We restarted aligning H path this week. It turns out that two 150mm ROC lenses need to be at the same location for best matching, which was impossible, so we just placed them as close as we can (~28mm apart). We don't seem to have 75mm ROC lens.
Mode overlap between measured and designed was OK (91% for PIT, 93% for YAW) if we pretend that the green beam leaving A:M3 is Gaussian, and we cannot do much better than this given that we L1 and L2 cannot be closer and we have zero freedom to move L3.
We're not quite satisfied with the beam shape in YAW, it seems as if it's clipped somewhere but we didn't find anything obvious. We need to measure the profile of the beam coming out of the fiber (i.e. transmission of H:BS1). We're also using 1064nm optics for relaying green light on the optics table downstream of A:M3, and it might help to temporarily change them with green mirrors though my hunch is that it's not going to make a large impact.
Friday 4th June. I replaced the 1064nm mirror with a 532nm mirror (NewFocus 5105). The 1064nm mirror is in a labeled case on the table.
Now you cannot see the beam though the back of the mirror (photo) but it has not changed the strange beam shape in yaw. Beam shape normally and with beam scanner at 45deg attached.
I retook the beam scan data by turning the light off to reduce the size of the shoulders which may have reduced the measured waist size. Data attached.