# Control Theory

1. Nov 15, 2015

### Davidak

Hi,
Is this Bode diagram tells that the system is stable? As I see it is, because of the φ>0. What doest it mean that the upper diagram has two zero-crossing?

#### Attached Files:

• ###### bode.JPG
File size:
15.4 KB
Views:
66
2. Nov 15, 2015

### BvU

Hi dave

I see $\phi < 0$ up to very high frequencies

What's the stability criterion in your context ? The amplitude never runs off to infinity, so you may well be right ...

3. Nov 15, 2015

### Davidak

Thank you for your respons!
As you see in the picture, the second zero crossing defines a phase, which is -90 and according to the stabilty criteria it is stable because -90>-180. Is it correct?

4. Nov 15, 2015

### Davidak

This is the step-respons diagram of the same system. Its stabel. So the Bode should be also stable, but i m not sure.

#### Attached Files:

• ###### upload_2015-11-15_14-46-6.png
File size:
52.7 KB
Views:
63
5. Nov 15, 2015

### BvU

I think you are right. This looks like a second order system with a $\zeta$ of about 0.3.

Apparently your criterion is $\ \phi > 180^\circ\$ when |response/input| = 1 which I don't really recognize. Your call.

6. Nov 15, 2015

### FactChecker

Yes. It has what is called 90 degrees of "phase margin" at that frequency because 180-90 = 90. All lower frequencies have more phase margin. The higher frequencies have what is called "gain margin" because the gain for those frequencies is less than 1. So it is well within the stable region at all frequencies, as you can see by the damping of the step input signal.
PS. Usually the gain margin of a system is defined as the margin at the frequency where the phase shift is -180. In this system, there is no frequency like that.

Last edited: Nov 15, 2015
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook