For a while the 227b dust monitor in this clean room has been reporting high counts. (see attached plot back to May 15th)

One theory was that this may be due to a missing prefilter on one of the HEPA units. (alog 6375)

This cleanroom is split in two, with a garbing area on the HAM2 side and cleaned parts on the other. Each half has two HEPA units. The HEPA unit with the missing prefilter is over the cleaned parts. The 227b dust monitor is also in this half.

This morning I took further measurements with one of the HHPC-6 particle counters. In the cleanrooms over the two test stands, and the larger one closer to HAM3, the counts were flat zero.
The counts I measured in the small cleanroom with the HHPC-6 appeared to confirm earlier counts with the 227b. The counts were elevated in both halves. I did not see a noticeable difference directly under the HEPA unit with the missing prefilter. However, inside the cleanroom near the floor of the entrance to the gowning side, the counts were repeatably and exceptionally high, I believe up to from 480 counts at 0.3 um to 160 counts at 5.0 um.

I plan to do more detailed measurements tomorrow.

The seismic BS front end h1seib2 crashed after it was "daq reloaded" (following a hand edit of its DAQ INI file).  This caused most of the corner station front end models to be marked as 0xbad by the DAQ. I did the following

• took h1seib2 out of the Dolphin fabric using h1seib1
• power cycled h1seib2
• performed a start_streamers on h1susb123

At this point all the front end models went into good 0x0 DAQ status. 

After power was restored, the h1iopseib2 model did not autostart, I restarted it by manually pressing the BURT button on the GDS_TP medm at the appropriate time. I was then able to start the models h1isibs and h1hpibs.

At this point the following models were reporting INI changes: h1isiitmy, h1isibs and h1isietmy with a DAQ status of 0x2000. I restarted the DAQ and all is now good.

Created new snapshots for mc1/mc2/mc3/etmy/itmy called "h1sus${optic}_lock.snap" representing the "lock" state of the suspension.

those can be recovered by going into the burtfiles directory

cd /opt/rtcds/userapps/release/sus/h1/burtfiles

and burtrestored with :

burtwb -f h1sus${optic}_lock.snap

I was up on the Mezzinine measuring for cables and noticed that the pumps were not turning.  No level trip, no VFD trip, servo running pushing max volts.  Richard reminded me of a problem that we should have corrected some time ago.  The output from the servo has 4 BNCs to the four pump stations.  There was a bad BNC that has shorted in the past and it did so again.  Once it shorts, no matter what the servo does, the outputs are shorted and the VFD gets zero volts hence no revs.  As soon as I unplugged the bad BNC they rev'd back up.  We took it back down and Filiberto replaced the BNC.  This should not happen again.  I may have bumped the cable and caused the short but I'm pretty sure it happened 10s of minutes before I was in the area.

The pumps went down at ~1818utc and the BSC2 HEPI moved a fair bit, several .001".  After the BNC repair I got the Pump Output back to operating pressure ~2040utc with the HEPI Actuators reaching equilibrium position a few minutes later.  The difference in before & after position at BSC2 was ~0 to ~600 counts, all less than 0.001".  Obviously a need for alarms--working on it!

(Chris, Alexa, Daniel)

We loaded the most recent LSC model from LLO. It includes inputs for the ALS channels as well as input from EY. It includes outputs to all suspensions controlled by LSC. The slow controls locking signals from the end ALS systems are now fed through the LSC input matrix and filter modules.

A new ISC model was loaded into EY. It has the following features:
- Removed remaining code related to green WFSs,
- Send slow arm feedback from the ALS REFL PHD locking to corner,
- Send transmitted power to corner,
- Added normalized sum for QPDs,
- Added communications library to interact directly with EtherCAT system, and
- Some filter and channel names have changed to make them more consistent between the slow and fast systems.

The EtherCAT computers ECATEX(Y) were renamed into ECATX(Y)1.

Duplicated channels in the EtherCAT system (SYS-COMM) were fixed. Some channel names had to be shortened as well. New DAQ ini files were generated and loaded for corner PLC2, EY PLC1 and EY PLC2. The auotburt files for ECAT systems were updated as well. They were copied into $(USERAPPS)/ecat.

