Craig Dan Danny Georgia + Jeff B and Ed

Over the past week we've taken a couple of steps with the ring heaters, trying to find a more optimal TCS configuration for 35 W input power. Craig and Dan previously set up the AWG_LINES guardian to inject frequency noise, intensity noise, and 9MHz RIN lines. I wrote a quick matlab script to demodulate these lines, and have made some plots of how these couplings change over time with the ring heater steps. This is building on an ipython notebook Dan had already written, but it cuts the data into chunks so it doesn't require the cluster to run. This work is ongoing but I'm posting what we've got so far.

It's a bit loose for now, but for interested TCS'ers, script lives here:

     /ligo/home/georgia.mansell/TCS/AwgLineDemod/plotTCSDemodLines.m

20190325 - Common decrease in ITM RH power

On Monday Jenne decreased the ITM ring heaters in common by ~250 mW. She saw:

• improvements in POP_RF18, POP_RF90, circulating power, and power recycling gain
• a decrease in the cavity pole and Kappa_C
• ~5 Mpc decrease in the BNS range

The first attachment shows an ndscope of the diagnostics, as well as the ring heater powers. The second attachment shows the demodulated lines, normalised to t ~ 700 s (sorry it's a bit dodgey for now). The time axis starts just after the first step down, at roughly the same time as the cursor in the first attachment (the lines were not on before this time), the lower plot shows the ring heater set_upper_power, which is controled by Danny's ring heater guardian. 

The interferometer was well thermalised before this change, having been locked for 19 hours. Things to note: the coupling of the 9 MHz line, and high-frequency frequency noise appears to grow with this ring heater step. The high-frequency intensity noise coupling drops, though seems to turn back up towards the end - maybe we overshot some minimum coupling? The ring heater time constant is very slow. The frequency and intensity lines in the bucket seem unaffected.

20190325 - Differential decrease in ITM RH power

Later the same night, we tried a slightly differential step down in the ring heaters, lowering ITMX by ~0.1W and ITMY by ~0.26W. We see similar behaviour in our diagnostics (third attachment), increase in RF18 and circulating power, decrease in the f_c and range. Note that this step was taken sooner after locking, and that the interferometer was still thermalising. The fourth attachment shows the demodulated lines. Changes before the step at t~7500 are due to the thermalisation of the interferometer. After the ring heater step, the high frequency intensity noise again takes a dip (with a pretty fast time constant), while high-frequency frequency noise and RF9 increase. Craig noted that the high frequency line and the 9MHz RIN line (at 72.3 Hz) are moving together. With this step intensity noise in the bucket seems to rise, and frequency noise in the bucket seems improved.

20190323 - Common increase in ITM RH power, followed by decrease

On Friday night Danny and Ed took a 0.1 W common increase in the ring heater power. Note earlier that night we had seen that a 0.5 W increase on ITMY only caused POP_RF18 to plummet potentially causing a lockloss. During this step the interferometer was still thermalising, see for example RF90 (purple trace in fifth attachment) dropping before the ring heater step. I originally got excited to see the range increase but probably that was due to some squeezer shenanigans happening at the same time? With the increase in power RF18 looks like it drops a little, but other diagnostics don't show much. I think the demodulated lines (sixth attachment) are dominated by the thermalising IFO.

Tonight (@ 10:57:43 UTC) we have tried increasing the power on ITMX ring heater only, we requested 0.8 W from the filter input, where nominally it is 0.5 W. If future operators or commissioners would like to return this value to nominal after lockloss (eg if ASC is problematic), this can be done in the command line:

     caput H1:TCS-ITMX_RH_INVERSE_FILTER_IN 0.5