40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  40m Log, Page 184 of 337  Not logged in ELOG logo
ID Date Author Type Category Subject
  7709   Tue Nov 13 22:12:15 2012 JenneUpdateIOOTurn off MCL path before doing MC spot measurement

[Jenne, Den]

We have discovered that the MCL loop squishes the length fluctuations that result from the MC spot measurement angular dither.  This is good, in that the MCL is doing what it ought to do.  However, we need to turn it off before doing a spot measurement.

  7708   Tue Nov 13 21:05:35 2012 DenUpdateAdaptive Filteringonline and simulation

For a last few days I've been working on oaf and simulink model to simulate it. First I did online subtraction from MC when MC_L path was enabled. Inside my code I've added a sum of squares of filter coefficients so we can monitor convergence of the filter.

coeff.png     online.png

To to this I've measured path from OAF output to input without AA and AI filters. Then made a vectfit using 2 poles and zeros. Foton command

zpk( [-2.491928e+03;5.650511e-02], [-4.979872e+01;-3.278776e+00], 6.011323e+00)

mag.png    phase.png

My simulink model consists of 3 parts:

  • cavity with seismic noise at low frequencies, 1/f^2 noise at medium frequencies and white noise at high frequencies
  • this cavity is locked using feedback compensation filters that we use to lock arms
  • locked cavity with adaptive filter

Adaptive filter in the model uses online c-code. It is connected to simulink block through an S-function. Sampling frequency of the model is 10 kHz. It works fairly fast - 1 sec of simulation time is computed in 1 sec.

overview.png       af.pngsim.png  sim_coeff.png

I've tested FxLMS algorithm and MFxLMS algorithm that is faster. I plan to test 2 iir adaptive algorithms that are already coded.

  7707   Tue Nov 13 21:05:22 2012 AyakaUpdateWienerFilteringAcoustic noise cancellation with MC

 In order to perform acoustic noise cancellation with MCL signal, I am trying to find sweet spots for microphones.

I set microphones at various places around MC chambers, and see how coherent microphones and MC signals are.
I had checked the half part of MC.

  • data set #1
    place where I set the microphones (left), MCL signal (blue) and its error (green) (right top), and coherence between microphones (original: fine lines, error: thick lines) (right bottom).
    microphone_pos1.pdfpsd_coh1.png
  • data set #2
    microphone_pos2.pdfpsd_coh.png 
  •  data set #3
    microphone_pos3.pdfpsd_coh.png
  • data set #4
    microphone_pos4.pdfpsd_coh4.png

The acoustic noise around the MC2 chamber is most critical so far. I could subtract the signal and the sensitivity got 2 times better.
I will see the acoustic coupling from the other side of MC.

Attachment 4: psd_coh.png
psd_coh.png
  7706   Tue Nov 13 20:13:07 2012 KojiConfiguration40m UpgradingEndtable upgrade for auxiliary green laser : Circularize the 1W NPRO beam profile

These fs are too short.

Quote:

I propose to use lenses with focal lengths f1 = 22.2 mm and f2 = 25 mm. The beam diameter after the lenses will be dx = 164.05 um and dy = 163.19 um.

 

  7705   Tue Nov 13 16:18:51 2012 ManasaConfiguration40m UpgradingEndtable upgrade for auxiliary green laser : Circularize the 1W NPRO beam profile

With reference to measurements made earlier: elog

Beam parameters for Innolight 1W NPRO are:

wx0 = 160 um 

wy0 = 181.1 um

z0x = -9.17 cm

z0y = -10.19 cm

The beam is clearly elliptical.  We will introduce an additional pair of cylindrical lenses to circularize the beam before it enters the faraday.

I made calculations for the beam divergence ratio and checked with thor labs catalogue of cylindrical lenses to find pairs that will match the ratio. 

I propose to use lenses with focal lengths f1 = 22.2 mm and f2 = 25 mm. The beam diameter after the lenses will be dx = 164.05 um and dy = 163.19 um.

  7704   Tue Nov 13 11:30:54 2012 ManasaConfiguration40m UpgradingEndtable upgrade for auxiliary green laser

I'm set on the mission to get the new bigger endtables setup for the auxiliary green laser; now that the tables are already here.

I want to have everything documented in this same thread for future reference. It has been a pain trying to filter relevant elogs. I'll be working on the layout redesign one at a time....starting with the ETMX end.

This is the simplest cartoon layout of  ETMX endtable (not the actual table layout):

ETMX_sch.png

I have been searching through the elogs for the beam parameters measured earlier. I'm assuming they would not have changed much and will make calculations based on them.

However, we will have to change a few not-so-good mounts and include/exclude some optics.

 

P.S. HR (steer) are necessary steering mirrors and HR are just folding mirrors for the drawing.

 

 

 

  7703   Tue Nov 13 08:34:11 2012 SteveSummaryVACpumpdown after 24days at atmosphere

 

 

Attachment 1: pd74d1.png
pd74d1.png
Attachment 2: 24datm.png
24datm.png
Attachment 3: pdd1.png
pdd1.png
  7702   Tue Nov 13 00:28:10 2012 AyakaUpdateIOOMC locked, spots centered

[Rana, Ayaka, Jenne]

We aligned the REFL beam to the center of PD.
Also we removed the small black parts from mirror holder so that the beam is not clipped. They are originally for holding the mirror, but the mirror should be held by the small screw on the side of the mirror mount. This screw was hidden by the label, so we moved the label on the right hand side of the mirror mount (See a picture below).  
DSC_4902.JPG

Also we removed the half-wave plates and PBS so that laser power is increased.
DSC_4904.JPG

Then I aligned the beam for PMC, locked MC, and centered the beam spots.

 spot1112.png

The MC2 pitch is a little bit high but still close enough to the center.

Jenne had also centered the beam spots on QPDs for WFS.

  7701   Tue Nov 13 00:27:50 2012 JenneUpdateElectronicsEthernet Illuminator Control

Quote:

 

 The ethernet power strips have been installed. 192.168.113.110 is on ETMX, 192.168.113.111 is on ETMY and 192.168.113.112 is on the vertex. I have also written an EPICS file "illuminator_control.adl" (currently stored in my named directory) that allows a user to turn individual sockets on and off at each of the three locations. Some short tests have indicated that everything is in working order.

