[Jamie, Jenne, Mirko]

New c1oaf model installed

We have installed the new c1oaf (online adaptive feed-forward) model.  This model is now running on c1lsc.  It's not really doing anything at the moment, but we wanted to get the model running, with all of it's interconnections to the other models.

c1oaf has interconnections to both c1lsc and c1pem via the following routes:

c1lsc ->SHMEM-> c1oaf
c1oaf ->SHMEM-> c1lsc
c1pem ->SHMEM-> c1rfm ->PCIE-> c1oaf

Therefore c1lsc, c1pem, and c1rfm also had to be modified to receive/send the relevant signals.

As always, when adding PCIx senders and receivers, we had to compile all the models multiple times in succession so that the /opt/rtcds/caltech/c1/chans/ipc/C1.ipc would be properly populated with the channel IPC info.

Issues:

There were a couple of issues that came up when we installed and re/started the models:

c1oaf not being registered by frame builder

When the c1oaf model was started, it had no C1:DAQ-FB0_C1OAF_STATUS channel, as it's supposed to.  In the daqd log (/opt/rtcds/caltech/c1/target/fb/logs/daqd.log.19901) I found the following:

Unable to find GDS node 22 system c1oaf in INI files

It turns out this channel is actually created by the frame builder, and it could not find the channel definition file for the new model, so it was failing to create the channels for it.  The frame builder "master" file (/opt/rtcds/caltech/c1/target/fb/master) needs to list the c1oaf daq ini files:

/opt/rtcds/caltech/c1/chans/daq/C1OAF.ini
/opt/rtcds/caltech/c1/target/gds/param/tpchn_c1oaf.par

These were added, and the framebuilder was restarted.  After which the C1:DAQ-FB0_C1OAF_STATUS appeared correctly.

SHMEM errors on c1lsc and c1oaf

This turned out to be because of an oversight in how we wired up the skeleton c1oaf model.  For the moment the c1oaf model has only the PCIx sends and receives.  I had therefore grounded the inputs to the SHMEM parts that were meant to send signals to C1LSC.  However, this made the RCG think that these SHMEM parts were actually receivers, since it's the grounding of the inputs to these parts that actually tells the RCG that the part is a receiver.  I fixed this by adding a filter module to the input of all the senders.

Once this was all fixed, the models were recompiled, installed, and restarted, and everything came up fine.

All model changes were of course committed to the cds_user_apps svn as well.

 The title of this post should of course have been " ... - comparison with ETMY" not " ... - comparison with ITMY"

Just to find out where we are currently, I plotted the ITMY and SRM oplev spectra along with the ETMY oplev spectra. ETMY seems to be very good, so comparing with this seemed useful, so we know how much we have to improve by. The SRM power spectrum appears to be around 2 orders of magnitude higher than ETMY over pretty much the whole measurement band. The ITMY power spectrum is not so bad as the SRM above about 60Hz. Next thing to do is to check the dark noise level for the ITMY and SRM QPDs.

[Mirko, Kiwamu]

Put up a temp. setup on the laser table to measure the RF modulation depths using the optical spectrum analyzer. First with a pickoff beam with about 2mW => SNR of 8 of 1 peak per FSR.

Then with a beam with about 100mW. Much better SNR on the single peak but still no sidebands visible. Modematching not too good in either case. Shouldn't matter.

 We have made a new plan for the ITMY and SRM oplev optical path which uses as few optics as possible. This should help to reduce coupling from vibrations of optics in the oplev path back into the GW channel. To get enough room for the turning mirror into the SRM it might be necessary to move the POY optics a bit nearer to the tank. 

Today I worked on getting the ITMY and SRM oplevs back in working order. I aligned the SRM path back onto the QPD. I put excitations on the ITMY and SRM in pitch and yaw and observed the beam at the QPDs to check for clipping. They looked clean from clipping.

The f2a filters were newly designed and installed on BS and PRM.

