J. Kissel

I've taken the standard health checks on H1 SUS ETMX after most of the (re)install work has been completed (completely replacing the test mass, replacing the ERM with an AERM -- and subsequently changing the PenRe, adding AMDs, BRDs, etc. etc.). I'm happy with all the results, and think we can begin the close-out process.

The only thing of interest: this is only the second instantiation of an AERM suspension (LLO is still in the process of their ETM upgrades). Most R0 modes show the expected shift in response from an ERM-type to AERM-type -- except for the second lowest pitch mode (in current AERM ETMX at 1.1 Hz, in former ERM ETMX at 0.8 Hz, and in current AERM ETMY at 0.7 Hz), which appears different from ETMY's AERM, and shows up drastically differently in the R to R transfer function. It's interesting, but I'm not alarmed by this: we know from ERM and ThinCP experience that depending on the details of the reaction chain cabling, these "middle" pitch modes can be wildly different between reaction chain type instantiations. There's also nothing special about the damping loops for the reaction chain -- in fact there's a resonant gain at 1.3 Hz; good for this instantiation -- so I don't expect any change in performance. I've confirmed that the damping loops close and are stable. I'll get a loops-closed TF to confirm this once doors are on and we have some more time. (A note to my recent controls class -- this is exactly why we don't do plant inversion!)

Otherwise, the reaction chain transfer functions look as expected.
The main chain looks virtually identical to it's previous incarnation -- a testament to (a) the amazing assembly work performed by Betsy and Travis, and (b) the incredible reproducibility of the QUAD's mechanical design.  

OSEM ASDs confirm that Fil and Besty's ground loop checking (LHO aLOG 42007) were indeed successful: although we see the typical mechanical resonances of a Locked ISI between 100-1000 Hz in the M0 and R0 BOSEMs, the overall noise floor is about at the expected BOSEM noise floor, and while > 1 kHz "humps" exist, there are no humps that are particularly worse than what we've see before. There're also no obvious combs.

I've said it before, but I'll say it again -- excellent work Team SUS!

Data Templates: 
Amplitude Spectral Densities:
/ligo/svncommon/SusSVN/sus/trunk/QUAD/H1/ETMX/Common/Data/
    2018-05-30_1606_H1SUSETMX_OSEM_ASDs.xml

Transfer Functions:
/ligo/svncommon/SusSVN/sus/trunk/QUAD/H1/ETMX/SAGM0/Data/
    2018-05-30_1517_H1SUSETMX_M0_Mono_WhiteNoise_L_0p01to50Hz.xml
    2018-05-30_1517_H1SUSETMX_M0_Mono_WhiteNoise_T_0p01to50Hz.xml
    2018-05-30_1517_H1SUSETMX_M0_Mono_WhiteNoise_V_0p01to50Hz.xml
    2018-05-30_1517_H1SUSETMX_M0_Mono_WhiteNoise_R_0p01to50Hz.xml
    2018-05-30_1517_H1SUSETMX_M0_Mono_WhiteNoise_P_0p01to50Hz.xml
    2018-05-30_1517_H1SUSETMX_M0_Mono_WhiteNoise_Y_0p01to50Hz.xml
/ligo/svncommon/SusSVN/sus/trunk/QUAD/H1/ETMX/SAGR0/Data/
    2018-05-30_1618_H1SUSETMX_R0_WhiteNoise_L_0p01to50Hz.xml
    2018-05-30_1618_H1SUSETMX_R0_WhiteNoise_T_0p01to50Hz.xml
    2018-05-30_1618_H1SUSETMX_R0_WhiteNoise_V_0p01to50Hz.xml
    2018-05-30_1618_H1SUSETMX_R0_WhiteNoise_R_0p01to50Hz.xml
    2018-05-30_1618_H1SUSETMX_R0_WhiteNoise_P_0p01to50Hz.xml
    2018-05-30_1618_H1SUSETMX_R0_WhiteNoise_Y_0p01to50Hz.xml

J. Kissel

After lots of work on H1 SUS TMSX during this spring's vent of EX (a repair of OSEMs, LHO aLOG 41675, and a physical rotation of the suspended stage to account for the QUAD's now-missing ERM wedge, LHO aLOGs 41942 and 41954), I've taken a full health check to make sure we're ready for doors in the next few days.

All health checks pass with flying colors.

In the transfer functions, you can see the V and P transfer functions are now much more similar in magnitude to other TMTSs, indicating that the repair, recentering, and recalibration has fixed the problems (FRS Ticket 6446).

