Improved peak fitting in omc_scan

Trent, Matt, Georgia, Jennie W

fit_peak.py takes a chosen field and mode peak and fits a lorentzian curve to it. The problem with it was that all important arguments like gps_time, duration, field, and mode were all hard coded. We wanted to edit this code so that you can input these values as command line arguments. This new file fit_peak_w_args.py lives in my forked branch of labutils tgayer.

This code also calculates the finesse of the cavity from the calculated full width half max and prints it out to the command line.

We would also like to further improve this code by changing the fit so that we factor in the baseline of the data so that the tails of the fit better match the baseline.

An example of what you run on the command line [python3 /ligo/gitcommon/labutils/omc_scan/fit_peak_w_args.py 1393542033 80 "two watt single bounce side bands nominal" "single bounce" "carrier" "00"]

[/ligo/gitcommon/labutils/omc_scan/fit_peak.py]

[/ligo/gitcommon/labutils/omc_scan/fit_peak_w_args.py]

Updating OMCscan to reflect transition to OMC001

Matthew, Jennie W, Gabriele

 

In the initialization of the OMCscan code (which gets OMC scan data, analyzes it and then plots it), I updated several values to reflect the transition from OMC003 to OMC001, so that omc analyses are accurately done. For example, several small changes include:

• FSR -> 264.97 MHz
• Finesse --> 426.96
• TMSv --> 58.05 MHz
• TMSh --> 57.94 MHz
• PZT2response --> 6.435 MHz/V (this change is from 5.9 MHz/V, which is puzzling because the previous PZT2response [12.73 nm/V] is quite close to the new one [12.92 nm/V])

The values were obtained from T1500060 Table 19, which report the OMC optical test results for OMC001; note: the conversion from nm/V to MHz/V is found by the relation delta(f)/f = delta(L)/L, where delta(L) is 2*PZTresponse in nm/V, L is the round-trip cavity length, and f is 1064nm converted to MHz

Dark offset script run, SDFed

Georgia, Jenne, Gabriele, Elenna

We ran the dark offset script, but this time made sure to SDF everything appropriately. Screenshots attached.

Tuesday OPS End Shift Report.

TITLE: 03/05 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Preventive Maintenance
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 8mph Gusts, 6mph 5min avg
    Primary useism: 0.01 μm/s
    Secondary useism: 0.11 μm/s
QUICK SUMMARY:

Since Lunch time the control room has been enthralled with trying to relock H1 for the first time since Jan16th!
FIlter Tube Cavity is still closed.

LOCKING Notes:
TMS Alignment was tampered with before we realized that the Shutters were closed.
Using Camilla's alog with instructions on dithering alog 62024.
ALS Trouble Shooting before running TMS ITM and ETM Baffle Scripts.
ITMY Baffle PD4 was "Broken?"
Got light onto ETMY Baffle 1 and turned OFF the offsets to bring the light back to center then realigned ETMY via Guardian.

TMSX Baffle align script worked fine.
ITMX Baffle align Script failed to find any light.
But we didn't notice this and continued onto ETMX Baffle Align  which found light.
Unfortunately this was NOT ALS light that we saw on ETMX. Sheila and I decided to Shutter ALS and the light we found on ETMX Baffle 1 did not go away when shuttering ALS shutter on X arm.
Troubleshooting ALS X some more: We then used the Oplevs from January to find a good spot to take ETMX. 
First Light on ASLX 22:44 UTC
TroubleShooting PLL Problems that are making the X arm hard to lock. Eventually I got the ALSX beam spot looking as good as I could and then paused while Ryan S checked out the PLL.
Ryan S found out that the Beckhoff reboot turned off ALSX fiber Servo compensation. Checking out the SDF there was a dead give away that the problem was caused by a Beckhoff reboot. The date of the Change was Sun Dec 31 16:00:00 1989 which was the big clue.

DigiVideo3 Needed to be rebooted again.