So the lock of PRMI will be more stable .

Once the SRM oplev project settles down, I will adjust the f2a filters on SRM too.

I changed a filter bank name (C1IOO-WFS1_PIT) in c1ioo model reverting it to its earlier name.  Had to restart c1ioo model and the fb

 The PRM sus damping restored.

Kiwamu, Keiko, Anamaria

We were able to lock PRC using REFL11I after improving the MICH dark fringe a bit (moving BS) and rotating AS55 and REFL11 such that the signal was maximized in the phases we were using. The dark port is not so dark... but the lock is stable.

I had finished the whole REFL path alignment, but I didn't have a good input beam reference at the time, which is why we had to realign the PDs and the camera. We only had strength to realign 11 and 55. Otherwise, we just need to tweak and center beam on 33 and 165, figure out what's up with 55 and be done with the AP table mods. I hope.

 Anamaria, Keiko

- We aligned MICH and were successfully locked MICH using AS55Q. The other mirrors were misaligned so that the other degrees of freedom didn't exist. AS55 was fed back to BS. The f2a filters on BS suspension were required to lock, because the pos feedback was unbalanced to angle degrees of freedom.

- We tried to lock PRCL next, however, because we aligned the MICH and the REFL beam paths were changed, REFL PDs didn't have the light anymore. The REFL paths were modified now, so we will try the PRCL locking next.

- We couldn't confirm REFL55 signals although we alined the REFL paths to REFL55 PD.

In addition to REFL 33 ans 165, I checked the orthogonality for the other existing three channels.

AS11

ABS = 1.025035  PHASE = -93.124929 [deg]

REFL11]

ABS = 0.920984  PHASE = -88.824691 [deg]

REFL55

ABS = 1.029985 , PHASE = -90.901123 [deg]

- - - - - - - - - - - - - - - - - -

The demodulated signal was set as 50 Hz (for example LO 11MHz and RF 11MHz+50Hz from function generators.) These AS11, REFL11, REL55, REFL33m REFL165 are the current available channels in terms of the connection to the data system from the demodulation board. I am going to estimate the error next.

I went to see what was wrong with the ITMs and found that people have been working on them and have left them in a broken state with no elog entry.

This is sad and unacceptable.

Whoever is working on these should post into the elog what the Oplev layout plan is, have someone check it, and ONLY THEN get to work on it.

The layout as it looks tonight is too complicated. With too many optics we will not have a low noise optical lever setup. The new layout should use a bare minimum number of optics and only use very stable mounts.

Although we did some of the Input Matrix diagonalization, we have not yet actually used this knowledge. As a result all of the optics are shaking all over the place.

Sunshine Task: Set the input matrices to their calculated values and then adjust the OSEM damping gains for all optics so as to get a Q ~ 5.

I've found that a few of the screens still have Whited-Out fields due to naming changes (OL SUM and ALS-> TM OFFSET). I attach a screen shot of it.

The OL screens have the wrong SUM names and the IFO ALIGN screen is pointing to the wrong SUS screens.

I have updated the wiper.pl script (/opt/rtcds/caltech/c1/target/fb/wiper.pl) that runs on the framebuilder (in crontab) to delete old frames in case of file system overloading.  The point of this script is to keep the file system from overloading by deleting the oldest frames.  As it was, it was not properly sorting numbers which would have caused it to delete post-GPS 1000000000 frames first.  This issue was identified at LHO, and below is the patch that I applied to the script.

