Thanks

[Kumar_B]

How to get comma(,) in the numeric control

-Kumar  B