J. Kissel, T. Vo, H. Yu, K. Kawabe, C. Vorvick, D. VanderHyde

Hang got us through low noise ASC, and we pushed through switching to low noise ETMX and into nominal low noise for the first time in DAYS. Hooorrrrray!! 

As we were reconciling all SDFs for all models against their OBSERVE files (which we were able to do and we blindly accepted everything), and clearing guardian nodes, in order to got to OBSERVING mode, the wind decided to pick up to 40 mph, and we lost lock. 10 minutes at ~75 Mpc...

J. Kissel

A big time suck after lock losses this evening, and last night, has been the FSS going into oscillation after a full IFO lock loss. This has been reported before (45551) but the loop parameters appear to only have been checked in upon last in late November (LHO aLOGs 45547, 45541). 

I'm not enough of an expert to commission the thing better... we've tried only the few superstitious things we know of -- namely, turning on and off the autolocker. We're typically waiting for the time scale it takes for the temperature loop to scan the entire cavity -- 5 to 10 minutes.

The LHO vacuum team spent the week with Pfeiffer staff to install a new screw pump in the mid-Y mechanical room and a temporary turbo station in the mid-Y VEA, not bolted to tube, but blanked off with pressure gauge. We were able to spin up a turbo, backed by scroll pump, while communicating with a live screw pump in order to demonstrate functionality of controls and logic. We even demonstrated a magnetically levitated bearing touching at 7000 rpm: E1800387

We left all three pumps on over the weekend and the air compressors in chiller yard are running to feed the screw pump purge air kit with 45 psi pressure.

J. Kissel, C. Vorvick, H. Yu, T. Vo

Just wanted to post log saying that we've recovered the IFO to the point we had functional yesterday evening. Thus, we're continuing to commission the low-noise ASC system in hopes to get toward nominal low noise.

Many thanks to the cast thousands that helped up recover from last night's computer / I/O chassis failures!

On Nov 10th people did a test of turning on the ITMX bias, and saw no noise added in DARM. 45183  Robert and I did the same test Dec 6th and we did see that there was a noise peak at around 58 Hz, which came and went with an on off test and was pretty similar for positive and negative biases. 45720

I've gone back to the times of the Nov 10th test and made the first attached plot, which shows that there really wasn't a peak around 58Hz at that time.  

It would be interesting to know what might have changed with the ITM ESD or other electronics around the ITMs.  

J. Kissel, K. Riles

Keith and Dave believe they've solved their issues with the hardware injection infrastructure, so I've been given the OK to restore the HARDWARE filter such that it passes all hardware injections through. The CW excitation path now has an open pipe to actuate PCALX and the ETM (though there's currently nothing running.)

TITLE: 12/14 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC

STATE of H1: Commissioning

INCOMING OPERATOR: Cheryl

SHIFT SUMMARY:

16:00 (8:00) Start of shift

16:01 (8:01) Jenne to PSL racks

16:10 (8:10) Jenne back from PSL

16:45 (8:45) Karen leaving Mid-Y

16:53 (8:53) Richard to LVEA -- switching cards in IO chassis

17:07 (9:07) Jeff to cleaning area

17:15 (9:15) Jeff back from cleaning area

17:25 (9:25) Nutsinee to ISCT6 rack

17:27 (9:27) Vanessa to LVEA

17:57 (9:57) Chris to MX

17:57 (9:57) Robert to EX -- look at electronics

18:05 (10:05) Karen to LVEA

18:21 (10:21) Jeff to CER -- reboot STS2 chassis

18:25 (10:25) Jeff back from CER

18:26 (10:26) Jeff to CER -- recenter STS sensor

18:29 (10:29) Karen out of LVEA

19:29 (11:29) Bubba to MY -- inspect vacuum work

21:09 (13:09) Nutsinee to ISCT6

22:35 (14:35) Gerardo, Pfeifer crew to EY, MY, EX, MX (not in VEA’s)

22:51 (14:51) Nutsinee to ISCT6

23:56 (15:56) Dave to MSR, EX, EY -- improve computer labeling

00:00 (16:00) End of shift, handing off to Cheryl

J. Kissel

Once we felt comfortably recovered enough (early this afternoon) to head toward initial alignment, Jenne noticed that ISI ITMX ST2 was moving MUCH more than ITMY. 
After poking around  -- (because of item (2) in my previous recovery aLOG from earlier this morning I was accutely aware that there should be NO sensocr correction in Z going to the ISI; LHO aLOG 45938) -- thus I found that the sensor correction path (which, again, is revamped and quite new) had all paths for ALL ground motion degrees of freedom being pumped into correcting the ST1 CPS without any filtering. That will definitely create some excess motion!

