It is tradeof between stability and gain. To avoid oscilations in feedback amplifier, phase must be less that 180 at unity gain frequency. You can make frequency compensation with less phase margin, to obtain bigger gain at lower frequency, but you risk to make oscilator, not amplifier.
It can be proved as a general case that given a phase margin of 60° the amplifier's step response will not have any peaking, thus very less possibility for it to become stable. Thus 60 degrees is normally chosen to be the phase margin.
45° degrees is taken as a tradeoff, to have faster risetime and less settling time.
In some situations when you need to achieve higher bandwidtha, smalll peaking of 4.3% can be tolerated. This response has phase margin 64, and is called Butterworth response. For higher phase margin, peaking much more disturbs signal.
i dont believe on this 60 degree PM concept..
for critical dampling, PM shd be close to 70 degree at Wcl...
and u can achieve that by setting ur nondom pole freq:
Pnon=3*Wcl (feedback factor=1)
Pnon=2.5*Wcl (feedback factor=3)
Pnon=2*Wcl (feedback factor>5)
thnkz everybody
one more question...........how do i define phase margin if mag plot crosses 0db more than once??
which systems usually hv such response??? can ne1 give an example