pase margins for servoloop problem

[mooman]

Hi all,
I have never before delt with controls, now I need to prepare project which I find quite difficult.

I need to find the proportional regulator gain for loop with servovalve to fit the specified reuirements for gain and phase margins, rise and settling time.
In my analized loop I generate Bode Plots for Open Loop state,
I check gain and phase margins to meet specified requirements.
Next I close the loop increase the regulator gain to meet the reauirements for rise and settling time.
The requirements are correct for OPEN LOOP gain and phase margins, CLOSED LOOP rise and settling time.

But when I generate Bode Plots for Closed Loop i see the Magnitude diagram starting from value of 0, and Phase diagram starting from -180
Which gives me -180 phase margin.
I am positive that each element transfer functions is properly calculated.
I think they are correctly combined since it is a straight loop and I deal only with multiplication of those.

Is it even possible to occur and if, how can I interprate such results ??

 

[LvW]

But when I generate Bode Plots for Closed Loop i see the Magnitude diagram starting from value of 0, and Phase diagram starting from -180
Which gives me -180 phase margin.

Hi mooma, you should know that the term "stability margin" applies to OPEN loops only.
That means: When the BODE diagram of the open-loop transfer function (normally called simply "loop gain") shows, for example, a phase margin of 50 deg, the meaning is as follows:
The CLOSED loop will be stable, unless there are unwanted phase excursions in the order of 50 deg. within the loop.
With other words: The closed-loop has a phase margin of 50 deg, but this value can be derived from open-loop simulation/measurement only.

- - - Updated - - -

Remark: The closed-loop phase has nothing to do with the phase margin - it is an indication of operation with signal inversion only.

 

[mooman]

Thank You for your response this really cheard me up.
I have another question:

my regulator gain, for which the requirements are satisfied is > 600
what do You thin about it, isn't this value to high ??
requirements for rise/seettling time are quite lifted, values are .25s/.5s

on the other hand, coefficients in analized transfer function have extreamly high values in denominator