The EX susrack magnetometer (all directions) is unsafe, meaning it picks up signal directly from DARM. Presumably it's picking up the electrical signal from the feedback of DARM to EX.

This was determined by Sheila's suggestion of looking at a calibration sweep. We're also analysing the detchar hardware injections, which should show the same thing. We see the calibration sweep appear in the magnetometers below about 100 Hz. The Y direction also sees something at double the frequency. I'm only attaching that spectrogram.

Edited to add: The calibration sweep is going through the Y end, so it's not just picking up that.

Ops Shift Transition: 04/16/2019, Owl Shift 07:00 – 15:00 (00:00 -08:00) - UTC (PT)

State of H1: NLN

Intent Bit: Observing

Weather: Low wind

Primary 0.03 – 0.1Hz: 0.02 um/s

Secondary 0.1 – 0.3Hz: 0.15 um/s

Outgoing Operator: Patrick

Quick Summary: In Observing for close to an hour when I took over. Microseism decreasing over the last 24 hours.

TITLE: 04/15 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 107Mpc
INCOMING OPERATOR: Niko
SHIFT SUMMARY: Rough night for locking. See log.
LOG:

23:36 UTC Lock loss
00:06 UTC Stopped at PREP_DC_READOUT_TRANSITION. Commissioning resuming.
02:01 - 2:05 UTC Sheila to LVEA, setting DC PD transimpedance back to normal
02:07 UTC Lock loss
02:41 UTC Lock loss from ENGAGE_ASC_FOR_FULL_IFO
02:42 UTC Initial alignment
03:54 UTC Initial alignment done
04:16 UTC Lock loss from DARM_TO_DC_READOUT
05:11 UTC Broke lock reverting SDF differences at NLN (alog 48518)
X arm green transmission keeps falling off
Going to try an initial alignment of just green arms
05:28 UTC Finished alignment and offloading of green arms. Relocking.
Initial alignment of green arms seems to have helped. Able to move past ALS without X arm transmission falling off.
Stopped at PRMI and aligned BS and PRM, but lost lock after a poor acquisition of DRMI
Another poor acquisition of DRMI but was able to adjust SRM in time to fix it before it broke lock
06:07 UTC NLN. No SDF differences.
06:08 UTC Observing
06:10 UTC Dropped out of observing by squeezer. CLF unlocked.
Fixed by hitting init, down, etc on various squeezer guardians and then 'caput H1:SYS-MOTION_C_BDIV_C_OPEN 1' when it was trying to lock OMC
06:19 UTC Observing

06:08 UTC No SDF differences.

DARM has odd spectrum (see attached). Suspect squeezer guardian is not at nominal.

Commissioning is done for the night. Running through initial alignment.

Sheila, Danny

Today we took about 5 hours out of observing in order to do a series of OMC scans. The measurements were heavily inspired by what Dan Hoak did in alog 22175.

What I wanted to see from these scans was a direct effect of the high spatial frequency optical path distortion on the carrier and 9 MHz and 45MHz sidebands. My attempt to accomplish this required the following:

• Prepare OMC for scanning while locked at 2W RF
• Set up and start 100 second period 70 V amplitude ramp excitation to PZT2
• gather a series of scans at 2W, (also grabbed some scans with 0 DARM offset for contrast defect)
• power up to 5 W after 1000 seconds (for the point absorbers to reach maximum optical path distortion)
• gather scans at 5W for at least 1000 seconds
• power up to 10 W
• gather scans at 10W for 1000 seconds (grabbed scans with 0 DARM offset for contrast defect)

Another separate but relevant measurement was to measure the field content at the AS port with a 35W thermal state. Because we did not want to damage the PD by scanning at 35W we let the interferometer thermalize at 35W and dropped down to 10W to take scans. It is important to note that it was not trivial to power up to 35W on DC readout and transition back to RF hence the large amount of time out of observing. Because of time constraints, the thermalization criteria had to be an hour and a half at 35W rather than 2+ hours which should be good enough for this measurement.

Results from the scans are pending and will be posted in a comment on this log entry.

This is just a summary so that this is written in one place in the log:

Koji measured the modulation depths for 9 and 45 MHz using an OSA after installing the new EOM: 41435  For the 9 MHz Koji measured gamma =0.191 for an epics setting of 23.6dBm, which is used during lock acquisition.  We reduce this in low noise to 20dBm, giving us the final modulation depth of 0.1352  (That is the whole story for 9MHz).

