I found that the diffraction power of ISS had decreased again.

This time, it seems that somebody changed the offset setting for some reason (or maybe it was related to Stefan's ODC activity, see alog 9222). I set it back such that the diffraction power stays at 10%.

H1:PSL-ISS_REFSIGNAL = -1.93 (had been set to be -1.95 by someone on 11th of January)

Note that this value is the same as what I set on this past Friday (see alog 9219)

I did some cleanup work for the PSL odc channels

- Updated the ODCV2.mdl library to use the MODAL_RATE part from RCG 2.8. This replaces the need to manually set the model rate. This library is in almost every model currently running, but there is no real urgency to push the change everywhere.
  Updated SVN: ODCV2.mdl revision 6837

- Added ADC/DAC overflow and Excitation monitoring to ODC for the PSL. Updated and restarted the following files (SVN revision 6839):
        h1pslfss.mdl
        h1psliss.mdl
        h1pslpmc.mdl


- Updated the PSL ODC medm screens. PSL_ODC.adl                    SVN revision 6838  (picture attached)
- Updated the ODC overview screen.  SYS_CUST_ODC_SITE_OVERVIEW.adl SVN revision 6842

- Updated the h1setODCbitstrings and h1setODCbitmask scripts in cds/h1/scripts to include the PSL settings. SVN revision 6841.

I went through the X arm alignment again, unfortunately the computer crashed before I saved the alog but the centered values for TMS are 147.5 PIT -234 YAW (+10 in pit compared to yesterday, +6 in YAW) ITM 72 PIT -62 YAW (basically the same as yesterday).

It was easy enough to find the pointing of TMS and the ITM using the baffle PDs starting from the last alingment in alog 9191.  However, I could not easily find the ETM pointing.  I did find the beams on the baffle PDs by doing a slow scan (Strip chat screen shot attached) and found:

with 40dB gain, PD3 has 2.1 V with ETM PIT 292 YAW 44.  REFL B then has 15400 counts

with 20dB gain PD1 has 2.6 Volts with ETM PIT 246 YAW 101 Refl B has around 1100 counts. 

From this the aligned position should be 269 PIT 72.5 Yaw.  This is remarkably different from where I found similar fringing yesterday, 384 PIT -77 YAW.  While there was some HEPI commisioning yesterday afternoon, I would have expected this to change the TMS and ETM alingments by the same amount, so either one or the other of these alignments is bad or something moved.  Also, the ETM oplev has moved quite a lot. 

The PD levels for these alingments seem suspicously low, if the mode matching was perfect I would expect the intensity on the diodes to only decrease by a factor of the ETM reflectivity compared to the straight shot from the transmon (alog 9191) Instead these are a factor of 100 and a factor of 10 lower. Either these are not the right beams or the mode matching could be wrong.  There was clearly some fringing on the refl PD and COMM (which I am using as an X arm trans readback for now). 

It is possible to find some light on the PDs for ETM alignments that don't make sense (PD1 on the wrong side of PD4 ect), probably a second bounce beam hitting the PD. 

I also changed the polarization into the X arm fiber (photo of settings attached). 

Once I saw some fringing I was able to "lock the arm"  and the "lock"  would stay for 10s of seconds although I'm not sure this was really locking to any reasonable mode of the cavity.  On the camera I could see that the beam motion is large, we have verry high winds.  I also could not go out to the end station to work on the locking because of tumbleweeds, photo attached. 

Keita, Koji, Stefan

I went out to ISCT1 and steered the REFL beam on to REFLAIR_A_9. After a little bit of alignment tweaking and closing the PRC loop (no MICH loop yet) we got some pretty decent REFL dips. More later this weekend.

Looking at the IMC transmitted light, I realized that the laser power occasionally dropped by 15% and came back to the nominal level in 10-ish seconds (see the attached screenshot). Since I knew this was usually associated with ISS, I took a look at ISS. Sure enough, the diffracted light was too low and it was about 5% or perhaps even less. According to the trend (see the second file attached), it seems that the diffraction power started decreasing after around 18th of December in the last year and recently dropped steeply. This too low diffraction was causing the frequent unlock of the ISS loop.

