The AA board shown in attachment 1 will be used in the seismometer hardware setup. A cartoon of this setup is shown in attachment 2.
BNC connectors are required for the seismometer breakout boxes. So the four-pin LEMO connectors present in the AA board were removed and panel mount BNC connectors were soldered to it. Red and blue colored wires were used to connect the BNC connectors to the board. Red wire connects the center of the BNC connector to a point on the board and that connection leads to the third leg (+IN) of the IC U### and the blue wire connects the shield of the BNC connector to the second leg (-IN) of the IC U###.
All the connections (including BNC to the AA board and in the AA board to all the filters) were tested using a multimeter by the beeping method and it was found that channel 10 (marked as C10) had a wrong connection from the point where the red wire (+ve) was connected to the third leg (+IN) of IC U91 and channel 32 (marked as C32) had opposite connections meaning the blue wire is connected to the third leg (+IN) of IC U311 and red wire is connected to the second leg (-IN) of IC U311.
We fixed the anti-aliasing board in its aluminum black box, the box couldn't be covered entirely because of the outgoing wires of the BNC connectors, so we drilled additional holes on the top cover to slide it backwards by 1cm and then screw it.
We had to fix the AA board box in rack 1X7, but there wasn't enough space, so we tried to move the blue chassis (ligo electro-optical fanout chassis 1X7) up with the help of a jack. We removed the blue chassis' screws but we couldn't move it up because of a piece of metal screwed above the blue chassis, then we weren't able to screw the two bottom screws again anymore because it had slided a bit down. Thus, the blue chassis (LIGO ELECTRO-OPTICAL FANOUT CHASSIS 1X7) is still not fixed properly and is sitting on the jack.
To accommodate the AA board (along with the panel-mounted BNC connectors) in rack 1X7 we removed the sliding tray (which was above the CPU) and fixed it there. Now the sliding tray is under the drill press.
Not exactly sure what the problem was, but I updated to the head of the SVN and rebuilt and it seems to be working fine now.
The WWF_M connector is the end of the STS2 seismometer orange cable and the S1 connector is the end of the gray 26-pin-cable
I made a quick sketch of how to include two more RF PDs on the REFL beam, given the space we have on the table. We want to install REFL33 and REFL165, 3f signals for the the two modulation frequencies we are using. The point is to make the distance from first beam splitter the same to all PDs so that we can use only one lens before this BS to make the beam the right size. Currently there are 2 PDs on the refl beam, REFL11 and REFL55, predictably. So the drawing shows 4 PDs. Drawing is to scale but is a bit coarse. Hopefully we'll take pictures once we're done.
Reference from current BS splitting beam to the existing PDs.
I edited the C1SUS_SUMMARY.adl file and set the channels in alarm mode to show the values in green, yellow and red according to the values of the thresholds (LOLO, LOW, HIGH, HIHI)
I wrote a script in python, which call the command ezcawrite and ezcaread, to change the thresholds one by one.
You can call this program with a button named "Change Thresholds one by one" in the menu come down when you click the ! button.
I'm going to write another program to change the thresholds all together.
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.
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.
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.
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.
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.
As promised, I have made a final AP table drawing, including the MC camera relocation changes by Kiwamu. I have posted it in the wiki on the tables list, and on the AP table page I've attached the inkscape .svg I used to make it, if someone needs to do small modifications.
Attached is a pdf version of it.
1) REFL beam has been split into 4, to go in equal powers and equal beam size to the now 4 REFL RFPDs, 11, 33, 55 and 165. A lens had to be added for REFL165 because it's a 1mm PD instead of 2mm like the other 3.
2) MC camera has moved.
3) I've cleaned up most of the random components on the table, put them away, and tidied up the cabling.
Script backup regularly runs on op340m by crontab,
but the true backup were not taken since Oct 16, 2010
as the backup program was looking at the old script directory.
/cvs/cds/script/backupScripts.pl was modified to look at the new script directory.
Script backup regularly runs on op340m by crontab,
but the true backup were not taken
$command = "cd /cvs/cds/caltech; /usr/sbin/tar cfX - $EXCLUDE_LIST scripts | bzip2 > $ARCHIVEDIR/scripts_$curdate.tar.bz2";
$command = "cd /opt/rtcds/caltech/c1; /usr/sbin/tar cfX - $EXCLUDE_LIST scripts | bzip2 > $ARCHIVEDIR/scripts_$curdate.tar.bz2";
I found that the MC WFS had large offset control signals going to the MC SUS. Even though the input switch was off, the integrators were holding the offset.
