Jamie, Hugo,

Tested the following HAM-ISI transition scripts, under the guardian interpreter:
READY      -> DAMPED
DAMPED   -> ISOLATED
ISOLATED -> DAMPED
DAMPED   -> READY

Minor adjustments appart, It was a very straightforward process, thanks to Sebastian and Charles help. 
 

HAM-ISI transition script are ready, and work under Jamie's new Guardian Interpreter 

Next steps for SEI guardian:

• merge safe.snap & alarm text files into guardsnap files (today)
• finish up recovery scripts & get them called at transition script failures (and not only when EPICS alarms are triggered)

The VME crate for the LVEA Y arm vacuum controls failed last night.  From initial investigation it looks like the power supply died.  I have swapped out the crate with a different one and re-installed all of the cards and cables.  Cyrus assisted in getting it back up and running. See his report.

I don't see any apparent way to attach the ISC cables through the top mass. 

There are four DB25 cables for ISC electrical components on the table/tele (i.e. picomotors, beam diverter, and QPDs). These are first attached to the top mass and then to the ISI table. Right now this is supposed to be done by using a standard PEEK clamp inside the top mass.

See the attached pictures, and a helpful cut-out  view drawing of the top mass from D0902163.

 In the pictures, you're looking down in a rectangular opening (or a "hole") in the top mass, and deep inside that hole there is a standard PEEK clamps tightened by thin hex head bolt. Now, we are supposed to somehow thread four DB25 cables through this hole, and then through these cable clamps, squeeze the clamps, and tighten them. The clamps are 3 inches or more down inside the hole, and you cannot access them unless you have 1/4" thick and maybe 5 inches long fingers. I even don't know if there's enough space to physically allow us to thread the DB25 heads through the top mass hole, let alone through the cable clamps. 

The only apparent possibility is to disassemble the top half of the top mass to expose the clamps, remove the clamps, thread the cables with bulky DB25 heads between the blades through the bottom half of the mass (I don't know if this is possible but let's assume it is), put the clamps on, then assemble the top half again, in-situ , and rebalance everything. That's not a real solution.

I, Corey and Stefan searched for a threaded hole or two on the top mass so we can attach some hold-down cable clamps, but couldn't find any. All threaded holes are already used by something else. And even if there is one, we cannot simply loop the cables around the mass, as now the table cloth encircles the mass completely, unlike the old design.

Someone please give me a procedure to make this work.

FYI, the old way to do this (i.e. at LHO EY): The cables are first fixed to the bottom of the top mass which has a cable clamp (e.g. see this picture of TMSY top mass taken from the bottom), then the cables are looped around the back side of the top mass like this picture, then attached to the top of the top mass which has the same clamp as the bottom, and then from the mass to the ISI table.

It was really clumsy to work with, so the ISC group requested an improvement together with other things, but until today I didn't appreciate that the mass structure was completely redesigned (as opposed to small changes here and there).

Found CP1 level @ 23" WC = 55% full at 0830 -> Determined CP1 LLCV fully closed via removing inst. air and observing no stroke -> Begin manual filling of CP1

The LY vacuum controls have been invisible since last night. I hope someone can correct this.

thanks

TMSX was balanced on the spot without using the balance bridge, and it was just fine.

First the top mass was fixed level using EQ stops, then table/tele was hung, balanced while half-supported by Genie lift using the movable masses, and I have to say that these are much smoother than the first article. Since the table/tele was measured 79.5kg, we added 2.5kg masses (3x500g, 5x200g).

Next the top mass EQ stops were all backed off, the table cloth was moved around a bit to allow the top mass to float freely, and of course we had to use the top slide masses to balance it.

As was observed at LLO, there were two spots where the clearance between the mass and the table cloth is uncomfortably small (less than a millimeter): One is the back right corner and the other is the front right EQ stop. Table cloth was moved to provide maximum clearance (but it is surrounding the mass so it's a zero sum thing).

Anyway, the suspension is hanging freely, the only thing is that the side BOSEM is almost touching, and that's why Caltech people sent us a modified BOSEM mounting plate, which I gave to LHO C/B today.

We haven't measured the height of the TMS, and we need a help from either AOS or SUS to measure this using optical level.

The following transfer functions are being measured overnight:

HAM2: Tilt decoupling frequency measurement (Interrupted last night by a DAC error)
HAM3: Tilt decoupling frequency measurement (Interrupted last night by a DAC error)
HAM6: Local to Local transfer functions for Initial In-Chamber testing (further inquiery)

Started a round of transfer functions overnight on ETMX top mass main and reaction chain with damping on and damping off. start time = 1060049009.

