SEI- Hugh running TFs on HAM6. Jim working on isolation filters on HAM5

SUS - work on quads for 3IFIO; last assembly almost complete. TMONS left to build.

TCS - X arm aligned. Viewport covers will be removed for more accurate alignment. Y-arm: wrong lens installed in vacuum - no point in trying to align at this time

CDS- cabling for aux systems continues. building RF box for 3IFO.

Commish - cont. alignment of DRMI this weekend.

I found the laser status screen indicating the "WARNINGS" had triggered.  Went to the diode room to investigate.  Found "NPRO switchoff during run"
illuminated.

I reset the laser, everything returned to normal.  I have to look up what the error message means as I do not recall what it was for.  There is
no evidence of the NPRO dropping out over the past 48 hours.  Attached is the trend data for the past month.  No signs of the NPRO tripping out.

Maybe a glitch set it off, or either the insertion of the reference cavity heater card or the outer loop power stabilisation card.  Will check
again on Tuesday to see if the condition crops up again.

no restarts reported

As preparation for automating the initial alignment, I did some things associated with the X arm locking.

(LSC-TRX_A_LF path was realgined at ISCTEX)

After the PSL light was locked to the X arm, I realigned TRX_A_LF path in order for LSC to be able to do triggered lock acquisition. The beam was already hitting the center of the bottom periscope mirror from the beginning. I steered the bottom periscope mirror and some optics in the down stream. I found that a lens had been taken out  presumably due to the previous QPD work in this past July (see alog 12672). There was a post holder and a base plate secured on the table in front of the bottom periscope mirror. Looking around, I found a 2 inch lens (PLCX-50.8-515.1-UV-1064) with a post laid on the table. So I inserted it to the post holder and looked this was the right lens as the beam nicely converged towards the photo-diode (Thorlabs PDA100A).

Also, I put a temporary steering mirror aside. This is one of the mirrors that Sheila placed in this past July (alog 12672) and has been blocking the PDA100A. The DC voltage coming out from the PDA100A was about 300 mV with a 00-mode transmitted light on it.  I did not check the gain setting of the PDA100A. The ADC count reached 1x10⁴ cnts when the arm was locked on a 00 mode. Probably we will need to decrease the gain of the PD or that of the interface box at some point in the future in order to aviod saturation.

In addition, I adjusted the trans camera. I needed to move a black-place beam dump closer to a beam splitter which is the one before the PDA100A. Then I steered the beam splitter and the angle of the camera to coarsely center the beam in the view. I could not identify the channel number for this signal at the video matrix in MSR. Looks like channel #53 is it, but it keeps flickering for some reason and therefore hard to tell. Though, a 00-mode beam was visbile ocasionally. We need to fix this in the day time tomorrow.

(IM4/PR2 alignment with REFL9 WFS)

Instead of aligning IM4 and PR2 by hand, I started doing it using the REFL WFSs. I used REFL_RF9_A and _B and they worked well. Once I zero-ed the electronics offsets in the RF9 demodulated signals,they started looking reasonable. I closed two loops by feeding the A_RF9_I signal to IM4 and the other to PR2. IM4 turned out to be a well-diagonalized actuator in the sense that it was able to actuate mostly on the WFS_A signals. On the other hand, PR2 had some cross-coupling which is not surprising at all. I need to work on this a bit more in order to have more diagonalized servos. I did not precisely adjust the gains. So these are the remaining tasks.

Anyway, closing the loops increased the transmitted light and reduced first-order coupling from angular fluctuation of some optics. Note that I had the DC centering servo running in order to keep the spot centered on both WFS_A and _B.

Baffle PDs, now with the correct labels:

Especially in pitch, these are rather different from what we found yesterday, which is not too suprising since we were hitting the ITM somewhere far from the center yesterday.  There is a beam at the end station coming back from the ITM, and some fringing at ISCT1.  To be continued....