I have disabled the ASCPIT outputs in the MC SUS. Suresh is going to fix the MC autolocker script to gracefully handle the OFF and ON and then test the script before resuming the WFS testing.
MCL data for OAF may be suspect from this morning.
I have edited (uncommented existing commands) the following scripts to enable WFS locking to come on when the MC is locked.
I have checked that the autolocker script switches off the mcwfs when mc loses lock and then switches it on after re-obtaining lock.
nodus:elog>w; who ; date
9:20pm up 44 day(s), 5:14, 5 users, load average: 0.29, 1.04, 1.35
User tty login@ idle JCPU PCPU what
controls pts/1 9:18pm 5 -tcsh
controls pts/2 2:37pm 6:39 25:02 25:02 /opt/rsync/bin/rsync -avW /cvs/c
controls pts/3 9:14pm w
controls pts/4 4:20pm 1:56 5:02 5:02 ssh -X rosalba
controls pts/8 8:23pm 47 4:03 -tcsh
controls pts/1 Nov 14 21:18 (pianosa.martian)
controls pts/2 Nov 14 14:37 (ldas-cit.ligo.caltech.edu)
controls pts/3 Nov 14 21:14 (rosalba)
controls pts/4 Nov 14 16:20 (192.168.113.128)
controls pts/8 Nov 14 20:23 (gwave-103.ligo.caltech.edu)
Mon Nov 14 21:20:48 PST 2011
I turned it back on, maybe around 11am? Definitely a little while before the 12:30 meeting.
EDIT by KI:
Sorry, it's me. I was checking if AC was doing something bad on the ALS noise.
While discussing the suspension hysteresis measurements, Koji, Kiwamu, and I realized that the suspension wire standoff is aluminum, whereas the standoff for the LIGO LOS are using quartz.
Using a soft aluminum standoff is bad. The movement of the suspension will slowly wear the groove and produce opportunities for mechanical upconversion and hysteresis.
In fact, the wire standoff as well as the clamping block on the top should be made of sapphire or ruby to prevent any such wearing issues. Steve is hot on the case.
I can't create a new page on the 40m wiki. The page that I was trying to create is
I get this message when I try to save the new page:
Page could not get locked. Unexpected error (errno=13).
This address for wiki is obsolete. Recently it was switched to https://wiki-40m.ligo.caltech.edu/
Jamie is working on automatic redirection from the old wiki to the new place.
The new one uses albert.einstein authentication.
Koji asked aloud tonight if we could measure the coating thermal noise of the refcav optics by beating the refcav light with the MC_TRANS light. Then we looked at our calculations for the noises:
Displacement noise of T=200ppm silica/tantala coating on a 1" silica substrate with a 300 micron beam spot = 1e-18 * sqrt(100 Hz / f) m/rHz.
Displacement noise from coating thermal in the MC is roughly smaller by the beam size ratio (1.8 mm / 0.3 mm). Some differences due to 3 mirrors and more layers on MC2 than the others, but those are small factors.
So, the frequency noise from the refcav should by larger than the MC thermal noise by a total factor of (1.8 / 0.3) * (13 m / 8 inches) ~ 400.
Another way to say it is that the effective strain noise in the RC is (1e-18 / 0.200) = 5e-18 /rHz. This translates into (5e-18 * 13) = 6.5e-17 m/rHz in the MC. (in frequency noise its 1.5 mHz/rHz).
I have measured the frequency noise in the LLO MC to be at this level back in 2009, so it seems possible to use our RC + MC to measure coating thermal noise by the length amplification factor and compete with Frank+Tara.
So today we set up the Jenny RC temperature setup to lock the LWE NPRO to the RC and then set up the beat note with the IFO REFL beam on the AS table. By using the 2 laser beat, we are avoiding the VCO phase noise issue which used to limit the PSL frequency noise at ~0.01 Hz/rHz. To do this we have reworked some of the optics on the PSL and AS tables, but I think its been done without disturbing the beams for the regular locking. Beat note has been found, but the NPRO has still not been locked to the RC - next we setup the lockin amp, dither the PZT, and then use the New Focus lock box to lock it to the RC.
You might think that its hard to measure this since the MC has ~1 MHz frequency fluctuations and we want to measure down to 1e-4 Hz. But, in fact, we can just use a 200 m MFD with a LT1128 preamp. Then we use the MFD to stabilize the MC length to the refcav and just use the control + error signal of the MFD setup as the coating thermal noise measurement.
Note: Beat found at ~40deg for the aux laser. The aux laser is on but the shutter is closed.
