TITLE: 09/05 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 113Mpc
OUTGOING OPERATOR: TJ
CURRENT ENVIRONMENT:
    SEI_CONF state: WINDY
    Wind: 10mph Gusts, 7mph 5min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.23 μm/s
QUICK SUMMARY:

• H1 in Claibrations when I arrived
• 23:29:46 - H1 in Observe
• unfortunately some violins had rung up quite high on ETMX, with modes 11, 13, and 16 all above 3.5RMS and rising
• 23:55:42 - H1 in Commissioning, I set violin mode gains
• plot atached shows which violin peaks were rung up (magenta), and current (red)
• there are a number of peaks which are not gone from the spectra

M. Ball, S. Dwyer, J. Kissel

We were able to collect the entire suite of calibration measurements (sensing function suite and each of the three actuation stage suites) in 
    - the current nominal configuration (i.e. Spot moved back to "July" positions, with a digitally requested SRCL offset; see LHO aLOGs 51592 and 51440), and 
    - in the proposed new actuator configuration (i.e. PUM L2A filters back on, and with the UIM longitudinal control boosted below 0.3 Hz; see 51717 and 51709).

Recall that we plan to update the front-end calibration next week Monday (Sep 9 2019), and in doing so, adopt the newest, latest, bestest configuration tested today and thus create a new epoch of O3. 
These measurements will inform and complement that model and its uncertainty nicely. 
Also, FYI, that update will resolve the discrepancy between front-end and GDS computed BNS ranges on the summary pages (not the reason we're doing it, but a nice side benefit).

Analysis and more details to come!

Here're the data:
Sensing Function measurements:
/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/FullIFOSensingTFs/
Nominal Config:
    2019-09-04_H1_PCALY2DARMTF_BB_3min.xml    | Newly tuned excitation to better facilitate studies like LHO aLOG 51654
    2019-09-04_H1_NominalConfig_DARM_OLGTF_LF_SS_5to1100Hz_15min.xml
    2019-09-04_H1_NominalConfig_PCALY2DARMTF_LF_SS_5t1100Hz_10min.xml

    2019-09-04_H1_L2ADecoupUIMBoost_DARM_OLGTF_LF_SS_5to1100Hz_15min.xml    
    2019-09-04_H1_L2ADecoupUIMBoost_PCALY2DARMTF_LF_SS_5t1100Hz_10min.xml

Actuator Measurements:
/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/FullIFOActuationTFs/
Nominal Config:
    2019-09-04_H1SUSETMX_L1_iEXC2DARM_10min.xml
    2019-09-04_H1SUSETMX_L1_PCAL2DARM_8min.xml
    2019-09-04_H1SUSETMX_L2_iEXC2DARM_12min.xml
    2019-09-04_H1SUSETMX_L2_PCAL2DARM_6min.xml
    2019-09-04_H1SUSETMX_L3_iEXC2DARM_12min.xml
    2019-09-04_H1SUSETMX_L3_PCAL2DARM_6min.xml

With proposed PUM L2A decoupling ON and UIM Boost ON
    2019-09-04_L2ADecoupUIMBoost_H1SUSETMX_L1_iEXC2DARM_12min.xml
    2019-09-04_L2ADecoupUIMBoost_H1SUSETMX_L1_PCAL2DARM_8min.xml
    2019-09-04_L2ADecoupUIMBoost_H1SUSETMX_L2_iEXC2DARM_12min.xml
    2019-09-04_L2ADecoupUIMBoost_H1SUSETMX_L2_PCAL2DARM_6min.xml
    2019-09-04_L2ADecoupUIMBoost_H1SUSETMX_L3_iEXC2DARM_12min.xml
    2019-09-04_L2ADecoupUIMBoost_H1SUSETMX_L3_PCAL2DARM_6min.xml

J. Kissel

At the tail end of today's calibration suite, I gathered a bit of time with PCALY driving near the 60 Hz power main line in the DARM ASD, in the region where we have been regularly performing offline linear and non-linear subtraction. The idea being that, if we apply the subtraction techniques, and exclude the PCALY excition channel from the list of subtracted witnesses, then we want to see the surrounding noise be subtracted away without impacting PCAL. No expects this to go awry, but it's nice gold-plated proof that the subtraction is doing good and no bad.