Currently, no illuminators are hooked up to the power strips. However, the power control will most likely be ready for use tomorrow, granted I can find and use extension cords so that the illuminators might reach their respective power strips.

 I'm sure Charles meant to also say that he connected the ETM power strips to the ethernet switches in those racks.  For the vertex, the ethernet switch is in 1X2, but there isn't space in there, so the power switch was installed in 1Y2.  The vertex ethernet cable is along the overhead inside cable tray.

I'm not sure what we want to do about connecting the new power strips to the illuminators.  No illuminator is close enough that its built-in cable can reach the power strip, so we'll need extension cables or some such.  Charles is going to ask Steve about the plan tomorrow.

  7700   Tue Nov 13 00:19:51 2012 JenneUpdateCamerasITMX is just fine

Quote:

In other, semi-unrelated news, the ITMXF camera has been not working for a while.  The bottom right quad on the test mass tv has been dark for at least a week or two.  Steve, when you have a chance (after the oplevs are all taken care of), can you see if there's something obvious that's wrong?

 While helping Charles string the ethernet cable (elog 7698) for the power strip in the vertex (from 1Y1 to 1X1), I looked at the ITMX power cord.  It was connected to the same power strip as the illuminator, which, since the illuminator was turned off by turning off the power strip, meant no power was going to the camera.  Since Charles is very close to having the new power strips set up, I unplugged the ITMX illuminator and turned the power strip back on.  ITMX camera is back to normal.

  7699   Tue Nov 13 00:15:08 2012 JenneUpdateIOOInput and Output PZTs turned on

I turned on the power supplies for the output PZTs and pitch and yaw for PZT2.  This is back to the condition that we had during atmosphere alignment, so after Ayaka has finished tweaking the MC, we can have a look at alignment of the interferometer.

  7698   Mon Nov 12 23:38:50 2012 CharlesUpdateElectronicsEthernet Illuminator Control

Quote:

Configured ethernet controlled power strips to have static IP addresses: 192.168.113.110, 192.168.113.111 and 192.168.113.112.

Wrote a python script to interact with the power strips that can turn individual sockets on or off via telnet.

This functionality will be implemented on the control room computer GUIs in short order.

 The ethernet power strips have been installed. 192.168.113.110 is on ETMX, 192.168.113.111 is on ETMY and 192.168.113.112 is on the vertex. I have also written an EPICS file "illuminator_control.adl" (currently stored in my named directory) that allows a user to turn individual sockets on and off at each of the three locations. Some short tests have indicated that everything is in working order.

Currently, no illuminators are hooked up to the power strips. However, the power control will most likely be ready for use tomorrow, granted I can find and use extension cords so that the illuminators might reach their respective power strips.

  7697   Mon Nov 12 16:52:03 2012 SteveUpdateVACpumpdown has completed

 The pressure is CC1 3e-4 Torr at 8 hours of pumping.  Status: vacuum normal

The IFO is ready for light

 

Attachment 1: pd74.png
pd74.png
Attachment 2: pd74at8h.png
pd74at8h.png
  7696   Mon Nov 12 09:21:10 2012 SteveUpdateVACpumpdown has started

I'm looking for IOO steering mirror power supplies. I found two so far: one at 1Y3 as Kepco  + 75V and one at OMC_North rack -150V

Are there any more? The OMC is off.

ANSWER: only one of the steering PZT can be driven, so see their power supplies

Attachment 1: beforepd.png
beforepd.png
Attachment 2: IMG_1800.JPG
IMG_1800.JPG
Attachment 3: IMG_1801.JPG
IMG_1801.JPG
  7695   Fri Nov 9 18:28:23 2012 CharlesUpdateSummary PagesCalendar

The calendar tab now displays calendars with weeks that run from Sunday to Saturday (as opposed to Monday to Sunday). However, the frame on the left hand side of the main page still has 'incorrect' calendars.

 

  7694   Fri Nov 9 17:15:05 2012 Manasa, Steve, AyakaUpdateGeneralWe're closed! Pumping down monday morning

Quote:

After a brief look this morning, I called it and declared that we were ok to close up.  The access connector is almost all buttoned up, and both ETM doors are on.

Basically nothing moved since last night, which is good.  Jenne and I were a little bit worried about how the input pointing might have been effected by our moving of the green periscope in the MC chamber.

First thing this morning I went into the BS chamber to check out the alignment situation there.  I put the targets on the PRM and BS cages.  We were basically clear through the PRM aperture, and in retro-reflection.

The BS was not quite so clear.  There is a little bit of clipping through the exit aperture on the X arm side.  However, it didn't seem to me like it was enough to warrant retouching all the input alignment again, as that would have set us back another couple of days at least.

Both arm green beams are cleaning coming out, and are nicely overlapping with the IR beams at the BS (we even have a clean ~04 mode from the Y arm).  The AS and REFL spots look good.  IPANG and IPPOS are centered and haven't moved much since last night.  We're ready to go.

The rest of the vertex doors will go on after lunch.

Jamie and Steve got the ETM doors on this morning.

We got the other heavy doors including the ITMs, BS and the access connector in place.

If nobody raises any concerns in reply to this elog, Steve will assume it as a green signal and will start pumping down first thing Monday morning after the final check on the access connector bellow screws.

 

Steve! 

Ayaka and I got the ITMY and BS door closed at 45foot pounds just now. 

  7693   Fri Nov 9 11:38:38 2012 jamieUpdateGeneralwe're closing up

After a brief look this morning, I called it and declared that we were ok to close up.  The access connector is almost all buttoned up, and both ETM doors are on.

Basically nothing moved since last night, which is good.  Jenne and I were a little bit worried about how the input pointing might have been effected by our moving of the green periscope in the MC chamber.

First thing this morning I went into the BS chamber to check out the alignment situation there.  I put the targets on the PRM and BS cages.  We were basically clear through the PRM aperture, and in retro-reflection.

