Unclear on stepped FRA Analysis

[Lloyd]

I’m clearly missing something while trying to use FRA analysis. I made an extremely simple test feedback circuit to try to measure open-loop gain:

When I use AC analysis I get exactly what I would expect, an integrator with ~10 dB of gain at 100 Hz and 90 degrees phase shift:

Bode analysis is the same:

But when I use stepped FRA analysis I get the following:

Which seems to be accurate to a certain frequency, then it diverges from the correct response.

I’m sure I’m overlooking something in the setup or some understanding of the limitations? Test file attached.

fra integrator test.qsch (5.9 KB)

[KSKelvin]

Check time domain simulation data, at fmeas=100k, .tran 40/fmeas is not enough to reach steady state. Here change cycles to 400.

fra integrator test (KSK).qsch (5.2 KB)

[Lloyd]

Of course, thanks again Kelvin!