There are two times in this study: first with the PCAL lines just outside the typical non-linear junk we typically see surrounding 60 Hz, and the second with the PCAL lines right on top of the junk. 
During these times, the IFO was otherwise in its nominal configuration, with no other tests going on. Indeed, we didn't have to modify any existing PCALY calibration lines in order to ask this extra amplitude of the PCAL's OFS AOM driver -- i.e. these extra lines didn't saturate the optical follower servo.

(All times are with the date 2019-Sep-04, and reported in HH:MM:SS)
    Time of Test                 Duration      PCALY f1   PCALY f2      Amplitude
                  (22:48:36 UTC, done with CAL measurements, request back to NOMINAL_LOW_NOISE)
                  (~22:50 to ~22:54 UTC tuning frequencies and amplitudes)
    22:55:00 to 23:10:00 UTC       15 mins       58.91 Hz   60.95 Hz      200 ct
                  (switched line frequencies at 23:10:13 UTC)
    23:11:00 to 23:28:00 UTC       17 mins       59.69 Hz   60.35 Hz      200 ct
                  (lines turned off at 23:28:34 UTC, Observing by 23:29:46 UTC)

The amplitude was chosen such that the amplitude of the PCAL lines roughly matched the amplitude of the 60 Hz line in DELTAL_EXTERNAL.
In retrospect, I guess to really test the algorithms, I should have chosen amplitudes roughly at the amplitude of the junk, but... c'est le vie.
The test is trivial, so easily repeatable.

To subtract these excitations from DARM / DELTAL_EXTERNAL / GDS-CALIB_STRAIN, use the TX or RX photodiodes measuring the light going to and from the test mass: 
    H1:CAL-PCALY_RX_PD_OUT_DQ
    H1:CAL-PCALY_TX_PD_OUT_DQ

You can track / trend the frequencies, amplitudes, and ON/OFF yourself using the following EPICs settings channels:
    PCALY f1                             PCALY f2
    H1:CAL-PCALY_PCALOSC5_OSC_FREQ       H1:CAL-PCALY_PCALOSC6_OSC_FREQ
    H1:CAL-PCALY_PCALOSC5_OSC_SINGAIN    H1:CAL-PCALY_PCALOSC6_OSC_SINGAIN
    H1:CAL-PCALY_OSC_SUM_MATRIX_1_5      H1:CAL-PCALY_OSC_SUM_MATRIX_1_6

TITLE: 09/04 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Calibration
INCOMING OPERATOR: Cheryl
SHIFT SUMMARY: 8.5 hour lock. Calibration measurements are finishing up.
LOG:

• 1545 Karen to MY
• 1559 Dust monitor pump started
• 1632 Karen done
• 1950 HW injection started
• 1958 HW inj. done
• 2000 Calibration starting, out of Observing

Since INJ_TRANS guardian failed to perform an injection the other day (alog 51682), we decided to do it manually using awgstream. I followed instructions by Adam.

Start time: 2019/Sep/04 19:50 UTC (that's when injection was enabled).

Actual start of the injection was tGPS~1251661857.6.

Injected waveform was prepared by Reed and committed to SVN by Adam as ${svncommon}/InjSVN/hwinj/Details/detchar/detchar-hwinj-schedule_LIGO-T1900555-PROPOSED_H1_INJECTIONS-1251586818-390.txt

What I did to enable injection and then inject:

caput H1:CAL-INJ_TRANSIENT_GAIN 1; caput H1:CAL-INJ_TINJ_TYPE 3 ; caput H1:CAL-INJ_TINJ_STATE 2

awgstream H1:CAL-INJ_TRANSIENT_EXC 16384 /ligo/svncommon/InjSVN/hwinj/Details/detchar/detchar-hwinj-schedule_LIGO-T1900555-PROPOSED_H1_INJECTIONS-1251586818-390.txt 1.0

CAL-INJ_TINJ_TYPE==[1, 2, 3, 4] means [CBC, burst, detchar, stochastic]. In this case it was already 3 but I did set to 3 anyway.

CAL_INJ_TINJ_STATE==[-1, 0, 1, 2, 3] means [stopped, idle, pending, streaming, completed]. It was 0 (idle) probably because INJ_TRANS guardian was and is in FAILURE_DURING_ACTIVE_INJECT state since the aforementioned failure,  and I set it to 2 (streaming).