After that Daniel and company moved the EOM driver out of the PSL enclosure and added an RF combiner/amplifier, (41889) which increased the drive power for 45MHz by +1.83dBm compared to the time of Koji's measurement.  The modulation depth measured with an epics setting of 27dBm (used for lock acquisition) should have a modulation depth of 0.243 assuming the EOM response measured by Koji. In low noise we use an EPICS setting of 24dBm, so we should expect 0.172 modulation depth for 45 in full lock.   (These numbers are a little different from Daniels because he used the EOM response from G1800724 rather than Koji's measurement).

Craig and Georgia also measured modulation depths by exciting RFAM and measuring intensity noise after the OMC (47113) but those results aren't agreeing well with the numbers that Koji and Daniel posted. 

Pep Covas, Jeff Kissel

We report on some tests of the 20 and 18 bit DACs performed in the test stand. We drive a signal of varying frequency and amplitude, and check if we see nonlinearities in the spectrum (i.e. lines at multiples of the injected frequency or at other frequencies). The reading is made at the output of the anti-imaging chassis with a break-out board connected to an SR785 (more details about the setup are shown in the first attached image). The following plots have been obtained with the 20-bit DAC, but similar results are observed with the 18 bit DAC.

The second figure shows the results obtained by injecting a signal of 100 Hz near the maximum amplitude (~2**19 cts). A comb of lines is produced, with spacing equal to the injected frequency, although only the odd harmonics can be seen. The next figure shows a wider frequency range, where we can see that for frequencies higher than ~4 kHz further harmonics are not visible (the last visible harmonic with an integration time of 10 s is around number 75, but longer integration times make larger harmonics also visible). The next two figures shows the same injected frequency at half number of counts (half amplitude). Many lines can still be seen, at a reduced amplitude than the first figure. In order to not see any of the non-linear lines, the signal amplitude must be lower than ~30000 cts, more than an order of magnitude away from the maximum (with longer integration times this number would be even lower). This comb of lines with spacing equal to the injected frequency is produced at any injected frequency and sampling rate of the user model.

By increasing the frequency of the signal, many lines start to appear in the low frequency range. For example, for a signal of 5950 Hz, lines with a spacing of 86 Hz appear, as shown in the 6th figure (again, only odd harmonics). These lines appear to come from aliasing of the high frequency harmonics of the main comb discussed previously. Harmonics which lie between 2**15 Hz (Nyquist frequency) and 2**16 Hz (sampling rate of IOP model) are ported to the main frequency range by 2**16 - n*freq, where n is the harmonic index. Harmonics between 2**16 and 2**17 are ported by n*freq - 2**16. This pattern is repeated for all the produced harmonics, which get transported to frequencies between 0 and 2**15 Hz. The 5950 Hz frequency has been chosen to replicate the problems discussed in https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=37139. Figures 7 and 8 show a wider range of frequencies and the same injection at half amplitude, where the low frequency lines dissappear.

The main comb and the high frequency aliasing explains some of the lines which can be seen, but not all of them. We make further tests by driving two sinusoidal signals of different frequencies at the same time. Plot 9 shows an example of a 100 and 125 Hz signals, where new lines which cannot be seen when only injecting separately 100 or 125 Hz appear. These new lines have a spacing of 25 Hz, suggesting some type of beating between the two injected signals. This beating may explain some of the unexplained lines which appear in the spectrums discussed before. This may also explain some lines seen in DARM at frequencies which seem to be a beating of violin modes and calibration lines, as discussed here: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=48161. 

Furthermore, different sampling rates of the user model have been tested. Plot 10 compares the output at different sampling rates. Although the main comb of spacing equal to the injected frequency is there in all the cases, different lines arise for each different sampling rate, with the situation being worse the lowe the sampling rate is.

Plots 11-14 show some readings performed at the excitation point with DTT (measurement point 3). Besides the main frequency, a comb of lines can also be seen. The spacing of this comb depends on the frequency of the main injected signal. This is a different feature than the comb which is seen after the DAC and AI, since the main comb of spacing equal to the injected frequency cannot be seen here, neither the beating between multiple injections.

Some readings have also been taken previous to the signal getting into the AI chassis (measurement point 2), and the same lines and nonlinearities have been observed as when measuring from point 1.

We would like to move several of the calibration line frequencies, in order to get them out of our most sensitive region, as well as to help give more accurate measurements of the interferometer properties. 