The BS was not quite so clear.  There is a little bit of clipping through the exit aperture on the X arm side.  However, it didn't seem to me like it was enough to warrant retouching all the input alignment again, as that would have set us back another couple of days at least.

Both arm green beams are cleaning coming out, and are nicely overlapping with the IR beams at the BS (we even have a clean ~04 mode from the Y arm).  The AS and REFL spots look good.  IPANG and IPPOS are centered and haven't moved much since last night.  We're ready to go.

The rest of the vertex doors will go on after lunch.

  7692   Fri Nov 9 08:11:57 2012 SteveUpdateAlignmentnothing moved - all good

 Oplevs and IP ANG are still centered. Why  do the SRM and PRM move 5X  more ?  I could not check the sensing voltages because the computer failed.

AS and REFL are looking the same as last night.

ALL LOOKING GOOD!

 

Attachment 1: opsingGOOD.png
opsingGOOD.png
Attachment 2: op8h.png
op8h.png
Attachment 3: 8amF.png
8amF.png
Attachment 4: keepfailing.png
keepfailing.png
  7691   Thu Nov 8 22:04:43 2012 CharlesUpdateElectronicsEthernet Illuminator Control

Configured ethernet controlled power strips to have static IP addresses: 192.168.113.110, 192.168.113.111 and 192.168.113.112.

Wrote a python script to interact with the power strips that can turn individual sockets on or off via telnet.

This functionality will be implemented on the control room computer GUIs in short order.

  7690   Thu Nov 8 20:54:08 2012 Manasa, AyakaUpdateAlignmentReconfirming on IPPOS, IPANG and oplevs centering

Quote:

" We found that IPANG was not on its photodiode, but determined that it was centered on all of the in-vac mirrors, and that it was just a little bit of steering on the ETMY end out-of-vac table that needed to be done."

Manasa took photos of all test mass chambers and the BS chamber, so we can keep up-to-date CAD drawings. 

Oplevs and IPPOS/IPANG are being centered as I type.  Manasa and Ayaka are moving the lens in front of IPANG such that we have a slightly larger beam on the QPD.

 

The lens in front of IPANG on the out-of-vac table was moved to get a larger beam giving reasonable signals at the QPD.

IPPOS did not need much adjustment and was happy at the center of the QPD.

All oplevs but the ETMY were close to the center. I had to move the first steering mirror about half an inch on the out-of-vac table to catch the returning oplev beam from ETMY and direct it to the oplev PD.

* We have taken reasonable amount of in-vac pictures of ETM, ITM and BS chambers to update the CAD drawing.

 

  7689   Thu Nov 8 20:11:54 2012 JenneUpdateAlignmentAS steering moved out of POY's way, 2 green beams onto PSL table

[Jenne, Jamie, Manasa, Ayaka]

 

Flipped mount of OM2, moved OM2 behind POY pickoff so we're out of the way of POY.  Adjusted and recovered rest of AS path.

We found that IPANG was not on its photodiode, but determined that it was centered on all of the in-vac mirrors, and that it was just a little bit of steering on the ETMY end out-of-vac table that needed to be done.

Got green flashes in Yarm, moved down periscope to the north by ~1 inch in order to get y green out to PSL table.  This also involved moving the steering mirror on the IOO table immediately after the down periscope to match.  We measured the MC spot positions before and after touching the periscope, and there was no significant change. 

Aligned X green to X arm (centered on ITMX, ETMX, although no flashes since we didn't move ETMX's biases around), then made sure it was centered on all of its steering mirrors, and came out of the vacuum.

Manasa took photos of all test mass chambers and the BS chamber, so we can keep up-to-date CAD drawings. 

Oplevs and IPPOS/IPANG are being centered as I type.  Manasa and Ayaka are moving the lens in front of IPANG such that we have a slightly larger beam on the QPD.

 

In the morning, Jamie is going to put apertures back on 2 of the suspended mirrors for one last check that moving things on the IOO table didn't do anything bad, but since the AS and REFL beams on those cameras didn't move significantly, we think things are fine. 

Heavy doors go on in the morning, and access connector at ~1pm, if not before lunch.  Then Steve will start pumping early Monday morning!  Hooray!

 

PS, for reference,

AS: AS_8Nov2012.png

REFL: REFL_8Nov2012.png

  7688   Thu Nov 8 10:11:58 2012 AyakaUpdateWienerFilteringPlay with wiener filtering

I will attach a document which describes how the noise affect the wiener filter and the noise cancellation ratio.

And I re-estimate the SN ratio in the microphone (but still rough):

R_model.png

The yellow line is modeled signal level, and cyan line is modeled noise level.

Then, the estimated filtered residual noise is:

estimate.png

The noise is already subtracted enough below 80 Hz even though there is still coherence.
Above 300 Hz, the residual error is limited by other noise than acoustic noise since there is no coherence.
I am not sure about the region between 100-300 Hz, but I guess that we cannot subtract the acoustic noise because primary noise (see the document), such as a peak at 180 Hz, is so high.

Attachment 1: document.pdf
document.pdf document.pdf document.pdf
  7687   Thu Nov 8 09:04:45 2012 SteveUpdateCamerasCanon T3i purchased

Canon EOS Rebel T3i has been purchased. See rating

Attachment 1: BH_409130400.pdf
BH_409130400.pdf
  7686   Wed Nov 7 23:22:45 2012 ranaUpdateAlignmentJamie's tip tilt proposal

 With Picos, we lose the ability to dither the input beam as well as align the beam with the IFO locked. And the active TT will still have hysteresis, but also actuators. Once in vacuum, I'm not sure how we adjust them - what's the error signal for PR2/PR3 ?

If the interferometer is aligned, why not just pump down now? I'm not sure that we have evidence of TT hysteresis issues once people stop touching them.

  7685   Wed Nov 7 19:07:45 2012 JenneUpdateCamerasNo beam seen on external camera views

I have written some scripts which collect photos, then average them together, and subtract out an averaged background (as Rana described in elog 7678). 

I am not seeing any beam spots on any of the resulting pictures. 