Adam told me that if I wanted to start injection at a specific time I could have done

awgstream H1:CAL-INJ_TRANSIENT_EXC 16384 /ligo/svncommon/InjSVN/hwinj/Details/detchar/detchar-hwinj-schedule_LIGO-T1900555-PROPOSED_H1_INJECTIONS-1251586818-390.txt 1.0 gpstime

instead.

After the injection was done, I did

caput H1:CAL-INJ_TINJ_STATE 1; caput H1:CAL-INJ_TINJ_TYPE 1; caput H1:CAL-INJ_TRANSIENT_GAIN 0

as per Adam's instructions.

FAMIS11233

ITMX Pit is > 10urad

J. Kissel

Since I put the the power into PRM on to the ops overview (see LHO aLOG 51612), in addition to the power out of the PSL -- and noticed it has been dropping by 0.1W every day or three, I plotted a trend of these two input power metrics over the past ~4 months of O3.

Recall that we "permanently" started observing at "37 Watts" of "input power" on June 10, which is marked by the cursor; see LHO aLOG 49783.
Also, note that when we say "input power" in terms of "35 Watts" and "37 Watts" during O3, we mean the power out of the PSL. 

The revealed three things to me:
(1) The power loss between in power out of the PSL the power in to PRM is significant; consistently around 4.5 W, or a 12% difference.
(2) Since we've initial requested "37 Watts," and the powers were 37.5 into the IMC and 33.0 W in to PRM, we've lost about a Watt: the power is now 36.7 W into the IMC, and 32.1 W, and thus a loss of 2-3%
(3) There was a substantial increase after maintenance day on Jun 26 05:20 UTC (Tuesday Jun 25 2019 22:20:00 PDT), and a similar level drop Aug 30 02:44 UTC (Thursday, Aug 29 2019 19:44:00 PDT; after the new ALS pick-off was installed in the PSL; LHO aLOG 51610). The most recent drop was (32.4 - 32.1) / 32.4 = 1% drop in input power to the PRM.

For now this is just at the level of "huh," but given that the sensitivity is roughly proportional to the square root of input power, perhaps we should pay a bit closer attention to these kinds of changes and drifts, if possible.

TITLE: 09/04 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 117Mpc
OUTGOING OPERATOR: Travis
CURRENT ENVIRONMENT:
    SEI_CONF state: WINDY
    Wind: 4mph Gusts, 2mph 5min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.21 μm/s
QUICK SUMMARY: Locked for 30min, calm environment.

TITLE: 09/04 Owl Shift: 07:00-15:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Observing at 117Mpc
INCOMING OPERATOR: TJ
SHIFT SUMMARY:  Back to Observing near the end of shift.
LOG:

9:59 SEI Config to EQ for Tonga EQ

11:51 Back to WINDY

12:28 Lockloss.  No obvious cause.

14:38 Observing.  No operator intervention needed except for the usual green arm ETM tweaks.  Took many attempts to get to NLN, but no consistent place where it would lose lock.  Did not do IA.

14:55 Vanessa to MX

TITLE: 09/04 Owl Shift: 07:00-15:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Observing at 116Mpc
OUTGOING OPERATOR: Ed
CURRENT ENVIRONMENT:
    SEI_CONF state: WINDY
    Wind: 12mph Gusts, 9mph 5min avg
    Primary useism: 0.03 μm/s
    Secondary useism: 0.22 μm/s
QUICK SUMMARY:  Lock is 9 hours old.  No issues to report.

TITLE: 09/04 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 117Mpc
INCOMING OPERATOR: Travis
SHIFT SUMMARY:

• The only thing to report was NUC5 crashed at 6:25UTC. It was restarted and is again running properly
• Handing off to Travis

LOG:

The squeezer took H1 out of obseverving and the guardian could not relock the CLF loop.  The attached screenshot shows that there was extra intensity noise out of the laser at this time, shown in the attached screenshot.  I turned on the noise eater, the noise went away, we were able to relock the squeezer, and when turning the noise eater off again we stayed locked. 

Keita and Daniel trended the power out of the fiber from the PSL and saw that the polarization changed around 9 am local time, near the start of maintence. We used the half wave plate on the pico motor to restore the power on the fiber beatnote.  

Back in observing now.

