2-stage cmos fully differential opamp designed for resistor feedback network for gain control. common-mode feedback using another diode-loaded differential pair feedback to 2nd stage of main opamp. standalone opamp ac simulation both diff mode and cm mode have enough phase margin. when connected in feedback network and simulating loop gain, it is found that common feedback becomes positive feedback. Phase at dc is 0 degree instead of 180 deg. I was suspecting it because main common feedback fighting with diff path through feedback network . But even I replaces common-mode feedback diff-pair with high gain vcvs, which has no bw limit and should overcome gain limit. Still I got common-mode feedback as polarity positive, 0 degree phase at dc.
Is it a real unstable issue? I checked step response of whole feedback network, it seems no ringing.
Otherwise, how to explain 0 degree phase instead of 180 for negative feedback system? Also, STB simulation results of phase margin number is wrong.