 

The script to get 500 pictures is

.../scripts/general/videoscripts/videocapture50

and it's inputs are {name of camera} {folder to save in} {noBeam or withBeam}, where noBeam and withBeam indicate whether or not the PSL shutter is closed.  For the saved photos to work nicely with the Matlab script, the folder to save in should be in the format (Month_day_year/CAMERA).  So today's ITMYF pics, for example, are in Nov_7_2012/ITMYF/ .

So, you run it once with the shutter open, and once again with the shutter closed.

 

To create the new picture, open ImageBkgndSubtractor.m in the same .../scripts/general/videoscripts folder, edit the top few lines (month, day, year, camera name).  Run it, and it will read through all the pictures and supply a background-subtracted output, and save the output (as well as a version where every pixel value is multiplied by 3) in the same folder as the 500 pictures.

The pictures are all saved in

/opt/rtcds/caltech/c1/scripts/general/videoscripts/photos

so really, for my example above, it would be /opt/rtcds/caltech/c1/scripts/general/videoscripts/photos/Nov_7_2012/ITMYF/, with 2 subfolders, noBeam and withBeam, and the final pictures are saved in /opt/rtcds/caltech/c1/scripts/general/videoscripts/photos/Nov_7_2012/ITMYF/ .

 

In other, semi-unrelated news, the ITMXF camera has been not working for a while.  The bottom right quad on the test mass tv has been dark for at least a week or two.  Steve, when you have a chance (after the oplevs are all taken care of), can you see if there's something obvious that's wrong?

Here are the background subtracted photos that I've taken today:

BS_PRM_7Nov2012_SpotImage_pixelsTimes3.png

ETMXF_7Nov2012_SpotImage_pixelsTimes3.png

ETMYF_7Nov2012_SpotImage_pixelsTimes3.png

ITMYF_7Nov2012_SpotImage_pixelsTimes3.png

MC2F_7Nov2012_SpotImage_pixelsTimes3.png

MC2F is included, even though you can see the spot usually, just to prove that I'm not trying to subtract away the spot!  You just can't see it in any other picture.

  7684   Wed Nov 7 17:20:01 2012 jamieUpdateAlignmentJamie's tip tilt proposal

Quote:

Steve's elog 7682 is in response to the conversation we had at group meeting re: Jamie's proposed idea of re-purposing the active tip tilts.

What if we use the active TTs for the PR and SR folding mirrors, and use something else (like the picomotors that Steve found from the old days) for our input steering?

I think we will still need two active steering mirrors for input pointing into the OMC, after the SRC, so I think we'll still need two of the active TTs there.

My thought was about using the two active TTs that we were going to use as the input PZT replacements to instead replace the PR2/3 suspensions.  Hysteresis in PR2/3 wouldn't be an issue if we could control them.

With static input pointing, ie. leaving PZT2/3 as they are, I think we could use PRM and PR2/3 to compensate for most input pointing drift.  We might have to deal with the beam in PRC not being centered on PRM, though.

Koji's suggestion was that we could replace the PZTs with pico-motors.  This would give us all the DC input pointing control we need.

So I guess the suggestion on the table is to replace PZT1/2 with pico-motor mounts, and then replace PR2/3 with two of the active tip-tilts.  No hysteresis in the PRC, while maintaining full input pointing control.

  7683   Wed Nov 7 15:51:44 2012 JenneUpdateAlignmentJamie's tip tilt proposal

Steve's elog 7682 is in response to the conversation we had at group meeting re: Jamie's proposed idea of re-purposing the active tip tilts.

What if we use the active TTs for the PR and SR folding mirrors, and use something else (like the picomotors that Steve found from the old days) for our input steering?

  7682   Wed Nov 7 15:17:15 2012 SteveUpdateAlignmentsteering option with pico motor?
We have two ready for vacuum 1.5" mirror mounts with pico motors in our hands. 


 
Attachment 1: IMG_1795.JPG
IMG_1795.JPG
  7681   Wed Nov 7 14:59:35 2012 SteveUpdatePEMhigh particle count ALART & oplevs

Quote:

The BS camber is open only. We should close ASAP

Outside air quality is 1.7- 2.2  million particles  / cf min of 0.5 micron

 Air is still bad and the chambers are closed. Before lunch  Jamie repointed the PRM oplev. Manasa and I reset oplevs: BS and ITMX.

ETMX and ETMY are fine.

SRM and ITMY oplevs needs more work.

Attachment 1: oplevs.png
oplevs.png
  7680   Wed Nov 7 11:43:29 2012 SteveUpdatePEMhigh particle count ALART

The BS camber is open only. We should close ASAP

Outside air quality is 1.7- 2.2  million particles  / cf min of 0.5 micron

Attachment 1: highpartcount.png
highpartcount.png
  7679   Wed Nov 7 09:09:02 2012 SteveUpdateAlignmentAlignment-
PRM and SRM  OSEM LL 1.5V are they misaligned?
Attachment 1: 9amNov7w.png
9amNov7w.png
  7678   Wed Nov 7 07:11:10 2012 ranaUpdateAlignmentAlignment- POY and oplevs. photos.

The way to usually do image subtraction is to:

1) Turn off the room lights.

2) Take 500 images with no beam.

3) Use Mean averaging to get a reference image.

4) Same with the beam on.

5) Subtract the two averaged images. If that doesn't work, I guess its best to just take an image of the green beam on the mirrors using the new DSLR.

  7677   Wed Nov 7 00:10:38 2012 Jenne UpdateAlignmentAlignment- POY and oplevs. photos.
Can we have a drawing of what you did, how you confirmed your green alignment as the same as the IR (I think you had a good idea 
about the beam going to the BS...can you please write it down in detail?), and where you think the beam is clipping? Cartoon-level, 20 
to 30 minutes of work, no more. Enough to be informative, but we have other work that needs doing if we're going to put on doors 
Thursday morning (or tomorrow afternoon?).

The ETMs weren't moved today, just the beam going to the ETMs, so the oplevs there shouldn't need adjusting. Anyhow, the oplevs I'm 
more worried about are the ones which include in-vac optics at the corner, which are still on the to-do list.

So, tomorrow Steve + someone can check the vertex oplevs, while I + someone finish looking briefly at POX and POP, and at POY in 
more detail.

