After installing the new filters on ETMY suspension, I took an open loop gain transfer function of the top mass in the longitudinal degree of freedom, with the loop closed, and compared the measurements against the model.

The first pdf attached shows the measurement in blue, and the model in green. Except for the first pitch mode, at around 0.5/0.6Hz, they match really well.

The second pdf attached shows the open loop gain transfer function for different damping configurations that have been used during HIFO-Y commissioning. The old filters mentionned are refering to the one Jeff K. designed to reduce sensor noise contribution to test mass displacement added to the extra boost at .43Hz and .99Hz

The new filters are clearly more stable, providing the same damping at the lowest longitudinal mode than the previous filters with a phase margin of 36 @ 3.54Hz.

Attached is a plot of ETMY and ITMY oplev signal + reflected beam signal while the cavity was locked. The motion @ .43 Hz is still high. Although, looking at the plots from the previous aLog it seems like the ETMY ISI could provide a better isolation, as it did in May. 

Today, while the cavity was locked, and the ISI was in "level 3 isolation" I took a spectra of ETMY M0 L1 and L2 stages in Long and Pitch. 

I plotted the spectra against previous ones with the suspension damped and the chamber under vacuum for the three of them. The comparison shows different configuration of the ISI

1) in blue, ISI damped in may

2) in green ISI isolated in may

3) in red ISI isolated today

Continued assembly of the Manifold Cryopump baffle.

Attached Class B Top Ring to Balancing Fixture.

Started to bend blades -
1. Weight Hanger is not long enough for the amount of weights needed to bend blades (160 lbs), see pic.
2. Screws are not long enough to secure tongue depressor on blade to Bend fixture (currently SCHS #10-24 x 7/8"; looking for SCHS #10-24 x 1". TMS has some at LLO).

Called it a week.

I did cleanup on the HAUX screens on the same lines as in alog 7049 for HTTS.

For all the screens in $(USERAPPS)/sus/common/medm/haux with gedit, I changed link arguments to look like the following:

args="%(read $(USERAPPS)/asc/common/medm/asc$(optic)$(num)_overview_macro.txt),USERAPPS=$(USERAPPS),SITE=$(SITE),site=$(site),IFO=$(IFO),ifo=$(ifo)"

Formerly some of the screens expected $(OPTIC) to be the full optic name, and some used $(OPTIC)$(num), so I changed everything to the latter convention.

Details:

No change:

Basic cleanup: SUS_CUST_HAUX_BIO.adl, SUS_CUST_HAUX_BIO_all.adl, SUS_CUST_HAUX_DACKILL.adl, SUS_CUST_HAUX_DC_ALIGN.adl, SUS_CUST_HAUX_M1_COILOUTF.adl, SUS_CUST_HAUX_M1_DRIVEALIGN.adl, SUS_CUST_HAUX_M1_EUL2OSEM.adl, SUS_CUST_HAUX_M1_LOCK.adl, SUS_CUST_HAUX_M1_OPTICALIGN.adl, SUS_CUST_HAUX_M1_OPTICALIGN.adl, SUS_CUST_HAUX_M1_OSEM_ALIGN.adl, SUS_CUST_HAUX_M1_OSEM_ALIGN.adl, SUS_CUST_HAUX_M1_OSEM_ALIGN_??ZOOM.adl, SUS_CUST_HAUX_M1_OSEMINF.adl, SUS_CUST_HAUX_M1_SENSALIGN.adl, SUS_CUST_HAUX_M1_WD.adl, SUS_CUST_HAUX_M1_WD_ACT_BANDLIM.adl, SUS_CUST_HAUX_M1_WD_OSEMAC_BANDLIM.adl, SUS_CUST_HAUX_MONITOR_OVERVIEW.adl, SUS_CUST_HAUX_MONITOR_OVERVIEW_all.adl, SUS_CUST_HAUX_OPLEVINF.adl, SUS_CUST_HAUX_OPLEVINF1.adl,

Other fixes:

SUS_CUST_HAUX_DAC_MONITOR.adl: I removed display of /cvs/cds/llo/scripts/utilities/nova_logo.gif, which was not only LLO-specific but shouldn't have been in a custom screen. Moved time string into visible area and corrected hardwiring to FEC=10.

SUS_CUST_HAUX_M1_DAMP.adl (and assorted other screens): corrected hardwired FEC.

SUS_CUST_HAUX_M1_DEMOD.adl, SUS_CUST_HAUX_M1_TEST.adl, SUS_CUST_HAUX_M1_DAMP.adl: I got the L1-specific links from various "GUARDIAN SET" buttons pointing to the apparently intended screen, $(USERAPPS)/cds/common/medm/FILTALH.adl, and made the asgs list more standard, but there was a $(channel) argument missing, so I couldn't getting it working fully. It doesn't seem to work at LLO either, so I leave it to the Guardian folks to take further.

SUS_CUST_HAUX_OSC.adl: added $(IFO) $(OPTIC)$(num) string to screen.

Everything has been committed.

WP4034. Dave and Jim.

