The DM first starting having issues at ~11am 2 weeks ago on April 22nd (Tues) for seemingly no reason, there wasn't any work or alogs that day that would made us suspicious of causing this.

Following my alog comments on alog84282, Dave showed me where the physical comtrol box is in the CER, restarting it brought the connection back but I then found that the dust monitor had no flow :(. This DM hasn't been used since we got it calibrated in April of 2024, so I swapped it for the last pumpless spare we had which has good flow but now it's not connecting again. I've tried powercycling the DM itself, the ioc, and the comtrol box then the ioc which is what got the other DM to reconnect, and it still getting "No reply from device within 1000 ms" for all its PVs. The DM it self is working and reading counts properly.

Doing a telnet network status (comand "ss -e") on h0epics for the diode room port yielded:

timer:(keepalive,38min,0) uid:1001 ino:20939307 sk:ffff8800b67bd500
ESTAB       0      0                                                                                        10.105.0.80:37347                                                                                    10.105.0.100:8000 

(CoreyG, MitchR, RandyT, JimW)

Previous work posted here.

Today started with Randy and Mitch getting the small green tractor to EX and then working on compacting some of the very soft sand around the wind fence (on the front side of the Wind Fence---closest to EX).

Started to remove the final 3-wind fende panels (of 6 total).  While driving the orange rental manlift (on the backside of the wind fence), it managed to get dug in some pretty deep sand a few times.  On the last dig-in, we kept the manlift here (dug out/compacted sand around it and focused work on the furthest away wind fence panel since we already had the orange manlift here).  The plan is to do everything for this panel location all at once:

1. Remove old (crappy) hardware,
2. Install better hardware,
3. Clamp horizontal wires to the end/furthest pole,
4. Start tensioning the 5-horizontal wires into place
5. Install new wind fence panel

Today we made it to step 4 and did the 1st/bottom horizontal wire.  

We ran out of the thick wire, so the trailer was brought back to the Corner Station so the 1200lb spool can be loaded into the trailer in the morning.  Then will continue getting that 1-panel installed....and then work on getting the orange manlift unstuck.

J. Kissel

It was brought to my attention that the DC centering loops for the REFL path weren't working, so we revisited the sign of the RM1 and RM2 signal chains to confirm the functionality. 

Recall, 
- 2025-04-21 Daniel Fil and Marc found that raw ADC input voltage for OSEM sensor readbacks for the RMs were negative, and have been for over a decade, discovered as they replaced the in-air, sat-amp to vac feedthru cable which flipped the sign for both RM1 and RM2 OSEM sensors. LHO:84027
- 2025-04-25 Daniel, acknowledging that this OSEM sensor sign flip should be accounted for somewhere, decided to stick the sign flip in the 4x COILOUTF gains, typically reserved for magnet polarity compensation. LHO:84123
He acknowledged that putting the compensating sign there would impact the REFL WFS DC centering loops, and we've later realized this will also impact the alignment sliders. 
- 2025-04-29 For the above three "don't put the sign flip there!" reasons, on 2025-04-29, I reverted Daniel's COILOUTF gain sign flip. LHO:84186
That means that regardless of whether RM1 or RM2 have the wrong magnet polarity (see LHO:40853 which concluded it was RM2***), the actuator chain's sign will be the same as it has been for a very long time, and no ISC or alignment loops should need adjusting.
I left that aLOG with a "we'll adjust the damping loop signs once the in-chamber work dust settles."
- 2025-04-30 Rahul, a day later, wanted to resolve the remaining issue. I then communicated to Rahul that we needed to flip the sign in the DAMPING loops, and *only* in the DAMPING loops, because these are the only loops that would need to be changed. He instead re-flipped the coil output filter COILOUTF gains again, see LHO:84204, and accepted them into SDF.