Also, the OSEM ASDs confirm that Fil and Besty's ground loop checking (LHO aLOG 42007) were indeed successful: although we see the typical mechanical resonances of a Locked ISI between 100-1000 Hz, the overall noise floor is about at the expected BOSEM noise floor, and while > 1 kHz "humps" exist, there are no humps that are particularly worse than what we've see before. There're also no obvious combs.

Great work Team TMTS!

Data Templates:
Amplitude Spectral Density check for Ground Loops:
/ligo/svncommon/SusSVN/sus/trunk/TMTS/H1/TMSX/SAGM1/Data/
    2018-05-30_1557_H1SUSTMSX_M1_OSEM_ASDs.xml

/ligo/svncommon/SusSVN/sus/trunk/TMTS/H1/TMSX/SAGM1/Data/
    2018-05-30_1520_H1SUSTMSX_M1_WhiteNoise_L_0p02to50Hz.xml
    2018-05-30_1520_H1SUSTMSX_M1_WhiteNoise_T_0p02to50Hz.xml
    2018-05-30_1520_H1SUSTMSX_M1_WhiteNoise_V_0p02to50Hz.xml
    2018-05-30_1520_H1SUSTMSX_M1_WhiteNoise_R_0p02to50Hz.xml
    2018-05-30_1520_H1SUSTMSX_M1_WhiteNoise_P_0p02to50Hz.xml
    2018-05-30_1520_H1SUSTMSX_M1_WhiteNoise_Y_0p02to50Hz.xml

(Note, the ISI was locked for these measurements.)

The transfer function templates needed some work:
    - Decreased the resolution from 0.01 to 0.02 BW, which is a new standard for non-QUAD suspensions, since the resonances can all be resolved and it improves the spped of characterization by a factor of 2.
    - Some prior templates (about half, but not all, strangely) had incorrectly been saving the OSEMINF_IN1 channels instead of the OSEMINF_OUT_DQ. This messes up the post processing scripts, and meant that we'd been plotting the OSEM basis incorrectly for some time (though the last time anyone measured this SUS in its entirety was back in Aug 2017).
    - For some DOFs where coherence around resonances was a struggle, I added a few resonant gain boosts to the excitation filter. 

Note to self -- next time you measure the TMTSs, improve the processing scripts to plot the phase, like recently you've done for other suspension types.

Jonathan, Dave:

the Debian8 framewriter h1fw0 crashed at 10:39 this morning, reporting a segmentation fault. I've included an image of the journald logs at the time of the crash. Systemd restarted the service correctly.

Looked at improving the reference cavity transmission by making adjustments to the mode matching.  I moved the last
lens before the reference cavity L12.  The reference cavity transmission was ~1.65 V when I started and was ~2.87 V
when I finished.
  L12 original micrometer position 3.48
  L12 new micrometer position 3.00

    When the reference cavity was unlocked and blocked there was a -1.6 mV offset in the output of the RFPD.  When the
reference cavity was simply unlocked, the RFPD output was 475 mV.  When locked, 80 mV.  This yields a cavity visibility
of ~83%.  Whis not bad but not great.  L12 could not be moved back any further because its base would hit a mirror mount.

    Previously the common gain was set to 20 dB.  The attached plot "tf.png" shows the transfer function for a few values
of the fast gain.  The higher the fast gain is increased, the more pronounced is the dip in the transfer function, which
in turn means the cross over between the fast actuator and the Pockels cell is not quite right.  I found that dropping
the fast gain to 3 dB improved this a little.  It could be that a finer adjustment would fix things but this is the best
I could do in the time available.  Finally the common gain was increased to 21.5 dB in order to push the unity gain up.
With a common gain of 21.5 dB and fast gain of 3 dB, the unity gain frequency is ~518 kHz with a phase margin of ~60
degrees.

    I'm not quite sure how robust these settings are since there's a large-ish resonance out at ~762 kHz which may
cause some problems.

    There is still some room for improvement.  The increased reference cavity transmission should be enough for ALS from
the end stations (hopefully).

We installed a secondary filter on the LEMI power supply and now all vault electronics are working, finally!

This closes WP7470

Jim W. - Transitioned End-Y to laser safe.

Dan, Jenne

We tested the AS port protection by setting the fast shutter threshold low. The shutter triggered and the sum on AS_A/AS_B dropped. The AS port protection MEDM correctly signaled that the shutter closed and the screen elements went red. We reopened the shutter and light is coming through but the fast shutter screen says there is an "error" but doesn't provide any more information, screenshot attached.

