[Nergis, Haocun]
After locking the OPO, we were able to get a locked SEED beam on the SQZT6 for homodyne work.
Main steps:
We did the visibility maximization process with PD2, getting 96%, but it was only 92% for PD1. Then we found that when the homodyne is balancing for the LO beam, the two diodes read very different values with the SEED beam. I checked the polarization of the two beams, and both of them are well s-polarized.
This led me to go back and recheck the diodes of the homodyne, finding that PD2's reading is only about half of value of PD1 when shining the same beam on them. Also, Daniel and I inspected with an IR viewer, and found that there is a bright reflection from PD2. We probably need a new diode for PD2, but this means the mode matching is good enough, and we can still use the current set up to work on the CLF and LO loop.
Some changes of the mode matching solution: (I will update the diagram later soon.)
This is very strange to me. Do you get the large reflection with the LO or seed beam? The reflection suggests the issue is with the diode, but this suggests something strange with the BS. If the diode has a large reflection, then perhaps to balance the LO, the BS is tilted strongly and the seed has the opposite R/T and so looks highly unbalanced. It sounds like the diode is so bad that you shouldn't be able to get the two diodes balanced from the BS though. If bad diode + unbalanced BS = balanced photocurrent were the case, then you shouldn't be able to get such good visibility in one of the diodes though I would think. Anyway, we can ship diodes, but I'm just wondering if somehow the BS or beamsize on the diodes could cause what you describe. I recall that the diodes aren't always perfectly flush with the board and so can hit at differing angles, which can also affect the reflection.
SQZT6 Table Layout Attached
The homodyne balancing problem was solved. Mainly by adjusting the beam size and angle shining on the diodes.