We were looking for optics which cause peaks in beat noise. Periscopes and mirror infront of RCAV are more sensitive

than other components. Also, cavity's suspension (pendulum) might be needed to be modified.

Mechanical peaks from optical components can be up/down converted and cause extra noise in 

beat measurement, thus we also need to eliminate them. By listening to the PLL feedback signal, we

can tap each optic component to verify which one is sensitive to acoustic noise. So far there are

periscopes, mirror in front of RCAV, and cavity suspension that are quite sensitive to acoustic noise

compared to the others. For the suspension, we might have to mount the cavity on a block instead of

hang it on springs.

After we worked around the table, RCAV temp jumped again. By kicking the table, we can bring RCAV temp

back to what it should be (fig1 top left panel.) ACAV and RCAV RCTRANSPD also changed as well (two right panels.) 

This might be electronics jump, or that the servo was oscillating before, and when the

kick brought it back to lock again. At least ACAV temp read out is quite immune to these jumps.