The AS camera seems to be hosed. Need a bit of alignment. (KA) ==> Fixed. (Jan 15)
I have slightly rotated the lambda/2 plate, which is used for attenuating the REFL beam's power on the AS table
because the plate had been at an unusual angle for investigation of the glitches since last Thursday.
It means the laser power going to the coating thermal noise setup has also changed. Just keep it in mind.
As usual, I noticed several bad things within 30 seconds of sitting in front of the workstation. Today its that there are OFF or missing filters on the MC TRANS.
is this the normal state? Screenshot attached.
Fri Feb 03 19:57:20 2012
Fri Feb 03 20:25:19 2012 : Aligned all SUS to center their OL beams
Fri Feb 03 20:29:21 2012: Aligned all SUS to make OL_PIT = 0.5
ND filter ND3 (which is at the REFL port to the REFL OSA) is removed. Don't forget to put it back when you restore PRM!!!
I don't know what tripped the PRM watchdog, but it was unhappy. I manually moved the sliders on the IFO align screen away from the positions of the save file before turning on the damping, to make sure that I wouldn't be sending oodles of power to the REFL port, since the ND filter is still removed. So PRM is damped now, but misaligned.
Kiwamu and Koji
We found that the intra-cavity mode of the PRC is not round although it was obvious even with the DARK and REFL port images.
We need to review the mode matching situation.
In order to look at the PRC intra-cavity mode, we reconfigured the POP CCD.
If we look at the beam reflected from the Michelson, the beam is round. However, the PRC intra-cavity mode can never be round
in any resonant conditions. (Pict 1, 2, and 3, for the sideband resonant, carrier resonant conditions and another carrier resonant
one, respactively). Particularly the mode of the carrier resonant case is very unstable and always changing.
By misaligning the PRM, we can compare between the spot directly reflected from the Michelson and the one after additional round trip in the PRC (Pic 4).
They looks round, but it was obvious the secondary reflection is dimmer and larger (Pic 5). The intensity difference corresponds to the factor RPRM RMI
(i.e. product of the reflectivities for the PRM and MI). It can be understand if the dimmer spot looks smaller due to the artifact of the CCD. But it is opposite.
This may mean the mode matching is not correct. We are not sure what is not right. This could be just an incorrect incident beam, the curvature error of the PRM,
beam is distortec by the TT mirrors, or some other unknown reasons.
More precise analysis can be done with quantitative analysis of those two spots with Beamscan. This could happen tomorrow.
From the mode measurement I and Suresh have done yesterday, I calculated what beam size we expect at ETM ((1) upper Fig.1) and at ETM after one bounce ((2) lower Fig.1).
In case of (1), we expect approximately w=6300 um (radius), and w=4800 um for one-bounce spot (2) from the measured mode, see Fig.2.
This roughly agree with what we observed on CCD camera. See, pic1 for (1) and pic2 for (2). The spot at the ETMY (1) is larger than the one-bounced spot (2). From the monitor it is difficult to assume the radius ratio. The observed spot of (2) is a bit smaller than the prediction. It could happnen when (A) the ETMY (as a lens) is slightly back of the ideal position (= the distance between the ITM and ETM is longer than 40m) (B) the real waist is farer than ITM position toward MC (I assumed roughly 5 m from Jenne's plot, but could be longer than that).
pic1 (left): beam spot hitting on the suspension frame. pic 2 (right): the one-bounced beam spot hitting on the suspension frame.
I've added the PEM_SLOW.ini file to the fb master file, which should give us the slow seismic RMS channels when the framebuilder is restarted. Example channels:
I also updated the path to the other _SLOW.ini files.
I will do it first thing in the am tomorrow, when Kiwamu is not busy getting real work done.
Here's is a the diff for /opt/rtcds/caltech/c1/target/fb/master:h
controls@pianosa:/opt/rtcds/caltech/c1/target/fb 1$ diff -u master~ master
--- master~ 2011-09-15 17:32:24.000000000 -0700
+++ master 2012-03-29 19:51:52.000000000 -0700
@@ -7,11 +7,12 @@
The framebuilder seems to have been restarted, or restarted on it's own, so these channels are now being acquired.
Below is a minute trend of a smattering of the available RMS channels over the last five days.
Today I tried to make the lms filter to work online. I played around with the signals (GUR1_X and MC_F) to pre-whiten them and in the end the following configuration worked out:
1. mu = 0.03, tau = 1e-5, downsample=8, nCoeff = 4000, delay = 5 (sample-and-hold delay is not included in the new code, it should be added here!)