Sheila, Kiwamu

We reconfigured most of the GigE cameras such that the analog gain is set to a high value in order to make the process of finding green light easier.

Now green beams as well as objects are quite visible on the cameras. For example, see the attached images below:

The camera image on ITMX looks similar to what Livingston has and therefore I am satisfied with the high gain setting.

(some details)

Sheila and I were having difficulty finding green light in most of the GigE cameras and therefore we had been wondering why they were not as sensitive as those in Livingston.

Looking at the camera initialization files in /ligo/cds/lho/h1/camera/ , we found a type which had been preventing the camera servers from changing the analog gain in its initialization process. In order to set the analog gain, the line should start from 'Analog Gain', but the line in all the files started from 'Default Analog Gain'. So we took out the 'Default' and reconfigured the servers by pressing the 'reload config' button in the medm screens. Due to this typo, all the cameras had been set to a default low gain regardless of what gain were requested in the file. According to the camera code (see camera_server.py in svn), the default analog gain is 100. We increased most of the camera gains to 1023 which is almost 10 times higher than they used to be. I am not sure if this is the highest, but certainly this was the largest number I saw among those files. In addition, changing the gain to some other higher value did not apparently make it brighter. So I guess that it should be already at the highest or close to the highest value. 

Additionally, we found another typo in the initialization file: 'Default Exposure' should be 'exposure' according to camera_server.py. We fixed both typos in all the existing files.

Peter K, Sudarshan, Gabriele

As stated in the previous alog entry, we finally understood the cabling of the picomotors, and we checked that we can move all the four axis of the two ISS motorized mirrors. 

We managed to get a beam on the QPD, with a total sum signal of about 12000 counts. Since the transimpedance should be 100k, this corresponds to something like 100 uW, which is even more that what we expect from 10 mW in input of the ISS array. So we are quite confident that we have the real beam on the QPD.

However, we still don't see much power on any of the eight photodiodes. The signal is something of the order of 5-6 counts at maximum for some of the diodes. Moving the picomotors while trying to keep the beam on the QPD didn't help much.

Searching for the maximum power on the photodiode is made difficult by the lack of a real DC output: the present board has a -20 dB gain at DC, rises with a first zero at 6 mHz, to about 60 dB at 10 Hz. This means that any transient in the power impinging on the photodiodes will easily mask the improvement we should get if the beam pass over the photodiode too fast.

So we modified the signal acquisition: using a DB9 break box we directly wired the test points after the transimpedance to the ADC, skipping the whitening. This modification has been implemented for PD1,2,3,4.

Moving around the picomotors we could finally get a beam on two photodiodes: PD1 sees something like 0.2 mW and PD3 something like 0.15 mW. This is still much lower than the expected ~1 mW per diode. PD2 and PD4 see some signal, but much smaller. We don't know what's happening to PD5-8, since they're still acquired with the whitening.

In this position, the beam is not hitting the QPD. 

Aidan. Greg.

I fine-tuned the alignment of the ITMX HWS probe beam this afternoon. Actually, I'm now getting a return beam from ETMX, which I don't want, so I'm a little too close to a perfect retro-reflection. I'll detweak it tomorrow.

The camera has also been moved close to the conjugate plane of the ITM HR surface.

All of this was done with the LEXAN cover on the viewport. Further refinement will be necessary when this is removed. It should also clean up the return beam image.

The ITMX return beam is shown here. You can see the normal shadow from the baffles and the earthquake stops. There are no noticable stray beams coming from anywhere else.

I also misaligned ITMX to prevent the return beam getting back. Here's the return beam from ETMX. I know this is from ETMX because it swings back and forth when I put an injection into ETMX YAW.