I adjusted the offset voltage to increase the diffraction power to 10% as was done by Doug (see alog 7810).

H1:PSL-ISS_REFSIGNAL = -1.93 ( which had been -2.00)

It seems this fixed the glitching issue as of now.

I updated the the LHO PSL code (FSS, ISS, PMC) to syrnchronzie with the LLO status, using all the common parts already in svn. This was slightly tricky, since there were some local changes at LHO that were never committed because the LLO files had changed. This also brings the PSL ODC code to the same status as LLO. Some more ODC threshold setup work will be required over the weekend to make sure the channel is green when it should be.

Files checked into svn: version 6832:
h1pslfss.mdl
h1psliss.mdl
h1pslpmc.mdl

I also re-implemented the local LHO changes to the common files, namely:
- $Id$ and $HeadURL$ tags for svn version keeping 
- Defining science channels by adding * in the DAQ block for the following channels:
   pslfss: FAST_MON_OUT, MIXER_OUT
   psliss: AOM_DRIVER_MON_OUT, PDA_OUT, PDB_OUT
   pslpmc: HV_MON_OUT, MIXER_OUT

Files checked into svn: version 6833:
pslfss.mdl
psliss.mdl
pslpmc.mdl

I guess LLO should syc those files to have the PSL run of the same library at both sites.


There is one more thing: psldbb.mdl I did not touch it today. LLO is up to date. LHO has some local mods (presumably from the MATLAB2012 move), BUT:
Both versions are so messed up with wiring that I do not dare touch it.

Lots of work on ETMX HEPI today. Because of the potential for HEPI pump glitches to feed into the BSC ISI's St1 T240s, we worked on installing position control on HEPI. These loops are now working and we have left the entire SEI platform (HEPI & ISI) in an isolating cofiguration. This does mean that commissioners will have to enter their offsets in a somewhat different manner, but Fabrice has explained it to Kiwamu. We have also attempted to put HEPI back in it's tilted state, it's not exactly where it was, but its pretty close. Monday, we will have to update safe.snaps for HEPI.

The ETMy has had two of its four new fibers rewelded. 

Daniel, Sheila

We measured some tranfser functions of the PLL servo board.  

Attached files are : X end PLL servo board 10kHz to 100kHz

AA1.txt , common compensation on, 0dB gain, generic filter on

AAA2.txt, 0dB gain, common compensation on, fast option on (this is a 292.48kHz notch)

AAA3.txt 0dB gain, nothing on  (we were using the common mode servo which has a high impedance output to drive the RF analyzer which has 50Ohm inputs, we also used a BNC T. This might be why there is strange 2dB feature around 200kHz which should be subtracted from the other traces).  

J. Kissel

After a few more tweaks of the individual HSSS screen, and creating the grouped RM and OM ALL screens, all HTTS modifications are complete and ready for deployment. Detailed instructions on how to install are listed below. Once LLO installs these changes (slated for next maintenance period), we can finally close out ECR E1300578. All necessary changes are committed to the userapps repo in various locations, again detailed below.

Installation Instructions
-------------------------
Capture alignment settings
(1) Make sure you either have alignment settings captured in a safe.snap, or at least handy for later.

SVN:
(2) svn update the sus corner of the userapps repository. (If you haven't accidentally updated already JAMIE, then) You should see a plethora of new files in the folders mentioned in LHO aLOG 9195. I've made further updates since that log, so be sure to update right when you're ready to install. Update the cds/h1/medm/ corner as well, so you get a feel for how the MEDM marco files need to be called from a sitemap.

Simulink:
(3) Create a new model ${ifo}sushtts.mdl (which you can probably copy from h1sushtts.mdl, to minimize labor). You'll need to 
- change the cdsParameter block to reflect the new ${ifo}, 
- change all of the IPC channel names,
(4) Create a new model ${ifo}susauxasc0.mdl (which again, you can copy from h1susaux.mdl, to minimize labor). You'll only to
- change the cdsParameter block to reflect the new ${ifo}
(5) Copy over changes (or simply overwrite) the ${ifo}susim.mdl model. You'll need to
- change the cdsParameter block to reflect the new ${ifo}, 
- change all of the IPC channel names,
(6) Compile all new/updated models.

