model restarts logged for Tue 16/Feb/2016
2016_02_16 13:58 h1tcscs
2016_02_16 14:27 h1isietmy
2016_02_16 14:29 h1dc0
2016_02_16 14:31 h1broadcast0
2016_02_16 14:31 h1nds0
2016_02_16 14:31 h1nds1
2016_02_16 14:31 h1tw1

Maintenance day. TCS and ISI-ETMY model changes. H0EDCU_VE.ini change. Related DAQ restart.

Wed 10/Feb/2016 - Mon 15/Feb/2016 No restarts reported.

TCS model change. WP5732

Dave, Aidan:

Aidan made model changes to h1tcscs.mdl 

Buried EY Seismometer added to h1isietmy model. WP5733

Jeff, Robert, Dave:

h1isietmy was modifed to read the external seismometer, these channels were added to the science frame.

Adding all Vacuum Controls channels to the DAQ. WP5731

Dave:

All the vacuum controls channels were added to the DAQ. Increased the number of channels from 280 to 1777

DAQ restart

Jim, Dave:

DAQ was restarted to support all the above changes. After several hours h1fw1 became unsynchronized to h1fw0 (we had not seen this before). Restarting h1fw1 fixed this.

GLIBC patching

Dave:

Patched and restarted critical servers and about half of the control room workstations.

I filled the PSL diode chiller with 350ml of water.  Attached is a 30 day plot showing where the diode chilller was filled on Feb 4th, and then the alarm level increasing starting around Feb 13th.

I have increased the LLCV valve setting to 18% (from 15%) to try and compensate for changing conditions.

Factors that are changing are:

The supply pressure is falling as the liquid level in the storage dewar falls.

Heat leak to the liquid transfer line is increasing with outdoor temperature. This results in more vapor in the line and less liquid to the pump.

Radiation from the beam tube as the beam tube warms.

The attached plot shows 100 days of the control valve setting. The flat line at 15% is due to the failure of the differential pressure sensing- we have had to resort to manual setting of this valve.

Here is a 45-day trend of HEPI Pump Pressures & Drives (thus closing out FAMIS # 4517).

A few note:

• EX Pressure & Voltage drive, made slight shift down on Jan 5th (and the noise levels became quieter)
• EY Pressure & Voltage drive had shifts on 1/21 & 1/27;  in between these dates the noise levels were noisier than before and after.

Attached is a time series of TMSX roll osems over the last year, which has been slowly drifting in the same direction the whole time.  There was a went in June, wich is the larger jump, you can see when the sat amp started to oscillate a few weeks ago, and me moving roll over the weekend. This drift is mostly seen in F1, I don't know if it is real or not.  

Rana has told us that at LLO the TMS QPD positions are sensitive to roll.  Before closing the soft loops we moved roll by 250 urad (accroding to the osems) and saw that this had no impact on the error signals of the soft loops.  

Here are optical lever 7-day trends; closing out FAMIS # 4392.

Masayuki

The ISS 2nd loop medm screen has been Improved. The digital path has been shown.

I have noticed that h1fw1 is lagging quite a ways behind h1fw0 (they are normally synced to the second for the fast data CRC and within a few seconds of writing frame files). h1fw1 was lagging 1 to 10 seconds on CRC and up to 20 seconds for frame writing.

17:47PST, restart h1fw1 daqd process. h1fw1 is now back in sync. The lack of CRC sync lead me to think this is a daqd problem and not a Solaris QFS slow raid.

1610 - 1700 hrs local -> To and from Y-mid + Y-end

Today took nearly 25 minutes to over-fill CP3.  It is as if the pump is really low.  Like this past Sunday's fill the ambient temperature is unseasonably warm.  Also, the BT cleaning crew have been leaving the Y1 BTE doors propped open overnight.  Could the extra warm BT be significantly changing CP3's consumption (see Mike Z.'s comments from Sunday's entry)?  Also, there has been a slight uptick in Y-arm pressure for the past day or so.  I confirmed that IP9 is programmed in FIXED voltage @ 7000V and isn't stepping between voltages.  

Next scheduled over-fill to be Thursday, Feb. 18th before 4:00pm

J. Kissel, R. Schofield
WP 5733

Robert and I moved around cabling for the recently re-installed GND STS that has been buried +X (at ~10 m away from the building) of the ETMY building. This STS (S/N 89921) was formerly in the vault during O1. Now cabling for this STS is such that the BNCs, from the hostbox outside, go into channels 13, 14, 15 of the PEM BNC AA chassis at the very bottom of the TCS-C1 rack. Out of the back of this chassis, we've installed a new (standard L-Comm) DB9 cable which now goes to the front of the HEPI AA chassis, in U30 of the SEI-C1, in the port marked "27-30." This corresponds to the channels ADC3_[26-28], which have been pre-wired up to go into STS_A at the top level of the EY model. In addition, I've unhooked the top-level GND_STS block from the library, and added three new sets of channels,
H1:ISI-GND_STS_EYBURIED_X
H1:ISI-GND_STS_EYBURIED_Y
H1:ISI-GND_STS_EYBURIED_Z.
Each above channel prefix set has 
- test points stored to frames as "_DQ" (stored at 256 [Hz]), e.g. "H1:ISI-GND_STS_EYBURIED_Y_DQ";
- an EPICs "_MON" channel, e.g. "H1:ISI-GND_STS_EYBURIED_Y_MON";
- the full BLRMS infrastructure with EPICs channels "_BLRMS" with associated BLRMS extensions for the given band, e.g. "H1:ISI-GND_STS_EYBURIED_Y_BLRMS_30M_100M."
The modified top-level model has been committed to the userapps repo.

