[Stefan, Jenne]

We have been looking at different PD signals to see if we can set up a guardian check that the fast shutter indeed closes upon lockloss.  Right now we don't have a good way to code it in guardian yet, so Stefan has made a template for the lockloss script that we can run every lockloss. 

Note that AS_C and ASAIR_B are both before the shutter, so should not see a change in light upon shutter close.  AS_A, AS_B, and OMC_DCPD are all after the shutter, and should lose light if the shutter successfully closes. 

For now, we have some DTT snapshots.  NDS2 is under maintenence, so not all plots have the same traces, sorry.

• The first one (2016_08_01_0843utc.png) shows an example of the shutter not closing. Signals are scaled arbitrarily to make them all visible. In this particular case, the wire was probably already broken, because we send the trigger signal requesting the shutter to close, but it doesn't actually close (NB the shutter state is readback by Beckhoff, so the timing of that signal can only be trusted to within ~1 sec or so).  All of the signals except for the DCPDs saturate, but it's clear (particularly in comparison to the next 2 plots) that the light doesn't get shuttered.  Note that the OMC was already offline at this time, so the OMC DCPD signal isn't super useful here.
• The second plot (2016_03_08_0430utc.png) shows a time from before the April HAM6 vent, so a time when the shutter was definitely working. In particular, note that the AS_A and AS_B sums get cut off at 5.4sec.  This is the shutter closing.  However, there is some kind of bounce or something, where the shutter allows light through for another 20 msec or so before closing and staying closed. 
• The third plot is the most clear version of the shutter working (2016_08_01_0020utc.png), showing again AS_A and AS_B sums getting cut off.  You can also see ASAIR gets the true peak of light after the shutter does the 'bounce and re-close' thing. 
• The fourth plot (2016_07_22_0745utc.png) shows a time when the shutter's trigger failed, and so the shutter didn't close.

Once NDS2 is back online, we'll post a few more plots of a few more locklosses, with both ASAIR and OMC DCPD included. 

Also, we don't want to use signals such as the OMC input pointing QPDs that are whitened for this fast information. 

J. Kissel
ECR E1500045
FRS 6014
WP 6051

I've completed the installation of the infrastructure for individual switch-ability for all QUAD's PUM / L2 actuation stage coil drivers. This stage should now have the same functionality as the BS M2, and all of the M3 stages of the HSTSs (the MCs, the PRs, and the SRs). I've created EUL2OSEM burt files for the walk through switching the state of each coil, so the only step that remains to take advantage of the new infrastructure is to modify the ISC_LOCK guardian's COIL_DRIVERS state to include the QUADs.

Details:
-------------
This update includes (this should double as an installation checklist for LLO)
- Updates to the ITM and ETM QUAD library parts: (see screenshots in LHO aLOG 28906)
      /opt/rtcds/userapps/release/sus/common/models/
      QUAD_MASTER.mdl                  << Modified the BIO block to use the new Individually Controlled 
                                          PUM library block, arranged connections from BIO to L2 
                                          blocks accordingly
      QUAD_ITM_MASTER.mdl              << (same as above) Modified the BIO block to use the new 
                                          Individually Controlled PUM library block, arranged 
                                          connections from BIO to L2 blocks accordingly
      STATE_BIO_MASTER.mdl             << Created new library block for individual control of PUM driver
      FOUROSEM_DAMPED_STAGE_MASTER_WITH_DAMP_MODE.mdl     << Modified COILOUTF bank to accept individual control
                                                       and switched EUL2OSEM matrix from static to ramping
- Corresponding updates to the impacted ITM and ETM QUAD MEDM screens,
      /opt/rtcds/userapps/trunk/sus/common/medm/quad/
      SUS_CUST_QUAD_OVERVIEW.adl
      SUS_CUST_QUAD_ITM_OVERVIEW.adl
      SUS_CUST_QUAD_BIO.adl
      SUS_CUST_QUAD_ITM_BIO.adl
      SUS_CUST_QUAD_L2_EUL2OSEM.adl
- Restored the nominal L2 EUL2OSEM matrix values, given the new channel names from the ramp matrix part
- Installed a ramp time of 8.0 sec for the EUL2OSEM matrix, as is the case for the other Triples that have been modified previously
- Restored the nominal coil driver state and test/coil enable for each coil individually (LHO likes these to be in State 2 (acq ON lp OFF) in the DOWN state, currently)
- Creating new burt configuration files for the EUL2OSEM matrices as each coil is turned off. These now live and been committed here:
      /opt/rtcds/userapps/trunk/isc/h1/scripts/sus/
      ?tm?_l2_out_ll.snap
      ?tm?_l2_out_lr.snap
      ?tm?_l2_out_normal.snap
      ?tm?_l2_out_ul.snap
      ?tm?_l2_out_ur.snap
- Reconciling the SDF system such that 
      - All new EUL2OSEM matrix channels are accepted and monitored
      - All new individual BIO channels are accepted and monitored
      - The EUL2OSEM ramp time has been accepted and monitored

This closes out the above mentioned work permits and FRS for H1.