There are enough constraints on what frequencies are okay to use for calibration lines that I made a little script to help visualize where we're proposing to put the calibration lines at the lower (below 20 Hz) frequencies.  Lilli has a nice summary of these constraints in alog 44590.  EvanG provides in alog 44892 the frequency regions that are okay from the viewpoint of puslar searches.  

Currently, our calibration line frequencies are [15.1 Hz, 16.7 Hz, 35.9 Hz, 36.7 Hz, and 331.9 Hz], and I am proposing to change those to [15.6 Hz, 16.4 Hz, 17.1 Hz, 17.6 Hz, and 410.3 Hz].  (We also have lines at 7.93 Hz and 1083.7 Hz, but we're not moving them).

In the top panel of the attached figure, I have horizontal bars for the okay ("non-vetoed") regions according to the pulsar group, assuming we need to be at least 0.1 Hz away from known pulsars for which we could potentially measure spindown limits.  I show the LLO calibration frequencies in blue (since we don't want to use the same frequencies, although we have been for the 15.1 Hz line since about January 14th); several frequencies in yellow that have been proposed in various alogs, including those from Lilli and Evan linked above; the beam splitter rigid body frequency (Lilli's alog quotes 17.54 Hz for this, but it looks like it's closer to 17.8 Hz for us); and the ADS frequencies that we are currently using, since we must not drown them out with calibration lines.  Finally, in red, I show my new proposed suspension calibration frequencies: 15.6 Hz, 16.4 Hz, 17.1 Hz, and 17.6 Hz. 

In the bottom panel, I show where these frequencies lie in relation to our DARM spectrum at a time that the calibration lines were off (so that it's clear that current cal lines at, for example, 16.7 Hz, would change and not interfere with these new frequencies). 

In the bottom inset, I also show the proposed frequency for the pcal line that is used to measure the DARM cavity pole.  LLO uses 434.9 Hz for this line.  We will use something lower, since our DARM cavity pole is currently much lower than L1's. 

Either tomorrow or Wednesday, Jeff and I will:

• Change our calibration lines to the new frequencies.
• Change the DARM2 FM4 notch from a 331 Hz bandstop to a 410 Hz bandstop.
• Change the matching filter module in the CAL-CS copy of DARM2.
• Change (check) the list of filters in the H1 modelparams file.
• Change / check the calibration line frequencies in the pyDARM models / modelparams files.
• Create new EPICS records for  time-dependent value calculations.
• Tell Maddie, so that she can update the GDS pipeline.

TITLE: 04/15 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Commissioning
OUTGOING OPERATOR: Corey
CURRENT ENVIRONMENT:
    Wind: 9mph Gusts, 5mph 5min avg
    Primary useism: 0.04 μm/s
    Secondary useism: 0.21 μm/s 
QUICK SUMMARY:

In commissioning for OMC scans.

TITLE: 04/15 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 107Mpc
INCOMING OPERATOR: Patrick
SHIFT SUMMARY:

6.5hrs of OBSERVING this shift, and then ~4hrs of Approved Commissioning (TCS OMC scans) started at the end of shift and will carry over to the evening.
LOG:

• 15:45 TCS Chiller & HEPI level checks (Ed)
• 19:49 Glitch seen (by Sheila/Jenne) on several ASC PITCH Control signals (seen on nuc6) drift for PRC2 & SR2 (latter took step down and took over 10min to return).  
• 21:05 Shut off circuit breakers for LVEA roughing pumps (this allows their removal later by VAC) [Richard WP#8172]
• 21:27 Out of Observing for APPROVED Commissioning: OMC scans for TCS for ~4hrs  (WP#8170, Danny Sheila)
• 21:27 (while out of Observing) Calibration Line frequency correction made (JeffK)
• 21:28 LOCKLOSS
• 22:26 Handing over to Danny after getting H1 to PREP_DC_READOUT_TRANSITION for his OMC scans.

My Locking Notes:  DRMI alignment tweaking & OMC Carrier

1)  If dropping out in steps between DRMI & DC Readout, tweak up alignment by stopping at CHECK_AS_SHUTTERS, and then:  tweak PRM & SRM to get POP18 up to 60 & POP90 down to 15. (Thanks to Jenne for the reminder she's told me several times before!)  :)  

2) If one gets a message for the OMC not finding the carrier, one can not do anything and wait for the 600sec sleep for ISC_LOCK, or open up the OMC_LOCK node and request FIND_CARRIER (it'll go DOWN & then find the carrier).  One can do this request eventhough this node is managed.