A third ADC card was installed into the h1oaf0's IO Chassis to support the expansion of the PEM system. The accelerometer channels from the PSL were attached to the second AA chassis.

I changed the h1iopoaf0 models to read the new ADC card. I changed the h1peml0 model to read the new accelerometers. We restarted both models. We found that there is a mismatch between the AA and ADCs for the second and third ADCs, we suspect an internal ribbon-cabling issue. For now we hooked the second ADC to the "third" ADC interface board to get the signals into the DAQ with the correct names for the weekend. We will fix the problem on Monday.

The DAQ was restarted to read the new PEM channels, frame size increased from 388MB to 413MB for a 32 second frame (now 13MB/s).

The corner station pem model has only been upgraded for the accelerometer channels. I will add the other signals on monday.

The DNS servers ns0 and ns1 have been updated to support local lookups of LHO GC and ligo.org addresses.  In other words, lookups for addresses in these domains are answered directly by ns0 and ns1, and do not rely on the LHO GC or ligo.org nameservers to be accessible for recursive lookups.  Any changes or updates to these domains however, are reliant on the GC nameserver being available so that the required zone transfer can take place.

08:00
Laser Safe
Alarms: CDs, and Dust #9

09:00 Continuous dust alarms on DM #9 (HAM2/3). Inform Patrick T.
09:03 Patrick T. move dust monitor #15 to HAM5/6 cleanroom
09:15 Kyle R. install and torque bolts on HAM1 east and west doors
10:00 John W. & Kyle R. close GV-5
10:05 Apollo crew craning ISI install fixtures and BSC repair arm over X-Arm
12:12 Dave B. Recycle H1OAF
12:45 Filiberto C. pulling cables at HAM1, 2, 3
13:45 Cory G. Working on ISC tables in East bay
14:40 John W open GV-5
14:35 Alexa S. to End-Y
15:58 Hugh P. and Dave B. recycle DAC to add HAM ISI channels

Installed the latest version of the HAM-ISI master model. This version features the addition of the ground STS signals as DAQ channels. Those new DAQ cannels will be used for the design of Sensor Correction filters.

HAM2-ISI and HAM3-ISI safe.snap were updated.

HAM-ISIs 2 through 6 were recompiled and restarted.

Isolation loops were turned back on on the IMC ISIs.

DAQ restart followed.

The bad hard drive in slot 6 of cdsfs1 has been replaced, with the new drive being set as a spare in the RAID.

I moved this dust monitor under the clean room over HAM 5 and 6.

Kyle - 
Added and torqued door bolts on HAM 1 East and West doors this morning in order to get doors to seal against their O-rings -> Pumped/vented HAM1/HAM2 annulus volume multiple times -> Helium signal 6.5 x 10-7 torr*L/sec (Fine Test Mode) before Pump/Vents and 4.5 x 10-7 torr*L/sec (Fine Test Mode) after Pump/Vents -> this may or may not be significant

Filiberto, Hugo

The 2 remaining ground STSs were connected to the HAM electronics racks. Signals are broadcasted to the ISI models. HEPI models need to be updated to receive those signals.

The chamber-side testing of HAM6-ISI was performed last week. HAM6-ISI Chamber-Side Testing Report is now available under the DCC.

Thursday, 11 July 2013

Gerardo Moreno, Scott Shankle, Mike Vargas, Nichole Washington

Weldments and Cylinder were unwrapped and brought into the cleanroom with the help of Apollo.

Weldments were assembled to Cylinder. 
Assembly was difficult because the holes are to small, no room for adjustment of parts. Hole diameter on Weldments and Cylinder is 0.209". Screw diameter is 0.190". All 35 locations were mated. 
Holes should be reamed out for future assemblies.

pics 1 and 2 - teamwork
pic 3 - buttoning up
pic 4 - end of the day

Time spent on actual assembly for the day was ~4 hours.

The /ligo file system on cdsfs0 was found in a read-only state.  I powered down cdsfs0, then restarted.

NOTE: cdsfs1 may have lost a disk from it's RAID. Investigation ongoing. 

Dan upgraded h1ldasgw0 to Solaris 11.1 SRU 8.4.0 (same as LDAS) and added multipathing support for the Nexsan SATAboy2. I have just restarted h1fw0. We will see if the upgrades help the instability problems with h1fw0.

We swapped out the IO Chassis for h1iscex with one having the correct number of ADC cards (built by using two TCS chassis). h1iscex now has the required 3 ADC cards and one 16bit DAC.

The IO Chassis for h1oaf0 was relocated from the LVEA teststand area to the PEM rack in the H1 CER. We used the onestop fiber cable already run for this Chassis.

(Stefan, Alexa)

We calibrated the PDH locking signal of the REFL beaming using Stefan's calibration scheme below (reference SUGWG lab logbook):

Motivation:
Easy way to calibrate a PDH error signal in Hz/Volt.