Screen preparation:
(7) Set up your sitemap to call up either the individual HSSS screen,
${userapps}/sus/common/medm/hsss/SUS_CUST_HSSS_OVERVIEW.adl
or either of the groupings of HSSS,
${userapps}/sus/common/medm/hsss/
SUS_CUST_OM_OVERVIEW_ALL.adl
SUS_CUST_RM_OVERVIEW_ALL.adl
/opt/rtcds/userapps/release/sus/common/medm/haux/
SUS_CUST_HAUX_OVERVIEW_all.adl
(Note, eventually we'll move the HAUX overview into the hsss folder, and rename it properly capitalized)

Installation:
(8) Ask your sysadmin to prepare for the new installs by
- Removing h1asctt from 
- Adding h1sushtts and h1susauxasc0 to
all EDCU, master, IPC, .ini, etc files.
(9) Make-install and start the models.
(10) Restart the DAQ
--------------------

Once LLO has successfully installed everything and confirmed all necessary functionality, we can get rid of the 
${userapps}/sus/common/medm/haux, and 
${userapps}/sus/common/medm/htts 
folders, and get of any leftover HTTS stuff in the ASC folder, like the screens in 
${userapps}/asc/common/medm/asctt

   Last set of TFs on H1-SR2 showed a longitudinal to yaw cross coupling at 1.51 and 2.77Hz. I adjusted the Coil Holder and recentered the BOSEMs.
 
   There were only two of the four Coil Holder Mounting Brackets installed, (which may be a source of the noise in this suspension and a contributor to the cross coupling) when the earlier TFs were taken. The missing two Mounting Brackets required modification due to misaligned mounting slots. These modifications were not completed in time to be installed on the suspension before it was moved to the LVEA. One of the Mounting Brackets was modified correctly; and has been installed on the suspension. The second Mounting Bracket was not modified correctly and will need to be reworked before it can be installed. Three of the four Mounting Brackets are now clamping the Coil Holder to the Weldment. 

   I have started a new set of TFs on H1-SR2, to see if the BOSEM adjustment and the addition of the third Mounting Bracket will improve the performance of the suspension over that revealed in the earlier TFs.        

[Alexa, Koji, Kiwamu]

We finally completed the in-vac work in HAM1 and the Apollo crews put the door back on. All the necessary beam comes out to ISCT1 and all the necessary beam hit the in-vac diodes. Yeah!

Some pictures will be uploaded in ResouceSpace later.

Here are a list of what we did today:

• TMSX was aligned to make sure that the green light hits the center ITMX (reported in alog 9203).
• Sheila aligned PR3 to get the green light on the iris on ISCT1 (reported in alog 9191).
• We placed the beam dump in the position where the beam should be.
  • This is the one we could not align yesterday (see alog 9196).
• We checked the spot position of every optic in the ALS path with the green light.
  • They looked good.
  • Therefore we didn't touch the alignment of these mirrors.
• We checked the spot position of every optic in the POP path with the infrared POP beam and with an IR viewer.
  • They looked good.
  • The beam still landed on the center of the bottom periscope on ISCT1.
  • Therefore we did't touch the alignment of these mirrors either.
• We opened the shutter of the PSL/ALS light pipe to check the alignment of this path.
  • It was a bit too high and therefore we steered the top periscope mirror of the PSL to lower it.
  • The horizontal alignment was good from the beginning.
• We moved onto the REFL path.
  • The REFL beam coming out from HAM2 toward HAM1 seemed a bit too low by probably 5 mm.
  • The horizontal alignment looked very good.
  • Even so, the beam on RM1 and RM2 and their downstream looked well centered. So we didn't change anything.
  • The beam was already on all three diodes from the beginning.
• The tip-tilts were found to be not damped.
  • There was no damping signals flowing through the model. I am not sure why we lost the last good settings for them.
  • Koji and Jeff Kissel worked on this and finally made them damped.
• We centered the beam on the REFL LSC diode by steering the steering mirror (which is, technically speaking, a BS) in front of the diode.
  • This was done by looking at the digitized DC signal.
  • Also we moved the black glass beam dump, which is supposed to catch the reflection off of the REFL LSC diode, by an inch or to the west because the beam had been on the edge of the black glass.
• We found that the reflected light off of the WFSs were hitting their steering mirror and not making to the black glass beam dump.
  • Note that WFS_A, which is the PSL side of the WFSs, didn't have a black glass beam dump. Therefore we asked Corey to give us a dump and we newly installed it.
  • We introduced some more rotation in the angle of the WFS detectors to get the reflected light spatially separated from the incident.
  • We took several pictures before and after changing WFSs' orientations.
  • Then we fine-adjusted the location of the black glass beam dumps to catch the reflections.
• We centered the beam on each WFS by steering the picomotorized mirrors on the sled.
  • This was done by looking at the digitized WFS_DC signals.
  • Also we had to switch the WFS interface D-subs with the ones of the analog WFSs to get the signal available.

