for a 2-pole system, you can use this idea.
And I think Step response is more important than PM for reality use.
for example when system has one dominant pole and 2 conjugate poles, or has zeros, PM may be good, but step response is very bad.
Very easy get PM from: calculator =>ac >function PhaseMargin
Added after 5 minutes:
Alan_Nesta said:
for a 2-pole system, you can use this idea.
And I think Step response is more important than PM for reality use.
for example when system has one dominant pole and 2 conjugate poles, or has zeros, PM may be good, but step response is very bad.