After the meeting, I aligned the IFO to the IR, and then I aligned the Ygreen to the Yarm. I then found the beatnotes and used ALS to hold the arms with CARM/DARM, locked the PRMI, and reduced the CARM offset until I had arm powers of about 3. Given that this was at 3pm, and people were tromping all over inside the IFO room, I feel positive about tonight.
So, IFO seems ready, carm_cm_up script was successful, and got me to arm powers of 1, and then I further reduced the offset by a bit to go a little higher.
I need to figure this out before it's worth trying any acrobatic AO path turn-on scenarios.
Also this evening, I went to the Yend and did another tweak-up of the green beam alignment.
Never mind. This is just the low pass filter that Den put in to try to deal with the moving cavity pole.
Before I realized this, just in case it was a computer quirk, Koji and I rebooted the end station front end machines. This had no effect other than to keep me searching and measuring until I figured it out. If I turn off the low pass, the phase pops back up to very close to the reference. The low pass currently comes on automatically as part of the carm_cm_up script.
REFL33, AS55, REFL55,REFL165,REFL11,POX11,POP22
There were quite a few more demodulator units labelled with PD names. Do any of them need to be included in the automated frequency response measurement system? Please let me know so that I can include them to the RF switch and check them for proper illumination, which i will do for all the above PDs next week.
In the order that makes more sense to me, it looks like you have:
REFL11, REFL33, REFL55, REFL165,
We don't really need POP22 right now, although we do want the facility to do both POP22 and POP110 for when we (eventually) put in a better PD there. Also, we want cabling for POP55, so that we can illuminate it after we re-install it. If we're working on 2f PDs, we might as well consider AS110 also, although I don't know that there was a fiber layed for it. The big one that you're missing is POY11.
I managed to recover c1sus. It required stopping all the models, and the restarting them one-by-one:
$ rtcds stop all # <-- this does the right to stop all the models with the IOP stopped last, so they will all unload properly.
$ rtcds start iop
$ rtcds start c1sus c1mcs c1rfm
I have no idea why the c1sus models got wedged, or why restarting them in this way fixed the issue.
In addition to needing obnoxiously regular mxstream restarts, this afternoon the sus machine was doing something slightly differently. Only 1 fb block per core was red (the mxstream symptom is 3 fb-related blocks are red per core), and restarting the mxstream didn't help. Anyhow, I was searching through the elog, and this entry to which I'm replying had similar symptoms. However, by the time I went back to the CDS FE screen, c1sus had regular mxstream symptoms, and an mxstream restart fixed things right up.
So, I don't know what the issue is or was, nor do I know why it is fixed, but it's fine for now, but I wanted to make a note for the future.
I'm not sure why the c1cal model didn't come up the last time c1lsc was rebooted, but I did an "rtcds start c1cal" on the lsc machine, and it's up and running now.
A few times this evening, I had been having trouble locking CARM and DARM with ALS, and holding it for very long. When it started happening again, I switched over to locking the individual arms with ALS. Yarm seems to be totally fine, but Xarm has something funny going on.
Rana and I have narrowed it down to being a problem with ETMX. We were watching ETMX's oplev and local damping error signals, and would see occasional glitch events. This happened when oplev + local damping were both on, both off, and when only local damping was on. We believe that this points to something weird with the coil driver and actuator chain.
We tried to watch for a while to see if it was a step event (something switching on and off periodically), or an impulse event (some transient oscillation in an opamp perhaps), but the problem went away again. We have come to no conclusions other than we have a problem that needs watching.
During our investigations, to more softly turn off the damping, Rana set the local damping gains, as well as the oplev gains to zero using a ramp time. We don't recall the precise numbers, and conlog doesn't have the gains recorded, so we made an educated guess. The local damping seems fine, but the oplev damping should be re-confirmed. Steve, can you please show Harry how, and have him help you measure the ETMX pitch and yaw oplev loops, and set the gains so that they match up to the references, and then post the measured bode plots when you're done?
c1auxex has forgotten who it is. Slow sliders for the QPD head were not responding, so I did a soft reboot from telnet. The machine didn't come back, so I plugged the RJ45-DB9 cable into the machine and looked at it through a minicom session. When I key the crate, it gives me an error that it can't load a file, with the error code 0x320001. Looking that up on a List of VxWorks error codes, I see that it is: S_hostLib_UNKNOWN_HOST (3276801 or 0x320001)
I'm not sure how this happened. I unplugged and replugged in the ethernet cable on the computer, but that didn't help. Rana is going in to wiggle the other end of the ethernet cable, in case that's the problem. EDIT: Replacing the ethernet cable did not help.
Former elogs that are useful: 10025, 10015
EDIT: The actual error message is:
boot device : ei
processor number : 0
host name : chiara
file name : /cvs/cds/vw/mv162-262-16M/vxWorks
inet on ethernet (e) : 192.168.113.59:ffffff00
host inet (h) : 192.168.113.104
user (u) : controls
flags (f) : 0x0
target name (tn) : c1auxex
startup script (s) : /cvs/cds/caltech/target/c1auxex/startup.cmd
Attaching network interface ei0... done.