If at all possible, no clamping / unclamping of anything on the in-vac tables. Let's try to use things as they are if the beams are getting to 
where they need to go.  Particularly for the oplevs, I'd rather have a little bit of movement of optics on the out-of-vac tables than any 
changes happening inside.

I made a script that averages together many photos taken with the capture script that Rana found, which takes 50 pictures, one after 
another. If I average the pictures, I don't see a spot. If I add the photos together even after subtracting away a no-beam shot, the 
picture us saturated and is completely white. I'm trying to let ideas percolate in my head for how to get a useful spot. 
  7676   Tue Nov 6 18:39:16 2012 DenUpdatedigital noiseifo checking system

Matlab version of ifo digital noise estimation code is almost ready. It estimates digital noise introduced by each filter bank in each model. I'm waiting for NDS group to complete function to download online data to Matlab. Now code downloads data from the past that is not great because not all _IN1 channels are recorded and some of them are recorded at lower frequencies.

There might be some useful functions in this code for other applications as I've heard during the meetings. This is what it does

  • reads model names from the input list
  • for each model
    • finds corresponding Foton file and extracts modules with sos filters and sampling rate of the model
    • finds corresponding MDL file and makes a search for subsystems with "top_names" tag and "biquad=1" tag
    • creates _IN1 channel names using module names and subsystems with "top_names" tag
    • for each channel inside the model
      • reads filter bank parameters (which filters are ON, switches, limit, offset...)
      • downloads data
      • calculates output and estimates digital noise
      • checks that output is less them limit if it is on
      • reports if something is wrong

NDS group plans to release the function to download online data this week. Hopefully, it will be possible to download ~30 channels at a time. Code will need a few minutes of data for each channel. So it will be possible to check the whole ifo during the night.

At this point I've checked 40m using DQ channels. We have ~40 IN1_DQ channels with non-empty filter banks. These are osems channels. Digital noise is low for them.

  7675   Tue Nov 6 17:22:51 2012 Manasa, JamieUpdateAlignmentAlignment- POY and oplevs

Right now, Manasa, Jamie and Ayaka are doing some finishing touches work, checking that POY isn't clipping on OM2, the second steering mirror after the SRM, and they'll confirm that POX comes out of the chamber nicely, and that POP is also still coming out (by putting the green laser pointer back on that table, and making sure the green beam is co-aligned with the beam from PR2-PR3.  Also on the list is checking the vertex oplevs.  Steve and Manasa did some stuff with the ETM oplevs yesterday, but haven't had a chance to write about it yet.

We were trying to check POY alignment using the green laser in the reverse direction (outside vacuum to in-vac) . The green laser was installed along with a steering mirror to steer it into the ITMY chamber pointing at POY.

We found that the green laser did follow the path back into the chamber perfectly; it was clipping at the edge of POY. To align it to the center of POY (get a narrower angle of incidence at the ITMY), the green laser had to be steered in at a wider angle of incidence from the table. This is now being limited by the oplev steering optics on the table. We were not able to figure out the oplev path on the table perfectly; but we think we can find a way to move the oplev steering mirrors that are now restricting the POY alignment.

The oplev optics will be moved once we confirm with Jenne or Steve.

 

[Steve, Manasa]

We aligned the ETM oplevs yesterday. We confirmed that the oplev beam hit the ETMs. We checked for centering of the beam coming back at the oplev PDs and the QPDsums matched the values they followed before the vent.

Sadly, they have to be checked once again tomorrow because the alignment was messed up all over again yesterday.

  7674   Tue Nov 6 17:07:04 2012 jamieUpdateAlignmentAS and REFL

AS: tmp6oTENk.png

REFL: tmplamEtZ.png

  7673   Tue Nov 6 16:38:37 2012 jenne, jamie, ayaka, manasaUpdateAlignmentAlignment back under control again

We had a big alignment party early this morning, and things are back to looking good.  We have been very careful not to bump or touch tables any more than necessary.  Also, we have removed the apertures from the BS and PRM, so there are no more apertures currently left in the chambers (this is good, since we won't forget).

We started over again from the PZTs, using the PRM aperture and the freestanding aperture in front of PR2, to get the height of the beam correct.  We then moved PZTs to get the beam centered on BS, ITMY, ETMY.  We had to do a little poking of PR2 (and PR3?) to get pitch correct everywhere.

We then went to ETMX to check beam pointing, and used BS to steer the beam to the center of ETMX.  We checked that the beam was centered on ITMX.

We went through and ensured that ITMX, ITMY, PRM, SRM are all retroreflecting.  We see nice MICH fringes, and we see some fringes (although still not so nice...) when we bring PRM and SRM into alignment.

We checked the AS path (with only MICH aligned), and made sure we are centered on all of the mirrors.  This included steering a little bit on the mirrors on the OMC table, in yaw.  Initially, AS was coming out of the vacuum, but hitting the side of the black beam tube.  Now it gets nicely to the table.

For both AS and REFL, we made sure there is no clipping in the OMC chamber.

I recentered the beams for AS and REFL on their respective cameras.

IPPOS was centered on the QPD.  This involved moving the first out-of-vac steering mirror sideways a small amount, since the beam was hitting the edge of the mirror.  IPANG was aligned in-vac, and has been centered on the QPD.

Right now, Manasa, Jamie and Ayaka are doing some finishing touches work, checking that POY isn't clipping on OM2, the second steering mirror after the SRM, and they'll confirm that POX comes out of the chamber nicely, and that POP is also still coming out (by putting the green laser pointer back on that table, and making sure the green beam is co-aligned with the beam from PR2-PR3.  Also on the list is checking the vertex oplevs.  Steve and Manasa did some stuff with the ETM oplevs yesterday, but haven't had a chance to write about it yet.

  7672   Mon Nov 5 20:37:01 2012 AyakaUpdateWienerFilteringPlay with wiener filtering

 I am trying to find what limits the reduction rate with wiener filtering.
I did some calculations below:

Reduction rate estimation by microphone noise
estimate.png

When the instrumental noise (noise in microphone) and noise injected to signal after the acoustic signal is injected exist, the noise cancellation rate is limited. (I will write a short document about it later.) I assumed that there is only instrumental noise and that the other noise in PMC is below enough, and calculated the cancellation rate. The instrumental noise is modeled according to the measurement before (ELOG).
The green line is the original PMC signal, the red one is PMC residual error, and the blue one is PMC residual error estimated by the cancelling rate.
Around 30 - 80 Hz, the wiener filtering seems to be already good enough. However, I do not know what limits the cancellation rate (such as 100 - 200 Hz).