So, today, 2025-05-06
I again reverted the sign back to how they've been since 2013 
    $ caget H1:SUS-RM1_M1_COILOUTF_UL_GAIN H1:SUS-RM1_M1_COILOUTF_LL_GAIN H1:SUS-RM1_M1_COILOUTF_UR_GAIN H1:SUS-RM1_M1_COILOUTF_LR_GAIN
    H1:SUS-RM1_M1_COILOUTF_UL_GAIN 1
    H1:SUS-RM1_M1_COILOUTF_LL_GAIN -1
    H1:SUS-RM1_M1_COILOUTF_UR_GAIN -1
    H1:SUS-RM1_M1_COILOUTF_LR_GAIN 1

    $ caget H1:SUS-RM2_M1_COILOUTF_UL_GAIN H1:SUS-RM2_M1_COILOUTF_LL_GAIN H1:SUS-RM2_M1_COILOUTF_UR_GAIN H1:SUS-RM2_M1_COILOUTF_LR_GAIN
    H1:SUS-RM2_M1_COILOUTF_UL_GAIN -1
    H1:SUS-RM2_M1_COILOUTF_LL_GAIN 1
    H1:SUS-RM2_M1_COILOUTF_UR_GAIN 1
    H1:SUS-RM2_M1_COILOUTF_LR_GAIN -1

and now have also flipped the damping loop sign, to account for the OSEM sensor readback sign change fixed from the in-air satamp to feedthru change
    $ caget H1:SUS-RM1_M1_DAMP_L_GAIN H1:SUS-RM1_M1_DAMP_P_GAIN H1:SUS-RM1_M1_DAMP_Y_GAIN
    H1:SUS-RM1_M1_DAMP_L_GAIN      1
    H1:SUS-RM1_M1_DAMP_P_GAIN      1
    H1:SUS-RM1_M1_DAMP_Y_GAIN      1
    $ caget H1:SUS-RM2_M1_DAMP_L_GAIN H1:SUS-RM2_M1_DAMP_P_GAIN H1:SUS-RM2_M1_DAMP_Y_GAIN
    H1:SUS-RM2_M1_DAMP_L_GAIN      1
    H1:SUS-RM2_M1_DAMP_P_GAIN      1
    H1:SUS-RM2_M1_DAMP_Y_GAIN      1

This *positive* damping loop gain is markedly different than any other suspension, which have damping loop gains negative. Now I'm pretty confident that means that RM1's magnet polarities have been installed incorrectly, rather than RM2 as declared in 2018 ***.
However, because we tried during this vent and can't access the magnets to verify the polarity, let alone change it (see LHO:84178), we will leave this situation as is because we don't want to have to think about the down-stream impacts of alignment and DC centering.

That being said, the ISC team is going to put a positive digital alignment offset in PIT and YAW now, and will confirm that this steers the beam according to standard SUS convention:
    :: with a right-handed coordinate system defined by +L perpendicular to, and coming out from, the HR surface, and +V pointing up,
    :: +PIT should push the optic "down" curling around the +T access, and 
    ::: +YAW should push the optic "counter clockwise, if you're looking from above the mirror, down the -V axis." 
If they find this backwards, the may request a change...

For now, in this sign configuration, I've
- Confirmed that the damping loops close and are stable, and
- Saved the current sign state into the userapps repo copy of the h1sushtts_safe.snap, to which the target area's safe.snap and OBSERVE.snap both point,
    /opt/rtcds/lho/h1/target/h1sushtts/h1sushttsepics/burt$ ls -l safe.snap 
        lrwxrwxrwx 1 controls advligorts 64 Oct  5  2017 safe.snap -> /opt/rtcds/userapps/release/sus/h1/burtfiles/h1sushtts_safe.snap
    /opt/rtcds/lho/h1/target/h1sushtts/h1sushttsepics/burt$ ls -l OBSERVE.snap 
        lrwxrwxrwx 1 controls advligorts 64 Jun 22  2018 OBSERVE.snap -> /opt/rtcds/userapps/release/sus/h1/burtfiles/h1sushtts_safe.snap


Attached are
    - a screenshot of me accepting the correct, above mentioned, COILOUTF and DAMP gains in SDF (I swear and promise I hit accept after I took the shot.)
    - a one-month trend of the RM1 COILOUTF gains, with the above associated timeline called out
    - a one-month trend of the RM2 COILOUTF gains, with the above accociated timeline called out

Jennie W, Sheila,

I have been working on a numerical model to check against some of our single bounce and full IFO measurements of the OMC to IFO mode-matching. Keita did a similar analysis using single bounce measurements and an alamode mode-matching simulation here alog #71145. In this entry while trying to check input parameters for the model, Sheila, Keita, Camilla and I worked out that the original; design for the OMC telescope has an extra 14 cm between OM1 and the input to the OMC.