Attaching network interface lo0... done.
Error loading file: errno = 0x320001.
Can't load boot file!!
We have decided to keep better track (using new-fangled digital "computers") of our modifications to electronics boards.
The idea will be to create a new DCC document for every electronics board (when we pull a board and modify it, it should receive this treatment) that we have, and that document will become a history of the board's life. Version 1 will be a copy of the original drawing. Version 2 should be a modified version of that drawing with the current situation. All future versions should be modified from the most recent version, to reflect any changes. Notes for each updated version should include an elog reference to the work, so that we know why we did things, and have a place to find photos of the actual modifications. Elogs should also include a link to the DCC version. DCC titles should include the phrase "40m Revisions" for ease of searching.
Patient Zero for this new system will be the PMC servo card. The DCC number is D1400221. As of this moment, this just has the V1 original drawing with no modifications.
This has been included in the 40m's DCC document tree that Jamie started back in November 2012.
I have put in a new nominal value for the FSS fast gain: 21.5 dB.
There is an oscillation peak in the MC error point spectra around 41.5 kHz if the FSS gain is set too high. I used the 4395 to have a look at the MC error point, and saw that if I set the FSS fast gain any lower than about 18 dB, the peak wasn't getting any smaller than -41 dBm. If I set the fast gain any higher than about 26 dB the peak wouldn't get any larger than about -34 dBm.
However, if I set the gain to 19.5dB, the PC RMS drive is consistently above 2 V, which isn't so good. If I crank the gain up to 27 dB or more, the PC RMS will stay below 0.9 V, which is great.
As a compromise, I have decided on 21.5 dB as the new FSS fast gain. This puts the oscillation peak at about -39.5 dBm, and the PC RMS around 1.6 V.
I changed the nominal gain by ezcawrite C1:PSL-STAT_FSS_NOM_F_GAIN 21.5. This sets the nominal value so that the FSS screen's fast slider doesn't turn red at the new value. And, since the MC autolocker reads this epics channel and puts that into the gain during the mcup script, the MC autolocker now uses this new gain. For reference, it used to be set to 23.5 dB.
ezcawrite C1:PSL-STAT_FSS_NOM_F_GAIN 21.5
I have put beam dumps in front of both of the end transmission QPDs so that I could measure the dark noise. They are still there.
I checked that the Yend QPD sliders and switches were doing things as I expected. I couldn't do the Xend since c1auxex is still lost (elog 10165). I'll post plots and actual information on this checkout, as well as my calculation of what this dark noise means in terms of meters for CARM when we're using 1/sqrt(trans) signals tomorrow morning.
Measurement of Yend transmission QPD dark noise
Since EricQ had already checked out the whitening filters (see elog 9637 and elog 9642), I didn't check on them. I just left them (the analog whitening, and the digital antiwhitening filters) on.
First, I checked the noise vs. transimpedance gain. There are a few too many settings to put them all on one plot, so I have them sorted by the original transimpedance: 0.5 kOhms vs 5 kOhms. It's a little tricky to see, but all of the spectra that begin with the 5k transimpedance have a little extra noise around 10 Hz, although I don't know why. In the legend I have made note of what the settings were. x1 x1 is my representation of the "inactive" setting.
I then looked at the noise with different whitening gain slider settings. All but one of the traces are the 20 kOhm setting.
These .xml files are in /users/jenne/Arms/TransQPDnoise_July2014/
Calculation of inverse sqrt transmission sensitivity
I used Optickle to give me the power transmitted through the ETMs. I first find the transmission in the FPMI case. I use that to normalize the full PRFPMI transmission, so that the output units are the same as our C1:LSC-TR[x,y]_OUT units.
I take the square root of the transmitted power (sum of transmissions from each arm) at each CARM offset point, add 1e-3 as we do in the front end model to prevent divide-by-zero problems, and then take the inverse.
I find the slope by taking the difference in power between adjacent points, divided by the CARM offset difference between those points.
In this plot, I have taken the absolute value of the sensitivity, just for kicks. I also display an arbitrarily scaled version of the log of the transmitted power, so that we can see that the highest sensitivity is at half the maximum power.
Calculate the QPD dark noise in terms of meters
Finally, I put it all together, and find the dark noise of the QPD in terms of meters. Since the spectra were measured in units where the single-arm transmission is unity, the already match the units that I used to calculate the sqrtInv sensitivity.
I take the spectra of the QPD dark noise for the 20 kOhm case, and multiply it by the sensitivity calibration number at several different CARM offsets. As we expect, the noise is the best at half-max transmission, where the sensitivity is maximal.
