Camilla and Corey have replaced the green mirror in reflection that was scratched (this was an optic reused from O3). This afternoon with the new optic in place I returned to finshing the G path, in reflection off the OP towards SQZT7.
The first attachment shows the measurement with no lenses between the OPO and the beam profiler. The vertical waist is 175um radius 278mm before G:M2 (which should be about 65mm from OPO M1). The horizontal waist is 204 um 201 mm before G:M2 (meaning 142 mm from OPO M1).
After placing the +150mm ROC lens about 50mm from G:M2, I remeasured the beam shown in second attachment. The resutling vertical waist of 323um is 364mm downstream of G:M2; horizontal waist of 263um is 248mm downstream of G:M2.
I didn't take care to block the OPO while making these measurements, which would have been a good idea to prevent passing through a resonance from disturbing the measurement. These measurents aren't in agreement with T1900649-v8 figure 22.
Ok, I looked back at my mode matching code. In that case it was calculating along an incorrect path for an obvious reason that I missed given all the many paths setup in that document. I fixed the issue and attached is a new Fig 22 Fig22_improved.png. It agrees much better with your measurements and indicates that you can keep G:L3 where it is. I also checked how the matching with the SK path changes. It also improves quite a bit, since the new PUMP-REFL beam is better collimated. The matching with the FCGS -> SK path beam is 91\%.
With this fired up, I'm looking into my OPO model to also look at the A: path measurements. I'm getting better matching to the A:L1-A:l3 measurements after some adjustments. I'll post more later, since I'll probably have to adjust the layout a bit to ensure good matching.