2. witness pass: AA32 = cheby1("LowPass", 4, 1, 32) AND 0.1:0
3. witness adaptation path: AA32 AND AI32 = cheby1("LowPass", 4, 1, 32) AND 0.1:0
4. error path: AA32 AND 0.1:0 AND anti_1Hz. Before I added anti_1Hz filter oaf did nothing. This filter tries to approximate the actuator transfer function. Note, it is not in the witness adaptation path. This is some sort of whitening.
5. correction path: AI32, gain = -1
Convergence time ~ 5 mins. The performance of the filter is far not perfect compared to the offline implementation. But it deals with a stack though.
Hey, folks. Please remember to commit all changes to the SVN in a timely manor. If you don't, multiple commits will get lumped together and we won't have a good log of the changes we're making. You might also end up just loosing all of your work. SVN COMMIT when you're done! But please don't commit broken or untested code.
pianosa:release 0> svn status | grep -v '^?'
I have setup a shared .bashrc for all the workstations that is symlinked to the normal location on all machines:
controls@rossa:~ 0$ ls -al /home/controls/.bashrc
lrwxrwxrwx 1 controls controls 23 2012-05-25 15:37 /home/controls/.bashrc -> /users/controls/.bashrc
This should help simplify maintenance considerably. Editing that file on one machine will edit it for all. Just edit this one file! Don't try to get fancy and add extra files!
I also added a bunch of aliases that had previously been missing. This should help with some of the problems that people had been having.
You can of course still write scripts in csh/tcsh or use tcsh in a shell if you wish. Just don't change the default shell for the controls user.
A nicer, better maintained version of tconvert is now supplied by the lalapps package. It's called lalapps_tconvert. I installed lalapps on all the workstations and aliased tconvert to point to lalapps_tconvert.
Today I REPOSITIONED THE SEISMOMETERS in order to triangulate noise sources (as Rana suggested).
I re-levelled all of them, locked them, and turned them on. They should be located out of sight, but just in case:
GUR 1 IS DOWN THE X-ARM, behind the interferometer.
GUR 2 IS BETWEEN THE TWO ARMS, BEHIND THE CABLE TRAP THAT RUNS PARALLEL TO THE X-ARM.
STS 1 IS DOWN THE Y-ARM behind the interferometer.
I'll wait a day for them to stabilize (continuing to reset STS-1 every hour or so) and then begin taking data tomorrow morning, depending on the condition of the signal.
Ideally, I'd like a few days' worth of data, so I'll update when I've changed the configuration back to the way it was prior.
Ideally, I'd like a few days' worth of data, so I'll update when I've changed the configuration back to the way it was prior.
Highlighting good, ALL CAPS LESS SO!
I checked the connections specified in the old Gulap Pin Map and found that they do not correspond to the current values. I mapped out the current connections (in this case, the letter refers to the labeled pin on the mil/spec while the number refers to the pin on the 37 pin DSub, labeled consecutively):
A-1, B-2, C-3, D-4, E-5, F-6, G-7, H-Unused, J-8, K-unused, L-9, M-10, N -11, P-12, S-13, T-Unused, U-14, V-15, W-16, X-17, Y-18, Z-Unused, a-Unused, b-19, c-20, UnlabeledPin-Unused.
There are 20 pins in use of 26 total, which is good because that means Jenne and I can use the ~70m long 24 wire cable to make a new Gurlap 1 cable.
Den and I moved the Streckeisen, Guralp 2, and Trillium seismometers to the isolation box in order to measure the noise of the Streckeisen while we have the Trillium.
Jan and I wanted to measure the ringdown at the IMC. Since the QPD at the MC trans is not fast enough for ringdown measurements, we decided to install a pickoff to include a faster PD while not disturbing much of the current MC trans configuration. The initial configuration had very little space to accommodate the pickoff. So the collimating lens along with the QPD were moved 2 inches closer to the incoming beam. A 50-50 BS was put in front of the QPD and the steering mirror was moved behind to reflect MC trans output to the new PD. The current configuration is shown below with the MC autolocker threshold mentioned in Jenne's elog
The hunt for a faster PD wasn't satisfactory and we found a couple of PDs that were good for measurements actually didn't work after installing them. The one currently installed is also not satisfactorily fast enough for ringdown measurements. We'll hunt for faster PDs at Bridge tomorrow and replace PDA400. Also the IMC unlocked from time to time....may be we were noisy and didn't master the 'interferometer walk' very well.
The PDA255 is a good ringdown detector - Steve can find one in the 40m if you ask him nicely.
We found a PDA255 but it doesn't seem to work. I am not sure if that is one you are mentioning...but I'll ask Steve tomorrow!
I double checked the PDA255 found at the 40m and it is broken/bad. Also there was no success hunting PDs at Bridge. So the MC trans is still in the same configuration. Nothing has changed. I'll try doing ringdown measurements with PDA400 today.