I'm starting to lock for the night, and I noticed that PRM is very, very pitched. Why? The PRM pitch slider is 5 full integer units higher than the backup (and the backup value is about where I like it, around -0.2).
I am not aware of any scripts that touch the PRM slider values. The PRM ASS (which I haven't used in ages) offloads the biases to the SUS screen fast channels, so even if someone turned that on and then saved the values, it wouldn't leave the PRM so very, very misaligned.
I have restored it, and relocked the PRMI, so all is well, but it's very weird to have found it so misaligned.
I took some data tonight for a quick look at what combinations of DC signals might be good to use for DARM, as an alternative to ALS before we're ready for RF.
I had the arms locked with ALS, PRMI with REFL33, and tried to move the CARM offset between plus and minus 1. The PRMI wasn't holding lock closer than about -0.3 or +0.6, so that is also a problem. Also, I realized just now that I have left the beam dumps in front of the transmission QPDs, so I had prevented any switching of the trans PD source. This means that all of my data for C1:LSC-TR[x,y]_OUT_DQ is taken with the Thorlabs PDs, which is fine, although they saturate around arm powers of 4 ever since my analog gain increase on the whitening board. Anyhow, the IFO didn't hold lock for much beyond then anyway, so I didn't miss out on much. I need to remember to remove the dumps though!!
Self: Good stuff should be between 12:50am - 1:09am. One set of data was ./getdata -s 1089445700 -d 30 -c C1:LSC-TRX_OUT_DQ C1:LSC-TRY_OUT_DQ C1:LSC-CARM_IN1_DQ C1:LSC-PRCL_IN1_DQ
./getdata -s 1089445700 -d 30 -c C1:LSC-TRX_OUT_DQ C1:LSC-TRY_OUT_DQ C1:LSC-CARM_IN1_DQ C1:LSC-PRCL_IN1_DQ
I removed the dumps in front of the trans QPDs. The Yend QPD needed re-normalization, so I did that.
I realized while I was looking at last night's data that I had been doing CARM sweeps, when really I wanted to be doing DARM sweeps. I took a few sets of data of DARM sweeps while locked on ALSdiff. However, Rana pointed out that comparing ALSdiff to TRX-TRY isn't exactly a fair comparison while I'm locked on ALSdiff, since it's an in-loop signal, so it looks artificially quiet.
Anyhow, I may consider transitioning DARM over to AS55 temporarily so that I can look at both as out-of-loop sensors.
Also, so that I can try locking DARM on DC transmission, I have added 2 more columns to the LSC input matrix (now we're at 32!), for TRX and TRY. We already had sqrt inverse versions of these signals, but the plain TRX and TRY were only available as normalization signals before. Since Koji put in the facility to sqrt or not the normalization signals, I can now try:
Option 1: ( TRX - TRY ) / (TRX + TRY)
Option 2: ( TRX - TRY ) / sqrt( TRX + TRY )
DARM does not yet have the facility to normalize one signal (DC transmission) and not another (ALS diff), so I may need to include that soon. For tonight, I'm going to try just changing matrix elements with ezcastep.
Since I changed the c1lsc.mdl model, I compiled it, restarted the model, and checked the model in. I have also added these 2 columns to the AUX_ERR sub-screen for the LSC input matrix. I have not changed the LSC overview screen.
We had a look at the RIN of the transmission signals TRX and TRY, when the arms were individually locked on IR. If the intensity noise is very bad, then the transmission signals aren't really a good option to use for locking. So, if the RIN is bad, we need to work on our intensity stabilization.
We need to understand what the situation is with the AOM, and why it isn't working as expected, so that we can reinstall it. We also need to decide if we're going to use the SR560 setup, or if the Chas ISS is sufficiently characterized for us to use.
The RIN is certainly bad. Also, I don't know why the Xarm's RIN is worse between 10 Hz and a few hundred Hz than the Yarm.
I looked at what the RIN contribution of the sqrtInv sensor is by locking the arms individually on IR using POX and POY. I then took spectra of the sqrtInv channels. For the Xarm, I had forced the triggering so that the QPD was being used as the transmission PD, while the Yarm was using the regular Thorlabs PD. I also had the green lasers locked to the arms, and took beatnote spectra to see what the sensing noise of the beatnotes is, all at the same time.
For the sqrtInv channels, I used the Optickle calibration from elog 10187. For today's plot, I am using the calibration at about 1nm, since that is about where we are when we transition to the sqrtInv Thorlabs signal usually.
For the ALS channel, I was using the _FINE_PHASE_OUT signal, which is in units of degrees of phase for a single green wavelength. So, since k * x = phi, I want the phase data to be converted to radians (2*pi/360), and use k = 2*pi / lambda_green. So, doing some algebra, this gives me x = phi_degrees * lambda / 360 for my calibration.
What I see in the plot is that the ALS sensing noise is pretty bad compared to the sqrtInv channels, so maybe we don't have to work so hard on the ISS this next week. Also, the Thorlabs PD is much better than the QPDs, which maybe isn't so surprising since we have them set so that they have good SNR at higher power.