This is a result of a bit more SDF confusion --
(1) On a normal day, the seismic platforms (both HEPI and ISIs) SDF systems are run comparing against the OBSERVE.snap (because there should be no change in the platform state between with the IFO is locked vs. unlocked). However, when a model gets restarted it restores to a safe.snap. 
(2) Further, any time you add new filter banks and channels to a model, you have to *initialize* them in the SDF system (go to the "CHANS_NOT_INIT" menu page, and hit the MON ALL button to not only accept the new values but accept them as monitored.) 

Jim had done a great job at initializing and capturing all the SDF changes in the OBSERVE.snaps before we lost the world this morning, and got *almost* all of the safe.snaps updated, but he missed ISI ITMX. C'est la vie, we're only human.

Thus, when ISI ITMX got restarted this morning the all newly revamped and renamed sensor correction filter banks, came up as they had when originally installed: with all filters OFF, the input and outputs ON, and a gain of 1.0. This is a standard of RCG that has been plaguing us for years. Dave agrees and will file bugzilla request.

I've since restored the sensor correction paths to match ITMY, and made sure to capture those settings in both the OBSERVE.snap and safe.snaps.

Added 50 mL to crystal chiller

Jeff K., TVo, Hang

Before the computer crash last night we tried to went to nominal low noise last night. However, we could not go through the low noise ASC state as we kept encountering some CHARD YAW oscillations at 0.4-0.5 Hz.

Things did:

    1). As the TCS preloading changed from 50 W to 30 W, our SRC ASC locking point (using AS72) needed to be adjust. Thus we went to 20 W and tuned the SRM alignment by hand to optimize the POP buildups and then set the SRM locking point there by phasing all the demodulated 72 MHz signal at DC into the I phase and thus left the Q phase signal which we used as error signal 0. This offset set at 20 W seemed to be working fine at 2 W as well. It was a bit strange that we could make the POP18 signal significantly higher and quieter by moving SRM, and it seemed to indicate the cross-coupling from SRM to BS. Thus we might need to optimize the MICH ASC input matrix to try to better decoupling SRC.

    2). We tried to went through the low noise ASC state by hand and had a lot of troubles with CHARD YAW. First we could not reduce its gain to a UGF of 3 Hz as set previously. If we reduce the gain by this much an oscillation grew at ~ 0.4 - 0.5 Hz. Such an oscillation could be suppressed if we instead made the CH Y UGF 3.7 Hz by putting the DC gain to 1.25.

    3). As a result of higher UGF, we could not engage CH Y FM2 (ELP10) due to lack of phase margin. Instead, we shifted the cutoff freq up to 18 Hz (FM3 in CH Y, ELP18) and reduced the amount of attenuation. This of corse degraded our ASC noise but the loop's margin was already ~ 20 deg with this less aggressive filter.

    4). We tried to check the CH Y blending by dithering a CH Y line @ 8.125 Hz and then matching the refl path and the QPD path gain. Our setting of CH_Y_A (DC refl sensor path) gain 1.4 and CH_Y_B (AC TR QPD path) gain of 0.77 seemed to match the two paths well at 8.125 Hz. However, as indicated by the CH Y blending study (LHO:45922), the refl senors' response seemed to behave well at >~ 8 Hz. At lower frequency, the CH_Y OLTF measured using only refl sensors experienced mysterious phase delay relative to the OLTF measured with the QPDs. This might be due to the cross-couplings of many dofs (CH, PRC1/2, INP1...) in the refl WFS. Thus our way of matching the responses at 8 Hz might not guarantee the good matching between REFL WFS and TR QPD at 0.4 Hz. In fact, by increasing the ratio between the CH_Y_A to CH_Y_B we could shift the CH_Y oscillation freq from 0.4 Hz to 0.5 Hz, indicating the instability was likely to be related to the blending.

    5). While higher BW CH_Y gain suppressed the oscillation initially, we could see this 0.4-0.5 Hz thing come and go at ~ 10 min time scale, and eventually we lost lock at ~ hour timescale. Thus our current CH Y loop was very close to the stability margin...

[Jenne, Keita]

After the computers were brought back up, the optics, and especially the input PZT on the PSL table, were brought back to old saved values for their sliders.  Even though the PMC was locked, with the PZT off so far, the IMC-PWR_IN_OUT PD was reporting 0.0W injected to the vacuum. 

I restored optics and especially the PZT to slider values from yesterday during a lock, and the IMC-PWR_IN_OUT read a sensible value of 1.8W.  However, we were still not seeing any image on IMC REFL despite MC1 definitely being back in its position. I went and used the input shutter at it's midway position to attempt to see if we were getting beam injected to the vacuum, and we definitely were not. 