Technical Notes:

ETMX Oplev positions AFTER we have good ALSX beam Spot:
H1:SUS-ETMX_L3_OPLEV_PIT_OUTPUT = -5
H1:SUS-ETMX_L3_OPLEV_YAW_OUTPUT = -12

ITMX Oplev positions AFTER we have good ALSX beam Spot:
H1:SUS-ITMX_L3_OPLEV_PIT_OUTPUT = -7
H1:SUS-ITMX_L3_OPLEV_YAW_OUTPUT = -1

ETMY Oplev positions AFTER we have good ALSY beam Spot:
H1:SUS-ETMY_L3_OPLEV_PIT_OUTPUT = -16
H1:SUS-ETMY_L3_OPLEV_YAW_OUTPUT = -13

ITMY Oplev positions AFTER we have good ALSY beam Spot:
H1:SUS-ITMY_L3_OPLEV_PIT_OUTPUT = -13
H1:SUS-ITMY_L3_OPLEV_YAW_OUTPUT = 9

Comm Beat note: 1.5
Diff beatnote : -14

The Beam splitter was touched up as well.
BeamSplitter Oplevs as they currently Stand:
H1:SUS-BS_M3_OPLEV_PIT_OUTPUT = 7
H1:SUS-BS_M3_OPLEV_YAW_OUTPUT = -9

Dark offsets by hand , for AS_C.

PR2 and IM4 were just touched up and then did Dark Offsets by hand for POP 45.
PR2  Witness sensors positions:
H1:SUS-PR2_M3_WIT_LMON = -29
H1:SUS-PR2_M3_WIT_PMON = 614

IM4 Witness Sensors postions:
H1:SUS-IM4_M1_DAMP_L_INMON = -41
H1:SUS-IM4_M1_DAMP_P_INMON =-2803

X ARM Sensor Correction has been turned off the Entire time due to Jim's BRS being rung up by him working on it.

Input Align Offloaded Successfully!!!
PRC_ALIGN OFFLOADED!!!
MICH_BRIGHT OFFLOADED!!!
INITIAL_ALIGNMENT COMPLETE!!!

                                                                                                                                 

Start Time	System	Name	Location	Lazer_Haz	Task	Time End
23:29	SAF	LASER SAFE	LVEA	-	The LVEA is LASER SAFE	03:14
16:03	FAC	Karen	LVEA	N	Technical Cleaning	18:15
16:04	Beckhoff	Daniel	Remotely	N	Slow controls.	17:19
16:05	CDS	Dave & Erik	MSR	N	Recovering DC0	17:12
16:06	EE	Ken	LVEA HighBay	N	Lamp replacement	20:06
16:18	EE	Fernando	LVEA	Beckhoff	slow controls work	18:18
16:26	FAC	Chris & Pest Con	LVEA	N	Pest control with pest control	17:21
16:29	FAC	Randy	LVEA West	N	Craning in the West bay.	18:08
16:36	FAC	Mitchel	EX	N	Taking Picts to document Wind Fence work	17:00
16:53	EE	Fil	LVEA	N	Measureing Cable tray lengths in the LVEA.	20:53
17:03	Contractor	ACE	EY	N	Effluent removal	19:03
17:10	FAC	Ken	FTCE	N	Getting lift taking to LVEA High bay	19:36
17:11	FAC	Mitchell	EX	N	Pulling Camera	19:48
17:12	CDS	Erik	HAM Shaq	n	moving compters	17:32
17:16	FAC	Christina	Water Tower and Overpass	N	Forklifting recycling OSB Receiving	18:46
17:20	PSL	Sheila , Trent, Matt	LVEA	N	Turning on side bands	17:29
17:21	FAC	Kim	EX	N	Technical cleaning	18:51
17:25	SEI	Jim Mitchel	EX	N	Camera removal	18:45
17:50	SUS	Austin	Control Rm	N	PR2 & MC2 TF	18:47
18:07	FAC	Tyler	Mid Y	N	Bearing Inventory	18:20
18:15	VAC	Janos	EX	N	OPENING GATE VALVE!!!	19:05
18:16	FAC	Karen	EY	N	Technical Cleaning	18:59
18:52	CDS	Jonathan Mike & tour	Roof	N	Tour going to the Roof.	19:00
19:01	FAC	Kim	Mid X	N	Technical Cleaning	19:50
19:06	VAC	Janos	LVEA	n	OPENING LVEA GATE VALVES!!!!	21:01
19:47	VAC	Gabriel Alena	LVEA	N	Taking Pictures Elena back First	20:15
20:31	Camera	McCarthy	End X	N	Installing PZT Camera	21:26
20:45	SEI	Jim	EndX	N	BRS Work	21:27
20:58	FAC	Eric	Mid Y Mid X	N	Running pumps	21:36
21:15	FAC	Tyler chris	X arm	N	Tumbleweeding wit Big Green	23:15
21:24	Dust	Ryan C	LVEA	N	Dust Mon Check	21:40
21:58	CDS	Marc	MY	-	Inventory	22:41
22:05	SQZ	Nutsinee Nioki	SQZT0	Local	SQZr Work , Nioki out early	01:05
22:16	VAC	Jordan	FCES	-	HAM8 RGA	23:16
22:50	EE	Marc & Daniel	Mid Y	N	Beckhoff mid Y	00:50
23:29	VAC	Jordan Gerado	LVEA	N	turbo pumps turning off	23:53