Summary

ALS path on the PSL table was realigned (it was totally misaligned after the main IO path was redone on May 09, alog 41924).

The beam is already going to the SHG in ISCT1 but we need to refine the alignment there. We don't have to go back into PSL to do this work, the rest should be done in ISCT1 in laser hazard some time next week (or earlier when there's an opportunity).

Details

See https://dcc.ligo.org/D0902114 for nomenclature.

At first the beam was blocked by the Faraday aperture (first attachment). I and Ed used ALS-M1 and ALS-M2 to clear the Faraday and center the beam on two irises between ALS-M3 and the periscope. After we were satisfied, however, Thomas and Dan couldn't find the beam in ISCT1. We also found that the beam was clipping on the top periscope mirror (ALS-M5) on the PSL table though the beam went into HAM1 (first attachment, in this picture two irises are almost closed so it's easier to see the beam position).

Dan found that the beam was already hitting the in-chamber steering mirror, and we regained the beam in ISCT1 by turning the top peri mirror mostly in YAW in the PSL room.

From there on, I tried to use the top (ALS-M5) and bottom (ALS-M4) peri mirror on the PSL table to walk the beam on ALS-M5 while keeping the beam in ISCT1. That worked  until the bottom peri mirror hit some hard stop in YAW. There's plenty of actuator threads left, but it seems like the mirror mount and the 45 deg mounting block had a mechanical interference. I ended up using ALS-M3 also. We just stopped doing this when the beam position on the top peri mirror on the psl table was not as terrible as before (attachment 2, same caveat as the first picture). Then we used the ALS-M5 to steer the beam at the center of the bottom peri mirror in ISCT1.

In the PSL room we moved the irises so the beams are centered on them. 

Gerardo M., Chandra R., Kyle R.

Today we removed the two paper clip jumpers that had been added to the terminal strip withing the vacuum rack back during the CP4 bake.  Filberto C. and Gerardo M. had added these in order to mimic the GV11 CLOSED and GV12 CLOSED signals required by CDS in order for us to utilize the CP4 regeneration PI loop.  The nominal wiring for both of these gate valves had been removed so that they could be baked along with CP4.  As such, these two signals had to be simulated via these jumpers.  Ultimately, this ended up being a mute exercise as we eventually abandoned this approach altogether, opting instead to go "old school" and using a manually controlled variac instead.  Today we also had to re-install the fuses associated with these two signals.  Following this, Gerardo noticed a discrepancy between the Y-mid Station-specific MEDM screen vs. the overall Site Overview MEDM screen.  One screen showed the correct valve status following today's signal restoration but other screen did not.  He and Patrick T. sorted this out later somehow. 

Chandra R. and I had both heard a noise last Friday when attempting to fully open GV11.  The sound seemed to have emanated from the limit switch area so we opted to not fully open the valve at that time until we could fully investigate (today).  The motor drive is entirely dependent upon proper setup of the limit switches and we wanted to confirm that nothing had been changed during the removal of the signal wiring.   So, this morning, Chandra R. contacted Ken D. (electrician) to confirm that he had only removed the wiring from the GV11 and GV12's limit switches when preparing for the CP4 bake out and had not changed any of the mechanical setup.  Ken D. confirmed that only the wiring had been removed and that the mechanical setup of the switches had been left alone.  Even so, I removed the limit switch box cover plate and observed the internal workings as GV11 was stroked up and down (a few inches in each direction).  I tested the OPEN and CLOSED limit switches by manually changing their states and observing that the CDS MEDM screens responded accordingly.  Additionally, I confirmed that the motor drive stopped driving the valve in response to my manually closing the OPEN limit switch.  Convinced that everything was normal, we fully opened GV11. 

Next, we re-installed all of GV12's annulus piping (new modified version) minus the gate portion.  We will finish this up tomorrow and, assuming no issues, start the hours long process of opening GV12. 

