I changed the label for button PICOMOTOR_H on the isc/commom/medm/ISC_CUST_PICOMOTOR_OVERVIEW.adl screen from "TCS HAM5" to "HWS table". I also changed the "name" argument associated with this button to "HWStable" so that this appears as the title of the picomotor screen that pops up when this button is pressed.

These changes were committed to the SVN.

Looking at the problems from last night I began looking at the ESD this AM.  Noticing that the main read back for driver was flickering on and off I by-passed the LV driver and looked only at the HV driver.  It was obvious the DC bias was not present.  Hoping it was not the Driver itself looked at the incoming DAC channels and all functioned except the DC bias channel.  Tracked it down to the output of the IO chassis.  Filiberto replaced the 18bit DAC, ribbon cable and interface card.  This cleared up the issue.  While working on this noticed the ESD can get into some odd states when the on/off switch is pressed.  I think we still have some binary issues that latch the unit up.  Was able to clear the problem by going to the BIO screen and changing states.

We've been working on a script to try improve the function of the rotation stage which controls the waveplate setting the power output to the CP.  The script is now working, but we had problems yesterday with stability of the rotation stage control. The Y rotation stage stopped responding to requests.  It's not clear what the cause of this problem is, but restarting the EtherCAT chassis for the rotation stages appears to fix the problem at least for a period of time.  After a first restart the X-arm rotation stage which is currently in use, started to exhibit problems.  The stage would correctly find home, but any angle request caused it to move randomly.

I re-calibrated the Y-arm rotation stage.  The X-arm one does not appear to be correctly calibrated which may be a BURT restore issue.  However this only affects calibration of power to angle, and should not affect operation for angle requests.

At this point I've completed the control script but don't plan on implementing it while the rotation stage is not working in a stable manor since the script asks the stage to move multiple times and may increase the tendency for it to fail.

Script is located at userapps/trunk/tcs/common/scripts/TCS_CO2P_SETPOWER.py

TJ, Laura

We've been working on tracking the RF45 glitches to veto them out of the astrophysical searches. We were initially tracking the control point of the RF45 MHz amplitude stabilization loop, H1:LSC-MOD_RF45_AM_CTRL_OUT_DQ, which seemed to be correlated with the severe glitching in DARM. However, as O1 progressed we noticed that the monitors we had built for this channel were missing some periods of severe glitching in DARM that we recognized as RF45 glitches. During these periods, the 45 MHz amplitude stabilization loop control point was completely quiet, but glitching was evident in MICH alignment degrees of freedom and DARM.

Since then, we've found that the most consistent witness to the RF45 glitches is a vertex alignment control signal, H1:ASC-AS_B_RF36_I_YAW_OUT_DQ. This signal is correlated with the glitches in DARM even when the 45 MHz amplitude control signal shows no discernable features. We've never seen any activity in the 9 MHz amplitude control signals during these glitchy periods. It doesn't seem like the amplitude stabilization of the 45 MHz sideband is the root cause of the noise in DARM, perhaps it's more of a witness to a broader issue in the RF electronics that commonly impacts generation of the 36 MHz signal.

19:42 UTC Started Conlog test code on conlog-test-master and connected it to the same 97,469 channels as the production code on h1conlog1-master.

We added and aligned the fast photodetectors (PDA50B) to the HWS table, per T1500579. The outputs from these photodetectors go into a whitening board (D1500430). The input channels in the ADC are adc0_22 and adc0_23 in the TCS CO2 model.

Alignment:

We used an IR card with mounted in a C-Ring and the Green Beam from the ALS to align to the center of the aperture on the photodiodes. See the before and after alignment photos:

Also attached is a photo of the whitening board.

Calibration:

The input channels in the real-time code are: H1:TCS-IFO_POLZ_S_HWS and H1:TCS-IFO_POLZ_P_HWS.

The filter banks in these channels:

1. Undo the whitening board (2x poles at 0.5Hz and 2x zeros at 10.9Hz)
2. Convert from counts to volts
3. Convert from volts to amps (+50dB transimpedance setting of PDA50B)
4. Convert from amps to Watts (divides by response = 0.56 A/W at 1064nm)

Also, the following DC offsets are added at the input based on measurements of the background value when the photodiodes were not illuminated:

• S_HWS_OFFSET = -28.76 counts
• P_HWS_OFFSET = 20.92 counts

From here, the outputs go into a 2x4 matrix for combination to produce S and P estimates for the power incident on the BS_AR surface. A rough estimate for these values are as follows:

• S_BSAR_INC = (1/R_BSAR_S) * (1/T_DCBS) * S_HWS= 1/0.027 * 1/300E-6 = 1.235E5 * S_HWS
• P_BSAR_INC = (1/R_BSAR_P) * (1/T_DCBS) *P_HWS = 1/34E-6 * 1/300E-6 = 9.8E7 * P_HWS

Based on the complex reflectivites of the BS for S and P, the ratio of S to P at the AS PORT increases by a factor of 39 from the BS_AR surface. We use the following parameters to calculate the polarization angle (in radians) at the AS PORT. 