PMC realignment

Naoki, Nutsinee, Daniel

While scanning PMC, we noticed a large higher order mode. We went to SQZT0 and found that it was due to yaw misalignment. After we aligned PMC, we measured PMC input and transmission power. The PMC input power is 820mW and the transmission power is 680mW so the PMC transmissivity is 83%, which is much higher than before. We calibrated the SQZ-LASER_IR_DC_POWERMON and the SQZ-PMC_TRANS_DC_POWERMON to 820mW and 680mW, respectively. After we locked SHG, the SHG GR power reached 100mW.

We also put some beam dumps in SQZT0.

OMC ASC signal comparison

From Minhyo & Preet 
After changing of OMC control, it might be good to check the signals.
(Refer to Jennie's alog: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=76126, https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=76103)

Figures are comparison between Mar. 5 and one of the date in O4a (Jan. 2. 2024): 
1) SUS-OM3_LOCK comparison with both PIT and YAW
2) SUS-OMC_ASC comparison with both PIT and YAW
3) OMC-ASC_QPD A comparison with both PIT and YAW
4) OMC-ASC_QPD B comparison with both PIT and YAW

Seems that the signal went up in high frequency (4 ~ 10 Hz), which is pretty similar with QPD signals.

Update on Mar. 11

After locking of detector, compared the signals between O4a (Jan. 2, 2024, 05:00:00 UTC) and last Sat. (Mar. 9, 2024, 10:00:00 UTC).

1) SUS-OM3_LOCK comparison with both PIT and YAW
2) SUS-OMC_ASC comparison with both PIT and YAW
3) OMC-ASC_QPD A comparison with both PIT and YAW
4) OMC-ASC_QPD B comparison with both PIT and YAW

QPD signals went down from the last time (Mar. 5), which is similar to SUS (OM3, OMC) signals.

3-5 vent vacuum diary

Today's activities:
- EX gate valve (GV20) was OPENED. The pressure there is already 7.8E-9

- At the Corner, after the large IP controllers were replaced yesterday, the pressure was dropping continuously. As the IP data is not in the CDS (yet, because of swapping the controllers), see the attached pictures of the pressures directly at the IP orifices, at different times
- The main GVs (GV5; GV7) have been OPENED. The pressure is dropping nicely, it is currently at 7.5E-8 at the center
- All the turbo pumps (HAM6, OMC, Y-manifold, X-manifold) were valved out before/during GV openings

- HAM8 RGA will be ready for taking scans tomorrow
- HAM7 turbo was valved out, and it was also valved together with the short FCT section between BSC3 and HAM7 (FCV1 OPENED). The pressure rose a bit, but it is still at ~3.2E-7
- The pumping of HAM7 RGA was stopped, the cart was taken away
- FCT further schedule: after HAM8 RGA scan was done - and it is satisfactory - the whole relay tube - HAM7 - HAM8 line will be valved together and into the main volume (opening RV1, FCV2, FCV3, FCV4, FCV7, FCV8), and HAM8 turbo will be valved out. This will happen tomorrow (3-6)

In the rest of the day, the vacuum team will switch off turbos and clean up in the LVEA.
And, there will be some pumpdown statistics and summary posted to aLog later on.