 

Filtering signals

Wfilter_notch.png

I hypothesized that the wiener filter is not good because of some peaks or other noise. So I filtered the PMC signal and mic signal to see the difference.
The red line is wiener filter with no filters, the blue one is with filters (low pass, high pass, and notch).
The wiener filter seems to get smoother but the PMC residual error did not change at all.

 

  7671   Mon Nov 5 19:38:52 2012 jamie, jenne, ayaka, denUpdateAlignmentmore alignment woes

Earlier this morning we thought things were looking pretty good.  IPPOS, IPANG, and the AS and REFL spots looked like they hadn't moved too much over the weekend.  Our plan was to do a quick check of things, check clearances, etc., tweak up the oplevs, and then close up.  This is when I made the ill-fated decisions to check the table levelling.

The BS table was slightly off so I moved one of the thick disk weights off of the other disk weight that it was sitting on, and on to the table next to it.  This seemed to improve things enough so I left it there.  ITMY didn't need any adjustment, and I move a couple smaller weights around on ITMX.  Meanwhile Jenne was adjusting the output PSL power back into it's nominal range (<100mW), and re-tweaking up the mode cleaner.

When we then looked at the vertex situation again it was far off in yaw.  This was clearly evident on PZT2, where the beam was no longer centered on the PZT2 mirror and was near the edge.  This was causing us to clip at the BS aperture.

We took some deep breaths and tried to figure out what we did that could have messed things up.

Jenne noticed that we had moved slightly on the PSL QPDs, so she adjusted the PSL output pointing to re-aquire the previous pointing, and realigned the MC.  This had a very small positive affect, but not nearly enough to compensate for whatever happened.

We spent some more time trying to track down what might have changed, but were unable to come up with anything conclusive.  We then decided to see if we could recover things by just adjusting the PZT input steering mirrors.  We couldn't; recentering at PRM, BS, ITMY, and ETMY was moving us off of PR3.

Jenne suggested we look at the spot positions on the MMT mirrors.  I had checked MMT1 and it looked ok, but we hadn't looked at MMT2.  When we checked MMT2 we noticed that we were also off in yaw.  This made us consider the possibility that the BS table had twisted, most likely when I was securing the moved mass.  Sure enough, when I manually twisted BS table, by grabbing it with my hand, very little force would cause the input beam to walk much of the way across PZT2, more than accounting for the offset.  The effect was also very clearly hysteretic as well; I could twist the table a little and it would stay in the new position.

At this point we had fucked things up enough that we realized that we're basically going to have to walk through the whole alignment procedure again, for the third time this vent.  We were able to recover the PRM retro-reflection a bit, but the tip-tilts have drifted in pitch (likely again because of the table levelling).  So we're going to have to walk through the whole procedure systematically again.

Lessons learned:  Many things are MUCH more sensitive than I had been assuming.  The tip-tilts are of course ridiculous, in that lightly touching the top or bottom of the mirror mount will move it by quite a lot in pitch.  The tables are also much more sensitive than I had realized.  In particular, tightening screws can twist the table hystereticly by milliradians, which can of course completely loose the pointing.  We need to be a lot more careful.

Assuming the table hasn't moved too much we should be able to recover the alignment by just adjusting the PZTs and tweaking the pitch of the tip-tilts.  At least that's the hope.    No more touching the table.  No more leveling.  Hopefully we can get this mostly done tomorrow morning.

  7670   Mon Nov 5 13:28:15 2012 jamieUpdateGeneral40m DCC document tree

Link to the new 40m DCC Document Tree: E1200979

  7669   Mon Nov 5 10:34:48 2012 SteveUpdateVACexisting vacuum documents

Quote:

Quote:

Quote:

Quote:

Apparently all of the ION pump valves (VIPEE, VIPEV, VIPSV, VIPSE) opened, which vented the main volume up to 62 mTorr.  All of the annulus valves (VAVSE, VAVSV, VAVBS, VAVEV, VAVEE) also appeared to be open.  One of the roughing pumps was also turned on.  Other stuff we didn't notice?  Bad. 

 Several of the suspensions were kicked pretty hard (600+ mV on some sensors) as a result of this quick vent wind.  All of the suspensions are damped now, so it doesn't look like we suffered any damage to suspensions.

CLOSE CALL on the vacuum system:

Jamie and I disabled V1, VM2 and VM3 gate valves by disconnecting their 120V solenoid actuator before the swap of the VME crate.

The vacuum controller unexpectedly lost control over the swap as Jamie described it. We were lucky not to do any damage! The ion pumps were cold and clean. We have not used them for years so their outgassing possibly  accumulated to reach ~10-50 Torr

I disconnected_ immobilized and labelled the following 6 valves:  the 4 large ion pump gate valves and VC1,  VC2  of the cryo pump. Note: the valves on the cryo pump stayed closed. It is crucial that a warm cry pump is kept closed!

This will not allow the same thing to happen again and protect the IFO from warm cryo contamination.

The down side of this that the computer can not identify vacuum states any longer.

This vacuum system badly needs an upgrade. I will make a list.

 While I was doing the oil change of the roughing pumps I accidentally touched the 24 V adjustment knob on the power supply.

All valve closed to default condition. I realized that the current indicator was red at 0.2A  and the voltage fluctuated from 3-13V

Increased current limiter to 0.4A and set voltage to 24V     I think this was the reason for the caos of valve switching during the VME swap.

 

 I asked Jamie to read these docs so we can do some troubleshooting while at atmosphere. We have to have the all valve closed DEFAULT condition working-tested.

I'd like to reconnect all valves after this test so the state identification software could work, including interlocks.

Posted the following documents on 40m wiki.