8:00 - Performing alignment work at End X – Sheila
9:30 - Roofing contractors/inspectors on site today 
9:40 - Work on HAM4 – Hugh/Jim W.
9:45 -Inspecting the Corner Station roof – Inspectors/John/Richard
9:50 - Continuing Baffle work at LVEA West bay – Mitchell
10:16 - Roof inspectors going to End Y
11:36 - Finished working on LVEA – Mitchell
1:05 - Back to Baffle work on  LVEA – Mitchell
1:16 - Transitioning End Y to Laser hazard – Operator
1:46 - Re-checking SR2 BOSEMs – Jeff B.
1:50 - Going to End X to do measurements – Sheila
2:00 - Finished working on HAM1 - Kiwamu
2:20 - HAM1 East door installation – Apollo
3:05 – Powering up Ring Heater chassis at End X – Filiberto/Aaron

While in HAM 1, we verified that the beam diverters were functional. Indeed they were; however, Corner Pop beam diverter (BDIV_B) was flipped in regards to open/close, so I have updated the PLC to reflect the proper positioning.

The minute trend files for H2 used to be in /frames/trend/minute_raw.  They have been moved to /frames/trend/minute_raw_old, which allows for somewhat easier searches for files in the 0-ff directories.  If anyone really wants to look for minute trends from the H2 system, the NDS configuration file will need to be modified to look in that directory.

I added a new WD button to the SITEMAP for the CDS Watchdog screens (dark blue, bottom left). The new Hardware Watchdog (attached to ITMY) and the IOP software watchdog screens are accessible via this new button.

After alinging TMS and the ITM using the bafle PDs, I was unable to find light on the ITM baffle PDs using the old ETM referece (alog 9126). 

I found some fringing at: ITM 75 -59 ETM 384, -77 TMS 136.8, -240.3.  The ETM was moving too much to make a reasonable alingment.  Hugh, Jeff and Jim helped to get both the ETM and ITM to level one isolation, so things are moving less now. 

By misalinging the ITM I measure 12830 counts on refl B.  If I misaling the ETM by 40urad, I can clearly see two return beams, a brighter one from the ETM and a less bright one from the ITM.  Misaligning the ETM by 80urad, the etm return beam is no longer on the refl B PD.  I adjusted the ITM alignment to maximize the power on the refl B PD, at PIT 75 YAW -59 there are 2850 counts on refl B.  Blocking the refl B PD completely we have 52 counts.  I also misalinged the itm to see the signal disappear, so I do believe this beam was really from the ITM. 

The ratio of the ITM return to the ETM return beam is (1-Retm)^2*Ritm/Retm.  Assuming that the ITM reflectivity is 99%, this gives Retm=63%.  This reflectivity will only get lower is the ITM reflectivity is lower than 99%.  With Retm=63%, our measurement of the return beam power from monday night (alog 9171)means that the TMS effiicency  (one way) is 78%, matching better with Keita's prediction (alog 3077) of 82%.  In the first version of this alog I made a mistake, and wrote that the reflectivity was 53%.  

