-
We installed a 2nd POP camera, a polarizing beam splitter and 1/2 wave plate to have access to two separate video feeds for s-pol and p-pol. Their power should be more or less comparable (up to the effect of upstream beam splitters).
In the process we also decided to remove the green-stop filter - this way the PPOL camera can double up a green reflected camera.
Note that the polarizing beam splitter only forks for IR, so the IR_PPOL camera gets basically all green.
We also decided to use this green reflected camera to compare the reflected beam spot from the x-arm if we inject p-pol instead of s-pol.
The beam spots look strikingly different: compare fig 1 (s-pol, after we returned) with fig 2 (p-pol, same expose) and fig 3 (p-pol, but lower expose).
The s-pol image did look slightly different before and after our polarization moves - we don't really know why - maybe we slightly moved the green beam position on ITMX.
Figure 4 shows the green single-bounce reflection before we started changing the polarization.
With that camera in place, we locked PRX.
Fig 5 shows the now much cleaner S-pol camera, as well as the dirty p-pol camera. There are two distinct modes contributing to the p-pol contamination in the recycling cavity - the "upper-left" blob and the "lower L".
Note that the mixture changes significantly with a PRX error point offset - figure 6 shows the same situation with a -600cts PRX error point offset. Note that this also coincided with a significant drop in the total circulating p-pol power (as temporarily monitored by K1:ALS-X_ARM_INPUT_INMON). The fact that the p-pol modes are not co-resonant suggests that changing the Gouy phase of the PRC might further help us in changing the overall PRC loss.
Finally, we convinced ourself that the conjugation effect of over-coupled arm cavities should significantly reduce the losses due to p-polarization generation once the arms are locked on the carrier - at least for the carrier losses and p-polarization light that is not in the fundamental mode, that is. That conjugation effect is known to make the carrier insensitive to ITM lenses.
Here is the updated table layout. (JGWdoc 9623)