 

http://www.ligo.caltech.edu/~ajw/40m/40mVaccum_T000054-00.pdf
http://www.ligo.caltech.edu/~ajw/40m/40mProcedures.doc 

https://dcc.ligo.org/DocDB/0015/D000338/000/D000338-00.pdf
  7668   Mon Nov 5 09:53:35 2012 JenneUpdateAlignmentAS, REFL camera shots

Today's photos:

AS:

AS_5Nov2012.png

 REFL:

REFL_5Nov2012.png

IPANG / IPPOS trends:

IPPOS_IPANG_weekend_trend.png

 

 c.f. screen caps from Friday:

Quote:

These don't show anything too interesting, but we're including them to show where the beams are right now on the cameras, so we can compare on Monday.

 

AS:

AS_2Nov2012.png

 

REFL:

REFL_2Nov2012.png

 

 

 

  7667   Sat Nov 3 10:31:59 2012 ranaUpdateAlignmentPOP, POX, POY, IPPOS, IPANG, REFL all coming out of vac

  That's good, but I request two things:

1) Check that the REFL beam is coming from the HR surface and not the AR surface. The real REFL beam should have as much power as the Faraday output. And where does the AR surface reflection go?

2) Use frame grabber to get as many images of the spot positions on the mirrors as is reasonable. Don't endanger bumping the tables again, but take what images can be gotten by remote camera views.

  7666   Fri Nov 2 21:40:04 2012 ManasaUpdateAlignmentAS, REFL camera shots

 

 To get the camera shot of AS, Y1 mirror on the path was replaced by a 99% BS and transmitted beam was directed to the camera via a 50-50 BS (ND filters were distorting the image on the camera introducing fringes).

  7665   Fri Nov 2 20:41:53 2012 JenneUpdateAlignmentAS, REFL camera shots

These don't show anything too interesting, but we're including them to show where the beams are right now on the cameras, so we can compare on Monday.

 

AS:

AS_2Nov2012.png

 

REFL:

REFL_2Nov2012.png

 

 

  7664   Fri Nov 2 19:59:54 2012 JenneUpdateAlignmentPOP, POX, POY, IPPOS, IPANG, REFL all coming out of vac

[Evan, Jenne, Jamie]

We used the green laser pointer technique to adjust the POP steering mirrors behind PR2 to get the POP backward beam out onto the table (rather, the mirrors were adjusted so that the green laser pointer, mounted on the POX table, was co-aligned with the beam between PR2 and PR3).

We were unable (why? I feel like it wasn't so hard last time) to see the POX beam, with a camera pointed at an IR card.  We ended up just waving a lens-free CCD camera around on the POX table where we expected POX to be, found the beam, and decided that if the beam was getting to the table, that was good enough.

We then waved the camera around on the POY table, and found the POY beam on the table.  We also moved ITMY up and down in pitch, and saw that the POY beam was moving, so we were satisfied that we had the correct beam.  We should go back and do this same check with POX, although I'm pretty sure that we already have the correct beam.  But checking is good.

We confirmed that IPPOS was coming out of the chambers.  I didn't end up touching any in-vac mirrors for IPPOS, since they all looked centered, and the beam on the table was already centered on the steering mirror on the out-of-vac table.

We got IPANG out of the chamber to the ETMY table.  IPANG has, after the pickoff window, an adjustable mirror, and then a fixed mirror on the BS table.  The beam was very close to the edge, in yaw, on that fixed mirror.  Jamie unclamped it and moved it so the beam was centered, then twisted it until I got beam back down at the end, centered on the first steering mirror down there.  Then Evan and I got the beam centered on the other steering mirror on the in-vac ETMY table, and got the mirror to ~the center of the first out-of-vac steering mirror.  Then Evan adjusted the other steering optics so the beam was hitting the QPD.

We then got the real REFL beam out of the chambers.  I still don't know what that ghost/fake beam is.  Anyhow, we moved PRM around, and saw that the real REFL beam moves, while the fake one doesn't.  We adjusted the adjustable REFL steering mirror in-vac such that the real REFL beam came out to the table.  Once on the AP table, we moved the PRM around again, just to be doubly/triply sure that we had the correct beam.  We put a beam splitter (found on the SP table) after the lens in the REFL path on the AP table, and put the camera on the reflected side of that BS.  This is because, like the AS port, the beam is too dim at the normal camera spot (which for REFL is the transmission through a Y1 mirror).

Jamie has centered IPPOS and IPPANG QPDs, so we should look at the weekend trend come Monday, to see what things look like, and how they drift, if at all.

 

On Monday, we should:

* Check the alignment, and the centering of beams on all mirrors one last time

* Remove all apertures from suspended optics (I think BS and PRM may be the only two that have them at this time)

* Check oplev paths for all mirrors

* Check all pickoffs / beams that need to come out of the vacuum

* Start putting on doors

  7663   Fri Nov 2 16:09:14 2012 SteveUpdateGeneralOpen-sided mount - why

Quote:

Quote:

Quote:

Suprema- SS clear edge mirror mount 2" diameter is modified for 40m vacuum use. One left and one right handed one. It's adjustment screw housing is bronze! It is not ideal for out gassing.

It will be baked and scanned. If it passes we should use it.

We may need these to bring out some pick-off beams.

 I vote against it. We don't know about the grease inside the screw bushings - scans are not everything if adjusting the screw loosens up the grease. If we need more pick off mirrors lets just make some of the kind that we already use inside for the 2" optics.

 I think Steve had these prepared in response to my question a few days ago of how badly do we need adjustability for the POX/POY mirrors?  We already have cleaned open-sided mounts that have no adjustment screws.  So as long as the beam reflects off the ITMs horizontally (which it should), we can do yaw adjustment by twisting the whole mount. We don't need super fine yaw adjustment, we just need to get the beam out, so this is probably good enough.

We should put the POY mirror on this open-sided mount (the one without screws) some time.  Perhaps even today.

 The hex adjustment screw  is removed from the brass bushing for degreasing, cleaning and baking. Newport says the  bushing is is brass not bronze.