--- wiper.pl.orig  2011-04-11 13:54:40.000000000 -0700
+++ wiper.pl       2011-09-14 19:48:36.000000000 -0700
@@ -1,5 +1,7 @@
 #!/usr/bin/perl
 
+use File::Basename;
+
 print "\n" .  `date` . "\n";
 # Dry run, do not delete anything
 $dry_run = 1;
@@ -126,14 +128,23 @@
 
 if ($du{$minute_trend_frames_dir} > $minute_frames_keep) { $do_min = 1; };
 
+# sort files by GPS time split into prefixL-T-GPS-sec.gwf
+# numerically sort on 3rd field
+sub byGPSTime {
+    my $c = basename $a;
+    $c =~ s/\D+(\d+)\D+(\d+)\D+/$1/g;
+    my $d = basename $b;
+    $d =~ s/\D+(\d+)\D+(\d+)\D+/$1/g;
+    $c <=> $d;
+}
+
 # Delete frame files in $dir to free $ktofree Kbytes of space
 # This one reads file names in $dir/*/*.gwf sorts them by file names
 # and progressively deletes them up to $ktofree limit
 sub delete_frames {
        ($dir, $ktofree) = @_;
        # Read file names; Could this be inefficient?
-       @a= <$dir/*/*.gwf>;
-       sort @a;
+       @a = sort byGPSTime <$dir/*/*.gwf>;
        $dacc = 0; # How many kilobytes we deleted
        $fnum = @a;
        $dnum = 0;
@@ -145,6 +156,7 @@
          if ($dacc >= $ktofree)  { last; }
          $dnum ++;
          # Delete $file here
+         print "- " . $file . "\n";
          if (!$dry_run) {     
            unlink($file);
          }

Jeff has changed our AC filters inside the lab this morning. Now he is checking on the main filters at CES.  He will finish the roof units tomorrow.

Met One #1 counter is on the top of IOO chamber.  It is measuring 1 and 0.5 micron size particles.  One year of lab condition plot below.

Forgot to attach a picture of the ITMY's face camera when it was locked.

The horizontal position of the spot looks good, but the vertical position is apparently too low, which agrees with the A2L result.

For the acquisition of the MC_F channel, I suggest taking the FAST_MON BNC output from the blue FSS interface card in the Eurocard crate in the PSL rack. This can then be piped into the 2-pin LEMO plug (Ch. 1) of the Generic Pentek DAQ card which used to acquire the MC_L signal from the MC Servo Board.

The PRM damping was restored at side sensor var 1050

[Jenne, Mirko, with supervision from Jamie]

We are starting to create the new OAF model, so that it works with the new CDS system. 

I created (and did an "svn add" for) a new c1oaf.mdl, in the same place as the current c1lsc.mdl . Since the oaf will kind of be working with ISC things, I decided to put it in that folder.  So far this new OAF model just has SHMEM/PCIE memory sharing things to get info from the LSC and PEM models.  The OAF model has dcuid=22 (the same as in the old system), and lives on the LSC machine with specific_cpu=4.  This is the CPU number that Yoichi was going to use, but he ended up putting his FF stuff directly into the LSC model for delay reasons.

I modified the c1rfm.mdl to take seismometer and accelerometer info from the PEM model, and give it to the "rfm" via shmem, and then using PCIE (dolphin) to get the channel to the OAF model.

I modified the c1lsc model to have shmem outputs that go from the degrees of freedom to the OAF, and shmem inputs from the OAF's output to sum into the DoFs, just like Yoichi's FF stuff.  I also removed the old OAF_OUT, because it would only allow me to select one DoF at a time, and I will eventually want the ability to do multiple amounts of OAFing at the same time.  Hopefully.

All of the above changes have been svn'ed with nice log messages into the cds svn.

I have not yet modified the PEM model to give the seis/acc information to the RFM model.

I will need to acquire the PSL's PZT input as a representation of the mode cleaner's length if I want to apply the OAF to the MC to recreate past work.

I have reconfigured the refl beam path on the AP table to include REFL33 and REFL165. Would be done if we hadn't prepared P BSs instead of S, which required some serious digging to find two others. And if someone hadn't stolen our two 3m SMA cables that Keiko and I made on our previous visit and I had left with the 3f PDs. I don't expect them to reappear but if they do, it would be grand.

