Our South and North cavities have different error signals Vrms when locked.  Why?

The North path error signal is around 25 mVrms when locked. The South path error signal is 10 times worse, with around 260 mVrms when locked.

To diagnose the source of the South path issue, we switched the North PD to be the South PD, and relocked the South.  The error signal noise was the same.  Our excess South path noise is not sensor noise.

Now that we had a known good PD on the South path, we switched the North FSS to be the South FSS and relocked again.  The error signal noise was the same.  Our excess South path noise is not FSS control noise.

This means the excess South noise must be optical.  Apparently small intensity fluctuations on the REFL PDs can be created by poorly tuned polarizations.  We've gone around the lab and tuned almost all the quarter and half wave plates to minimize amplitude modulation, but still we see excess South path noise.

We may try switching the resonant EOMs, which are particularly senstive to polarization changes due to temperature fluctuations.  The North path has the Broadband EOM that Andrew made a driver for, while the South path has the resonant EOM designed by some company.

We really want to avoid switching the lasers, since realignment and acquistion will be hard, and the lasers are very different.

If the noise in the South path is still the same after all this, we might think more carefully about the polarizations going in and out of all our EOMs, AOMs, quarter and half wave plates.  Antonio wants us to rebuild some intensity stabilization servo but I am not convinced we need this yet.