Because the buried STS readout system has two factors-ot-two differences between the nominal STS readout system (one factor from the fact that only the positive legs of the host box signals are carried in via BNC, and then the other is from reading out this single-ended, half-signal with a differential ADC), I had to create new calibration filters (in FM2 of the H1:ISI-ETMY_ST1_GNDSTSINF_A filter banks, called "Cal_Buried") which multiply the nominal gain (10.1725 [(nm/s) / ct]) by 4 to get the calibration right. Thus, all of the derivatives of the above mentioned channels are calibrated into (nm/s), as is the case for all other GND_STSs.
The modified filter file has been loaded, committed to the repo, and the new FMs have been turned on.

As of Feb 17 2016 at 00:00 UTC, these channels are calibrated and awesome.

Calibrated DTT Template saved as
/ligo/home/jeffrey.kissel/2016-02-16/2016-02-16_H1ISIETMY_GND_STS.xml

P.S. Robert suspects that the best position for the STS is roughly at ~20 [m] from the building because it balances high-frequency drop off of coherence and coupling to building tilt. I.e. 10 [m] is a little too close, so the building tilt pollutes the ~40 [mHz] signal just as much as the STS on the slab, where as at 40 [m], the ~0.5 [Hz] motion is no longer coherent enough to do viable subtraction. So, he may dig one more hole and move it to 20 m, but it'll likely stay put until the next time Robert's in town.

Jeff, Darkhand, Kiwamu,

We have been meaning to do this, but kept missing a chance to do. Today we have fliiped the sign of the ESD bias voltage both on ETMX and ETMY.

This alog only describes what we did on the digital front end models. We will perform a set of confirmation measurement and verify the H1 DARM model later.

See alog 22135 for a concise summary of the bias flip.

[Bias flip on ETMY ]

• We flipped the bias request voltage from +9.5 => -9.5 in SUS-ETMY_L3_LOCK_INBIAS
• No other change required in SUS-ETMY because there is a if statement in ISC_LOCK which automatically checks the change in the bias sign and compensate for it by changing the sign of L3_DRIVEALIGN_L2L_GAIN.
• In CAL-CS, we have changed DARM_FE_ETMY_L3_DRIVEALIGN_L2L_GAIN from +30 to -30.
• Similarly, we have changed DARM_ANALOG_ETMY_L3_GAIN from -1 to + 1.

[Bias flip on ETMX]

• We flipped the bias request voltage from -9.5 to +9.5.
• We flipped DRIVEALIGN ellements as follows.
  • SUS-ETMX_L3_DRIVEALIGN_L2L_GAIN, from +1 to -1
  • SUS-ETMX_L3_DRIVEALIGN_L2P_GAIN from +0.021 to -0.021
  • SUS-ETMX_L3_DRIVEALIGN_L2Y_GAIN from 0.007 to - 0.007
• We changed the force coefficient of the linearization from -124518.4 to + 124518.4 in SUS-ETMX_L3_ESDOUTF_LIN_FORCE_COEFF
• All the the above changes are now written in ISC_LOCK with a if statement, so that it automatically checks the bias sign and apply a proper set of parameters.
• We have not updated CAL-CS for ETMX because it is currently not in use.

I've added a block to the TCS model to estimate the real-time thermal lenses in each of the test masses and the surface curvature of each of the test masses.

OPTIC_LENS.mdl

The main new library block, shown in the attached figure, is OPTIC_LENS which takes as inputs:

• Ring heater power
• CO2 laser power (N/A for ETMs)
• IFO ARM power (calculated from H1:ASC-TR_X_B_SUM_OUTPUT and H1:ASC-YTR_X_B_SUM_OUTPUT)
• Also, the estimated surface absorption of the optic

Each of these inputs goes into a filter that contains (or will contain) the transfer function for power to diopters for that specific actuation. These are:

• H1:TCS-ITMX_SIM_SUB_DEFOCUS_RH   (single-pass substrate thermal lens from the ring heater)
• H1:TCS-ITMX_SIM_SUB_DEFOCUS_CO2  (single-pass substrate thermal lens from the CO2 laser)
• H1:TCS-ITMX_SIM_SUB_DEFOCUS_SELF (single-pass substrate thermal lens from self heating)
• H1:TCS-ITMX_SIM_SUB_DEFOCUS_STATIC (single-pass substrate static lens from construction)
• H1:TCS-ITMX_SIM_SURF_DEFOCUS_RH (single-pass surface lens from RH: inverse of the ROC change from the RH)
• H1:TCS-ITMX_SIM_SURF_DEFOCUS_SELF (single-pass surface lens from self heating: inverse of the ROC change from self heating)
• H1:TCS-ITMX_SIM_SURF_ROC_NOM (nominal ROC of the optic)

which are summed together to yield. These are outputs for the model:

