In alog 51366 I showed that the BS yaw is repeatedly wandering by about 1.5urad, and in looking for the source of that wander, I've discovered that the PSL Ante-Room temperature (blue) predicts this BS yaw behavior (orange).  Of course, it's the diurnal temperature changes, but what is more interesting is that I can see the beam being steered on the PSL, with Camera 09 (red), that is on the IMC_IN beam.  I can also see that there is often an increase in noise on the FSS TRANS PD (pink) signal during these temperature swings, though the increased noise is most easily seen with the FSS REFL PD (gray).  The PMC TEMP (brown) appears to be effected by the temperature swing.

Of course the question of why the Ante-Room?  I started to recognize the connection between temperature and BS yaw, and then looked for signals in the PSL that would show if the beam was being steering in the PSL.  This led to Camera 9, and from there, I talked with Jason, and he suggested the PMC heater, and the PMC TEMP turned out to have the best signal to use.  I also looked at the FSS, and both TRANS PD and REFL PD signals showed the effect of temperature changes.

Having found evidence of the beam being steered from the PSL, I looked for the best temperature sensors, and the ACN (AC North, a sensor in the ceiling) was a very good predictor, but not 100%, the ACS (AC south) was also only sometimes changing when the beam was being steered, and the PSL table temperature sensors are basically flat, so they are not showing the effect of the temperature swing. The signal that led me to the Ante-Room temperature was the dust monitor temperature, from the dust monitor under the PSL table, which has a resolution of 1F, but was often showing that level of change on a diurnal cycle.

This behavior started around April, so I'm not sure if that was the change in weather, or something else. 

I was concerned that the PSL AC units were "acting up" so flipped the breakers off, and did not see a change (which is actually good, but doesn’t explain the IMC_IN and BS yaw changes).

Given how well the PSL Ante-Room temperature predicts the IMC_IN beam and BS yaw changes, and the contradicting PSL table temperatures, it's clear that either there's something we have not yet identified as influencing the thermal stability of the PSL, or something we have identified has changed, and we're seeing the effects now.

There are 12 plots from April to August in the attached PDF.  Recently, there have been a number of lock losses during the fastest change in BS yaw during its wander, but it does not always break the lock, so I've included plots that show both.

Plot times and FSS TRANS PD value:

• 21-Apr,   00:56UTC                 FSS TRANS PD=3.4                                                    

• 5-May,   23:57UTC                  FSS TRANS PD=3.9                                  

• 21-May,  00:35UTC                  FSS TRANS PD=2.8                                                    

• 10-Jun,  00:56UTC                  FSS TRANS PD=5.2                                  

• 29-Jun,  01:28UTC                  FSS TRANS PD=4.4                                  

• 5-Aug,   01:50UTC                  FSS TRANS PD=4.1                                  

• 6-Aug,   00:43UTC                  FSS TRANS PD=4.1                                  

• 10-Aug,  02:11UTC                  FSS TRANS PD=2.7                                  

• 12-Aug,  23:32UTC (X = -440000s)   FSS TRANS PD=2.1                                  

• 14-Aug,  22:59UTC (X = -280000s)   FSS TRANS PD=2.3                                  

• 18-Aug,  01:45UTC                  FSS TRANS PD=4.1                                  

• 19-Aug,  23:32UTC                  FSS TRANS PD=4.4                

I've also separately uploaded two of the plots from the pdf, which are recent examples, one where the H1 lock survived, and one where it did not.