I went back to the end station today to look into the beam quality.  At Keita's suggestion I put a retro reflector in to look at the beam without having to worry about what happens in the chamber.  With the retroreflector after the last steering mirror on the table, I set up the nanoscan in the path to WFS B and the thorlabs beam scan in the path to the HWS, whic doesn't double pass the Faraday.  Using the two profiles, I tried to move the Faraday around to find a position where both the beam in the HWS path  (the first pass beam) and the beam in the WFS B path (double passed) were good.  Although there were many positions that gave a good beam after the single pass, I didn't find anything that gave a nice gaussian profile in the rejected path.  I swapped the Faraday with a simliar one, and the beam quality was immediately better.  With a small adjustment I found beams that look guassian in both the WFS path (measured at 3 different positions) and the HWS path, using the retroreflector.  Reflecting off the ETM the beam is a little bit worse, but still bassically a guassian beam.