• H1:TCS-ITMX_SIM_SUB_DEFOCUS_FULL_SINGLE_PASS: the total single pass thermal lens in diopters
• H1:TCS-ITMX_SIM_SURF_DEFOCUS_FULL (the inverse of the current ROC, including thermal effects)
• H1:TCS-ITMX_SIM_SURF_ROC_FULL (the current ROC, including thermal effects)

h1tcscs.mdl

There are four of these blocks in the main TCS model h1tcscs.mdl - one for each test mass. Although the ETMs contain inputs for CO2 lasers, these are grounded. See the attached figure.

The outputs from the four blocks are currently terminated. I plan on updating the model to create channels for COMMON and DIFFERENTIAL MICHELSON and ARM lenses and others, as required. This will be after I've verified that the model and filters are all working correctly.

• Title: 2/16 OWL Shift: 16:00-00:00UTC (08:00-16:00PDT), all times posted in UTC
  

• State of H1: Locking
  

• Shift Summary: Typical maintenance day

• Incoming Operator: Corey

• Activity Log:

• 15:15 Chris S to EX for beam tube sealing
• 16:03 John to LVEA to Join Bubba
• 16:15 John soft closing GV5&7 to crane genie lift over beam tube
• 16:27 Ryan B restarting alog
• 16:34 Christina/Karen to EY to clean
• 16:51 GVs opening
• 17:01 SNEWS test in
• 17:03 Fil to LVEA to check inventory
• 17:24 Coca-Cola truck on site
• 17:40 Norco truck on site
• 17:40 Hugh, Robert to LVEA
• 17:54 Christina, Karen to LVEA to clean
• 18:12 Fil to EX to install chasis
• 18:46 Hugh to ends for Hepi fluid level check
• 19:08 Fil to EY to install chasis
• 19:12 Kyle to MY
• 19:27 Hugh back
• 19:31 Kyle to EX
• 19:40 Fil back
• 20:23 Kyle back
• 21:56 Hguh moving seismometer into LVEA
• 23:03 Corey resetting PSL WD
• 23:31 Corey into optics lab
• 23:56 Kyle to MY to overfill CP3

At 23:03UTC (3:03pmPST) I reset the PSL Watchdog this afternoon & CLOSED FAMIS Request #3585.

Bubba and I were able to crane the Genie manlift over the Y beam manifold this morning in order to continue the crane rail work.

GV5 and GV7 were soft closed for the duration from UTC 16:15 to 16:53.

The crane behaved normally.

{Rana, Evan}

This evening we looked at the coupling of DCPD bias voltage into DARM.

Each DCPD is biased with +12 V from a linear regulator with a 1 Ω output resistor. We wanted to inject extra bias noise, so we exposed the bias lines with a breakout board at the DCPD chassy in the HAM6 rack (see D1300502). Then we summed in the output of an SR785 across an impedance of 10 kΩ in series with 20 µF (see diagram; green shows the normal DCPD electronics and red shows the addition). The 10 kΩ gives a 1:10⁴ ratio of bias fluctuation to drive voltage, and the capacitor ac-couples the drive so that it does not pull on the bias at dc.

With the SR875 we drove a 1 Vpk line first at 187.3 Hz and then at 62.4 Hz (i.e., the bias fluctuation was 70 µV rms). We looked at the response in DARM (at 2 W dc readout, with DARM still controlled by EX). We had 10 mA dc on each PD, in the 400 Ω transimpedance configuration.

For 187.3 Hz (from 02:03:00 to 05:22:00 Z), the magnitude in DARM was 1.13×10⁻¹⁸ m rms, which implies a coupling of 1.6×10⁻¹⁴ m/V. The noise of the regulator at this frequency was 0.9 µV/Hz^1/2, which implies a DARM noise of 1.4×10⁻²⁰ m/Hz^1/2.

For 62.4 Hz (from 05:33:00 to 05:56:30 Z), the magnitude in DARM was 1.4×10⁻¹⁸ m rms, which implies a coupling of 2.0×10⁻¹⁴ m/V. The noise of the regulator at this frequency was 1.9 µV/Hz^1/2, which implies a DARM noise of 3.7×10⁻²⁰ m/Hz^1/2. Note that there is some room for error here because the DARM control configuration here is not exactly the same as the low-noise configuration. However, the ugf and phase margin should not be too different in the two configurations.

This is very close to the current low-noise DARM noise floor. However, if DARM is limited by voltage noise of the regulators, we should either expect that the DCPD null stream is equal in magnitude to the DCPD sum (it isn't), or the regulator noises are coherent (they aren't).