A new subsystem directory was added to svn, $(USERAPPS)/als. All ALS medm screens have been moved over from the isc directory. So, more clean up work is still required.

New substitution rules were defined for ISC to allow parametrization of the end station screens. A list of example replacement strings can be found in $(USERAPPS)/isc/common/medm/isc_macro_subst.txt.

A large effort is underway to update all ALS and end station screens to use the new substitution rules. This will allow screens to work in both end stations without change. It will allow screens to work on both sites. And it will bring some order into the parameter names, so they can be used more consistently.

Attached are plots of dust counts requested from 5 PM June 10 to 5 PM June 11.

Both the dust monitor at location 14 in the LVEA (H2 PSL enclosure) and the dust monitor at location 16 in the LVEA (H1 PSL anteroom) are indicating calibration failures.

Attached are plots of dust counts requested from 5 PM June 9 to 5 PM June 10.

Both the dust monitor at location 14 in the LVEA (H2 PSL enclosure) and the dust monitor at location 16 in the LVEA (H1 PSL anteroom) are indicating calibration failures.

8:42 Praxair delivery
Cleaning at end X
The BSC9 ISI was moved onto the test stand at end X.
10:00 Kiwamu I. transitioned the LVEA to laser safe to allow Filiberto C. to pull power cables to ISCT1. (WP 3966)
11:12 Paradise water delivery
11:59 Mark B. restarted the h1susauxh2 model. (WP 3967)
Cyrus R. and Jim B. went to end Y to connect the reflected memory switch to the corner station. Dave B. changed the IOP models on iscey and lsc0 to turn on the RFM DMA. Chris W. modified the h1iscey and h1lsc models to use the RFM IPC. Dave B. then restarted all the models on these frontends. (WP 3963)
Dave B. reconfigured broadcast0 to add channels to DMT monitoring for HIFO Y. (WP 3961)
12:34 The DAQ was restarted.
1:04 The DAQ was restarted to add Ethercat channels.
1:15 The DAQ was restarted after removing duplicate Ethercat channels.
Kyle replaced a leaking 2.75 CFF gasket on the GV1 plumbing. (WP 3968)
1:37 The DAQ was restarted.
Jim B. went to mid X to remove the IRIG-B from h1pemmx and move the timing signal from the fan-out to the IO chassis. Did not work, reverted to slightly modified original configuration. (WP 3969) (see alog below)
Mark L. went to end Y to see if power cabling needed to be removed tomorrow.
Thomas V. to align the optical levers for ITMY and PR3.
Michael R. transitioned the LVEA to laser hazard.

Last week, I tried to evaluate the best configuration to use the sensor correction (HEPI vs ISI) and the influence of the ISI stage 1 blend frequency (CPS-T240-L4C) on the isolation once the sensor correction is engaged on the ISI.

The stage 1 motion was measured in different control configurations:
- HEPI controlled in position (UGF 5Hz), ISI – Isolation Level 3 (40Hz UGF on stage 1 – 15Hz on stage 2) – Blend 250mHz (with T240s in stage 1 super sensors)
- HEPI controlled in position (UGF 5Hz) with sensor correction from STS-2 to HEPI, ISI – Isolation Level 3 (40Hz UGF on stage 1 – 15Hz on stage 2) – Blend 250mHz (with T240s)
- HEPI controlled in position (UGF 5Hz), ISI – Isolation Level 3 (40Hz UGF on stage 1 – 15Hz on stage 2) – Blend 250mHz (with T240s) and sensor correction from STS-2 to Stage 1
- HEPI controlled in position (UGF 5Hz), ISI – Isolation Level 3 (40Hz UGF on stage 1 – 15Hz on stage 2) – Blend 750mHz (with T240s) and sensor correction from STS-2 to Stage 1
- HEPI controlled in position (UGF 5Hz) with sensor correction from STS-2 to HEPI, ISI – Isolation Level 3 (40Hz UGF on stage 1 – 15Hz on stage 2) – Blend 750mHz (with T240s)
- HEPI with IPS-L4C blend (UGF 10Hz) with sensor correction from STS-2 to HEPI, ISI – Isolation Level 3 (40Hz UGF on stage 1 – 15Hz on stage 2) – Blend 250mHz (with T240s)

