Yesterday we went hunting for residual AM from the PM phase modulators. This is due to misalignment of the polarization going into the RF phase EOMs and is a potential source of lock point offset/noise in the PDH locks of the FSS.
I found that both North and Souths paths had quite a bit of residual AM. I dropped the locks on both cavities and offset the laser frequency to well below the TEM00 mode. Craig and I also installed a flipper on the north path (just before the periscopes) to return light without interacting with the cavity at all. We can see the AM by looking at the 14.75 MHz beat coming out of the RF port of each PD.
Noise power from both detectors was in excess of 85 dBm/Hz (0.4 mV) on both detectors. Maybe I should have just used units of dBm here, but this is the setting I chose at the time: we just need to see to minimize for now.
To reduce the AM, I put a PBS in before the EOM and walked the lambda/2 and lambda/4 to minimize transmitted light. This gave maximum linearization of the signal. To minimize the AM the PBS was removed and the beat on the reflection PDs (with lasers off resonance) was observed while only the lambda/2 was adjusted.
One problem with this optimization is that the output of the EOM then goes through a FI to separate linear ingoing/outgoing fields. I found that by walking the pre EOM lambda/2 WP and correcting the reflected cavity DC level with the lambda/2 before the FI it was possible to find a sweet spot where power through the FI. The adjustments of WP angle only needs to be very small, but it can be a fine tune to get it very close to totally minimized.
North path was reduced to level of -96 dBm/Hz above a SA dark noise background of -130 dBm/Hz. South path was reduced to a residual AM of -97.5 dBm/Hz on a dark noise background of -122 dBm/Hz. It should be noted that the noise floors of the two detectors are about an order of magnitude different looking on the Aglient SA. Something we need to check is the relative noise of the two 14.75 PDs as they are not the same build.
Optimizing polarization through BBEOMs
To ensure that the BBEOMs used for very fast actuation were aligned properly I put a ±10 V @ 1234 Hz sine wave into them and looked at the reflected BN out of the RF of the PDs. It was very fast to align the BBEOM polarization to reduce to below the noise floor of the Aglient.
Changes to North optical path
I found that there was no space in front of the BBEOM, AEOM and resonant resonant/(amplified) EOM in the north path to put a PBS. It is not super critical that these modulators sit exactly at the waist of the beam so I moved them back 3 cm so that a PBS on a narrow mount can be fitted into the gap for diagnostic purposes. I also temporarily removed a lambda/4 wave plate from after the AEOM as we are not using this for any power modulation at the moment.
Powers incident of cavities
After making various adjustments for the two paths I set the power going into the north path refcav to 1.27 mW. For the south path the power incident was 1.22 mW