No resets needed, all saturation counts 0.

FAMIS# 7066

   Took pet swipe in HAM6 on 08/02/2016 when doors were removed. Swipe was from -X side of OMC weldment to side of table, (see photo). 

   At the same time, removed 4" witness wafer from HAM6. Sample has been prepared for analysis (DCC T1600336). See photo for location of wafer.   

Krishna

I'm driving a piezo stack placed under the c-BRS platform with a function generator (drive frequency 5-200 mHz, 10V pk-pk) located on the floor in the CER. This setup is temporary and will be removed in a few hours.

There are no significant changes from last week. OSC DB3 still showing incorrect current measurement and AMP chiller flow rate showing very slkow steady decrease.

[Betsy TJ Koji]

We measured the optical powers in HAM6 before OM1-3 and OMC as well as at the transmission of OM1 and OM3.
Taking the ratio of these powers and AS_C QPD SUM, we get the calibration of the optical powers in HAM6.
This will be used for throughput calculation of the entire IFO output optics.

The ratio is

(Power into HAM6)/(AS_C QPD SUM) = 0.0162 +/- 0.0005 [mW/cnt]

Caveat: Our measurements were dominated by the systematic error of the power meter (as usual). Estimated size of the error is +/-3%.
Without having a better power meter, I don't expect I can improve this number.

As you can see in the spread sheet, each measurement has small statistical error like 0.1~0.5%. The numbers for AS_C QPD SUM was taken by the data downloaded from CDS.

However, two measurements of the optical power before OM1 (blue) showed 5% discrepancy. Also tracing the power along the main HAM6 path, the fluctuation of the power measurement (blue plot) exceeds the expected reduction of the power (orange plot) by measured transmissions. (Note that OM1 and OM3 has 800ppm and 1.6% transmission, that is consistent with the measurement in 2015.)

Therefore, this big fluctuation was concluded to be a systematic error of the measurement, particulary of the power meter (Ophir diode type sensor head), considering the stability of the AS_C QPD SUM.
From the two measurement of the optical power before OM1, and the plot, the systematic error was estimated to be +/-3%.

See LHO aLog 28649.  The voltage regulator has been replaced and this original unit returned to service.

WP 6063  FRS 5948

J. Kissel

There are several front end models with out-standing SDF diffs prior to the timing system upgrade boot fest. I attach screenshots because there's no chance we can reconcile these before all front ends go down. 
Models with the biggest collection of DIFFs:
h1pslfss
h1psliss
h1pslpmc
h1asc
h1lsc
h1sysecatc1plc1
h1sysecatc1plc2
h1sysecatc1plc3

Hope this helps us in the recovery!

Travis S, Darkhan T,

Overview

On August 4th PcalX maintenance measurements were taken at EX. During maintenance we adjusted AOM bias. The offset value for the OFS drive was also adjusted, but incorrectly (see details). Today we re-ajusted OFS offset value to +5.14V.

Details

During Pcal maintenance measurements of the Pcal laser power in the Transmitter module were taken to assess Pcal drive range. Analysis of the measurement will give us an efficiency of an AOM driver (max diffracted power), assessment of pick-off power levels for the transmitter PD, OFS PD, maximum available output power modulation range. At the moment the measurements have not been fully analyzed.

The AOM bias voltage was adjusted for it to give ~50% out of max available AOM drive when the OFS loop is open. The OFS offset was adjusted accordingly in the front-end. Originally this was not done correctly, cause the OFS offset was adjusted when the loop was not closed.

Later when Evan tried to close the loop, the error signal, i.e. sum of the OFS PD signal and the OFS offset (when excitations and injections are turned off), saturated the AOM driver.

This is now fixed by adjusting the OFS offset to +5.14V.

We turned back on the excitations (currently one line with 39322 ct amplitude at 3501.3 Hz) and hardware injections.

[Betsy, TJ, Fil, RichM, Peter, Koji, Calum, RichA (remote)]

Progress

- The OMCT path (Y+) was aligned. Using 10W input instead of 2W helped us to find the leakage cavity flash through the PZT mirrors with 50ppm transmission. In the end we confirmed that the beam is well close to the center of the viewport (simulator) using a sensor card+an IR viewer and a CDD setup.