Can you explain more what "broken/bad" means? Is there no signal? Is it noisy? Glitch? etc.
The PD saturates the oscilloscope just by switching on the power; with no real signal at all. But Steve helped locating a PD that is not being used at the AP table. So I will check it and replace the current one if it works!
The PD (pda255) at the AP table, close to the MC refl , which Steve mentioned to be not in use, has been removed from the table for testing.
The PD installed at MC trans to make ringdown measurements has been replaced with the above PDA255.
Koji opened up the PD and found that the screw connecting the PD to the pole was doing an additional job as well; connecting the power cable to the PD output in the inside. The PD is now fixed! Yippie...we have two PDA255 s at 40m now!!
I noticed that the IFO restore scripts have some problems. They use burt request files to store and restore the settings. However, the request files contain old channel names.
Especially channels with _TRIG_THRES_ON/OFF are now _TRIG_THRESH_ON/OFF, note the extra "H".
These scripts reside in /opt/rtcds/caltech/c1/burt/c1ifoconfigure/.
I fixed the PRMI_SBres and MI scripts. Someone should fix all other files.
As promised in previous e-log, this log is all about the current online seismic noise classification system.
While we had the BLRMS system already in place (which I helped make), Den realized that we would need better filters for the BLRMS channels, as we wanted a strong cut-off, but we also wanted a short step-response so that we could quickly classify seismic signals. Likewise, having a step response which oscillates is also undesirable as this could lead to false classifications of post-truck signal as trucks as a filter adjusts and then dips back down. Thus, after experimenting with many different filters, Den chose to use a combination of
chebyl("LowPass", 1, 1, 0.03)*chebyl("LowPass", 1, 1, 0.03)
as our low-pass filter. The step response and bode plot are below.
The next step was to write C code that would implement the feedforward neural network with my newly generated weights.
Next, I had to implement the code in the c1pem model, and normalize the inputs. Below is an overview of the model, and a close up of the C block section.
The above close-up includes the process of normalization (dividing by the square of the incoming signal), feeding through the neural network, and classifying.
Each seismometer channel set (GUR1X, GUR1Y, GUR1Z, GUR2X, GUR2Y, GUR2Z, STS1X, STS1Y, STS1Z) now has channels (and corresponding DQ channels) of the following form:
SEIS_CLASS : The class of seismic noise 1.0 means Earthquake, 0.5 means Quiet, and 0.0 means Truck. (There are only these 3 digital values).
SEIS_CLASS_EQ, SEIS_CLASS_TRUCK, SEIS_CLASS_QUIET: These channels represent the confidence of the neural network's classification. The class of the current signal will have an output of 1, where the other two channels will have an output between 0 and 1 representing the ratio of the neural network's output in that class neuron to the output in the classification vector neuron. To simply - suppose the neural network classified an earthquake. Ideally, the neural network output neurons would have the value [1, 0, 0], and SEIS_CLASS would equal 1.0 for earthquake. However, the output neurons probably read something along the lines of [0.9, 0.3, 0.5] - SEIS_CLASS is still 1.0, but SEIS_CLASS_EQ would be 1.0, and SEIS_CLASS_TRUCK would be 0.5 / 0.9 and SEIS_CLASS_QUIET would be 0.3 / 0.9. The lower the other two signals are, the better - this means that we are more confident in our classification.
The MEDM screen for this system (in the RMS system) has the following form for all seismometer channels (this one is GUR1X):
These are the screens I edited earlier in the summer, with modifications. The bottom filter banks represent the norm of the seismometer signal, which we use to normalize the inputs to the neural network.
Here a close-up of the most important part:
The orange meter on the right points to the current signal type. Here it reads truck - this is ok because it's the middle of the day, and there are a lot of trucks around. The left side represents our confidence in the signal - the signal is classified as a truck, so the "Truck" bar is saturated. The quiet signal bar is very low, which is good since it means that the neural network thinks that it's definitely not quiet. The earthquake bar has some magnitude, since earthquake signals and trucks have some degree of linear non-separability.
How has this been performing? Firstly, all of the seismometer channels have the same classification readout, which is good. Last night, all of the classes were "quiet", with an "earthquake" which occurred when Den jumped around GUR1 to simulate an EQ. This morning it was on "truck" as expected. The filters are still not fine enough to detect individual trucks, but I will continue to monitor the performance over the coming days.
If anyone has ideas on how better to represent this information, please let me know. This was the first thing that came into my head that would work with my MEDM monitor options, and I'm open to suggestions!