LVEA is Laser Hazard

08:21 Hugh – Running transfer functions on HAM6 HEPI
08:52 S. Sachder – Going to End-Y to check PEM microphones
08:54 Danny – LVEA west bay to work on 3IFO Quads
09:06 Filiberto – Cabling ISS Array Picomotor (PM5) to HAM2 feed-through
09:15 Travis – LVEA west bay to work on 3IFO Quads
09:33 Karen – Cleaning at Mid-Y
09:42 Richard – Working on cabling for PM5 at HAM2
10:00 Adian – Going into the LVEA
10:15 Cyrus – Going to all outbuildings to property tag equipment
10:15 Betsy – LVEA west bay for 3IFO Quad work
10:51 Richard – Going to LVEA to work on cameras with Filiberto
12:58 Danny – LVEA west bay to work on 3IFO Quads
13:15 Travis – LVEA west bay to work on 3IFO Quads
14:15 Greg – Attaching flex tubing between HWF and HAM4 viewport
15:17 Sheila – Focusing cameras on both X & Y arms 
15:42 Adian – Working on flex tubing installation at HAM4

All auto running under guardian, watchdogs if you please--H

Gabriele, Peter K., Daniel, Sudarshan

HAM2 Picomotor Channel List:

From our capacitance measurments (LHO: alog 14148), and the cable specification drawings: D1000581  and D1101515, we concluded that the picomotor on HAM2 are connected to the following channels:

Channel        Picomotor

1                     ISS (PM1-closer to the array)

2                     IM4 Transmission QPD 

3                     Faraday Half Wave Plate

4                     IMC Input Pointing

8                     ISS (Before the telescope) 

5, 6, 7             Empty

This is consistent with the channel assignment at LLO alog 12174

During the process we moved the Faraday HWP (Channel 3) and IMC Input Pointing (Channel 4) quite a bit, but we brought it back to its original position, atleast on theory. We moved these the same number of steps back as we moved forward and viceversa. :)

With the last TF data section taken last night, ran through the scripts to load generic controllers.  They closed no problem.  I believe troubles before were related to changes in the local to cartesian matrices.  Started the Guardian node and if functions fine.

Attached is a before and after spectra plot.  It looks like there is some increases with the loops closed on vertical dofs around 0.7 to 3hz but the broad peak at 4+hz is reduced.  The ISI has only damping loops.

Gabriele, Peter, Sudarshan

HAM2 Picomotor Test Using Test Setup :

We tested the picomotor 3 (supposedly ISS H1) and 5 (supposedly ISS pickoff) using the picomotor test setup and did not see any signal change in the StripTool window except for the 3X channel. 3X channel showed some increase in signal amplitude but would settle back down after the drive was complete. 

Picometer Cable Capacitance Measurement:

We measured the capacitance across each pair of connector and got the following results:

1X: 209 nF  (J1: 1-14)                         5X: 4.0 nF  (J2: 1-14) 

1Y: 208 nF  (J1: 2-15)                         5Y: 3.9 nF  (J2: 2-15)

2X: 211 nF  (J1: 3-16)                         6X: 3.9 nF  (J2: 3-16)

2Y: 212 nF  (J1: 4-17)                         6Y: 3.9 nF  (J2: 4-17)

3X: 217 nF  (J1: 5-18)                         7X: 3.9 nF  (J2: 5-18)

3Y: 4.3 nF   (J1: 6-19)                         7Y: 3.9 nF  (J2: 6-19)

4X: 201 nF  (J1: 7-20)                        8X: 211 nF  (J2: 7-20)

4Y: 211 nF  (J1: 8-21)                         8Y: 211 nF  (J2: 8-21)

Summary:

           1, 2, 3X, 4 and 8 has picomotor connected

           3Y, 5, 6, 7 are empty

We can now monitor the SR mirros and also can use the TV monitors as outputs. See the attached:

Daniel and I have found that the optical lever sign convention for ETMs and ITMs was changed in May/2013, and now it's opposite of what I remember.

I'd like to know why it is/was a good idea.

Right now, the dot on the screen goes up when the OL beam goes down.

