Starting this weekend, the slow down of the front end epics systems started on the H1 systems as h1boot was being backed up by cdsfs1. This is identical to the H2 problem which started July 4th.

Investigation is continuing, but for now the short term "fix" is to only rsync h1boot at 05:20 each morning. h2boot is rsynced at 05:40 each morning.

Changed gds software from gds-2.15.2 to gds-2.16.3 for linux workstations in control room.

Restarted IOP and x1sushxts27 models after the timing cable to the I/O chassis was replaced.  Complete power down, power up restart.  Burt restored 20120622-x1sushxts27_PR2.snap, restarted data concentrator.

( Aidan, Thomas, Elli)  Recently Aidan noticed fringes on the beam spot on the HWS down at EY.  He also noticed a coupling between beam movement and the spherical power the HWS senses.

In an attempt to clean up the fringes, yesterday I replaced the polarizzing beam splitter which sits just upstream of the HWS with a mirror.  This reduced the fringes on the HWS, improving the image quality.

This morning between 14:14 and 14:30 UTC Thomas and I gave a periodic excitation of PZT_1 and _2 in pitch and yaw and measured the spherical power fluctuation.  We missaligned the ITM to ensure the cavity was unlocked while taking measurements.  It appears we have reduced the coupling between beam motion and spherical power fluctuations.

The intensity noise eater on the Innolight laser is NOT turned on. Once we get it locking again we will turn it on to see the improvement.

Alberto and Bram.

With the strange increased power readings on the REFL_PD, which seemed to be linked to the Beckhoff PLL temperature feedback, and not being able to lock the PLL we headed to EY. After trying to find the beatnote (fiber output with the local laser), it turned out there was no light coming out of the fiber (even after Alberto relocked the RefCav). Keita re-relocked the RefCav.

With light coming out of the fiber, no beatnote was found until we tuned the crystal temperature from the original 31.27 deg C to 38 deg C (which seems to be it maximum). Even at that setting, it was at a few hundred MHz. All Beckhoff settings were turned 'off/cleared', as indicated on the screens.

To get the beatnote within range of the EY laser, I changed the RefCav to the next FSR (up in laser crystal temperature), with only a small change in beatnote frequency.

Adjusting the Lightwave (= RefCav) temperature down by ~0.5 deg C, requires the Innolight (EY) laser to adjust by 0.2 deg C (down), to follow the beatnote.

In the end I reduced the RefCav laser down by 5 deg C (to ~34.6 deg C), but tuning the laser temp of the Innolight laser did not to bring the beatnote down to ~40 MHz (still just above 100 MHz).

There is something fishy going on, but so far we didn't manage to find it.

Next we will unplug the temp feedback cable from the laser controller, just to be sure we don't have a secret offset.

Last week:
- we designed and installed new high blends in all 12 directions of the the BSC-ISI, in order to reduce low frequency gain peaking (both units)
- we designed and installed a .250Hz low blend in the longitudinal direction (Y) (both units)
this provided interesting improvement as reported in the last cavity measurement.

However,the measured isolation was not as good as expected (it did not follow the low blend as it should).

The problem seems to be in the blend filter conversion to the foton format: The low blend filters first coefficients are small and rounded.

The plot attached show the improvement after fixing the low blend in BSC8:
- Red curve is the reference (Stage 1 Damped)
- Blue Curve is Stage 1 Controlled, with rounded coefficients
- Green Curve is Stage 1 Controlled, with better filter coefficients

We'll work tomorrow on a solution to prevent this from happening again.

I leave both units up and running for Bram to possibly run cavity measurements. 







 
 

Jeff K, Jeff G, Thomas V


We thought it might be necessary to map out the quadrants of the ETMY and ITMY optical levers via offsets induced onto the suspension stages, namely M0.

After centering the beam on the optical lever QPD on ETMY, we offset the pitch on the M0 stage by a positive amount and expected the two lower quadrants of the QPD to have a spike in intensity (as deduced by the coordinate system set by SUS to determine which direction is positive and negative pitch and yaw).  Then offset yaw in a positive amount, which we expect to see the the two left quadrants to have a spike in intensity.  This was different than was what showing as pitch and yaw in the SUS medm and also is different than what I had used in my earlier calibration parameters.  This being said, the outcome of the mapping is as follows as viewed facing the QPD for ETMY:
13
42 

Repeating the test with the same process on ITMY OptLev, we found that the mapping didn't correspond once again to what was previously recognized as the mapping and is as follows:
42
13

The plot thickens:
As you can see, the two QPD's are not the same rotated about the center axis, which begs a confusing question of what might be going wrong. The most obvious answer is that the pitch signs are flipped, but where might this come from?  It could be a wiring issue in the hardware connecting the QPD to the ADC, a problem with a few negative signs somewhere in the readouts, or a difference in polarity of the output coil drivers F2 and F3 on the M0 stage of the ITMY.  The last option was found when we were trouble shooting and found that aLOG 2109 written by Jeff K. had addressed this issue but we couldn't find a log that resolved it.  It might be a combination of these things, still a work in progress.  It might be worth it to put in the glass viewport protectors and remove the enclosures to do a laser pointer test again on these QPDs.  Either way, the discrepancy between the two scenarios via the same process is perplexing.