- The other OMCT beam (Y- side) was found to be blocked by one of the panels of the black glass shroud. We needed to shift the Y- panel more than the allowed range by the design. In order to allow this shift, we flipped a PEEK bushing and removed some of the washers and viton O-rings on the captive compliant fixing screws. Two screws to hold the glass holder were also removed to widen the sliding range. (See attachment 1, 2)

- The light powers in HAM6 were measured at various positions. This will be analyzed in a different entry.

- The funtionarity of the picomotors were confirmed (cf. channel assignment ALOG LHO 16422). Also we confirmed that the two beam diverters are working.

- We found that failure of the fast mechanical. We could not close the shutter because the coil wire has an open connection. Close inspection revealed that the coil wire has a kink (attachment 3). By jiggling this kink we actually could fluctuate the coil resistance from 20Ohm to infinity. The replacement shutter was immediately sent from CIT. It is coming here tomorrow morning.

To Do (Tue afternoon)

- Install the replacement fast shutter.
- Ground loop check
- Other SUS/SEI exit procedure

Evan G., Jeff K.

We investigated the time delay of an injection made at H1:CAL-PCALY_SWEPT_SINE_EXC to H1:CAL-PCALY_TX_PD_VOLTS. The measured phase delay is 100.2 deg at 960 Hz (or equivalent delay of 290 usec).

Attempting to account for the phase from timing diagrams (see attached diagram, figure 1, DCC: G1501170), we find:

15.05 deg ("43.5 usec") 16-64k IOP upsampling
21    deg (61 usec)     DAC "processing" delay
13.11 deg ("37.9 usec") AI (analog)
--- PCAL --- (negligible)
13.11 deg ("37.9 usec") AI (analog)
21    deg (61 usec)     ADC "averaging" delay
15.05 deg ("43.5 usec") 64-16k IOP downsampling

Total = 98.3 deg ("284.8 usec")

We wanted to repeat the measurements made by Shivaraj at LLO (aLOG 27207) and verify the measurements yield consistent timing results.

We injected a 960 Hz signal in H1:CAL-PCALY_SWEPT_SINE_EXC and recorded the following time series (see figure attached):
H1:CAL-PCALY_TX_PD_VOLTS_OUT
H1:CAL-PCALY_RX_PD_VOLTS_OUT  (measured to be equivalent to the TX_PD channel)
H1:IOP-ISC_EY_ADC_DT_OUT
H1:CAL-PCALY_FPGA_DTONE_IN1
H1:CAL-PCALY_DAC_NONFILT_DTONE_IN1
H1:CAL-PCALY_DAC_FILT_DTONE_IN1
H1:CAL-PCALY_SWEPT_SINE_EXC 

The phase of the H1:IOP-ISC_EY_ADC_DT_OUT signal (measured at 65k) is determined from the amplitude of the signal at time T=0 with respect to the overall peak amplitude. We measured -2551 cts amplitude at T=0, where the sine wave has a peak amplitude of 4090 cts. The arcsine of the ratio gives the phase of -38.6 degrees in the range of -90 deg. to +90 deg. Since the true phase is outside of this range (more negative) we wrap instead to 180+38.6 = 218.6 degrees. This is the reference phase of the DuoTone signal.

We repeated the measurements to find the phases of all of the channels:
H1:IOP-ISC_EY_ADC_DT_OUT           = 218.6 deg <---- marked as time T=0
H1:CAL-PCALY_FPGA_DTONE_IN1        = 203.8 deg <---- also T=0 but filtered by the digital AA
H1:CAL-PCALY_SWEPT_SINE_EXC        = 219.4 deg
H1:CAL-PCALY_DAC_NONFILT_DTONE_IN1 = 182.7 deg
H1:CAL-PCALY_DAC_FILT_DTONE_IN1    = 100.2 deg
H1:CAL-PCALY_TX_PD_VOLTS_OUT       =  96.5 deg
H1:CAL-PCALY_RX_PD_VOLTS_OUT       =  96.4 deg

The difference in phase of H1:CAL-PCALY_SWEPT_SINE_EXC to H1:CAL-PCALY_TX_PD_VOLTS_OUT is 122.9 degrees at 960 Hz (= 355.6 usec)

So now there is a weird discrepancy in our measurements. The H1:CAL-PCALY_SWEPT_SINE_EXC to H1:CAL-PCALY_TX_PD_VOLTS measurement says 290 usec. The timing diagram predicts 285 usec. The phase of the waves measured in a time series is 355.6 usec. The difference is about 61 usec.

