Regarding open loop gain simulation of single ended op-amp

[whompah]

Hi all, I am new to analog IC design. Currently I am doing a simple 2 stage op-amp to see its open loop gain behaves. I found out that there are many ways of simulating it. So I tried few of them as shown in the 5 figures below.
1) inject AC=1 directly into one of the input
2) use large R and C values feedback from output back to -ve input (2 different ways of injecting AC=1)
3) use large L and C values feedback from output back to -ve input (2 different ways of injecting AC=1)

Based on all the simulation results, 2) and 3) gives similar result in terms of gain dB, UGF as well as "phase margin". All of them gave me 61dB gain, 77MHz UGF (forget about phase margin bcoz it is very bad)

However for 1), i get slightly different result of 60dB, 75MHz UGF
Can anyone enlighten me which method is more "robust" and more widely used ?
Thank you

 

[FvM]

All DC feedback simulation setups can be expected to give identical results - if you manage to make the feedback act for DC only and the operation points are identical.

The most "popular" error in open loop setups is to get a wrong operation point, particularly with high gain amplifiers.

 

[LvW]

As a straight-forward method, the most "logical" way to simulate the open loop gain of an opamp is
* first to determine the dc input offset voltage (dc transfer analysis), and
* secondly, to perform an ac analysis for the offset-corrected device (without any feedback).

 

[whompah]

All DC feedback simulation setups can be expected to give identical results - if you manage to make the feedback act for DC only and the operation points are identical.

The most "popular" error in open loop setups is to get a wrong operation point, particularly with high gain amplifiers.

I get what you mean, based on your advice and then by referring to my diagram again, the first diagram which does not have any RC/ LC "loop" has different DC operating point compared to the rest.