Attached are plots of dust counts > .5 microns in particles per cubic foot.

The h1isiham3 model running on h1seih23 was quite old, from June 8th. Before we start ISI testing we (Hugo and I) upgraded this model by bringing in the latest L1 HAM3 model and converting L1->H1. From now onwards we should maintain the h1isiham3 model as separate from the l1isiham3. The new model required new Dolphin IPC channels from the h1susmc2 model, so that model was rebuilt from L1  but was not restarted on h1sush34. I had some problems recompiling the h2suspr2 model, there are IPC errors between MC2 and PR2 which are currently unresolved.

There are still some errors with the H1 ISI HAM3 medm screens. There appears to be residual links to M1 and S1 systems. I'll work with Hugo to resolve the H1 screens.

   We have been working on resolving a post creek bake misalignment problem with the lower blade break off heights on H1-MC3. We have tuned out a lot of the difference, but the break off heights are still slightly out of spec. We are going to take transfer functions of the M1 level to see if MC3 performance is within tolerance despite the misalignment of the lower blade break off heights.  
  
   These are the open light values for BOSEMs on HSTS MC3. The offsets and gains, listed below' have been entered into MEDM. The BOSEMs have been adjusted to 50% light. Plan is to take a set of TFs tonight when the lab is quiet.    

OSEM    S/N    OLV       Offset     Gain

 T1     571    30360.8   -15180     0.988
 T2     565    28718.9   -14359     1.045
 T3     312    30896.5   -15448     0.971
 LF      74    27143.9   -13572     1.105
 RT     313    30112.7   -15056     0.996
 SD     694C   25573.4   -12787     1.173 

This afternoon we obtained the spectra of the lower stages (M2 and M3) of MC2 in both the damped and undamped states (using plothsts_spectra.m).  The spectra were taken over the lunch hour so the LVEA activity should have been pretty quiet.  Data was committed to svn.  WIll post a PDF shortly.

As well, I used awggui to try to verify that I could see drive on the coils of these lower stages (for example, exc H1:SUS-MC2_M2_COILOUTF_UL_EXC while watching the OSEMINF_OUT_DQ channel for UL).  This will need to continue later as at first pass, I couldn't see much signal on the M2 AOSEMs.

I really messed up but at least I remembered!  For the past couple weeks as we've been doing initial alignment and Actuator connection, I failed to correct the elevation numbers in HAM3 for the Xarm slope.  Oh boy...In case you care to know---The LIGO coordinate frame is not co-planar with the gravity field in most places and in fact there is a 2.4mm shift between them moving from BSC2 (where we set them equal) just to HAM3.  This shift is what I forgot to include in my elevation calcs when positioning HAM3 ISI Optical Table.
So today we disconnected and relocked the eight HEPI Actuators and cranked the HEPI DSCW Springs to pull the platform up to position.  Jason (IAS) was called in ~1400 and he gave us a bit of horizontal correction.  Tomorrow we'll reconnect the Actuators and ask for another IAS shot to put this back to bed.
At least we are getting better at Actuator connecting!

Thanks to Eric, Mitchell and Jason.

Activities of the day of which I was informed:

- Elli K. to visit Y-END station, ALS table work, 8:57 - 11:18 am.

- Karen R. clean floor in the LVEA, 9:00 - 11:58 am.

- Hugh R. work in and around HAM03, all day.

- Filiberto C. work in the corner station VEA, briefly around 10:38 am.

- Karen R. work @ Y-END, clean different areas, including the floor inside the vacuum support equipment room, 1:45 - 3:45 pm.

- Cheryl V. work at Y-END, TMS, 2:51 - 3:55 pm.

- Jason O. alignment work, 2:55 - 3:50 pm.

We found that the reference cavity had lost lock.

I relocked it by disconnecting the fast and slow inputs, tweaking a bit the temperature knob and re-connecting the input cables.

The cavity is now likely locked to a different FSR from before. This means that the beat note at the end station might now be falling out of the PD bandwidth.

We have to go to end Y and adjust the ALS laser temperature to get the beat note in the PD range again.

I have taken the calibration data and analyzed it. The plot of the linear response curve is attached.

I used a polynomial fitting program in python to get these numbers.  This should give you the conversion from volts to micro-radians.

           Slope            Y-intercept
Pitch   [  1.01902201e+06   8.42886118e+00]
Yaw     [  9.83122943e+05  -1.19329010e+01]

[Bram, Alberto]

Both the ALS REFL PD and the TMS green QPDs are showing larger counts than before the weekend. It looks like the trend started sometime between Friday and Saturday night.

See attached plots. The REFL PD is set with a negative gain so that it reads 0 when the arm was unlocked, and it beceomes postive when the arm is locked.

Now we read around -4000 cts when the arm is unlocked.

Also the green QPD segments, both A and B read around 2200 cts, vs about 2000 as they did before.

We're not yet sure if this means that there's actually more physical power in the beam going. If that was the case, then it would be quite strange.

Could it be a DAQ artefact?