# Controller for infinite phase margin

#### diegohelpme

Joined Jan 7, 2022
2
Hello, I have a transfer function of an active suspension system of a quarter of a vehicle, they ask me to calculate a controller from the Bode diagram that complies with overshoot of 33% and settling time 1. I have the problem that the margin of phase gives infinity, if someone could guide me I would appreciate it. Another question is because when graphing it, it gives me a curve of 10 ^ -5. Is this normal for active suspension systems?

#### LvW

Joined Jun 13, 2013
1,773
I suppose your graph should represent the loop gain (magnitude and phase) - correct?
This display of the frequency-dependent gain of the open loop is necessary for finding the stability margin.
As I can see, your loop gain phase starts at zero deg.
This is an indication for a wrong loop gain simulation - it always must start at -180deg (DC negative feedback for a stable quiescent point). It seems that you have forgotten to inlclude the phase inversion (-180deg) at the summing node for input and DC feedback signal)

#### diegohelpme

Joined Jan 7, 2022
2
I suppose your graph should represent the loop gain (magnitude and phase) - correct?
This display of the frequency-dependent gain of the open loop is necessary for finding the stability margin.
As I can see, your loop gain phase starts at zero deg.
This is an indication for a wrong loop gain simulation - it always must start at -180deg (DC negative feedback for a stable quiescent point). It seems that you have forgotten to inlclude the phase inversion (-180deg) at the summing node for input and DC feedback signal)
How can I reverse the phase? .
Yes, the Bode plot represents the gain and the phase. What I did to get the diagram was to enter the system transfer function in matlab using the sisotool command.
If it is possible to contact you by mail I would appreciate it

#### LvW

Joined Jun 13, 2013
1,773
How can I reverse the phase? .
Yes, the Bode plot represents the gain and the phase. What I did to get the diagram was to enter the system transfer function in matlab using the sisotool command.
If it is possible to contact you by mail I would appreciate it
It is not correct to use the transfer functions Bode diagram for determining the phase margin.
As I have mentioned, you must use the LOOP GAIN: Open the loop at a suitable point and inject a test signal.

#### tindel

Joined Sep 16, 2012
938
As LvW correctly points out - a lot looks wrong about this. It looks like you're providing the plant transfer function - but seems strange that it would be that small of magnitude. With that much overshoot, it sure doesn't seem to be controlled at all, again looks like the plant. LvW is right, break the loop, inject a test signal... measure your open loop gain.