Note: Refl beam from ifo looks a bit high, ~1cm on the lens 20'' from output port. Not sure what that means about ifo alignment change, I've left it as is. When we know we have a good alignment, we should be able to easily realign the beam path if necessary. If it remains the same, we might want to change the lens height.

Done:

1) REFL11 and REFL55 are now hooked up and aligned in a low power beam. (I set the power as low as I could by eye to not risk burning the PDs during alignment)

2) The required BSs and REFL33 and REFL165 are in place, powered.

3) I have set them in a configuration such that the beam is the same distance from the main beam, to adjust beam size easily for all 4.

4) Camera has been moved from main beam to behind a steering mirror, ND filters removed, centered on camera.

To Do:

1) Find one more longish SMA cable.

2) Align beam on REFL33 and REFL165.

3) Check beam size carefully. (I get a plateau on the scope, and I can "hide" the beam on the PD, but it could be better. The path has become longer by ~5-8inches.)

4) Adjust power.

5) Redo layout diagram, post in wiki.

The spot positions on ITMY and ETMY were measured using the LOCKIN modules in C1ASS when the Y arm stayed locked.

The beam was successfully aligned such that it hits the center of the ETMY mirror.

However on the other hand the angle of the beam is pitching and it's going upward as the beam propagates to ETMY.

/***** RESULTS ******/

Here is a summary of the measurement :

Also a cartoon is shown below.

The scale is not quite true, but at least it gives you a 3D information of how the beam is pointing down to the Y arm.

  /***** MEASUREMENT *****/

 In order to measure the spot positions the standard technique, namely A2L, was used.

Since the C1ASS model was made for doing the A2L measurements on each arm cavity, the LOCKIN modules in C1ASS were used.

First the Y arm was locked with AS55 (#5398), and then the C1ASS was activated by calling some scripts from C1ASS_QPDs.adl.

In order to calibrate the signals from LOCKINs, an intentional coil imbalance was introduce.

This is the same calibration technique as Valera explained before (#4355) for measurement of the MC spot positions.

Keiko, Jamie , Kiwamu

The I and Q orthogonalities of REFL33 and 165 demodulation board were measured by "orthogonality.py"  Python package scipy were addied on Pianosa to run this code. Please note that "orthogonality.py" can be run only on Pianosa.  

The results were:

REFL165

ABS = 1.070274, PHASE = -81.802479 [deg]

if you wanna change epics values according to this result, just copy and execute the following commands

ezcawrite C1:LSC-REFL165_Q_GAIN 0.934340 && ezcawrite C1:LSC-REFL165_PHASE_D -81.802479

- - - - - - - - - - - - - - - - - -

REFL33

ABS = 1.016008 , PHASE = -89.618724 [deg]

if you wanna change epics values according to this result, just copy and execute the following commands

ezcawrite C1:LSC-REFL33_Q_GAIN 0.984244 && ezcawrite C1:LSC-REFL33_PHASE_D -89.618724

Fig.1 and 2  are the resulting plots for 33 and 165 MHz demod baoards, respectively.You should look at the 3Hz in x axis, as the demodulated signal frequency was set as 3 Hz.

Fig. 1 REFL33 I and Q orthogonality at 3 Hz.

Fig. 2 REFL165 I and Q orthogonality at 3 Hz.

The Y arm has been locked with AS55.

A next thing is to check the spot positions on the ETMY and ITMY mirrors so that we can evaluate the recent beam pointing.

- - - parameter settings - - -

C1:LSC-YARM_GAIN = -0.03

AS55 demod phase = 0.2

WF gains = 21 dB

C1:LSC-TRY_OUT = 0.57 (maximized by steering PZT2)

[Suresh, Jamie]

The new versions of these three models have been committed to the svn.  I copy below the svn log I wrote over there.

I edited three models,namely c1ioo, c1mcs and c1rfm to bring four channels from C1MCS model into the
 C1IOO model through the RFM.

