Logan, Michael
We found the source of the problem regarding beam size. Before the PBS there is a telescope consisting of two 100mm lenses but they were quite poorly centered. We took them out and adjusted the reference alignment to lie at 74mm height above the usual designated set of screw holes on the breadboard, then put in the lenses, both times centering them by recovering the reference alignment, approximately. After I fixed them, the beam waist moved a bit closer towards the source than before, but we could see from measurements of the beam divergence that now we have the beam diameter changing by approximately 150 um per screw hole of z displacement, going to and from the beam waist. The beam waist diameter is approximately 250-300 um, we didn't measure super precisely but it seems enough for the purpose of using modulators to the OPO.
Basically, always set a reference target before inserting optical components that require precise centering.
Next step is to construct the beam path to test the OPO. The beam size as is is probably fine for modulators but the beam size increases when going to the Korean Faraday isolator experiment. Guessing from Gaussianbeam.exe rough calculations and looking at the size of the sensor card, we can also get about 1.5mm diameter collimated beam from small fiddling of the mode matching telescope. I should check again what is the clear aperture of Korean FI but I suppose they can always put a lens before their periscope if they really need a smaller beam.
I originally designed some beam size parameters to put a 2.3mm diameter beam into the fiber SHG. However, the Taiwan New Years visitors used a completely different setup to what I wrote down. In some sense my design was not very good (I left the bare minimum of space for QWP/HWP/Faraday/Mode matching after the laser, for no good reason), although I asked a few people about it and no-one pointed out these issues. The 2.3mm beam diameter was a parameter given to me by Chien-Ming Wu for the fiber input, and given he used this type of fiber SHG before I just trusted him on the number. But it seems the people who set it up had other ideas. I need to check.
I did some check based on fiber equations that you can find on Thorlabs or Newport.
The fiber to the SHG is PM980 (polarization maintaining 980nm in-fiber wavelength) with a mode field diameter 6.6 um. This gives f/D = pi*MFD/(4 lambda) = 4.8. Thorlabs FC/ACP connector fiber collimators are sold with f = 2.0, 4.6, 7.5, 11.0, 15.3, 18.4mm focal lengths, giving input beam diameters D = 0.41, 0.94, 1.53, 2.26, 3.14, 3.8 mm. The clear apertures are 2.0, 4.9, 4.5, 4.4, 5.0, 5.5 mm. In retrospect, maybe I should have picked the f = 4.6mm. But it's maybe not super important. exp(-2*(r/w0)^2) = 0.00051 (510 ppm diffraction loss) for the current coupler. r > 2.7 w0 gives < 1 ppm diffraction loss.