LDO pz analysis

hi

i am simulating the open loop response of an LDO regulator using 2 methods and i expected identical result.

1st, i place a iprobe in series with the feedback signal that went back to the error amplifer reference. this gave me a gain and phase plot. from it i could roughly tell where the pole and zeros are.
all is good. for example, dominate pole is at 1KHz.

2nd, with the same circuit, i manually break the loop (at the same place where i placed the iprobe in 1st case) using an inductor and inject an AC source into the left of it. then run pz analysis.
i expect pz analysis to return the numerical location of the pole and zero. HOWEVER, the result i got had pole/zero location that was way small. about 1E-6 Hz.
all dc bias is preserved.

questions:

1. am I simulating the open loop response in 2nd case correctly?

2. if so, why does the 2nd case result is so odd. i use simple RC circuit to test this setup and the result is correct.

thanks please help.

Frank