We plan to use the MC2-Trans-QPD as a third sensor in addition to the WFS1 and WFS2 in the ASC system so that we can sense all the six degrees of freedom of the MC.  However the MC2-Trans_QPD is a new incarnation of the old MC2 Oplev QPD and its four segments are are picked up in the c1mcs model running on the C1SUS machine.  Since we need these    in the C1IOO machine we    had to port these signals via the RFM.  The    changes    I made to these    models are described below:
1) c1rfm.mdl -->  I added several cds parts to carry these four    channels from c1mcs to c1ioo via c1rfm

2) c1mcs.adl -->  In the MC2 block of c1mcs I cut out the structure which converts the four segments of the  QPD in the pitch, yaw and sum signals and copied it to the c1ioo model.     In the    c1mcs I  grounded the input of the oplev signals just as they are in    the MC1    and MC2    suspension blocks. I then routed the ADC channels associated with the oplev QPD signals to the rfm model.

3) c1ioo.mdl --> I have made extensive changes to the c1ioo model to include the third sensor in the WFS feedback loop.  The six lockins in the model are used to excite the MC mirrors and the signals from WFS and QPD can be demodulated to determine the input or output matrices in the WFS feedback.  A sensor_demod_array permits us to select the signals we wish to demodulate. The MCL signal, used in the MC_ASS and A2L measurements, is retained as one of the signals.  The MC_TRANS_SUM was used several in  scripts and to preserve the continuity with this channel name we inserted three EPICS parts to generate a slow channel by the same name from this model.

Since the MC_TRANS_SUM is used in the MC-autolocker script we checked to make sure that the script is running after we recompiled all the three models and restarted them.  It is all fine.

I restarted the frame builder at the following times

Tue Sep 13 14:53:49 PDT 2011

Tue Sep 13 16:46:32 PDT 2011

Tue Sep 13 17:24:16 PDT 2011

 Old Coherent diode laser was replaced by Uniphase HE/NE 1125P at ITMX.

Newport 10B20NC.1 broadband beam sampler set up as beamspitter to dump 90% of the light into beam trap. Beam in to test mass  0.7 mW ,  returning to qpd 0.07 mW  = ~ 3,700 counts

f 1.5 m lens placed into ingoing beam path to reduce the spot size on the qpd.         ITMY oplev will be done in this manner tomorrow.

* IP-ANG is lost with the Piezo Jena PS

* IP-POS cable have to be found at the LSC rack

Whitening filters for the REFL33 & 165 demodulated channels were measured and confirmed that they are working. They can be turned on and off by un-white filter switches on the MEDM screen because they are properly linked. The measured filter responses are showen below. (Sorry, apparentyl the thumbnails are not shown here. Please click the attachments.) 

Attachments: (top) Whitening filter for REFL33 demodulation board. (bottom) Whitening filter response for REFL 165 demodulation board.

The shift of MC2 which Rana noted caused the beam spots on the MC mirrors to decenter. I used the mcassUp and mcassOn scripts and checked the output of the C1IOO lockins to get the spot positions.  I first tried to realign just the MC2 to recenter the spots.  But this was not sufficient.  I then worked on the pitch of all three optics since it is easier to align.   By the time this was done the offset in yaw also reduced, probably due to cross coupling between pitch and yaw in the coils.  At the end of the process I obtained all decentering around 1.5mm or less, then I went over to adjust the MC2TransQPD beam path so that we center the spot on the QPD.  This action shifted the stack,  I had to iterate this two more times before the successive corrections grew sufficiently small.  I think it may shift again if we touch the chamber (the image of MC2Face is still inverted).

The new spot positions in mm (MC1,2,3 pit MC1,2,3 yaw):
    1.3212   -0.8415    0.6795   -1.4292   -0.3574   -1.5208

- Further improvement of beam centering can be done but first I would like to be sure that the MC is stable.  The MC2Trans light is centered on the QPD as a reference.