One useful additional info: both for GV5 and GV7 50 psi was required to open up

The Turbo stations in the LVEA and at EX are now switched off. The details:
X-manifold:
Bearings: 100%
Turbo hours: 1948
Scroll hours: 1945

Y-manifold:
Bearings: 100%
Turbo hours: 950
Scroll hours: 2278

OMC:
Bearings: 100%
Turbo hours: 5963
Scroll hours: 5902

EX:
Bearings: 100%
Turbo hours: 1091
Scroll hours: 7141

CDS Maintenance Summary: Tuesday 5th March 2024

WP11737 Guardian upgrade

TJ, Jamie, Erik:

The guardian system was upgraded and h1guardian1 was rebooted. No DAQ restart needed.

WP11705, 11710 Beckhoff Optical Levers

Fernando, Daniel, Fil, Marc, Dave:

The new AUX Beckhoff PLC code was installed for the AUX systems at CS, EX and EY. New INI files were added to the EDC, a DAQ restart was required. The csaux, exaux and eyaux SDFs were restarted with the new channel lists. New channels have been accepted and monitored.

WP11720 Install Standard Front End Computer for h1cdsh8

Erik:

When originally installed, h1cdsh8 got a 14core W2275 front end computer because it was the only one on hand at the time. Erik replaced this with a standard 8core W2245 this morning, to free up the larger machine for other duties.

WP11739 Add new Pulizzi POE power controller channels to the DAQ

Dave:

H1EPICS_CDSACPWR.ini was upgraded. The new POE controller Pulizzi channels were added. EDC and DAQ restart was required.

WP11667 Add RACCESS channels back to DAQ

Dave:

H1EPICS_RACCESS.ini was added back to the edcumaster. It had been temporarily removed when we had EDC issues with an upgraded cdslogin. EDC and DAQ restart was required.

WP11668 Add CDS_LOAD_MON channels for cdslogin and cdsssh to DAQ

Dave:

cdslogin and cdsssh's CDS Monitor channels were added to H1EPICS_CDSMON.ini. EDC and DAQ restart was needed.

Update CDS SDF with Triplite and Pulizzi channels

Dave:

The h1cdssdf system as missing the latest Triplite and Pulizzi channels. These were added to the monitor.req and safe.snap files, sdf process was restarted.

DAQ Restart

Dave:

The DAQ and EDC were restarted for the above changes. Systems and num channels added are:

Pulizzi POE power control: +8

RACCESS remote access: +51

CDS load Mon: +24

Beckhoff CS AUX: +113

Beckhoff EX AUX: +35

Beckhoff EY AUX: +35

Total +266

EDC num channels

was	56,655
now	56,921 (+266)

DAQ restart was unremarkable, no GDS restart was needed.

 

Tuesday OPS Mid Day Pre LOCK report.

TITLE: 03/05 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Preventive Maintenance
CURRENT ENVIRONMENT:
    SEI_ENV state: MAINTENANCE
    Wind: 5mph Gusts, 2mph 5min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.12 μm/s
QUICK SUMMARY:

Erik and Dave recover the DC0

Erik and Dave replace front End Computer for HAM8 SEI

IMC Lock Node in Error 17:09 UTC
PSL Enclosure temperatures  
Landscapers came by to land scape
Guardian reboot at 17:35 UTC and is back at 17:40 UTC
Austin took some PR2 & MC2 TF
PSL Dust alarm 17:55 UTC

Gate Valve GV7 is now OPEN! Thank you Vacuum team.
Please Start Checking your Sub System's SDF Diffs.
I will start the baffle align script and Start trying to get ALS X going while the VAC team gets the last Gate Valve opened.

SDFs for OMC

I took a pic of the OMC SDF changes we made before SDF revert erases everything when the IFO goes through the locking stages which could happen today.

We might want to keep these changes but I am not sure we won't have to change some of them once the whole IFO is aligned, this also might neccesitate some guardian code changes as some of the OMC values are hard coded in.