M. Ball, S. Dwyer

During maintenance, we measured length drive to pitch and yaw transfer functions for our rerouted L2A control signal (LHO alog 51709). Earlier, we activated our L1 boost after rerouting the L2A signals and later added a 2Hz cutoff filter to the L2 length to L3 pitch decoupling filter, so we wanted to characterize this filter with the interferometer unlocked.

The first attached plot (Figure 1) shows the coherence between the drive signal in L2 and the pitch and yaw oplev signals in L3 before turning on our rerouted L2A filter (dashed lines), after rerouting (solid lines), and after loading a modified L2P filter (dotted lines). We see strong coherence between our drive signal and both pitch and yaw from 0.6 to 4Hz without our changes, and from 0.6 to nearly 5Hz with our changes.

The second attached plot (Figure 2) shows the transfer function magnitude and phase for L2 length to L3 pitch and yaw at the same three times. Again, dotted lines denote both rerouted L2A and the modified L2P filter, dashed lines denote no L2A filters, and solid lines denote just the rerouted L2A with no L2P cutoff. Between around 1.3 and 2.2Hz, the drive to pitch transfer function increases in magnitude after turning on the rerouted L2A then increases further after enabling the new L2P cutoff filter. However, above 3Hz, the transfer function with the new L2P cutoff filter drops off to below the rerouted case but above the original L2A case. This is the frequency range where we saw the impact on the ASC signals (LHO alog 51604). Between 0.6 and 1.5Hz, the drive to yaw transfer function also decreases after turning on the rerouted L2A and remains there after enabling the new L2P filter. Between 1.6 and 2.7Hz, the drive to yaw transfer function increases when turning on the rerouted L2A filter and remains there after turning on the new L2P filter.

M. Ball, S. Dwyer

Before lock was lost during maintenance this morning, we turned on our L2A rerouting as originally tested in LHO alog 51604. This time, we activated a boost (below 0.2Hz gain in the L1 L2L filter) and still maintained lock. With the rerouted L2A signals, we saw an increase in the actuator control signals between 3 and 5 Hz. With the boost on, we saw a decrease in these actuator signals below 0.3Hz.

We then tested a "new" (modified) L2 length to L3 pitch decoupling filter that included a cutoff above 2Hz to reduce the 3-5Hz signal increase. Little else was changed by this filter.

Attached are time series of the L1, L2, and L3 actuator control signals at the time the boost was activated. The L1 signal is largely unchanged, but some spikes appear larger. The L2 signal shows a net decrease in amplitude (this may help with locklosses triggered by saturation of these signals, such as this one from 9/3). The L3 signal remains largely unchanged.

Also attached are spectra of one of these signals before turning on the L2A rerouting and boost (dotted lines), between turning on the boost and loading the new filter (dashed lines), and after loading this filter (solid lines). Timestamps are as follows:

Rerouting turned on at 15:37:43 UTC.

Boost activated at 15:38:28 UTC.

L2P filter loaded at 15:51:30 UTC.

This has not been implemented yet, but we plan to do calibration measurements for it tomorrow.

I spent the morning mining the seemingly random locklosses from the weekend. Then Sheila pointed me towards the ndscope useful channels list that Jenne made - indeed the dither system (ADS) wanders off in many cases.

I don't know what they mean, but the lockloss tool shows the following:

39: 2019-09-01_12:13:59Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

H1:ASC-CHARD_Y_OUT_DQ starts to really wander off t-2 sec before lockloss, SRC channels wander at -2 sec.

H1:OAF-RANGE_RPL chans all WANDER at -7 sec.

38: 2019-09-01_16:15:40Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

H1:ASC-CHARD_Y_OUT_DQ starts to really wander off t-10 sec before lockloss, SRC channels wander at -1- sec.

No DARM glitch in the OAF RANGE RPL chans.

35: 2019-09-01_20:19:04Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

~WINDY - ends at ~16-20MPH

H1:ASC-CHARD_Y_OUT_DQ starts to really wander off t-3 sec before lockloss, SRC channels wander at -2 sec.

ADS PIT and YAW DOFs and DEMOD chans WANDER 10 sec before LL.

H1:OAF-RANGE_RPL chans all WANDER at -5 sec.

30: 2019-09-02_03:19:04Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

H1:ASC-CSOFT_Y_OUT_DQ starts to really wander off t-6 sec before lockloss, SRC channels wander at -5 sec.