• S_ASPORT = 38.8*S_BSAR_INC = 4.79E6 * S_HWS
• P_ASPORT = P_BSAR_INC
• AS_THETA = ATAN( SQRT( S_ASPORT/P_ASPORT ) ) ~ SQRT( S_ASPORT/P_ASPORT )

We made some changes to the TCS front end models yesterday morning:

- tcs/common/models/TCS_MASTER.mdl

Added two relative intensity noise channels for the IN-LOOP and OUT-LOOP photodiodes. They are constructed from the ISS_AC divided by ISS_DC (if ISS_DC <= 0, the denominator of this division defaults to 1).

New channels are: H1:TCS-ITMX_CO2_RIN_INLOOP, H1:TCS-ITMX_CO2_RIN_OUTLOOP and are saved to frames at 1024.

- tcs/common/models/TCS_POLZ.mdl

This is a new library block that takes the input from the IFO polarizaation sensors and creates channels that are S and P measurements at the HWS table, incident on the BS AR surface and at the AS PORT. There is also a channel that is an estimate of the polarization angle at the AS PORT. These new channels are all saved to commissioning frames at 1024

- tcs/h1/models/h1tcscs.mdl

Added adc0 channels 22 and 23 to the adc input. Injected these into the polarization block.

All these files were uploaded to the SVN.

State of H1: down more than 24 hours, ESD drive EY is restored to previous state (low voltage only)

Summary:

• Richard and Filiberto have been back and forth to EY since 7AM PT
• ESD drive at EY is now restored, however...
• ESD drive at EY only works in low voltage configuration, which is a bug...
• Richard and Filiberto are looking into why high voltage is not working
• ETMY optic, TMSY, BSC10 HEPI, and BSC10 ISI are restored, so...
• I am starting an alignment

Powered off h1susey to allow I/O chassis to be power-cycled to try to reset an 18 bit DAC which has a channel that is not outputting values. On power up, the IRIG-B timing acted as if it were disconnected, the time was absurdly off and the GPS time was not accurate.  Powered off the computer, let it rest for a minute, and powered it back on.  This time the IRIG-B time came back with a good time.

The h1susauxey computer needed to be power-cycled because it's I/O chassis was shut off by mistake.

The channel on the 18 bit DAC is still dead, Richard is hunting down parts to replace it.  Expect the h1susey computer to be powered off again.

I have been trying to reconcile the difference in laser room temperatures as measured by a temperature
sensor on the PSL table and the PSL environmental temperature sensor.  Back tracking through the
schematics ... etc. the signal H1:PSL-FSS_DINCO_REFCAV_AMBTEMP_OUTPUT corresponds to approximately
-2.207 V.  That would imply that the ambient temperature is over 40 degC, which clearly it is not.
An infrared camera image suggests the ambient temperature is around 22 degC with a couple of people
inside the enclosure.

    What may account for the discrepancy?
 - broken AD590 temperature sensor
 - non-zero voltage coming out of the DAC card, which is the ambient temperature reference signal
 - broken LT1021-7 in the temperature box
 - something else I cannot account for (always a distinct possibility)

Will try and sort this out in the next few weeks.

Jenne, Evan, Jim, Sheila

After the maintence period today the readbacks for the bias out of the ETMY ESD were zero.  Fil went out to EY to check on this, and saw that there was no voltage coming from the monitor channel, but did not know if there was actually bias going to the ESD.  

We tried to lock but failed to make the transition to ETMY.  After that we set up ETMY to be in high range, high voltage, linearization on (so that it would match ETMX) and drove them both in pitch.  The attached spectrum shows ETMY moving about 1/10 as much as ETMX, although they have the same drive state and the same output counts. This means we can't really noise hunt tonight. 

Liyuan Zhang (CIT), PeterK, JasonO, RickS

LLO has had some pre-modecleaner (PMC) issues.   S/N 08, their original unit, was removed due a "glitchy" PZT and the replacement unit, S/N 10 degraded quickly, apparently due to hydrocarbon contamination.

