According to the beam parameter of entry 1046 and the telescope design from yuefan(entry 1564), I simulated the beam situation again. But it seems not the same with yuefan simulation. But anyway I checked the robust for the NF telescope. The FF telescope is quite different from yuefan design, I will confirm with yuefan what makes this difference. And do the robustness check soon.

Difference one: the distance from L1 to NF quadrant is 0.535 (in my case) and 0.65 (in yuefan case)

Difference two: I used the same distance with yuefan design, but the gouy phase is 87.277(in my case) and 90.2(in yuefan case)

Notice: from the simulation, the beam size for FF quadrant (if I use yuefan design) is 2.179mm (radius). This is 2 times larger than the requirement. We should also check this.

I confirmed with yuefan that we are using the same initial beam parameter. Also, yuefan told me we have a tolerance of several degrees for the gouy phase. So I think with our hand we can have a precision within 1 cm, and in this case, it should be fine if we don't use the first lens on the rail. However, if we consider the distance estimation error from lens1 to NF quadrant PD, we may have another 1cm of error. In this case, maybe we should use rail.

Or we can just measure the beam parameter with beam profiler after BS and find out where is the beam waist. Then put quadrant PD in the position of the waist.

Or we can tell if we are seeing the decoupled motion of cavity by looking at the output signal of quadrant PD. Then we can decide the position of quadrant PD.