OMC scan with nicely (newly) aligned OMC with PSL beam

Matt, Trent, Jennie W , Sheila, Craig, Elenna

Executive Summary:
    - We found that the new scan indicates the OMC may have new mode spacing between sagittal and tangential modes. We took a 100s scan of the new OMC and without sidebands on, and we zoomed in on the C02 mode peak to examine the astigmatic peak separation, but we found none, as shown in image 3.

The first image is the entire scan, and we are curious to know if the astigmatic separation is much larger than before, leading to a MUCH smaller second peak that our fitting function skips over. The OMCscan fit with nonlinear corrections can be found here

How we did the scan:
  - We did 2 100s scan, sweeping H1:OMC-PZT_EXC and reading H1:OMC-DCPD_SUM_OUT_DQ

  - We added a Dark offset of .0034 mA to DCPD

  - the scan template required that we set the PZT2 offset to - 50

Progress on pcal migration to python - KDVM investigations.

D. Barker, T. Sanchez, D.Bhattacharjee, R. Savage, F. Llamas

Characterized the latency of a Keithley digital volt-meter (KDVM or DVM) by making measurements of a Martel voltage source using kvm_read.py.

The objectives:

• To see the difference between measuring with the "auto-zero" and "auto-gain" functions set to "ON", or "ONCE" for a given measurement.
• To characterize the latency when asking the DVM to give us data from both, the "ON" and the "ONCE", configurations.

The Martel was injecting 1 V to the KDVM. The python script took 1000 *immediate* samples from the DVM.

The pdf "kvm_ConfigComparison" illustrates the results from our measuremnt. The discretization of the data comes from rounding the time to 3 significant figures and the voltage to 5 significant figures (my bad). The top plot shows a voltage timeseries for both tests. Notice that the measurement labeled as "Once" finished acquiring the thousand samples in ~ 3 seconds before the measurement labeled as "On". The bottom plot shows the time difference between the samples. It is clear however that when the DVM is periodically using the auto-gain and auto-zero functions, the sampling will have inconsistent latencies.

A third measurement was done to characterize, with a better resolution, the latency of the DVM sampling function when the auto-zero and auto-gain are set to "Once". The results are shown on the pdf "kvm_LatencyTest1". A mean of the time difference between samples yield ~ 18.5 ms, which is interesting since 1 PLC takes 1/60Hz ~ 16.6 ms. The histogram shows that, from the Keithley itself, there is a variation of 0.012%.

In summary:

• The auto-zero and auto-gain functions are not needed for our application and are rather hurting instead of helping take accurate measuremens
• For an integration time of 1 PLC, the latency to acquire data from a DVM is of 18.5 ms and the variation is of 1.2 hops

It is probably helpful to follow up with a technical note on how to (will be) configure(d) the Keithley before making a measurement and to document the jargon of the useful configurations and functions used for pcal responsivity ratio measurements.

OMC Sensing matrix measurement done + ASC converges

Jennie W, Sheila, Matt Todd, Trent Gayer, Craig

Summary: Measured summing of QPDs on OMC and remeasured the sensing matrix and updated the input matrix for the OMC ASC. We also locked the OMC with no sidebands and took an OMC scan. The AS loops look good but we might need to check the summing of the QPDs offline as we weren't sure if the yaw quadrant summing of QPD B has the correct sign.

---

First we checked that the quadrants of the two OMC QPDs are summed the correct way.

When OM3 pitch is stepped up, as in image 1:

• the quadrant 1 and 4 increase and 2 and 3 decrease. This is the same for both QPDs.
• Though the A QPD responds more to OM3 than B which should be the correct way round.

Image 2 shows the same thing when we increase the yaw signal on OM3 segment:

• 1 and 2 increase and segment 3 and 4 decrease for QPD A.
• 1 and 2 decrease and segment 3 and 4 increase for QPD B.
• As for pitch the response of QPD A seems slightly less sensitive to OM3 than B which we expect.

---

See image 3 and 4 for the original input and output matrices for the OMC ASC before our measurements started today.

We measured the sensing matrix from pitch of OM3 to  pitch and yaw of A and B by moving OM3 up in pitch by 50 counts. See image 5. You can see here that yaw on the QPDs is cross-coupled  to pitch which we ignore for the purposes of this analysis but is good to note.