During the past week we have assembled a setup in the "Triples" lab (upstairs in the Staging building) to evaluate the performance of PMCs and, eventually, to repair them.

The setup is show in the photographs below.  Basically, we modematched a 2-W Innolight NPRO to the cavities which were removed from the tanks.  We used a Pcal integrating-sphere-based power sensor to measure incident, transmitted, reflected, and leakage power (through the two flat miirrors, M3 and M4).  We have ordered an amplitude-modulating EOM for making transfer function measurements to measure the bandwidth, but for these tests whe just measured the FWHM of the Airy profile by sweeping the laser frequency.

Removing the PMC from the tank turned out to be pretty challenging because the PMC bodies were stuck to the viton sheets at the bottom of the tanks.  We devised a method for removing them that works pretty well.  One can use the tank mounting blocks, the PMC body clamping bolts, and some Newport or Thorlabs "B2" bases to put upward pressure on the PMC.  On the one cavity we tested, it released within a minute or two. (See attached photo).

The measured performance of the "glitchy" PMC is broadly consistent with what Liyuan measured at Caltech.   Here we measured average losses per mirror of about 80 ppm.   We don't yet have an uncertainty estimate for this setup yet.

The estimated losses for the "dirty" PMC is about 1400 ppm per mirror.

Measurement details are in the .pdf file, attached.

We plan to repeat these measurements and estimate uncertainties.  However they clearly indicate that the "dirty" cavity is much more lossy than the "glitchy" cavity.  About 15% of the incident power is lost inside the cavity, vs. about 1% for the "glitchy" cavity.

Visual inspection of all four mirrors (from the outside) on both cavities did not reveal any noticable contamination or damage.

[JimW, Jenne, EvanH, Kiwamu, Sheila, Gabriele]

The new version of DTT seems only capable of getting a single channel of data at a time. 

We have tried several different kinds of channels, and DTT will only give you the first channel in the list.  No good.

I removed the following frequently changing channels from Conlog by adding them to the exclude list and regenerating the channel list.

- H1:ALS-X_CAM_ITM_PIT_POS
- H1:ALS-X_CAM_ITM_SUM
- H1:ALS-X_CAM_ITM_YAW_POS
- H1:IMC-ODC_CHANNEL_OUTMON
- H1:IMC-ODC_CHANNEL_PACK_PRE_PARITY
- H1:IMC-ODC_CHANNEL_STATUS
- H1:LSC-ODC_CHANNEL_OUTMON
- H1:LSC-ODC_CHANNEL_PACK_PRE_PARITY
- H1:SUS-SR3_M1_DITHER_P_OFFSET
- H1:TCS-ITMX_HWS_SLEDPOWERMON
- H1:TCS-ITMY_HWS_SLEDPOWERMON
- H1:VAC-EX_X4_PT525_PRESS_TORR
- H1:VAC-EX_X5_PT526_PRESS_TORR
- H1:VAC-EY_Y4_PT425_PRESS_TORR
- H1:VAC-EY_Y5_PT426_PRESS_TORR

As part of the prep work for firing up the high power oscillator, the optical surfaces were
inspected.  Some time back in 2014 there was a water leak over head 3 resulting in some of
the surfaces getting wet.

    The high power oscillator was opened.  A fragment of lens tissue was found - presumably
from the last wipe down of the laser that was done to accommodate a request from LLO - near
the lid switch close to head 4.  Some metal shards were in the location.  Small blue coloured
metal shards were also found near the Faraday rotator.  These we know to be from the blue
anodising of the lid of the oscillator.  There were signs of possible galvanic corrosion after
removing the side of the laser, and at the base of an internal support member.

    Signs of water drops on all four 4f lenses.  These were drag wiped off successfully:
Drag wiped 4f lens #1.  Mostly clean, small spot around 2 and 4 near the edge of the optic.
Drag wiped 4f lens #2.  Clean.
Drag wipes 4f lens #3, mostly clean, small spot around 2-2:30 near the edge of the optic.
Drag wiped 4f lens #4.  Possible dust spot between 11 and 12 o'clock, ~two-thirds from centre
to edge.

    After performing the drag wipe, the inside of the oscillator was wiped down again so as
to not kick up any dust during the remaining dust removal steps.

  The two quartz rotators looked fine.  All the laser rod surfaces appeared okay.  The dust
on the 45 degree dichroic mirrors was successfully blown off.

  There was a possible dust spot on output coupler at 3 o'clock as looking towards exit port.
This was blown off.