Anyhow, here's the plot:
Also, here is the Thorlabs PD only, with single arm locked on RF, with the noise calibrated to different CARM offsets:
As Koji pointed out, I messed up the calibration. However, fixing it doesn't change things that much.
From this calibration by Yuta, the Xarm ALS calibration is 54 deg / MHz, or 19.17 kHz / deg. So, I multiply my data which is in these degree units by 19.17e3 to get Hz. Then I use delta_f / f = delta_L / L to convert to meters. f = c / lambda_green, and L = 37.5 meters.
This only changes the calibration by about 10-15%. It still looks like the ALS noise is well above the RIN level of the sqrtInv signal.
Q is working on fixing the "save offsets" script for the ASS, because that has lost me my alignment two more times in the last few hours. But, right now I have both arms locked with transmitted powers of about 0.9! To get this, I ran the ASS scripts, and hand-tweaked the bias sliders of some of the optics to relieve the ASS outputs. Then I turned the ASS gain to zero, and by-hand turned off the oscillators. So, the ASS outputs are just frozen.
I haven't seen IMTX suspension kicks, I think since Q did the front end reboot earlier. There has been ITMY activity, however. I think I'm going to be bold, and try locking ALS.
After aligning the arms to IR, I aligned the Y green beam to the arm. Also, the X green beatnote was very small, so I aligned the PSL green for X.
The ALS system is iffy tonight.
After putting the cable back to the RF spectrum analyzer (it had been taken to test the frequency counter setup, and not put back), I had a good Yarm beatnote, but again this evening the Xarm beatnote is small. I touched up the PSL table alignment (very, very little needed, but it did double my peak height). I *think* that this is happening because we haven't settled into a good IFO alignment place, so the arm pointing keeps changing very slightly, which means that the PSL ALS alignment needs touching. Anyhow, even after alignment the Xarm beatnote is only -36 dBm at 81 MHz. It should be at least -25 dBm or so, although I haven't seen it any larger than about -35 dBm since the IFO beam was lost last Friday.
I am not able to hold ALS lock long enough to scan the arms and find the IR resonances. The only optics that I am actuating on this evening are the 2 ETMs. When I lose lock and look at the watchdogs, the ETMs are the only optics that have largeish numbers, which comes from the ALS lockloss. So, I don't think I am suffering from the ITM suspension kicks tonight. Rather, I think that it's that the ALS system isn't tuned up nicely.
I think that it is past time we tuned up and checked out the ALS PDH setup. Q: Can you please measure the loop TFs for both of the ALS PDH boxes tomorrow? At the very least we want to know what we're working with.
Evan: What is the status with the ISS?
I am going to try tomorrow to look at the suspensions, and see if I can track anything down. I feel like I see the kicks more often when the arms are locked, i.e. we are sending an LSC signal to them. The LSC POS signal is a factor of a few hundred larger than the damping SUSPOS signal is. Are we saturating something somewhere? Why is this a new thing? We certainly do see kicks when the LSC is not engaged, so this may not be the right path, but it is something concrete to look at.
The MC has been unstable and unhappy for the last several hours. When I looked, I saw that the FSS_FAST monitor has been hovering around 1 V, when it is supposed to be closer to 5ish.
I changed the C1:PSL-FSS_INOFFSET from -0.08 to -0.8537, and will see if the MC sticks around for longer this time around.
Dang it, I completely forgot. Well, anyhow, it pulled itself back down to less than 1V, and the MC stayed happy for several hours. I'm not totally sure what changing the offset did, but the MC seems happy for right now. I should take a quick look at the error point to make sure that I didn't mess up your tuning.
Yesterday, Q helped me look at the DACs for some of the suspensions, since Gabriele pointed out that the DACs may have trouble with zero crossings.
First, I looked at the oplevs of all the test masses with the oplev servos off, as well as the coil drive outputs from the suspension screen which should go straight out to the DACs. I put some biases on the suspensions in either pitch or yaw so that one or two of the coil outputs was crossing zero regularly. I didn't see any kicks.
Next, we turned off the inputs of the coil driver filter banks, unplugged the cable from the coil driver board to the satellite box, and put in sinusoidal excitations to each of the coils using awggui. We then looked with a 'scope at the monitor point of the coil driver boards, but didn't see any glitches or abnormalities. (We then put everything back to normal)
Finally, I locked and aligned the 2 arms, and just left them sitting. The oplev servos were engaged, but I didn't ever see any big kicks.
I am suspicious that there was something funny going on with the computers and RFM over the weekend, when we were not getting RFM connections between the vertex and the end stations, and that somehow weird signals were also getting sent to some of the optics. Q's nuclear reboot (all the front ends simultaneously) fixed the RFM situation, and I don't know that I've seen any kicks since then, although Eric thinks that he has, at least once. Anyhow, I think they might be gone for now.