I measure 12uW of 532 arriving on ISCT1 with the ITM misalinged (single shot beam, about 34mW input from ISCEX) ), there is 1.6uW on the DC PD monitoring the X trans.  The DC PD does not seem to be working, it is cabled (R4 ISC1:2) but changing the gain on the medm screen does not change the monitor output for this channel.  I also alinged the single shot beam onto the Comm BBPD, there are 3.5mW there with the ITM misalinged. The slow controls readout of this PD is also not working, but I can see that the PD is working in H1ALS-C_COMM_A_LF.  I added a -62 counts offset (the dark offset) now get around 40 counts with 3.5 uW on the PDs.  

J. Kissel, R. Bork, B. Abbott, D. Barker

In order to determine a good threshold for the new SUS Hardware Watchdog, we've shaken the ITMY chamber vigorously in various states of excitation. 
In summary, with 
- the HEPI and SUS user watchdogs disabled (actuator and inertial sensors set to un-tripably high values)
- ISI damping, SUS damping, and HEPI loops OFF (I just left the ISI tripped the whole time) and 
- driving HEPI at the limit of its DAC range (in Y, RX, and RZ), 
we shook the suspension (and ISI) enough that I would say "Wow, that's a *lot* of motion; it's giving me the willies. We should cut off SEI excitation after 10-20 minutes of this." 

This level of excitation produces roughly 110 [mV] RMS* from the RMS circuit of the SUS Hardware Watchdog. 
* The Hardware Watchdog currently does not expose the output RMS variables, either to analog or digital, so we've replicated what we believe is the SUS Hardware Watchdog's signal chain in the IOP model for the h1susb123's computer, based on the limited documentation available (namely Fig 4 of T1200306), and the DC gain of 31.2 [V/V] from the SUS Hardware Watchdog Chassis, measured on an identical setup in the H1 DAQ Test Stand (thanks Ben/Rolf!).

My figures of merit were 
- the ISI BLRMS Performance matrices, in which 
     - ST1 was solid red above the 0.3 [Hz] band in XYZ, 
     - ST1 was solid red in all frequency bands, and 
     - ST2 was solid red in all bands in all DOFs
- the QUAD M0 OSEM speed dials (the raw ADC input version), in which (for these particular DOFs of excitation from HEPI)
     - F1, F2, and F3 (which are the L, P, and Y sensors) were swinging quickly over *most* (not all) of their range (which is 0.7 [mm])
     - LF, RT, (which are the V and R sensors) showed signs of abnormally large movement, say 25% of their range
     - SD was moving minimally
- The DAC saturation counter screen for HEPI, which was showing several channels off-and-on saturating
- The ITMY Optical Lever, which was not swinging out of its range, but moving abnormally large, say 50% of its range



For future forensics, the rough times that I had this large excitation going was roughly between 
GPS: 1073411801 and 1073412182
UTC: Jan 10 2014 17:56:25 - Jan 10 2014 18:02:46
PST: Jan 10 2014 09:56:25 - Jan 10 2014 10:02:46 

Drive was 0.1 to 10 [Hz] "uniform" white noise, using three independent awggui sessions, driving channels
H1:HPI-ITMY_ISO_Y_EXC    100000 [cts]
H1:HPI-ITMY_ISO_RX_EXC   100000 [cts]
H1:HPI-ITMY_ISO_RZ_EXC   100000 [cts]

The QUAD, ISI, and HEPI survived admirably (as expected), and after the excitation was turned off, SUS and ISI damping loops were immediately functional again, and it quieted down to normal damped motion with ambient input motion from an unisolated chamber. All software watchdog values have been set back to their nominal values.

(Alexa, Kiwamu)

We noticed that the FSS was osillating significantly. I first lowered the resonant threshold from .8V to .5V. I also decreased the Common Gain from 30dB to 10.4dB. This seemed to help. It's possible the oscillation was due to instability in the MC servo.