I measured the beam jitter induced by AOM frequency scan by using beam profiler.
I set the requirement of beam jitter as 0.05mrad/MHz .
This value corresponds to 0.1mm shift at folded cavity input mirror though there exists 1MHz frequency shift (I assumed that the distance from AOM to folded cavity is 2m).
This beam jitter is small enough from the view point of alignment during the DC locking.

The measured beam shifts at 0.5m distance from the AOM was about 0.2mm with 15MHz frequency shift.
So the beam jitter is about 30urad/MHz with current configuration.
This seems small enough for the moment.

The next step is implement another double-pass AOM and measure the beam jitter.
Then construct ISS using AOMs.