(Preparation of Y arm locking)

(A) The f2a filters were newly designed and applied to ETMY (see the attachment)

(B) Once the Y arm is aligned such that the TEM00 mode flashes, the transmitted light is visible on the ETMYT CCD camera.

(C) With the newly installed resonant EOM circuit the PDH signal from AS55 looks healthy.

(some notes)

(A) To design the f2a filters there is a handy python script called "F2A_LOCKIN.py" in /scripts/SUS.

The script measures the coil imbalance at high frequency and low frequency using a LOCKIN module and then gives us the information about the imbalance.

The script hasn't yet been completed, so it doesn't return the intuitive answers but returns something non-intuitive. I will modify it.

(B)  To see the transmitted light from the Y arm I was going to align the CCD camera on the Y end table.

However I found that once the green light is blocked, the transmitted light can be visible on the camera without any re-alignment.

Therefore I haven't rearranged anything on the Y end table, but I just blocked the green light.

Perhaps we still need to align the photo diodes for the transmitted light.

(C) While Suresh was working on MC, I looked at the signal from AS55 with all the optics misaligned except for ITMY, ETMY and BS.

The signal from the Y arm looked very PDH signal, and the demodulation phase seemed to be about 45 deg to maximize the I signal.

I tried locking it by feeding the signal back to ETMY but failed due to a too much POS to angle coupling in the ETMY actuators.

I was momentarily able to capture a higher order mode with a negative gain in LSC-YARM_GAIN, but it was quite difficult to keep it locked.

This was because once I increased the gain to make it stable, the angle instability became more significant and lost the lock immediately.

This was the reason why I had to do the f2a filter redesign. Tomorrow we can try locking the Y arm.

[Mirko / Kiwamu]

 The resonant box has been installed together with a 3 dB attenuator.

The demodulation phase of the MC lock was readjusted and the MC is now happily locked.

(Background)

We needed more modulation depth on each modulation frequency and so for the reason we installed the resonant box to amplify the signal levels.

Since the resonant box isn't impedance matched well, the box creates some amount of the RF reflections (#5339).

In order to reduce somewhat of the RF reflection we decided to put a 3 dB attenuator in between the generation box and the resonant box.

(what we did)

 + attached the resonant box directly to the EOM input with a short SMA connector.

 + put stacked black plates underneath the resonant box to support the wight of the box and to relief the strain on the cable between the EOM and the box.

 + put a 3 dB attenuator just after the RF power combiner to reduce RF reflections.

 + readjusted the demodulation phase of the MC lock.

(Adjustment of MC demodulation phase)

 The demodulation phase was readjusted by adding more cable length in the local oscillator line.

After some iterations an additional cable length of about 30 cm was inserted to maximize the Q-phase signal.

So for the MC lock we are using the Q signal, which is the same as it had been before.

 Before the installation of the resonant box, the amplitude of the MC PDH signal was measured in the demodulation board's monitor pins.

The amplitude was about 500 mV in peak-peak (see the attached pictures of the I-Q projection in an oscilloscope). Then after the installation the amplitude decreased to 400 mV in peak-peak.

Therefore the amplitude of the PDH signal decreased by 20 %, which is not as bad as I expected since the previous measurement indicated 40 % reduction (#2586).

When I run StripTool on Pianosa, I get the following message

==== StripTool Xt Warning Handler ====
warning:         Axis: minVal is greater than or equal to maxVal

And the y-axis scale reverts to 0 -100 regardless of what ever I set in the controls panel

I ssh'ed into rosalba and ran striptool from there and did not face this problem.  So I think pianosa has a problem with Striptool.

Before we install the REFL 3f PDs I made a drawing of the current table layout, since there has been no update lately. Once I've incorporated the two extra PDs (now seen sitting bottom left), I will update the drawing and post in the wiki as well.

LSC-REFL33-I-IN1(IN2, OUT) and LSC-REFL165-Q-IN1(IN2,OUT) channels are back!