We measured the sensing matrix from pitch of OMC to  pitch and yaw of A and B by moving OMC up in pitch by 500 counts (as it was hard to see a rerponse for 50). See image 6.

For both we divided the step on the QPD by 50, this might make the matrix coefficients differently scaled but we can fix this later by changing the loop bandwidths if neccessary.

This must be done with the ASC off so the step response can be seen.

Then we inverted this sensing matrix to get the input matrix after zeroing the output values for ANG Y and OM3 and POS Y and OMC sus.

We also have changed the output matrix in pitch to feedback only to feedback to OM3 for POS DOF and OMC only for ANG DOF to simplify the loops. See image 7.

Next we measure the yaw with steps of 50 counts again in the same way. Shown in image 8 and 9.

Then we changed output to decouple yaw drives as for pitch so pos drive only goes to OM3 and ANG drive only to OMC sus.

---

 

Then Craig and Matt had to re-measure the OMC  in yaw as had measured the OMC offset drive wrongly.

The new input matrix after inverting our sensing matrix for yaw and scaliing it by a factor of 50 is shown in image 10.

New output matrix (no value changes, just turned off POS feedback to OMC sus and ANG feedback to OM3) is shown in image 11.

ASC now converges. Image 12 shows it working.

Matt locked the OMC with ASC on shown in image 13.

I've put in a "fake dark offset" OMC DCPD A to make the data from the OMC scans never be less than zero.  This is still engaged at the moment.

EX Windfence repairs

Randy, Mitchell

Yesterday Randy and I were able to start and complete minor repairs needed at the EX windfence. One section had a horizontal wire break at the upright bracket. A section on cable was secured into the bracket and clamped to either end of the broken cable. The cables were tentioned before cable clamps were added. Two horizontal cables below on the same upright needed to be retentioned as well.

On a different upright the nuts had backed completely off the horizontal bracket. New nuts were put inplace and locktite was used. Hog rings were installed as needed.

New minor Guardian release and h1guardian1 rebooted

WP11737

Jamie, Erik, TJ

We released a new Guardian revision 1.5.1 then quickly found some small bugs and then pushed 1.5.2. The h1guardian1 machine was rebooted today to take in the new changes, after Erik pulled in the new guardian and guardctrl packages. All nodes restarted without any major issues. Release notes:

• improve MEDM "STATES" screen control interface
  • Added more node and manager information
• add MEDM button to access code history through gitk interface
  • This removed the "related" button that has been broken for some time
• remove default 100 line log limit
• bug fixes (1.5.2)

 

Seeding Round 2

Dhruva, Nutsinee

Now that we are able to sqz with 40 times more CLF power than before we did a "quick" seeding measurement again. First we went out to measure seeing with a 1611 and saw nothing. We then repeated the measurement that was done in alog67050. The steps are

1) offset SQZ laser VCO by -1000 Hz
2) Demodulate LO with 3.124 or 3.126 MHz. This requires switching out RF output in the CER using the IFR set to 12dBm. 
3) Unplug the RF output that is in the rack left to the IFR, U4 channel 1. Send the new RF demod into ISC SQZ 30.
4) Back in the control room. Lock LO loop. 
Note that this is a Homodyne measurement.

On Friday we didn't see anything. Later I realized we could have lowered CLF power by half at the time so I cranked in back up on Monday to 0.7 mW on the CLF launch and repeated the test. The result was we saw 36 nW of seeing at 0.7 mW CLF output measured after AOM2. The plot attached is in V/sqrt(Hz). Shotnoise ~10^-7 V/sqrt(Hz) corresponds to 0.5mW of LO per PD. The seeding peak at 1 kHz measured 2.78*10^-6 V/sqrt(Hz). Solve for current then use the responsivity of 0.8 A/W to get 36nW. 

 

Attached the plot and a Mathematica notebook for more detailed calculations. Daniel suggested there shouldn't be a sqrt(2) in the seed power calculation but maybe a factor of 2. Seeding might be slightly higher or lower. Will update this alog again once I've done the math.

h1cdsh8 front end server replaced

The h1cdsh8 front end server, which was a 14-core ws2275 was replaced with a standard 8-core ws2245. 

The ws2275 is still operational and is on a shelf in the MSR.