Still open questions:

1. Why doesn't the transfer function agree with the phase counting? Phase counting appears to have an additional 16k clock cycle that the transfer function does not have.
2. Why does the NOFILT_DUOTONE path not have 2 16k clock cycles delay?
3. This still does not answer why the DARM open loop gain model is missing ~45 usec delay (15 usec in sensing chain and 30 usec in the actuation chain)
4. And we still don't understand the timestamp of the ADC "averaging" delay when it comes to these kinds of transfer function measurements

Returned laser power to 2.4W, see alog 28931.  Jenne set the guardian LASER_PWR node to 2.0 W.  Note, the PSL rotation stage is operational and not locked out.

Krishna

The low-frequency (<0.1 Hz) noise in c-BRS looked much higher (~10X) than normal and we also noticed that the frequency and Q of the system were lower than that measured at UW. We suspected that the flexures suspending the beam-balance might have been damaged. Therefore, today I replaced them with spares that are fairly close to the original ones in size.

As a consequence of the flexure replacement, the mass distribution of the beam-balance has changed, hence the 'd' value is likely to be much larger than the previous value (~ 3 micrometers). The balance has currently been tuned to ~50 mHz frequency, which is close to the original resonance frequency. As is normal after a new flexure replacement, the mean angular position is drifting in one direction and will likely settle in 1-2 days.

If the low-frequency noise improves, I will likley measure a tilt transfer function tomorrow and try to improve the 'd' value, if needed.

ECR E1600230-v1

WP 6053

We analyzed the transfer function through the ETM ESD Driver before and after the capacitor and TVS were applied (see ECR).  Using the Dynamic Signal Analyzer (SR785) set to sweep from 1kHz to 100kHz at 1000mV, the driver performs as expected when the waveform is applied to the PI input.  I have attached a plot of the transfer function displaying modified HV, modified LV, and pre-mod HV values.

[Corey, TJ, Betsy, Koji]

RESOLVED ISSUES:
- The issue of the suspension TF was resolved by adjusting one of the OMC EQ stop holders [LHO ALOG 28900]
- The sign-flip issue of the DCPD signals has been explained by the accoustic noise of the OMC [LHO ALOG 28901]

PROGRESS:
- The lock nuts of the upper mass EQ stops were fastened (by fingers)
- The black glass panels of the OMC shroud were restored. We checked the input/reflection apertures to be well aligned against the beams.
- The OMCT steering mirror was restored to the optical mount.

- The alignment of the WFS path was reviewed. It was found that the path is still well aligned. Just the WFS QPDs were manually aligned by the mounts with picomotors.

- We attached the viewport simulator on the flange to check the alignment of the in-air paths. The AS AIR path was very well centered on the viewport without touching. The OMCR path is about 50% towards East (-Y) from the center of the viewport. We could not find the OMCT spot no matter how we flash the OMC. This could be just the beam was too dim, or was blocked by the shroud. We'll investigate this on Monday again.

TO DO:

- Continue to work on the OMCT path alignment. This requires ~10W IMC input. Try to find a dim flashes. Use a CCD camera on a tripod to find the spot around the viewport cover.
- We want to check the calibration between AS_C QPD SUM, the incident power on OM1, and the incident power on the OMC breadboard.
- Other ISC items function check (picos, beam diverters, fast shutter - see Rich's comment below, etc)
- Ground loop check
- Other SUS/SEI exit check

[Betsy, Corey, TJ, Koji]

- Now the OMC cavity is flashing!

- The OM1~3 and OMC suspensions have been debiased, and the OSEMs were tweaked to have the flags at the center of them.

- The optical paths such as the OMC incident path, OM1 transmission path, OMC refl path ahve been aligned.
  We still need to align the WFS and OMCT paths and confirm viewport paths if the beams are hitting the viewport.

- Currently the OMC PZT HV is ON.

=== Details ===

Shutter mirror inspection

The surface of the shutter mirror was checked with the green lantern. We didn't observe any sign of degradation.

Mass balance

- We adjusted the weight of the new OMC to match with the old one as much as possible. Then the mass was moved to have reasonable flag positions in the OMC suspension.

- The lateral position of OMCS RT OSEM was adjusted to have the flag at the center in the OSEM.
- The centering of the OMCS OSEMs were checked to be within the tolerance range.
- Along with the mass balancing, the positions of the EQ stops for the OMC breadboard were checked. We found several EQ stops were too close or too far. The EQ stop holders were adjusted to have them reasonable gaps to the glass breadboard.