ADS PIT and YAW DOFs WANDER 8 sec before LL.

H1:OAF-RANGE_RPL chans all WANDER at -7 sec.

24: 2019-09-02_13:10:16Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

H1:ASC-CHARD_Y_OUT_DQ starts to really wander off t-5 sec before lockloss, SRC channels wander at -4 sec.

ADS PIT and YAW DOFs WANDER 8 sec before LL.

H1:OAF-RANGE_RPL chans all WANDER at -9 sec.

19: 2019-09-03_01:52:57Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

H1:ASC-CSOFT_Y_OUT_DQ starts to really wander off t-8 sec before lockloss, SRC channels wander at -10 sec.

ADS PIT and YAW DOFs WANDER 12 sec before LL. H1:OAF-RANGE_RPL chans all WANDER at -12 sec.

13: 2019-09-03_10:17:41Z ISC_LOCK NOMINAL_LOW_NOISE -> LOCKLOSS

H1:SUS-ETMX_L3_MASTER_OUT - glitch 1 sec before LL, but other glitches of same size/shape earlier not an issue... SRC/SOFT/HARD loops loop nominal in seconds prior to LL.

ADS NOMINAL before LL.

No DARM glitch in the OAF RANGE RPL chans.

________________________

on command line:

userapps

lockloss -c ADS_3_4_5.yml select (to show ADS)

lockloss -c darm_blrms.yaml select (to show glitch info on DARM, not sure what math is happening but large wander from -10sec on H1:OAF-RANGE_RPL_1_OUTPUT and all other similar chans, seems there was a glitch in DARM which maybe saturated the Dither line in DARM so that ADS couldn't work correctly... or something like that in commissioning person words...)

So, is it the old story that a glitch causes havoc in with the ADS - and does it seem worse this weekend? Dunno...

Ed called about difficulty locking, since I don't see anything obvious I looked at BRS channels I foudn trended here: 51338.  It looks like the low frequency BLRMS of BRS X has been growing exponentially.

Pursuant to LIGO-T1900555 [1] and discussions on the DetChar and HW injections mailing list [2,3], I've scheduled DetChar safety injections to begin at 23:00:00 UTC (18:00:00 CDT, 16:00:00 PDT) 3 Sep 2019. These should last for 390 seconds and contain 75 injections spaced 5 sec apart.

The scheduled injections have been added to GraceDb [4]. We will update the events in GraceDb to record whether these injections were successful post-hoc.

**Please note**, we have schedule coincident (but not necessarily coherent) injections at LLO for the same time. We include the LLO aLOG describing those injections in a comment below.

More details on how these injections were generated and added to the HW injection SVN repo are available via a README [5].

[1] https://dcc.ligo.org/LIGO-T1900555
[2] detchar@ligo.org
[3] hw-injections@ligo.org
[4] https://gracedb.ligo.org/search/?query=Test+INJ+gpstime%3A+1251586818+..+1251587400+instruments%3A+%22H1%22&query_type=E&results_format=S
[5] https://ldas-jobs.ligo.caltech.edu/~detchar/hwinj/LIGO-T1900555/README

Pursuant to LIGO-T1900555 [1] and discussions on the DetChar and HW injections mailing list [2,3], I've scheduled DetChar safety injections to begin at 21:00:00 UTC (16:00:00 CDT, 14:00:00 PDT) 3 Sep 2019. These should last for 390 seconds and contain 75 injections spaced 5 sec apart.

The scheduled injections have been added to GraceDb [4]. We will update the events in GraceDb to record whether these injections were successful post-hoc.

**Please note**, we have schedule coincident (but not necessarily coherent) injections at LLO for the same time. We include the LLO aLOG describing those injections in a comment below.

More details on how these injections were generated and added to the HW injection SVN repo are available via a README [5].

[1] https://dcc.ligo.org/LIGO-T1900555
[2] detchar@ligo.org
[3] hw-injections@ligo.org
[4] https://gracedb.ligo.org/search/?query=Test+INJ+gpstime%3A+1251579618+..+1251580100+instruments%3A+%22H1%22&query_type=E&results_format=S
[5] https://ldas-jobs.ligo.caltech.edu/~detchar/hwinj/LIGO-T1900555/README