My model takes the parameters for the bow-tie OMC cavity design from LIGO-T1300189, page six.

distance between two flat mirrors = distance between two curved mirrors =  0.2816m

diagonal distance between one curved and one flat mirror = 0.2844m

The RoC for each curved mirrors is take from LIGO-T1500060

Mirror C6 in table in appendices is curved mirror 1 in OMC design.

RoC CM1  = 2.57321 m

Mirror C5 is CM2

RoC CM2 = 2.57369 m.

Using the ABCD matrices for each mirror and the spaces between them I solved for the q parameter of the circulating mode just inside the input coupler.

This is  q = -0.1408 + 0.7102i m.

This gives a beam waist between the two flat mirrors of 4.9044e-4 m. Note: The following analysis is all for the vertical modes, since the OMC has some astigmatism due to a non-zero angle of incidence in the horizontal direction we simplify our analysis by only considering the vertical modes in the mode propagation, this is especially easy for our particular OMC (OMC001) as it has a very small degree of astigmatism (108 kHz  difference vs the fwhm of the resonance at 667 kHz) between the vertical and horizontal modes.

I then start just in front of OM2 using a q parameter (q_pre_OM2 = 0.8902 + 0.8034j) Sheila worked out from propagating the q from table 1 of LIGO-T1200410 for just after SRM using her alamode model from (alog #84089) entry.

I construct some 70 x 70 grid of possible q parameters.

I then propagate this towards the OMC using both the hot (1.75m RoC) and cold (2.1m RoC) that OM2 can have in its two PSAMS states (for calibration of the heater see alog #65276 and for the radii of curvature see alog #71145).

The q parameter can be used to define the electric field amplitude transverse to the beam propagation direction z:

U = 1/q ( exp( -jk ( x² + y² ) / 2q ))

where k is the wavenumber of the light and x and y are the transverse distances from the center of the beam in the horizontal and vertical directions.

Using the overlap integral:

O(q₁,q₂) = | U₂ U₁^* |² / ( |U₁|² |U₁|²​ )​​​​​ 

where U₁ is the field amplitude for TM 00 mode of the OMC and U₂ is the field amplitude of the pre-OM2 mode after propagation to the OMC.

for the OMC field and each of the fields propagated from before OM2 I work out two mode mis-match plots, where mode-mismatch (MM) is:

MM % = (1 - O(q₁,q₂)) * 100,

in terms of the imaginary and real parts of the q parameter I started with upstream of OM2. See image 1 for the grid of pre-OM2 modes against the mis-match each mode would have with the OMC mode at the OMC after propagating through a cold OM2 and OM3. See image 2 for the same plot but with the mis-match the pre-OM2 grid would have after propagating through a hot OM2 and OM3.

On both images the original mode before OM2 is shown by a white star and the coloured countours show the mode-mis-match percentage at the OMC. The white contour in each plot is a measurement of the single bounce mode-mismatch measured in alog #79229 for the cold state and in alog #78727 for the hot state.

Sheila verified these mode-matching values for the hot and cold state of OM2 by propagating q_pre_OM2 in her model to the OMC and overlapping it with the OMC q parameter from my model.

The points where the two measurement contours overlap give us two possible q parameters for the real mode pre-OM2 in the single bounce state. See the third image, where the blue contour is the mode-mismatch measured with the OM2 in the cold state and the red contour that measured with the OM2 in the hot state. The light blude star is the pre-OM2 q used in my numerical model obtained from Sheila's alamode model.

The two possible qs that match experiment are q1 = 0.593 + 0.925j and q2 = 1.626 + 0.967j.  This gives a beamwaist of 5.597e-4 m located 0.681 m behind OM2 or a beamwaist of 5.722e-4m located 1.714m behind OM2.

Terra Hardwick and others did some on-bench beam profiler measurements in HAM6 (alog #41490) and their model for between OM1 and OM2 seems to be consistent with q1 rather than q2.