Electrical functionality check

- The DCPDs were illuminated by a white flash light to check which DCPD responds to which channel. DCPDA and DCPDB are related to the DCPD on the transmission anfd reflection sides of the BS prism, respectively. (DCPD(T) = DCPDA, DCPD(R) = DCPDB). This seemed opposite to the case with the previous OMC. It was found that the difference in the internal cabling on the OMC caused this difference. This will be noted in the OMC testing procedure document (T1500060) athough this does not affect the calibration.

- The OMC QPDs were illuminated by the flash light. QPD1 (short arm) and QPD2 (long arm) correspond to QPDA and QPDB, as nominal.

Incident beam alignment & suspension debiasing

- Prep: IMC was locked at 2W. The beam was aligned on to the center of AS_C QPD.
- At this point, we already could observe the beams were on the OMC QPDs. Very good reproducibility.
  The signal ratios between OMC_QPD_A/B_SUM and AS_C_QPD_SUM (0.56 and 0.59 today) were confirmed with the ones with the numbers on Jul 28 (0.48 and 0.45).
  These ratios were enough similar to convince ourselves that they are the real spots.

From this point, we could follow the procedure in T1400588 (sec 2.3.3 and later).

- OMCS and OM3 were debiased to have (0,0), and used OM1 and 2 to align the spots on the OMC QPDs.

- This made OM1 Yaw ~-2000, and the OMC2 Pitch ~1500.

- The OM1 suspension cage was twisted to remove the OM1 Yaw bias.
- The OM2 suspension pitch adjustment screw was adjusted to remove the OM2 Pitch bias.

- The resulting offsets were: OM1 (116.9, -229.0), OM2 (94.5, 113.0), OM3 (0,0), OMCS (0,0) => Requirement <250 = 1/10th of the full scale => OK!

- We quickly checked some suspension transfer functions for OM1/OM2 and OMCS. OM1 and OM2 showed consistent TFs as the previous measurements. OMCS had the same resonant structures as before except for the resonant frequency of the lowest frequency mode. JeffK is checking the TFs more carefully.

- We turned on the PZT HV and scanned the PZT2 voltage. We confirmed that the OMC DCPDA and DCPDB were observing the OMC flahses.

Optical path check

- Main path: The spot positions on OM1/2/3 were checked. They looked fine.

- Shutter path: The mechanical shutter path was checked. It is still nicely aligned.

- OM1 trans path: The beam alignment in the OM1 transmission was checked. They looked fine. We still need to check the AS AIR viewport path with the viewport emulator.

- OMCR path: The beam spots on the OMCR steering mirrors were checked. The beam was not on the center of the steering mirrors. The first steering mirror (so-called M8) in the OMCR path was moved. The reflection path for 90:10 BS and the beam diverter path was checked. They looked just fine. M10 and M11 were used to align the spots on the OMCR QPDs. We didn't use M9 this time. We'll check this path again once the viewport emulator is attached on the chamber tomorrow.

- WFS path / OMCT path: We will work on these paths tomorrow.

Next steps

- Restore OMC blackglass shroud if the OMCS TFs look OK.

- Restore OMCT steering mirror

- Place the viewport emulator

- Confirm spot locations on the viewport

- We want to check the calibration between AS_C QPD SUM, the incident power on OM1, and the incident power on the OMC breadboard.

- Ground loop check

- Other SUS/SEI exit check

(Betsy, Corey, Keita, Koji & also some photos by Richard M)

Following up on alogs from Keita & Koji, photos from yesterday's work has been uploaded to resourcespace and can be found here:

https://ligoimages.mit.edu/?c=1696

(There's 91 photos.  I did not upload blurry photos.)

I'm attaching highlight photos to this alog & they are roughly numbered in the order taken:

1. M14 (was removed to blow particle off)
2. Fingerprints on NW corner of shroud
3. V-beam Dump (on/near QPD sled on northern part of table) with major damage on both pieces of glass
4. Burn hole at entrance aperature on the shroud glass for OMC
5. Photos of damage on V-Beam Dump after pulled out and on inspection table.
6. Photos of damage on V-Beam Dump after pulled out and on inspection table.
7. Photos of damage on V-Beam Dump after pulled out and on inspection table.
8. Action shots of Betsy/Koji & OMC removal
9. Action shots of Betsy/Koji & OMC removal
10. Action shots of Betsy/Koji & OMC removal
11. OMC PZT mirror with burn mark
12. OMC PZT mirror with burn mark
13. Particle contaminants on OMC Breadboard
14. Delaminated epoxy on one of the "tomb stones"