Loop filter issue with practical Amplifier

I indeed found that the PLL loop itself take care of the biasing part. I have a model to analyze the stability, in that model I have tested the 4th order active LF with a VCVS as amplifier. After stabilizing the loop, I put my LF in Transient model and it worked fine. But how to do the stability analysis with practical amplifier ? As it wont have any kind of gain in AC simulation, due to initial case (the op-amp will be in linear). Also in AC simulation I cant set any initial fixed value in op-amp. I have attached my ckt...