When I got back to the lab, there was enough water that it was seeping under the wall, and visible outside. Physical plant says it will take an hour before they can come, so I'm getting dinner, then will let them in.
The guy from physical plant came, and turned off the water to the kitchen sink. He is putting in a work order to have the plumbers come look at it on Monday morning. It looks like something is wrong with the water heater, and we're getting water out of the safety overpressure valve / pipe.
The wet things from under the sink are stacked (a little haphazardly) next to the cupboards.
The short cable from the slab to the sensor has been assembled and installed for the Trillium slab at the corner station. The corner still needs the sensor and the long cable, both of which are in use by the gyro experiment.
The STS-2 cable that was running to the Xend was pulled, and the new long Guralp cable that Den made was installed with help from Andres. The Xend just needs the sensor itself, which is also in use in gyro-land.
So, once we get the 2 seismometers and the one cable back from Zach, we should have 3 sensors nicely on the slabs that Den and Steve designed.
Here's the game plan for things that we need to do to get this IFO locked up.
Red is for things that should be done today, or tomorrow if they don't get finished today (eg. laser mode hopping temperature check). Orange is for things that will become red once the current red things are gone (eg. inferring the POP QPD gouy phase, and moving it to minimized PRM information). Green is for things that we'd like to do, but aren't high priority (eg. X green mode matching). Blue is for things that we should remember, but not plan on working on soon (eg. putting PZTs on the Yend table for green).
TODAY so far:
Q already did the tweak up of the PSL SHG crystal alignment. HE SHOULD ELOG ABOUT THIS. What was the final power of green that you got? Do we have any record of a previous measurement to compare to?
Q helped me install PDA55s on each of the lasers (I did the ends, he did the PSL) so that we could do the mode hop temperature check. For the Yend, I took the leakage transmission through the first Y1 steering mirror after the laser. This beam was dumped, so I replaced the dump with a PDA55. For the Xend, the equivalent mirrors are too close to the edge of the table, so I put in a spare Y1, and reflect most of the light to a beam dump. The leakage transmission then goes to a PDA55. Note that for both of these cases, no alignment of main laser path mirrors was touched, so we should just be able to remove them when we're through. For the PSL, I believe that Q took the rejected light from one of the PBSes before the PMC. He mentioned that he bumped something, so had to realign the beam into the PMC, but that he was able to get the transmission back up to 0.802, when we were seeing it in the mid 0.7's for the last several days.
The end temporary PDs are using the TRX / TRY cables, so we will be looking at the C1:LSC-TR[x,y] channels for the power of the end lasers. The PSL's temporary PD is connected to the PMC REFL cable. For the end PDs, since I had filter banks available, I shuttered the end lasers and removed the dark offset. I then changed the gains to 1, so the values are in raw counts. The usual transmission normalization gains are noted in one of the control room notebooks.
I did a slow ezcastep and ramped the temperature of all 3 lasers over about an hour. I'll write a separate elog about how that went.
This afternoon Q helped me put in some temporary PDs for checking for any mode hopping behavior in our 3 main lasers.
Q helped me install PDA55s on each of the lasers (I did the ends, he did the PSL) so that we could do the mode hop temperature check. For the Yend, I took the leakage transmission through the first Y1 steering mirror after the laser. This beam was dumped, so I replaced the dump with a PDA55. For the Xend, the equivalent mirrors are too close to the edge of the table, so I put in a spare Y1, and reflect most of the light to a beam dump. The leakage transmission then goes to a PDA55. Note that for both of these cases, no alignment of main laser path mirrors was touched, so we should just be able to remove them when we're through. For the PSL, I believe that Q took the rejected light from one of the PBSes before the PMC.
The end temporary PDs are using the TRX / TRY cables, so we will be looking at the C1:LSC-TR[x,y] channels for the power of the end lasers. The PSL's temporary PD is connected to the PMC REFL cable. For the end PDs, since I had filter banks available, I shuttered the end lasers and removed the dark offset. I then changed the gains to 1, so the values are in raw counts. The usual transmission normalization gains are noted in one of the control room notebooks.
I did a slow ezcastep and ramped the temperature of all 3 lasers over about an hour. Since we usually use the PSL around FSS slow slider value of zero, I swept that from -10 to +10. Since we usually use the Xend laser at around 10,000 counts, I swept that from 0 to 20,000. For the Yend laser, it is usually around -10,000 counts, so I swept it from -20,000 to 0. ezcastep -s 0.2 C1:ALS-X_SLOW_SERVO2_OFFSET +1,20000 C1:ALS-Y_SLOW_SERVO2_OFFSET +1,20000 C1:PSL-FSS_SLOWDC +0.001,20000
ezcastep -s 0.2 C1:ALS-X_SLOW_SERVO2_OFFSET +1,20000 C1:ALS-Y_SLOW_SERVO2_OFFSET +1,20000 C1:PSL-FSS_SLOWDC +0.001,20000
I was looking for something kind of similar to what Koji saw when he did this kind of sweep for the old MOPA (elog #2008), but didn't see any power jumps that looked suspicious.