After the survey of the large LN2 dewar jacket pressures (alog 84028) we used an ISP500 scroll pump to pump the jackets on CP2 and CP3.

CP3 took ~4 days of pumping to bring the pressure from 220 mtorr to <10 mtorr. CP2 was much faster with about 8 hours of pumping from 18 mtorr to 9 mtorr.

All other tanks (excluding CP4 not in use) had jacket pressures around 10 mtorr, so no additional pumping needed.

Closing WP 12498

Debasmita, Preeti, Gaby

Robert did some injections at 0.2 Hz in the BS HEPI which showed a slight increase in DARM. We wanted to check further if the microseismic ground motion (0.1-1.0 Hz) is creating any noise in LHO as it does in LLO. For this, we chose the glitches in DARM picked up by omicron, having frequency in the range 10-60 Hz and SNR in the range 5-20.

• The first image shows the variation of glitch rate and the microseismic ground motion during December 2024. It looks like the glitch rate was higher on the days when the microseismic ground motion was high. 
• The second image shows a scatter plot of glitch rate vs ground motion and from this we can see a correlation between the two. 
• We calculated the Pearson and Spearman correlation coefficients between the glitch rate and the microseismic ground motion in two frequency bands 0.1-0.3 Hz and 0.3-1.0 Hz. The correlation values are tabulated below. Ground motion in both the bands show good correlation with the glitch rate. But, the ground motion in 0.1-0.3 Hz is better correlated with the glitch rate. 

• The third image shows the frequency and SNR distribution of the glitches that we chose. There are a total of 11111 glitches. 
• The fourth image compares DARM spectra during a quiet time and a time of high microseismic ground motion and there are some excess noise visible during the latter.  

The attached pdf shows qscans of some of the glitches chosen from high microseismic days. It is not very clear from the qscans if the glitches have arch like morphology, which is usually the case for glitches caused by scattered light. But, some of the glitches do show repetition within short time interval which make us suspect if they are produced by scattered light. We are still working on this to figure out if it is the process of scattering which is upconverting the low frequency ground motion to create noise in the 10-60 Hz band.

In summary, it looks like the microseismic ground motion is affecting the DARM sensitivity at LHO, maybe not as much as it does at LLO.

WP12510 Add Geist environment sensors

Jonathan, Dave:

Last Friday I started five EPICS IOCs for the four new out-building sensors and one for the existing MSR sensor.

Yesterday I installed the MX and MY Geist watchdog 1250 sensor units in the CDS VEA racks (details in forthcoming alog).

Today we added the new EPICS Channels to the EDC+DAQ. A new H1EPICS_ENV.ini file was created, it is generated by running generate_env_ini.py

This new INI was added to the edcumaster.txt file.

WP12511 Add h1daqfw2 EPICS LOAD MON channels to DAQ

Jonathan, Dave:

A new H1EPICS_CDSMON.ini was created with the addition of h1daqfw2's channels.

New JAC Slow Controls channels

Daniel, Dave:

A previous slow controls upgraded generated new INI channels for the JAC system. Back then we elected to not add them to the DAQ immediately, rather wait for the next target-of-opportunity, which was today.

DAQ restart

Jonathan, Erik, Dave:

The DAQ was restarted to install the new channels for:

• JAC
• FW load monitoring
• ENV (EX, EY, MX, MY, MSR)

No major issues, GDS0 needed a second restart. Due to vent activities we did not take the trend writers offline for this restart, rather just restarted the EDC following the 0-leg restart.

Jonathan tested his new restart-less frame writer during this restart.

EDC channel count 59188 + 84 = 59,272 

The chilled water pumps at Mid X and Mid Y were run for several minutes today in order to exercise the bearings and re-wet the seals. This caused a drop in the temperature trending in NDSCOPE but is not an indication of a problem. 

Tue May 06 10:12:46 2025 INFO: Fill completed in 12min 43secs

The first four attached plots shows the regression of the PR3 damping signals from C/DHARD dofs (residual motion CHARD_P_IN1 etc).

The last two plots shows the regression of the HAM2 ISI signals from PR3 damping signals - the coherence and spectra.