Potential increses in the ALS bandwidth make acoustic coupling in the hundreds of Hz region interesting. The figure shows the vertex microphone signal during HIFO-Y and compares it to the signal during S6 science mode. We may well do better than S6 because there will be fewer fans in the LVEA during aLIGO. I include a spectrum from the vertex mic at LLO, also during S6 science mode, because LLO differed from LHO during S6 in that most electronics racks  were outside the LVEA. Thus, in the hundreds of Hz region, the LLO curve may be closer to what we will achieve during advanced LIGO science mode at LHO. At low frequencies LLO is worse than LHO during S6 because we did not reduce the sound from the HVAC as much at LLO as we did at LHO.

The acoustic features above 100 Hz in the HIFO-Y spectrum are driven by sound, so we can expect these features to be at least an order of magnitude smaller. 

During the red beam lock, the 2 HEPIs and ISIs (BSC1 and BSC6) were supposedly in the state given by alog https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=7235. By the time, the HEPI at BSC1 was locked. Some information were extracted from the 15-minute lock. ISIs provide great isolation above 100mHz but the motion increase below 100mHz is not negligible and can become problematic to lock the cavity. The results present the ISI’s contribution in the arm’s length variation.

The arm was kept locked using the red beam.  The green beam is used as a witness sensor.
 

1- State of the ISIs
The attached spectra (H1_BSC_ISI_ST2_GS13_Y_20130726.jpg) show the motion of stage 2 in the Y-direction of the 2 BSC-ISIs (ITMY and ETMY) during the lock.
On ITMY (red curve):
- It is not sure if the sensor correction was engaged properly (no amplification at low frequency and no and important motion at 700mHz)
- The resonance of the HEPI pier is clearly visible at 11.5Hz

On ETMY (Blue curve)
- Above 2 Hz, the motion is much larger than on June 14^th - Reference in black (same configuration?)
 

2- ASD of the arm’s length seen by the green beam
The ASD of the green beam is presented in attachment (H1_GB_Cavity_Length_Red_Lock_20130726.jpg).
The large motion seen around 40mHz is created by the ISI (mainly ETMY cf Coherence). The aggressive filters used for the sensor correction strongly amplify the ISI motion around 50mHz. It’s a tradeoff between amplification below 100mHz and a “0 degree phase” above 100mHz.

3-Coherence from the ISI drives to the cavity length
Once the HEPI and the ISIs are controlled (ground excited only), I used the control drive (Isolation filters output, damping filters output negligible) to evaluate the coherence from the drive to the length of the arm. Coherence in the bandwidth of interest (0-200mHz) are presented in the video at: https://svn.ligo.caltech.edu/svn/seismic/BSC-ISI/H1/Common/Misc/Video/H1_ISI_Coherence_Cross_Coupling_20130725.avi, channels are:
- 1 to 6: ETMY stage 1 drive in the direction X, Y, Rz, Z, Rx, Ry
- 7 to 12: ETMY stage 2 drive in the direction X, Y, Rz, Z, Rx, Ry
- 13 to 18: ITMY stage 1 drive in the direction X, Y, Rz, Z, Rx, Ry
- 19 to 24: ITMY stage 2 drive in the direction X, Y, Rz, Z, Rx, Ry
- 25: Y-arm length

NB: Index 1,1 is at the bottom left (The diagonnal is going up). The matrix is symmetric.

From 40mHz to 60mHz where the motion is maximal, the coherence matrix shows:
- the 2 ISIs were maybe in a different state (block matrix 1:12,1:12 very different from 13:24,13:24). Was the sensor correction engaged properly on ITMY? It is difficult to evaluate at posteriori. The sensor correction might be also less effective on ITMY since the STS-2 installed in the beer garden is relatively far from the chamber (in comparison with ETMY). Is there a reaction from the HEPI (locked in for ITMY and unlocked for ETMY)?

- ETMY
    -The coherence is important from stage 1 Y drive to length, stage 1 Ry drive to length and stage 2 Y drive to length.
   - The cross coupling drives are important
   - There is an important Z to Rz cross coupling (bad positioning of the horizontal CPS)

- ITMY
   - There is some coherence from stage 1 Y drive to length, stage 1 Ry drive to length and stage 2 Y drive to length. The coherence from the drives to the length is globally lower than ETMY.
   - The cross coupling drives are moderate
   - There is an important Z to Rz cross coupling (bad positioning of the horizontal CPS)

Some measurements are currently running to evaluate the tilt (X->RY and Y->Rx). The translation to tilt coupling will be used to correct the position sensor signals. This correction is used to take care of the CPS misalignment.