Here is the PSL:
And the Yend:
* Decided that earlier mode hop scan won't give us the information that we were hoping for. We need to think about where we can actually see the frequency change. Can we use the IR beatnote that we will soon have to do this? We'd only be able to scan one laser temp at a time, but that's okay. Leave, say, the PSL temperature alone, and scan one of the end laser temps. Using the PSL as the reference, we will be able to see if the frequency of the end laser goes crazy and jumpy as we pass through a certain temp. Then, repeat while holding the end laser constant and scan the PSL. Thoughts?
* Meditated on PSL oplev servo, but I need to make a Matlab script that can evaluate different loops according to a cost function based on elog 9690.
* Aligned IFO to IR, then greens to arms (got back to 0.9 for GTRY, but only about 0.5 for GTRX, with the PSL green shutter closed). Then aligned green beams on the PSL table, since the PSL green pointing had changed a bit from Q's crystal alignment tweak-up earlier today. Beatnotes are nice and big (see elog 10381 - The Yarm is the larger beatnote, and the Xarm is the smaller one.)
* Was not able to lock ALS comm/diff and hold long enough to get both arms to IR resonance. Also, saw that TRY's RIN was more than 50%(!!!). We took a look, and there seems to be much more low frequency noise than there was when the spectrum in the control room was taken for the multicolor metrology paper:
* Tried to balance the ALS comm/diff input matrix, with not a lot of success. First of all, it looks like the Xarm has overall about 10 times more noise! We were exciting MC2 in position (~88 Hz, about 130 counts I think), and then looking at DARM_IN1 for the peak. When DARM_IN1 was just one of the 2 ALS error signals (i.e. one matrix element set to zero), versus when both matrix elements were set to 1, we saw a factor of only about 3 in reduction of the peak height. We were hoping to have better cancellation of this pure CARM signal in the DARM channel. The Xarm green PDH loses lock every ~5 or 10 minutes, and when we relock it, this cancellation seems different, so we want to try again tomorrow when the ALS is locked on comm / diff, rather than just the free running ALS that we have now. Although, if the balance of the input matrix changes lock-to-lock, we may need to consider redoing the green PSL table layout so we get a pure DARM beatnote signal like they have at the sites.
* We want to change how the watch script for ALS works, although this is a low-priority task. Rather than looking at the control signal, we should maybe look at the sum of all the coil outputs, multiplied by a pendulum TF, and use that as a rough displacement sensor. We want to be careful of pushing too hard at low frequencies, but we want to allow higher frequency actuation without having the watch script shut things down.
* Also, I should put on the to-do list the revamp of the ALS find IR resonance script.
(Updated as of 4pm)
End PDH UGF improvement / post mixer LPF investigation (with in 2 weeks)
Riju measured the MC REFL PD transimpedance. See ELOG and related.
Why do we want to see less PRM motion? I thought PRC motion was causing
LSC issue of the central part. We wanted to maximize the PRM effect, don't we?
(Or is this to supress ETM motion during full lock?)
End PDH - good point, thanks.
ASC - Yes, this is so that we can use the POP QPD to feed back to the common ETMs after the CARM offset is already quite small. We will not use POP DC QPD for PRC any more.
Also, for future PRC ASC, I keep coming back to this in my head, but maybe it is less painful to install oplevs for PR2, PR3 than it would be to make an RF QPD. Neither is going to be trivially easy. But if we had sensors of the tip tilt motions, we could feed all of that back to the PRM to stabilize the PRC.
- Oplevs for PR2, PR3 => Almost impossible.
Because of the limited table space inside? That's the main reason I can think of that this method is hard. Am I missing something?
Last night, and again just now, I used the ./MC2_spot_[direction] scripts to center the MC2 spot on the trans QPD. The MCWFS handled overall alignment to correct for the fact that the ratios in the script aren't perfect. When I was finished, I ran the MC WFS relief script from the WFS screen. Last night, and again today, things had drifted until the yaw spot was more than 0.5 counts off.
The instigator of this was that we were seeing ring-ups of ETMs during our ALS locks this evening. We measured the ETMY violin resonance to be 624.10 Hz, and Rana found an elog saying that the ETMX was around 631 Hz, so we made a 2 notch filter and added it to FM4 of the LSC-SUS filter banks for both ETMs.
For the ETMY resonance, we measured the frequency in the DARM spectrum, and when we looked at the FINE_PHASE_OUT channels, the resonance was only in the Yarm sensor. So, we conclude that it is coming from ETMY.
Also in the realm of filter modules, the FM3 boost for CARM, DARM, XARM and YARM was changed from zero crossing to ramp with a 1sec ramp time.