Brass outgassing rate is 10x higher than copper that is truth. The surface area is small so  Krytox grease is ok if it is really needed in the 40m. ( It was ok in the past and it is still there )

  7662   Fri Nov 2 14:37:36 2012 JenneUpdateGeneralOpen-sided mount - why

Quote:

Quote:

Suprema- SS clear edge mirror mount 2" diameter is modified for 40m vacuum use. One left and one right handed one. It's adjustment screw housing is bronze! It is not ideal for out gassing.

It will be baked and scanned. If it passes we should use it.

We may need these to bring out some pick-off beams.

 I vote against it. We don't know about the grease inside the screw bushings - scans are not everything if adjusting the screw loosens up the grease. If we need more pick off mirrors lets just make some of the kind that we already use inside for the 2" optics.

 I think Steve had these prepared in response to my question a few days ago of how badly do we need adjustability for the POX/POY mirrors?  We already have cleaned open-sided mounts that have no adjustment screws.  So as long as the beam reflects off the ITMs horizontally (which it should), we can do yaw adjustment by twisting the whole mount. We don't need super fine yaw adjustment, we just need to get the beam out, so this is probably good enough.

We should put the POY mirror on this open-sided mount (the one without screws) some time.  Perhaps even today.

  7661   Fri Nov 2 13:20:35 2012 MashaUpdateMachineLearningFeedback controller

Quote:

Quote:

I have uploaded to my directory a directory neural_plant. The most important file is reference_plant.c, which compiles with the command

 We would appreciate some plots. Learning curves of recurrent NN working as a plant are interesting. For harmonic oscillator your RNN should not contain any hidden layers - only 1 input and 1 output node and 2 delays at each of them. Activation function should be linear. If your code is correct, this configuration will match oscillator perfectly. The question is how much time does it take to adapt.

Does FANN support regularization? I think this will make your controller more stable. Try to use more advanced algorithms then gradient descent for adaptation. They will increase convergence speed. For example, look at fminunc function at Matlab.

Hi everyone,

I've been on break this week, so in addition to working at my lab here, I've done some NN stuff. In response to Den's response to my last post, I've included learning curve plotting capabilities, 

I've explored all of the currently documented capabilities of FANN (Fast Artificial Neural Network - it's a C library) (most likely, there are additions to the library floating around in open-source communities, but I have yet to look into those). There is extensive FANN documentation on the FANN website (http://leenissen.dk/fann/html/files/fann-h.html), but I'll cut it down to the basics here:

FANN Neural Network Architectures

standard: This creates a fully connected network, useful for small networks, as in the reference plant case 

sparse: This creates a sparsely connected network (not all of the connections between all neurons exist at all times), useful for large networks, but not useful in the reference plant case, since the number of neurons is relatively small

shortcut: This creates some connections in the network which skip over various hidden layers. Not useful in the harmonic oscillator case since there are no hidden layers. Probably won't be useful in a better-modeled referrence plant since this reduces the non-linear capabilities of the model.

FANN Training  

TRAIN_INCREMENTAL: updates the weights after every iteration, rather than after each epoch. This is faster than the other algorithms for the reference plant.

TRAIN_BATCH: updates the weights after training on the whole set. This should not be used on batches of data for the reference plant, seeing as the time history dependence of the plant is smaller than the size of the entire data set. 

TRAIN_RPROP: batch training algorithm which updates the learning parameter.

TRAIN_QUICKPROP: updates the learning parameter, and uses second derivative information, instead of just first derivative, for backpropagation. 

FANN Activation Functions

FANN offers a bunch of activation functions, including a function FANN_ELLIOT, which is essentially the "signmoid like" activation function Den and I used this summer, which runs in the order of multiplication and addition. The function parameters (steepness) can also be set.

FANN Parameters

As usual, the learning parameter can be set. While over the summer we worked with lower learning parameters, in the case of the harmonic oscillator reference plant, since the error is low after the first iteration, higher learning parameters (0.9, for example), work better. However, this is a very isolated case, and, in general, lower parameters, though convergence is slower, produce more optimal results. 

The learning momentum is another parameter that can be set - the momentum factor is a coefficient in the weight adjustment equation which allows for the difference in weights beyond the previous weight to be factored in. In the case of the reference plant, a higher learning momentum (0.9) is optimal, although in most cases, a lower learning momentum is optimal so that the learning curve doesn't oscillate terribly. 

 

FANN does not explicitly include regularization, but this can be implemented by checking the MSE  at each iteration against the MSE at the n previous iterations, where n is the regularization parameter, and stopping training if there is no significant decrease (also determined by a parameter). The error bound I specified during training was 0.0001

The best result for the reference plant was obtained using FANN_TRAIN_INCREMENTAL, a "standard" architecture, a learning rate of 0.9 (as explained above) and a learning momentum of 0.9 (these values should NOT be used for highly non-linear and more complicated systems). 

I have included plots of the learning curves - each title includes the architecture, the learning algorithm, the learning parameter, and the learning momentum if I modified it explicitly.

All of my code (and more plots!) can be found in /users/masha/neural_plant

On the whole, FANN has rather limited capabilities, especially in terms of learning algorithms, where it only has 4 (+ all of the changes one can make to parameters and rates). It is, however, much more intuitive to code with and faster han the Matlab NN library, although the later has more algorithms. I'll browse around for more open-source packages. 

Best,

Masha

standard_BATCH_0p35_ref_plant_lc.pngstandard_QUICKPROP_0p35_ref_plant_lc.pngstandard_RPROP_0p35_ref_plant_lc.pngstandard_INCREMENTAL_0p35_ref_plant_lc.png

standard_INCREMENTAL_0p9_0p9_ref_plant_lc.png

  7660   Fri Nov 2 03:28:54 2012 ranaUpdateGeneralbronze bushing for 40m vac

Quote:

Suprema- SS clear edge mirror mount 2" diameter is modified for 40m vacuum use. One left and one right handed one. It's adjustment screw housing is bronze! It is not ideal for out gassing.

It will be baked and scanned. If it passes we should use it.

We may need these to bring out some pick-off beams.

 I vote against it. We don't know about the grease inside the screw bushings - scans are not everything if adjusting the screw loosens up the grease. If we need more pick off mirrors lets just make some of the kind that we already use inside for the 2" optics.

ELOG V3.1.3-