The pump light lenses looked okay with the possible exception of head 3, which looks like it
has some scratches.  The scratches were there before drag wiping, and it would see that the
laser operated previously with them.  O. Puncken was consulted.


Jason, Peter

Executive summary: 

In regard to narrow lines, the (nearly) full O1 H1 data set is little changed from what was reported for the first week's data: a pervasive 16-Hz comb persists throughout the CW search band (below 2000 Hz), accompanied by a much weaker and more sporadic 8-Hz comb; there remain several distinct 1-Hz and nearly-1-Hz combs below 140 Hz, along with other sporadic combs. The 1459.5 hours of 30-minute FScan SFTs used here span from September 18 to the morning of January 3. The improved statistics make weaker and finer structures more visible than in the 1st week's data. As a result, many new singlet lines have been tagged, and it has become apparent that some previously marked singlets actually belong to newly spotted comb structures. The improved statistics also make it more apparent that the originally spotted combs span a broader bandwidth than marked before 

Details: 

Using 1459.5 hours of FScan-generated, Hann-windowed, 30-minute SFTs, I have gone through the first 2000 Hz of the DARM displacement spectrum (CW search band) to identify lines that could contaminate CW searches. This study is very similar to prior studies of ER7 data, ER8 data and the first week of O1 data, but for completeness, I will repeat below some earlier findings.

Some sample displacement amplitude spectra are attached directly below, but more extensive sets of spectra are attached in a zipped file. As usual, the spectra look worse than they really are because single-bin lines (0.5 mHz wide) appear disproportionately wide in the graphics

A flat-file line list is attached with the same alphabetic coding as in the figures.

Findings:

 A 16-Hz comb pervades the entire 0-2000 Hz band (and well beyond, based on daily FScans)
 A typically much weaker and sporadic 8-Hz comb (odd harmonics) is also pervasive (all harmonics are labeled in figures, even when not visible)
 A 1-Hz comb with a 0.5-Hz offset is visible from 10.5 Hz to 133.5 Hz (previously was visible from 15.5 to 78.5 Hz)
 A 1-Hz comb with zero offset is visible from 16.0 Hz to 102.0 Hz (previously was visible from 20.0 to 68.0 Hz)
 A 99.9989-Hz comb is visible to its 11th harmonic (was previously visible to its 8th harmonic)
 The 60-Hz power mains comb is visible to its 9th harmonic (was previously visible to its 6th harmonic)
 There is a sporadic comb-on-comb with 0.088425-Hz fine spacing that appears with limited spans in three places near harmonics of 77, 154 and 231 Hz (ambiguity in precise fundamental frequency)
 There is a doublet 31.4127 and 31.4149 Hz comb visible to their 2nd harmonics (previously marked as only a 31.4149-Hz comb)
 There are three near-1-Hz combs not previous appreciated:

 0.99999-Hz comb from 19.2500 to 50.2497 Hz
 0.99816-Hz comb from 30.9430 to 60.8878 Hz
 0.9992-Hz comb from 30.9738 to 48.9594 Hz

 There is a doublet comb with spacings of 2.07412 and 2.07423 Hz, visible from their 9th to 32th harmonics (18.7-66.4 Hz) 
 There is a hint of another doublet comb with spacings of 99.9735 and 99.9784 Hz, but they are marked as singlets for now


Line label codes in figures:

b - Bounce mode (quad suspension)
r - Roll mode (quad suspension)
Q - Quad violin mode and harmonics
B - Beam splitter violin mode and harmonics
C - Calibration lines
M - Power mains (60 HZ)
s - 16-Hz comb
e - 8-Hz comb (odd harmonics)
O - 1-Hz comb (0.5-Hz offset)
o - weaker 1-Hz and near-1-Hz combs (various offsets, including zero)
H - 99.9989-Hz comb
J - 31.4127 and 31.4149-Hz combs
K - 0.088425-Hz comb
t - 2.07412 and 2.07423 combs
x - single line (not all singlets in the vicinity of quad violin modes are marked, given the upconversion)

Figure 1 - 0-2000 Hz 
Figure 2 - 20-50 Hz sub-band (shows complexity of combs below ~70 Hz)
Figure 3 - 50-100 Hz sub-band (shows continuation of strongest 1-Hz comb with offset of 0.5 Hz)
Figure 4 - 1300-1400 Hz sub-band (shows how clean the noise floor is away from 8-Hz, 16-Hz lines at high frequencies

Attachments:
* Zip file with miscellaneous sub-band spectra (with line labels)
* Flat-file list of lines marked on figures