We can also notice a good coherence between the two ISIs on stage 1 in the Y  and Z direction and stage 2 in the Z direction.

4- Coherence between the STS-2s and the arm
The coherence between the STS-2 and the arm is presented in the video https://svn.ligo.caltech.edu/svn/seismic/BSC-ISI/H1/Common/Misc/Video/H1_STS2_Coherence_Cross_Coupling_20130725.avi Channels are:
- 1:3: ETMY STS-2 X,Y,Z
- 4:6: ETMY STS-2 X,Y,Z
- 7: Y-Arm

The coherence at 60mHz is shown in attachment H1_STS2_Coherence_Cross_Coupling_20130725.jpg. Around 60mHz, the coherence between the corner station and the EY is important in the Y and Z directions. The coupling beween the ground Y direction is important. The ground motion is amplified by the sensor correction of the ISI then visible in the cavity.
 

5- Contribution of the drives in the arm's length
I am currently trying to evaluate what is the contribution of each drives in the arm.

At 10:45 last night (local) the DAC outputs of h1seih23 went to zero (and latched there) and the WD error bit was set on the IOP STATE_WORD. We worked with Rolf to diagnose the problem. For this error to have occurred there must have been a DAC FIFO error. He looked at the /proc/h1iopseih23/status file and it is showing no current FIFO errors. He has not seen a DAC FIFO error come and go before. With no futher online diagnostics needed, I restarted all the models and handed the system back to Hugo for HAM2 commissioning at 10:54 this morning.

A trend of the IOP STATE_WORD shows the WD bit being set at 10:45 Wed evening, which was latched on until this morning's restart. At two occasions overnight the DAC bit was also briefly set for less than a minute each time.

Rolf is looking into the code to see if additional diagnostics can be applied in future releases of the RCG.

Kyle, Gerardo 

Today, after John had notified Rai of our intent, we applied a few drops of Vacseal (SPI Supplies, Space Environment Laboratories, original formula 0502-AB) to the top of the previously identified leaking 1.33 CFF on GV6 (North gate annulus port).  The leak seams to be fixed now.  

I stopped h1fw0  to allow Dan to grow the SATABOY disk system controlled by h1ldasgw0. He merged the H1 and H2 SATABOYS into one disk system, doubling H1's capacity.

here are the times between which no frames were being writen by h1fw0

EX -----------------------
- ETMX alignment - Jason in AM
- TMS - Keita
- Betsy - at ETMX late AM and after lunch

EY -----------------------
- electronics - Filiberto

Corner ----------------------
- HEPI - Hugo and Hugh
--- Hugo - any electronics work on HAM2/3? - unable to drive due to unknown issue with IOP watchdog - see alog from Dave/Rolf

- H1SEIH23 IOP STATE WORD - actual real problem related to Hugo's 
--- Dave and Rolf looking into it - MEDM colors wrong (green when channel is really red) - see Dave's alog

- Travis - at ITMX

- DUST LAB1 and LVEA16 calibration failure

MX -------------------
CP6 filled - alarmed, but is OK now

Travis, Jason, Betsy

Over the last 4 days, we have worked to tune all 12 DOFs (6 each lowest mass on 2 chains) to within the IAS tolerances.  As usual, since all DOFs crosstalk, we took many roundabouts, but finally think we have acheived it.  Jason will append final pointing/position numbers.  TFs taken tonight will exhonorate this "final" pointing.  After TMS joins us on the table (still a few days out) and finishes their IAS alignment, we will finish payloading the QUAD.

Note, the alignment process takes many days because after each adjustment, many mechanical brackets and structural items need to be realigned such that the BOSEMs are in range and we can turn the damping on in order to make another IAS measurement of pointing.  As well, we have to maintain the floating OSEM/magnet pairing alignments on the lower stages which cause the chain to go out of alignment due to their own weight.  Then, IAS makes a measurement, we make the pointing adjustment in the chain, readjust the floating OSEM/magnet pairs (often at the cost of some of the point adjustment we just made), readjust many mechanical brackets, reenable damping everywhere, repeat IAS alignment, cycle around again.