Below 100mHz
- The amplification at low frequency seems to be more important when the sensor correction is implemented on HEPI (probably due to the tilt - twist of the HEPI structure). The effect is more visible in the horizontal directions.
  red & cyan (Sensor correction on the stage 1 ISI) vs purple & gold (Sensor correction on the HEPI)
- With the sensor correction implemented on the ISI, the amplification seems to be lower when the blend frequency (CPS-T240) is higher
  cyan (750mHz blend) vs red (250mHz blend)

[100mHz;2Hz]
- When the T240s and the CPSs are blended at 250mHz, sensor correction implemented in HEPI or ISI performs similarly (with SC on the ISI, isolation seems slightly better)
  red (SC on the ISI) vs purple (SC on the HEPI)
- The SC performs better when the blend frequency (CPS-T240) is increased at 750mHz. Are we reinjecting more noise or tilt when the CPS-T240 are blended at 250mHz? The Q of the low pass filter at 750mHz also has lower than the Q of the 250mHz low pass filter. It might also explain the difference of isolation around 200mHz.
  cyan (750mHz) vs red (250mHz)
- The smallest stage 1 motion is obtained when the HEPI is controlled with the IPS-L4C blend
  gold

Above 2Hz
-  Isolation is slightly better (at 2Hz) when the HEPI is controlled with the IPSs and the L4Cs.
   gold
-  When the T240s and the CPSs are blended at 750mHz, the isolation is limited due to the lack of “roll off” of the CPS low pass filters.
   cyan

Regarding these results, it seems that the sensor correction works better at low frequency (below 2Hz) when it is implemented on stage 1 (vs HEPI). Further measurements need to be performed to determine the effect of the blend filters with sensor correction on the isolation.

Tested a configuration for h1pemmx in which the timing for the I/O chassis was connected directly to the master fanout with a single-mode fiber, and the IRIG-B card removed from the front-end computer.  This would have allowed us to not use a timing fanout and IRIG-B fanout at the mid station for a single computer.

This test failed, running the front-end computer with no IRIG-B card installed caused the IOP model to have a 0x4000 status (timing error) and the state word also indicated a timing error. 

I reinstalled the IRIG-B card in the computer, and restored the timing system connections to their original state.  Single-mode fiber to the timing fanout at MX, multi-mode fiber connect the I/O chassis and the IRIG-B fanout, and coax cable connecting the IRIG-B fanout to the IRIG-B card in the front-end computer.  Had difficulty getting the IRIG-B card to read the correct time, eventually moved the connection to the middle card in the fanout where it sync'd up without problem.

WP3963. Jim, Cyrus and Dave. The 5595 RFM switches in the MSR and EY computer racks were connected to each other with single mode fiber optics cable. The ports used on the long haul patch panels were in conformance with Richard's drawing. The order of the front ends on the Y arm loop and node ids were verified. To permit DMA transfer of RFM data, the IOP models h1iopiscey and h1ioplsc0 were modified to add the rfm_dma flag to the CDS part. All models on these two front ends were restarted. Chris W modifed the h1iscey and h1lsc models to use RFM IPC and verified this is working. One problem he encountered was a build error with the h1lsc model, it uses RFM CARD=1 while h1iscey uses RFM CARD=0. A work around suggested by Rolf is to force h1iscey to use CARD=1 even though there is only one physical card, this fixed the error. This closes WP3963.

WP3961. The DAQ broadcaster was reconfigured to add a large number of DAQ channels needed for the DMT monitoring of the H1IFOY run. This closes WP3961.

ECAT EDCU. The DAQ was reconfigured to add three ECAT EDCU INI files, adding about 3,300 slow channels to the DAQ.