I don't know why, but TRY has somehow gotten a 0.3 count offset in the last hour.
Rana and I are witnesses for each other that neither of us has gone into the IFO room in the last several hours (and we're the only ones here). For some reason though, the TRY PD now has a 0.3 count offset. We have been doing some ALS locks, but we have not run the offset script in the last several hours. Closing the green shutter doesn't change things, and we still see the offset when the MC loses lock, so it's not to do with the end or the PSL laser. We haven't been in there, so there hasn't been a change in the room lights.
I have added a few things to the ASS model, and the ASC sub-block, so that we can send POP QPD information down to the ETMs for CARM angular control after we've reduced the CARM offset and gotten some carrier buildup. I did not remove our ability to actuate on PRM, so that we can still play with it in PRMIsb cases.
The input matrix has been expanded so that it can send signals to new CARM_YAW and CARM_PIT filter banks. The corresponding filter banks have been created. The output matrix was also expanded to take in the 2 new servo outputs, and so it can send signals to both ETMs, pitch and yaw. I did not include any triggering logic for this new CARM situation, since I assume we'll just turn it on and off with our scripts. (We haven't really been using the triggering capability of the PRM ASC either lately, although it's all still there). I added the inputs and outputs of the CARM servos to the list of acquired channels.
The ASC sub-block:
I also modified the top level of the ASS model. This was just a simple addition of summing nodes for the ETMs, similar to what was already in place for the PRM, so that we can send both the ASS dither alignment signals and the ASC servo control signals to the optics.
The ASS top level:
I also quickly modified the ASC screen to expose all of the new options:
The ASS model was compiled, and restarted. As usual, this temporarily removes the biases on the input pointing tip tilts, but the pointing seems to have come back without any trouble.
Not sure why, but Rana and I didn't see the super high Xarm noise with ALS that we reported last night (elog 10382).
The in-loop ALS noise seems fine. The out of loop measurement while the ALS is locked is a little tricky, since ALS hold the arms within the POX/POY linear ranges.
Here is the in-loop noise:
The game plan graffle file is now in the 40m dropbox, so anyone can edit it. Please just make sure to keep the date in the top right corner accurate.
Again unknown, but about 6 hours ago (so ~8am) the offset disappeared.
Here's a 1-day trend:
Here is a plot of last night's data with both the control and the error point on the same plot, in Volts. Q is still working, so I don't have a calibration number yet to get these to Hz.
Note in the control spectrum that we have very significant 60Hz lines.
EDIT: I also added a new branch to the DCC Document Tree, and 2 leafs (one for each end). Here's the ALS PDH servo branch: E1400350
* (JCD) Think about this box's purpose in life. What kind of gain do we need? Do we need more / less than we're currently getting? NPRO freq noise is 1/f and is 10kHz/rtHz at 1Hz (this is from a plot of an iLIGO NPRO from Rana's thesis, but it's probably similar). Talk to Kiwamu; the noise budget in the paper seems to indicate that we had some kind of boost on or something. Also, if we need much more gain than we already have, we'll definitely need a different box, maybe the PDH2 box that they have over in WBridge.
It's not so impressive yet, but here's a plot that shows (a) Rana's guess for laser frequency noise, (b) The inferred in-loop version of that noise, (c) The CARM linewidth FWHM, translated to Hz.
For (b), I take the loop that Rana and I measured last night, and I assumed that it continued on forever as 1/f toward low frequency. Then I do 1/(1+G) to get the closed loop version of the loop (which is a measurement with an artificial line tacked on the end), and multiply this with the laser freq noise, which is also totally artificial.
For (c), I do df/f = dL/L, with f = c/lambda_green, since the rest of the plot is meant to be in green frequency units.
This is my beginnings of trying to come up with a requirement for our green PDH boxes. We weren't very clear in the MultiColor paper about the nitty-gritty details (obviously), but then Kiwamu didn't expand on those details in his thesis either. He talks a lot more about the design considerations for the digital ALS loop, which isn't what I want today. I will send him an email to see if he had any notes that didn't make it into his thesis.
I calibrated the control signal from Volts to Hz using the rough PZT calibration of 5MHz/V for the Yend NPRO.
For the error signal, Q said that the Yarm PDH peak-to-peak height was about a factor of 100 smaller than the Xarm, so I used a calibration of 1.9e7 Hz / V.
Then, from Q's Mist simulation including the high Xarm loss, and the plot that he posted in the control room, the CARM linewidth looks like it is about 2pm. This is the number that I have included on today's plot. Note though that yesterday I was using a linewidth of about 30pm, which I got from an Optical simulation about a year ago. I do not know why these numbers come out an order of magnitude different! The CARM linewidth is actually about 20 pm. Both Q and I failed at reading log-x plots yesterday. I have corrected this, and replotted.