Input:
* Cavity pole frequency f0 (See next entry on how to measure this.)
* Maximum (peak) voltage of the error signal in a slow cavity scan: Vmax

Result: gain at lock point = f0/(2*Vmax)

Description:
The cavity resonance amplitude transfer function can (for high finesse cavities) be accuracy described by a Lorentizan:

TF = 1/(1+if/f0)

where f is the frequency offset from resonance. A phase modulated carrier field

Psi = 1 + i*exp(iwt) +i*exp(-iwt)

thus becomes

Psi_r = 1/(1+if/f0) + i*exp(iwt) +i*exp(-iwt)

The DC and 1-w part of |Psi_r|^2 are therefore

|Psi_r|^2 = 1/(1+(f/f0)^2) * ( 1 + [(1+if/f0)*(+i) + (1-if/f0)*(-i)]*exp(iwt) + C.C. )

Therefore the DC PDH error signal is

error/2 = [(1+if/f0)*(+i) + (1-if/f0)*(-i)] / (1+(f/f0)^2) = -2(f/f0) / (1+(f/f0)^2)

This has a maximum of 1 for f=f0 (see attached hand-written note). Therefore the full calibrated error signal is (sign omitted)

PDH_error = 2*Vmax*(f/f0) / (1+(f/f0)^2)

which has the slope 2*Vmax/f0 [Volts/Hz] around the locking point.

I created a script that imports the transmitted light (H1:ASC-Y_TR_A_NSUM_OUT_DQ) and reflected light (H1:LSC-REFLAIR_A_RF9_I_OUTMON) data. With this data, we follow the above calibration and create a fit. The script and results are attached. The graphs include the collected data and the calibration fit. The graph of the REFLAIR channel also includes the gain locking slope in red. From fitting the calibration to the data we succeeded in making a cavity pole meaurement of 82Hz. We also found the peak of REFLAIR to be 550 counts.

I did a major cleanup of the Tip-Tilt screens and linked them to the LHO SITEMAP.

I created argument macro files $(USERAPPS)/asc/common/medm/asc???_overview_macro.txt with all the optic-specific arguments (analogous to $(USERAPPS)/sus/common/medm/sus???_overview_macro.txt for SUS). I then went through all the screens in $(USERAPPS)/asc/common/medm/asctt with gedit, changing link arguments to look like the following:

args="%(read $(USERAPPS)/asc/common/medm/asc$(optic)$(num)_overview_macro.txt),USERAPPS=$(USERAPPS),SITE=$(SITE),site=$(site),IFO=$(IFO),ifo=$(ifo)"

Formerly some of the screens expected $(OPTIC) to be the full optic name, and some used $(OPTIC)$(num), so I changed everything to the latter convention.

Details of which files were touched are at end. I want to do one more round of checking for errors, so I haven't yet committed the changes yet.

No change:

ASC_CUST_TIPTILT_DAC_MONITOR.adl, ASC_CUST_TITPTILT_SAFE_SNAP_CMD.adl

Basic updates:

ASC_CUST_TIPTILT_BIO.adl, ASC_CUST_TIPTILT_DC_ALIGN.adl, ASC_CUST_TIPTILT_M1_COILOUTF.adl, ASC_CUST_TIPTILT_M1_DAMP.adl, ASC_CUST_TIPTILT_M1_DRIVEALIGN.adl, ASC_CUST_TIPTILT_M1_EUL2OSEM.adl, ASC_CUST_TIPTILT_M1_LOCK.adl, ASC_CUST_TIPTILT_M1_OPTICALIGN.adl, ASC_CUST_TIPTILT_M1_OSEM2EUL.adl, ASC_CUST_TIPTILT_M1_OSEM_ALIGN.adl, ASC_CUST_TIPTILT_M1_OSEM_ALIGN_counts.adl, ASC_CUST_TIPTILT_M1_OSEM_ALIGN_??ZOOM.adl, ASC_CUST_TIPTILT_M1_OSEMINF.adl, ASC_CUST_TIPTILT_M1_SENSALIGN.adl, ASC_CUST_TIPTILT_M1_WD.adl, ASC_CUST_TIPTILT_M1_WD_ACT_BANDLIM.adl, ASC_CUST_TIPTILT_M1_WD_OSEMAC_BANDLIM.adl, ASC_CUST_TIPTILT_M1_WD_OSEMDC_BANDLIM.adl, ASC_CUST_TIPTILT_MONITOR_OVERVIEW.adl, ASC_CUST_TIPTILT_OPLEVINF.adl, ASC_CUST_TIPTILT_OPLEVINF1.adl, ASC_CUST_TIPTILT_OSC.adl, ASC-TIPTILT_Main.adl, newDisplay.adl

More complicated fixes:

ASC_CUST_TIPTILT_DACKILL.adl: fixed broken link to GDS_TP screen.

Broken - needs redesign:

ASC_CUST_TIPTILT_M1_DEMOD.adl, ASC_CUST_TIPTILT_M1_TEST.adl: need to fix links to $(USERAPPS)/cds/common/medm/FILTALH.adl (2 places)

ASC_CUST_TIPTILT_BIO_all.adl, ASC_CUST_TIPTILT_MONITOR_OVERVIEW_all.adl, ASC_CUST_TIPTILT_OVERVIEW_all.adl: only allows for four (not five) TTs.