EOM PD restored:

• the Thorlabs PD that was originally installed on the EOM pick off path, has been restored to this location
• it was replaced by the ISS PD for a while, but the ISS PD is back in the ISS box
• this Thorlabs PD will be replaced with a large area PD

IO GigE camera 2 installed:

• IO_AB_M12 was swapped, the super polished mirror removed, and a 98% beam splitter installe
• at 7.6W incident on the bottom periscope mirror, there is 6uW on the camera
• since this camera is downstream of the rotation stage, it's SUM will vary as IMC INPUT power varies
• I've set the gain to 1850
• when the IMC IN power is 20W, the camera 2 gain will, in theory, produce a SUM of around 2.7e8, and closely match camera 1's SUM
• camera 1 incident power is 15uW, with the PMC transmitted power of 42.2W
• camera 2 incident power is 6uW, with 7.6W incident on the bottom periscope mirror

IMC power recalibrated:

• before any recalibration,  measured the power before the bottom periscope mirror was 7.6W, with a requested power on the rotation stage of 6.9W
• Keita recalibrated the IMC IN PD
• his notes are attached, as a snapshot

J. Kissel, J. Warner

Jim and I went to B&K hammer EX’s PCAL periscope this afternoon and came across a new bug I hadn’t seen before: the hammer channel’s BNC input spigot was constantly red while connected, and after activating a known functional template (which showed the 3-axis accelerometer working as normal), the hammer’s sensor input bar was flashing red and reporting that it was saturating / overloaded. However, between the ~1sec cadence flashes, you can see the glimpses of a normal hammer behavior. We tried:
	- Power cycling the satellite box
	- Closing and re-opening a known functional template (and a different template)
	- Physically disconnecting the hammer from the satellite box, (the spigot stays red — but that happens normally when a sensor is disconnected)  and reconnecting it
	- Swapping out the hammer’s B&K BNC with a normal BNC
	- Shifting all of the inputs down by one (doesn’t work at ALL! Sad… why have 6 channels then?)
	- Clearing and Re-selecting the hammer species in the template (8206-003)
	- “Clear and Detect Hardware” in the Hardware setup
	- Physically disconnecting the hammer’s accelerometer from the head and reconnecting it (grasping at straws … )
	- Changing the input gain on that channel in the template’s hardware properties (really grasping for straws… )

Attached are pictures of the satellite box in error, and the activated template in error.

I've got emails out to the experts and we'll continue to debug the hardware outside of the clean room, but this may mean we forgo B&K hammering of this periscope. 

Turned on the OPO TEC with the same loop shape as in air. Takes about 50 minutes to reach equilibrium.

We set up the ETMY HWS to run over the weekend to gather background noise data. The SPHERICAL_POWER looks normal for the first 17 hours until it suddenly drops down to a large negative value. It turned out that the WATCHDOGS tripped on TMSY at this time. The HWS continued to run without any other issues over the long weekend. 

TJ reset the watchdogs this morning and the HWS spherical power returned to the nominal zero value (without any interaction with the physical HWS hardware). We will continue to run it like this for the time being (until work needs to be done at ETMY).

Intensity distributions

The intensity distribution illuminating the HWS with the beam from the 50 micron core fiber is shown here. Spatial units = pixels. There is some high spatial frequency structure on the return beam. No doubt from clipping somewhere in the system. The circular aperture is consistent with the annular reaction mass aperture. 

When the Hartmann plate is put on the CCD, the result is the following:

200 micron core fiber

We switched to a 200-micron core fiber and achieved more transmission and similar HWS performance with a beam showing noticeably less high spatial frequency intensity fluctuations.

And with the Hartmann plate in place:

ALS cross-coupling

We turned on the ALS beam and found that, although s-polarization is incident on the PBS going into vacuum system, there is significant leakage through the PBS for the return beam. The HWS CCD looks like this when the ALS beam is turned on (and the Hartmann plate is removed).

We measured the amount of ALS light and found that at least 750 micro-Watts is getting through the PBS. This is a couple of orders of magnitude larger than the HWS beam (which is difficult to measure accurately because of the low power level.

Background wavefront changes

The background wavefront changes are analyzed below:

• The first (4.5 minute change) compares two times shortly after the HWS was initialized on Friday
  • The wavefront shows a small tilt but not much high frequency noise. The time scale for a point absorber is less than 4.5 minutes and the OPD from the ITMX point absorber was of the order of 80nm (about 50 to 100x larger than the background noise). Provisionally, without doing a full analysis to determine SNR, this looks sufficient to detect point absorbers.
• The second (1000 minute change) compares two times this morning
  • Ditto result from first plot.
• The third (4 day change) compares a reference from Friday with the HWS measurement from this morning
  • There is some clear medium spatial frequency structure to the HWS measurement here. This looks like diffraction fringes across the HWS probe beam. Further investigation is needed. However, we do know that if there are diffraction fringes on the beam we should see these become more pronounced as the time between two HWS measurements increases (assuming that the TMS and other optics slowly drift over time).

Background wavefront change: 4.5 minute change

Background wavefront: 1000s change

Background wavefront: approximately 4 day change