No water was added. No water was needed.

    Laser Status:
    Front End Power is 32.31W (should be around 30 W)
    70W Output Power is 70.23W
    Front End Watch is GREEN
    70W Watch is GREEN

    PMC:
    It has been locked 0 days, 0 hr 0 minutes (should be days/weeks)
    Reflected power = 9.893Watts
    Transmitted power = 54.94Watts
    PowerSum = 64.83Watts.

    FSS:
    It has been locked for 0 days 12 hr and 50 min (should be days/weeks)
    TPD[V] = 3.965V (min 0.9V)

    ISS:
    The diffracted power is around 2.2%
    Last saturation event was 0 days 13 hours and 51 minutes ago (should be days/weeks)

    Possible Issues:  Looks OK.

I have begun some prep work for tomorrow's maintanance day installation of updated CALCS and SUS quad models under work permit 8166.  This will address FRS 12335, as laid out in ECR E1900107, to eliminate the problem of timing slips on the supposed-to-be-synced oscillators causing problems with the time dependent calibration kappa calculations.  JoeB successfully completed this at LLO last Tuesday (LLO alog 45007), so here we are just pulling his changes.

So far, I have:

• Confirmed that there were no changes locally in any of the sus or cal userapps common and h1 folders that needed to be checked in.
• Pulled the common library blocks from the svn.
  • <userapps>/cal/commom/models/CAL_CS_MASTER.adl
  • <userapps>/cal/commom/models/CAL_LINE_MASTER.adl
  • <userapps>/sus/commom/models/QUAD_MASTER.adl
  • <userapps>/sus/commom/models/QUAD_ITM_MASTER.adl
  • <userapps>/sus/commom/models/FOUROSEM_DAMPED_STAGE_MASTER_WITH_DAMP_MODE.mdl
  • <userapps>/sus/commom/models/FOUROSEM_STAGE_MASTER_OPLEV_TIDAL.mdl
• Modified h1calcs to get the receiver parts for the calibration lines from the ends (3 for EX, 3 for EY).
• Modified h1susetmx to make sender parts for the 3 calibration lines.
• Modified h1susetmy to make sender parts for the 3 calibration lines.
• Compiled h1susetmx, h1susetmy, h1susitmx, h1susitmy, h1calcs
  • Note that I did not need to change the H1 ITM models, since they don't need sender / receiver parts.  Joe had already modified the QUAD_ITM_MASTER to have grounds for the subblocks that, in the ETM models, push signals up to the topmost level.
• Checked h1calcs, h1susetmx, and h1susetmy into the svn.

To do tomorrow during maintanance:

• Install all 5 models.
• Confirm SDFs for all 5 models are ready to go (they should be, but will check anyway).
• Take suspensions to their SAFE states, and restart the models, then bring the suspensions back to their ALIGNED states.
• Pull the updated matching medm screens for the calibration suspension demodulation lines.
• Set several EPICS values for some new switches, and for the demodulation phase that will account for the clock cycle delay.
• Once relocked, ensure that everything in kappa-generation land looks sensible.

Craig, Danny, Georgia

Today we stepped the SR3 heater down in 0.5 W steps, from 5W to 3.5W, according to plan. Here's what we found as we reduced the heating on SR3:

Buildups, cavity pole

• the DARM coupled cavity pole increased by a few Hz
• the optical gain (Kappa_c) decreased by ~2%
• POPAIR_B_RF18 decreased by ~5%
• The BNS range went up ~4 MPc, maybe this peaked at 4W on the SR3 heater
• The changes in arm power were only slight - maybe went up a little ar 4.5W, but decreased after that?

These values are shown in the first attachment, The kick in many signals just before we stepped down to 3.5 W is explained below.

DARM plant and laser noise couplings

We also monitored the DARM plant, frequency noise coupling, intensity noise coupling, and RF9 RIN coupling at each step.