Keita and I looked at the cameras that Cheryl has recently installed on the PSL table, and trended the centroid values for the last day or so.  In particular, IO Gige 2's image was different from where Keita remembered, so we trended the centroid values for that camera.  We moved the PZT sliders until the centroid values for IO Gige 2 were back to where they had been, and we started seeing light on IMC REFL camera.  Hooray for the IO path pointing reference cameras! 

After this, IMC recovery was easy and normal, just moving the optics a teeny bit more until the IMC locked on the 00 mode, and letting the guardian and WFS take over.  I did increase the WFS master gain up to 0.2 (5x the nominal value of 0.04), and then put it back after we were closer to converged. 

[Follow up from alog45846] After waiting for thermal related things stabilize overnight, I went out and took another measurement of CLF launch and SHG GR (first diode looking at green from the SHG). CLF is now much better intensity noise wise. Not sure why CLF RIN wouldn't just fall on Mephisto level (IR DCPD) at <1kHz. SHG_GR_DCPD however, isn't doing so well. Note that data from Dec 6 were from the before PBS cube was installed.

On Wednesday Daniel closed the AOM loop (alog45883) and here's pump intensity from SHG launch before and after. Much improved.

Sadly, TTFSS fastmon doesn't seem to improve. These were in-loop measurement (same loop configuration).

Intensity loop has a UGF at ~15kHz.

And here's how we lock it (more details on alog45883). Guardian for this is not in yet.

LO IMON looks a bit better (locked with PSL LO), but not significantly better. The rms is probably dominated by peaky features that didn't exist on October 12 (alog44575). Turning the half waveplate on CLF path helps with the peaks. One thing to test is to shoot in max CLF power and see if the features disappear.

Note to self, more to do:

- LO IMON compare between before and after PBS was installed

- Look at PSL ref cav trans noise

- Project SHG launch intensity noise and see if it compares with the measured IMON. This should tell us whether or not the access noise comes from intensity noise. 

- Compare OPO length noise then and now

- see squeeze with PSL LO to see if we have the same noise that we saw in DARM <-- Will probably do that one first.

Richard and I are power cycling the system now.

Last night we had a lot of problems with fast ALS glitches. Today Richard and co went around and inspected the fibers, tightened connections, and reduced strain in some areas. We've had no problems with this, but since its so tough to diagnose, it would be good to develop some code to find this glitching issue and report it on the ALS screen as a kind of Fault condition. The next time that the wind blows some sand on the fiber or a coyote bumps the laser safety tag, the glitching might come back.

In the second-trend plot, you can see that there are a lot of downward dips in the Green Y transmission, but not the X. We think that this is a fast phase shift in the fiber which is recovered by the FSS at the end.

In the DTT time series plot:

lower left: 3 minutes of glitches

upper left: time series AC coupled at 80 Hz

upper/lower right: zoomed in time series

From the zoomed in time series, you can see that the glitch has a time scale of ~2-5 ms.

It would be a very useful thing to make a tool that could run while the arms are locked on green and report on the number of such glitches in each arm. As a second step, being able to find the glitch in the ALS-X/Y_FIBR_ERR_OUT_DQ would be good. If we assume the glitch corresponds to 1/2 of the green linewidth...

By comparison, the free running NPRO noise should have an RMS of a few kHz at this timescale, so it may not be possible to see it with simple bandpass filtering.

Just like the work I did a couple weeks ago on HAM 4, today I tried to reduce cavity lengths by sending gnd z to HAM3 X. Seems like it works, although the high microseism has made getting good transfer functions difficult today. First attached plot is one of the plots I used to design the filter, but I literally just used the same zpk from the HAM4 z-x filter and matched the gain to the new transfer functions. The solid blue (or purple? I'm not sure) is the  gnd Z to MCF transfer function over the HAM3 x drive to MCF transfer function. The red line is the filter I'm using at HAM4, the gold line is the filter for HAM3. The dashed line is the expected reduction. I tried several times to get the driven transfer function, looking at HAM3 x drive to IMC-F, PRCL, PRM and MC2 M3 length drives, and the filter design ended up being the same. 

The second plot shows an on/off test with only PRMI with no arms. In this state PRCL and MC both see an improvement around the microseism. The dashed lines are with the Z-X sensor correction on, solid lines are with it off. Brown and green are the ground, which didn't change during the test. PRCL went from pink to orange (about 3x lower), MCF went from red to blue (almost 4x lower). Rana doesn't like my colors, and he has a calibrated spectra, so I might post that in a bit.

If people suspect this is causing problems it can be turned off like the SRCL sensor correction at the command line with:

caput H1:ISI-HAM3_SENSCOR_GND_STS_X_WNR_GAIN 0