To support the above changes and other user changes, the DAQ was restarted multiple times today.

May also have a leak in GV1's bonnet annulus seal -> will follow up

Changed etmy model per request by isc team :

• deleted the recently created link H1:ISCEY-SUS-ETMY_LONG in etmy model coming from iscey model (cf picture 1)
• replaced it by the copied H1:LSC-ETMY_L_SUSETMY from lsc model (cf picture 2 and 3)

Per WP 3967, I edited the h1susauxh2 model to remove blocks PR3OL and H2OL which were generating spurious signals 

H1:SUS-PR3OL_PR3OpLev_UL/LL/UR/LR

and 

H1:SUS-H2OL_H2OL_UL/LL/UR/LR

These had always been nonsense because the physical OL quadrant signals weren't even plumbed into the h1susauxh2 chassis for the h1susauxh2 model to read them. Rather, they were going into a different, non-aux chassis, h1sus2b, and the correct digital signals were already being generated by user model h1susim as H1:SUS-PR3_M3_OPLEV_SEG1/2/3/4_IN1 and H1:ISI-HAM2_OPLEV_QUAD1/2/3/4_IN1.

ETMY:
I have attached spectra of ETMY stage 2 motion in the Y direction measured during the single arm and a month ago while working on a new set of filters. Isolation was improved except around 1Hz (Blend filters are different) but stage 2 motion is close from the requirements. It is a better tradeoff.

The conditions of the measurements are specified below:

OAT:
- ISI
    - Damping
    - UGF 15Hz on all DOFs
    - Stage 1 Blend frequency at 100mHz with T240s in super sensors in X, Y and blend frequency at 750mHz in the other directions
    - Stage 2 Blend frequency at 250mHz in X, Y and blend frequency at 750mHz in the other directions

- HEPI
    - IPS-L4C blend frequency at 800mHz in all DOFs
    - UGF 10Hz on all DOFs
    - Sensor correction in X, Y and Z directions

2013 05 14 - Measurements starting at 1052610960 - 100s sample - 10 averages - 50% overlap
- ISI
    - Damping
    - UGF 40Hz on all DOFs of stage 1 and 32Hz in all DOFs of stage 2
    - Stage 1 Blend frequency at 250mHz with T240s in super sensors (all DOFs)
    - Stage 2 Blend frequency at 250mHz (all DOFs)
    - Sensor correction from ground (STS-2) to stage 1 in the X, Y and Z directions
    - Sensor correction from stage 1 (T240s) to stage 2 in the X, Y and Z directions

- HEPI
    - Pure position control - No blend
    - UGF 100mHz on all DOFs
    - No sensor correction

ITMY:
Damping and isolation filters installed on BSC1 (ITMY) are those installed during the OAT (minor tweaks were necessary since HEPI is locked). Isolation performance is similar to OAT. New set of filters will be designed when the HEPI will be unlocked.

Found x1fe3 and x1seiham front end computers to be non-responsive.  Rebooted both, they appeared to start normally and run their models OK.

These will be used during Initial Alignment.  It isn't a big deal if these get bumped for now but we don't want them damaged.  Please be careful.

No issues to report here.  The ISI is on the Test Stand shifted south one set of holes (7".)  Like EndY, this is to accommodate the TMS installation.  Let us know soon if this is not appropriate.

This is a log-log plot with a 1/t curve superimposed with exponent of 0.85. Also shown is a zoom in of the last 10 days.

It appears that we have cleared up some virtual leaks(annuli?) in our pumping exercises.

Attached are plots of oplev signals of ETMY and ITMY, as well as the arm cavity reflected power (ALS-Y_REFL_B_LF_OUT).
Plot 1 shows the signals in lock.
Plot 2 is a time series in lock.
Plot 3 shows the same signals as plot 1, but out of lock.
Note that neither channel is properly calibrated yet. (TBD)

Conclusion:
- The limiting peaks are ~0.44Hz and ~0.55Hz.
- Both ETMY and ITMY show the motion.
- Our slow feed-back of the length signal to ETMY M0 makes them worse for ETMY in-lock.