We disconnected and connected again the AA filters then the channels are fixed. Apparently the AA filters just before the digital world were somhow charged and not working... Thank you Kiwamu!

Wait. I am checking the whitening filters of the 33 and 165 demodulation boards.

Also, LSC-REFL33-I-IN1(IN2, OUT) and LSC-REFL165-Q-IN1(IN2,OUT) channels may not be working??

I also quickly checked the orthogonality of the demodulation board for REFL33 and REFL165 using function generators and oscilloscope. I checked the frequencies at 1,10,100,1K,10KHz of the demodulated signals. They are fine and ready for 3f signal extraction.

Filters at the RF inputs of REFL33 and REFL165 demodulation boards were measured again. The filters will be totally fine for 33MHz and 165MHz.

Last time I forgot to calibrate the cable lengths, therefore the phase delay of the measurement included the cable lengths. This time the measurements were done for REFL33 and REFL165 demod board with calibration. As the cable lengths were calibrated, the shown plots (Fig.1 and Fig.2) do not include the phase delay dues to measurement cables. Please note that the x-axis is in linear. The phase delays of both boards seems to be not too steep (it will not affect anyway, as Kiwamu pointed out in his comment on the previous post). You can see that the two filters do not filter 33MHz and 165MHz component out.

Fig.1 A response of a filter which is placed just after the RF input of the demodulation board for REFL33. X-axis is shown in linear (~50MHz).

Fig.2 A response of a filter which is placed just after the RF input of the demodulation board for REFL165.

A comment :

Since the LSC RFPD have a long cable of more than 6 m, which rotates a 33 MHz signal by more than 360 deg, so the delay has always existed in everywhere.

The circuit you measured is a part of the delay existing in the LSC system, but of course it's not a problem as you said.

In principle a delay changes only the demodulation phase. That's how we treat them.

RA: Actually, the issue is not the delay, but instead the dispersion. Is there a problem if we have too much dispersion from the RF filter?

I re-aligned the beam onto the MC TRANS QPD since Kiwamu had centered the spots on the mirrors. However, I then inspected the MC2F camera. After coming back into the control room I noticed that the MC transmission had gone down by 50% and that the MC2 OSEMs showed a large step. My guess is that somehow the opening and closing of the can shifted the suspension. So I adjusted the MC2 alignment biases to recover the transmitted power (its now ~50000 instead of the ~33000 from Friday).

 The 2nd beam from this laser is for the SRM's OpLev, so that shouldn't be changed.

For better or worse, we didn't do anything to the ETM OpLevs, because they don't have any in-vac steering optics.  We did however go through and check on all the corner OpLevs.

I lined up the Y Arm for locking and then centered the oplevs for ETMY and ITMY.

* The ITMY OL has still got the old style laser. Steve, pleaes swap this one for a HeNe. Also the optical layout seems strange: there are two copies of the laser beam going into the chamber (??). Also, the QPD transimpedance needs to be increase by a factor of ~10. We're only getting ~500 counts per quadrant. Its worth it for someone to re-examine the whole ITMY OL beam layout.

* The ETMY OL beam was coming out but clipping on the mount for the ETMY OL HeNe. This indicates a failure on our part to do the ETMY closeout alignment properly. In fact, I get the feeling from looking around that we overlooked aligning the OL and IPPOS/ANG beams this time. If we're unlucky this could cause us to vent again. I undid part of the laser mount and changed the height on the receiving mirror to get the beam back onto the QPD.

I noticed that there is significant green light now getting into some of the IR PDs; beacuse of this there are weird offsets in the TRY QPD and perhaps elsewhere. We had better purchase some filters to tape over the front of the sensitive IR sensors to prevent the couplling from the green laser.

* There is a beam on IPPOS, but its too big for the detector (this has always been the case). We need to put a 2" lens with a weak focusing power on this path so as to halve the beam size on the detector. Right now its clipping and misleading. There is also a 0.9V offset on the SUM signal. I'm not sure if this readout is working at all.