Keita, Sheila, Gabrielle

• We have a single shot beam from TMSY onto ISCT1, with 10uW of light, we have a beam from the ITM in the end station, and we have a second beam from ETMY on ISCT1. 
• Positive pitch is in the opposite direction on the ETMY alignment slider than on the TMSY alignment slider (I believe that TMS is the incorrect one).
• There seems to be a mix up with the baffle PD cabling.
• The ALS Y GigE camera has a softaware problem: the exposure cannot be changed from 2168317875, even though auto exposure is off.

Now, in chronological order:

result of TMSY to IMTY baffle PD search:

Based on D1200657-v4, (assuming the view shown is looking at it from the BS), this does not make sense with the arrangement of baffle PDs.  These votlages are low compared to what we see on the X arm, so we suspected that this was a second bounce or ghost beam, however, they do not move when the ITM or ETM are moved, so they are not reflections off of those optics.  If we think that if these are the real beam, the baffle PDs are mislabeled somewhere, so we used the TMS suspension alingment sliders to try to figure out which PD is which.  (Unfortuately, the TMS alignment slider also seems to be wrong.)

Based on our reinterpreation of the labels, we thought that the beam should be roughly centered on the optic when TMS is at -35 PIT, -31 Yaw.  However, we saw no sign of our beam at this location, nothing on ISCT1 and nothing returning to the ETM when we scan the ITM.  Keita trended the old values, and we saw the beam go by on SRM when he restored them.  Gabriele and I went to ICT1, and found a badly clipped beam by exciting the BS.  Later Keita and I returned and moved the TMSY alignment to get a somewhat better beam, with a TMS alignment of -87 PIT and -51.4 YAW and BS at 187 PIT and -258 YAW, we have 10 uW on ISCT1.  this was the same in the spring (alog 11296)  giving us some confidence that we really have the single shot beam on ISCT1.  (This is our first clue that the TMS alingment slider is wrong)

We went through the baffle PD game by moving ITMY and looking at the ETMY baffle PDs, using this TMS pointing which isn't necessarily centered on ITMY:

This seems to make sense, but the power is low again.  We think this beam should be roughly centered on the ETM with the ITM slider at 156.5 PIT -123 Yaw. 

Since exciting the ETM (around its saved alignment) and watching on the MC1 camera and at ISCT1 didn't show us any return beam, we redid the baffle PDs for the ETM pointing:

This beam also did not make sense with the baffle PD arrangement, or our interpretation of the actual PD arrangement based on the TMS alingment sliders.  This means that either ETMY or TMSY has a sign flip in the pitch alignment slider (our second clue): 

Once we understood this I decided that our ETMY interpretation of the cabling must be the correct one, since the TMS pointing that actually gets a beam through the optic is very different in pitch than the pointing we calculated based on the baffle PDs.  Using this information, our original TMSY centered pointing should have been -113 PIT, -35 Yaw.  Indeed, this pointing does result in a beam that doesn't look clipped on ISCT1.  I can also see two other beams on ISCT1 which come from the ETM, but haven't been able to easily align the arm cavity.  Although I see some fringing, the mutliple bounce beams seem to get clipped as I try to align them.  I think the next step is to start fresh with baffle PDs, now that we think we know which one is which.

QUAD 06 (Q6) Phase 1B transfer function plots are attached.   We had a hard time obtaining good coherence in the Transverse TF, so it is a bit hashy.  Will try again.  

Most notably is that, like Q8, the second pitch mode frequency is unexpectedly pushed upward on the main chain.  Recall, we never found the mechanism to fix it on Q8.  Interestingly, both the Q8 and Q6 assemblies are of the same batch of wires and are fresh builds, but by 2 different assembly teams, and on 2 different solid stack/test stand units.  Q8 is an ETM type of QUAD while Q6 is an ITM QUAD, but both main chains have the same pendulum parameters - both are detailed in the 'wireloop' model.

The Q6 data is plotted as QUADTST.