The data is from one hours low noise H1 lock on March 27, 2025 6:00 UTC.

The corresponding results for L1 are in 76513

Betsy, Daniel, Fil, Elenna, Jennie, Sheila, Ryan S, Camilla

Day 1: 84193, Day 2: 84228, Day 3: 84230 and 84239

• The majority of the day was spent with Betsy and myself fixing the broken REFL ASC connectors, details in 84265.
  • Daniel confirmed all signals now look good.
  • Need to confirm what S/N cable is now in what position.
• Betsy installed the pico motor cable from feedthrough to cable bracket and we plugged in all picos. Fil and Daniel installed the in-air cable.
  • Betsy and I tested that the piocmotors on the SLED (controller 2, motors 3 and 4) worked, however the medm labels or cables are swapped (ASC REFL A centering =  ASC REFL B centering), need to decide which to change.
• Betsy and I aligned ASC REFL A and B diodes, we confirmed that all segments saw the signal (block, unblocked beam) and used the picos to center the beams, plots attached.
  • The reflected beam from ASC REFL A was was dumped
  • After we finished centering I couldn't  find that ASC REFL B reflected beam,  so we need to tilt the diode more to dump the beam and realign the diode. 
• Jennie and Elenna (84254) and then myself and Elenna (84266) finished the last REFL path beam profiling measurements needed.
  • Betsy and myself measured the distances between optics for the beam profiling measurements: 84263
• Sheila, Ryan and Elenna looked into isolating the ISIs and the pointing of optics to get POP flashes: 84261
  • All isolated apart from HAM3 kept tripping so was left in ISI_DAMPED_HEPI_OFFLINE but PR2 alignment was adjusted a few urad to compensate for this.
  • I could see POP flashes with the 200mW beam using the IR card AND IR viewer, it was missing the periscope by ~1" in -Y.

To do's from this alog: Confirm ASC REFL RF cable numbers, decide whether to swap SLED pico cables or medm names, tilt ASC REFL B diode to dump reflected beam and realign.

Other things that still need to be done in REFL path: Move LSC diodes to correct to distance (so beam has 0.2 to 0.3mm beam raduis) and check alignment of beam on these diodes.

After this we will be ready ot move onto the POP path, Keita is checking the expected separation of POP and ASLS beams as it's important we don't center the shared optics on the POP path only.

Dry air skid checks, water pump, kobelco, drying towers all nominal.

Dew point measurement at HAM1 -42.1 °C.

FAMIS26041 (I meant to do this famis task last Friday)

Most chambers are still elevated due to the vent.

Closes FAMIS26377

For the CS, everything looks fine. MR_FAN5_170_1 is the noisiest fan at ~ 0.4.

For the OUT buildings, MY_FAN1_270_1 looks fairly noisy and got a little worse 2 days ago.

Laser Status:
    NPRO output power is 1.828W
    AMP1 output power is 70.35W
    AMP2 output power is 140.2W
    NPRO watchdog is GREEN
    AMP1 watchdog is GREEN
    AMP2 watchdog is GREEN
    PDWD watchdog is GREEN

PMC:
    It has been locked 13 days, 19 hr 33 minutes
    Reflected power = 23.39W
    Transmitted power = 105.5W
    PowerSum = 128.9W

FSS:
    It has been locked for 0 days 17 hr and 26 min
    TPD[V] = 0.8035V

ISS:
    The diffracted power is around 3.6%
    Last saturation event was 1 days 23 hours and 16 minutes ago

Possible Issues:
    PMC reflected power is high

TITLE: 05/06 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
    SEI_ENV state: MAINTENANCE
    Wind: 9mph Gusts, 5mph 3min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.11 μm/s
QUICK SUMMARY: Vent work continues today with more HAM1 alignment and wind fence work.

Keita, Elenna, Jennie W.

We have taken three beam profile measurements along the REFL path on HAM1: one in the location where REFL WFS B will go, one in the location where REFL WFS A will go, and one measurement further "downstream" where we placed a steering mirror after where WFS B will go and steered back towards the edge of the table. We will post further details later, more measurements to be made along this path tomorrow.