* I couldn't find any beam on IPANG at all. Not sure what's changed since Kiwamu saw it.

The phase delay due to the RF input filter on the demodulation board will not bother the resulting PDH signals.

I quickly calculated the below question (see the blue sentence in the quote below). I applied an arbitrary phase delay (theta) due to the filter I measured, on the detected RF signal by the photo detector. Then the filtered RF signal is multiplied by cos(omega_m) then filter the higher (2 omega_m) freqency as the usual mixing operation for the PDH signal. As a result, the I signal is delayed by cos(theta) and the Q signal is delayed by sin(theta). Therefore the resulting signals and its orthogonalitity is kept ok. From the sideband point of view, theta is applied on both upper and lower and seems to make the unbalance, however, as it is like a fixed phase offset on both SBs at the modulation frequency, the resulting signals is just multiplied by cos or sin theta for I and Q, respectively. It won't make any strange effect (it is difficult to explain by sentence not using equations!).

One of the reasons that conlog seems so slow lately is that its been writing 100's of MB of .log files every day since early summer. It looks like the people who have been working on the mdl builds have not been properly adjusting the conlog channel lists. When this is not done conlog just gets filled up with non-control channels like OUT, OUTPUT, OUTMON, etc.

Peter Shawhan has supplied us with many scripts in the conlog directory to clean up these bloated files and fix the channel list.

There is a LP filter just after the RF input of an demodulation board (its schematic can be found as D990511-00-C on DCC). I have checked if the 3f freq, 33MHz, can pass  this filter. The filter TF from the RF input to RF monitor (the filter is between the input and monitor) on REFL33 demo-board was measured as shown in Fig. 1. At 33MHz, the magnitude is still flat and OK, but the phase is quite steep. I am going to consider if it is ok for the PDH method or not.

 Fig. 1 Transfer function from the RF input to RF monitor on the REFL33 demodulation board. At 33MHz, a very steep phase is applied on the input signal.

 To check the demodulation boards for REFL33 and REFL165, a long cable from ETMY (SUS-ETMY-SDCOIL-EXT monitor) is pulled to the rack on Y side.

(1) A filter just after the RF input and (2) transfer function from the RF input to the demodulated signal will be checked for the two 3f demod boards to confirm that they are appropriate for 33 and 165 MHz.

Now that we are in a moderately stable condition, its time to design the optical lever feedback transfer functions. We should think carefully about how to do this optimally.

In the past, the feedback shape was velocity damping from 0-10 Hz, with some additional resonant gain around the pendulum and stack modes. There were some low pass filters above ~30 Hz. These were all hand tuned.

I propose that we should look into designing optimal feedback loops for the oplevs. In principle, we can do this by defining some optimal feedback cost function and then calculate the poles/zeros in matlab.

How to define the cost function (? please add more notes to this entry):

1) The ERROR signal should be reduced. We need to define a weight function for the ERROR signal: C_1(f) = W_1(f) * (ERR(f)^2)

2) The OL QPDs have a finite sensing noise, so there is no sense in suppressing the signal below this level. Need to determine what the sensing noise is.

3) The feedback signal at high frequencies (30 Hz < f < 300 Hz) should be low passed to prevent adding noise to the interferometer via the A2L coupling. It also doesn't help to reduce this below the level of the seismic noise. The cost function on the feedback should be weighted apprpriately given knowledge about the sensing noise of the OL, the seismic noise (including stack), and the interferometer noise (PRC, SRC, MICH, DARM).

4) The servo should be stable: even if there is a negligible effect on the ERROR signal, we would not want to have more than 10 dB of gain peaking around the UGFs.

5) The OL QPDs are dominated by drift of the stack, laser, etc. at some low frequencies. We should make sure the low frequency feedback is high passed appropriately.

6) Minimize transmitted power rms in single arm lock etc.

All the optcs were excited

Sat Sep 10 02:14:11 PDT 2011
999681266