• Craig will post the DARM plant and frequency noise measurements
• The intensity noise coupling at high frequency got worse as we reduced the heater power (second attachment shows darm during noise-buget-style ISS second loop intensity noise injections)
• The RF9 RIN coupling reduced slightly at 4.5W and 4W but was increased again when we measured at 3.5W (third attachment shows the 9MHz RIN excitation to DARM transfer function)

SRC ASC

While we were sitting at 4W on the SR3 heater we checked for detuning in the SRC ASC. This is the big peak in RF18 and RF90 in the first attachment; the 4th attachment is zoomed in during this time. We opened the loops and moved the sliders (top right plots), found nothing too interesting in pitch, but while aligning yaw we saw:

• some gains in RF18, but
• increase in RF90 (maybe we were aligning the SRC to compensate for a misalignment in the PRC?)
• a 2% increase in kappa_C (!)
• a 25 Hz drop in the cavity pole (!!)
• a corresponding dip in the range

We were surprised that the optical gain increased, that doesn't seem to hang together with the other pieces of information here. Maybe we should consider operating with 4W on the SR3 heater, and re-phasing the SRC ASC for this. [Edit: notes for attachment 4: at t = -4500s we turned the SRC ASC back on, which is why the alignment went back to its nominal level. The calibration lines were off before t = -4800s, and the calibration values before this time are not to be trusted.)

Other notes:

• The SR3 cage servo (really an oplev servo) was on for this measurement, this kept SR3 servo'd in pitch, but uncontrolled in yaw
• At 3.5 W on the SR3 heater we encountered a very slow (~2 minute period) oscillation in the corner DOFs. This settled when we returned the SR3 heater to 5W. We spent a bit of time tongiht with the ADS YAW DOFs off, to stop them from oscillating.
• We spent 2 hours and 34 mins out of observing today to do these tests, comfortably within our allotted 3 hours :-)
• While we were out of observing to make measurements we chopped the OMC ASC master gain for Sheila.
• To run our excitations we connected two LSC AO extra channels at the PSL racks to the common mode board excitation, and the 9MHz EOM excitation port, these need to be disconnected again.

We've measured the DARM plant to have a prospring at 6 Hz and a Q of about 4. About the spring frequency, the phase goes through around -90 degrees, and not the expected +180 degrees alog 48083 

This +180 degrees is not typical for causal control systems, but it is expected from SR IFOs as discussed in Section II B from BnC.

Using the Ward DARM model (Eq 3.83) I was unable to achieve a satisfactory fit of the H1 DARM plant at low frequencies, so I made some sliders to see if I could get a heuristic match.  I was not able to for reasonable IFO parameters.

I discovered that our Q is far too low given our optic transmissions.  One can lower the Q of the optical spring by increasing the SRM transmission or reducing the ITM transmission.  However, changes to these parameters also change the frequency of the optical spring and the DARM pole.  The pictured plot shows the best Ward model I was able to come up with to explain the current plant, featuring detuning of -0.5 degrees.

DARM Plant Measurement
I began questioning the measurement itself, but the procedure is pretty simple.  A PCAL to DARM measurement gives C/(1 - G), where C is the DARM plant, and G is the DARM OLG.  Then a DARM OLG is taken to get 1/(1 - G), and these two measurements are divided to give the DARM plant C. 
 
The PCAL calibration into meters is just two real poles at 1 Hz.  This has phase of -160 degrees at 6 Hz, i.e. there is some dynamic phase rotation happening at LF due to the calibration which may not be real.  

Craig, Georgia

Today we locked the OMC on single-bounce carrier with 20 W input and injected large sideband intensity noise to measure the modulation depths for locking and low noise settings.

Results

Sideband     Locking EOM Driver Slider Value [dBm]     Low Noise EOM Driver Slider Value [dBm]     Locking Mod Depth     Low Noise Mod Depth
---------------------------------------------------------------------------------------------------------------------------------------------
9 MHz                                   23.4                                        20.4                       0.190                   0.160
45 MHz                                  27.0                                        24.0                       0.219                   0.182


Method
A simple comparison of the carrier and sideband RINs when injecting.  

RIN0(f) = Γ2 RINsb(f)

We know that we can trust the 9 and 45 MHz RFAM monitors when we inject high broadband noise in the EOM driver.  
At the same time, if we inject enough sideband RFAM, the carrier RIN should show through in the OMC transmission.

All injections were amplitude=10, ellip("BandPass",4,1,30,5000) through the LSC EXTRA AO 3, which is connected to the front of the drivers.

Reason
We've been running several 9 MHz RIN tests.  It is good to know, for some high level of sideband RIN, the level of carrier RIN we expect. 

Old EOM Driver alogs
Koji installation
Daniel boost