Anyhow, here's the Yarm noise spectra calibrated plot:
I have emailed Kiwamu, but haven't heard back from him yet on what the original design considerations were, if he remembered us ever using a boost, etc. What this looks like to me is that we need to do some serious work to get the noise down. Maybe fixing the gain peaking and triggering the boost will get us most of the way there?
[Rana, Jenne, EricQ]
We did several things tonight. First, a list (so I can remember them all), and then some details.
(1) Jiggled ETMY SUS cables, removed kicks.
(2) Locked X and Y ALS, looked at POX, POY as out of loop sensors.
(3) Measured stuff (?) at the Yend.
(4) Reconnected REFL DC to SR560.
(5) Attempted CARM offset reduction.
When Rana and I started locking this evening, we saw (as Q has been witnessing for a while now) the ETMY kick a lot. However, it seemed to be kicking even more than usual. Since Q had been down at the end station recabling things, we wondered if a SUS-related cable got bumped. Rana went down to the end and pushed all the cables into their receptacles. One of the last sets that he pushed was the satellite box. We didn't have walkie-talkie communication, but the DC offset of the ETMY oplevs changed just a minute or two before he returned to the control room. So, we guess that it was the satellite box cables that were loose. Unfortunately, there is no clear way to strain relieve them, which is why they can so often be troublesome. Anyhow, the ETMY hasn't kicked since.
We locked the arms with ALS. We saw that the POX signal was about 20% of the full pk-pk height of the PDH signal, so it's mostly within the linear range, but not entirely. It is what it is, however, and we took measurements assuming that it's okay. I calibrated POX by putting an excitation onto ETMX, and matching the height of the peak in POX and BEATX_FINE_PHASE_OUT_HZ.
Q and Rana had also [remembered / put in / something] a digital readback for the end green PDH error point. Q went down to the end and gave me a number of 2600 Hz/V for the err mon port of the PDH board, which is what is connected to the ADC. With that and 20/2^16 V/cts, I had a calibration of 0.8 Hz/ct.
What we see in this plot is that the green end PDH is not the limiting noise for the POX out of loop measurement of the residual arm motion. Also, in the multi-color metrology paper, Fig 7 (which is posted in the control room), we see at about a little over 1 Hz a ratio of about 4.5 between the residual motion and the AUX PDH error signal. In today's plot, I see a ratio of about 20. I infer from this that the green PDH for the Xarm is fine, and that we may want to re-look at the ALS digital loop, but we should leave the X PDH alone.
Here is the Xarm plot:
Q took the data for the Yarm plot, so hopefully he can give it to us in the morning. What we did notice was that the noise was much worse for the Yarm. This prompted Item 3, measuring the loop.
Q and Rana went down to the Yend and measured some things. They came back, and said that they hadn't changed anything in analog while they were down there. One thing that Q did note was that we have almost 90 degrees of phase margin (since it's a 1/f loop), and about 10 dB of gain margin, above the UGF. So, we're in good shape for being able to try triggering the boost on the PDH box. Q will give us more notes on this work, as well as plots, in the morning.
At some point, I remembered that Q and Gabriele had repurposed the SR560 that we had been using for the REFLDC input to the common mode board. So, Q went and put it back, so that REFL DC goes into the SR560, and so does a DAC channel so that we can remotely set the offset. The A-B output goes to the REFL11I whitening channel, since real REFL11I goes into the input of the CM board. I think that today, the SR 560 was left at a gain of 1.
We decided to carry on and try to reduce the CARM offset some. An annoyance is that the Yarm still has pretty significant low-frequency noise, but the idea is that if we can get over to the sqrtInvTrans signals, it will be fine.
So, we didn't get much farther than we had in the past, but it was nice to get there at all again. I ran the carm_cm_up script (many times). One of the times, all I wanted to do was see how much I could reduce the CARM offset. CARM was on sqrtInvTrans, DARM was on ALS diff, and I was able to get the arm powers up to about 2.5. I don't know why I lost lock. The sqrtInv signals should be good until at least arm powers of 20 or so.
I was able to see the REFL DC dip, but only a teensy tiny bit. It went down by maybe 1 count. Q suggested looking at how deep it could get while leaving CARM and DARM both on ALS, and setting both offsets to 0. We were seeing arm flashes of about 50 counts, and REFL DC went from 0 to -800. So, I wasn't seeing much of a REFL dip, but it was definitely there when I went to arm powers of 2ish.
We tried looking at different sqrtInv options for DARM, and haven't come to any real conclusion. In the plot below, we are looking at a swept sine between DARM_IN1 (ALSdiff) and either MC_IN1 0.3*(sqrtInvX - sqrtInvY) or SRCL_IN1 (TRX - TRY / sqrt(TRX + TRY) ):
We have a few things to add to the to-do list:
* Put UGF servos for LSC loops in place.
* Implement UGF "servos" (per Koji's suggested method) for phase trackers.
* Write a lockloss script that is run by the ALS watch scripts - print a PDF of error and control signals for every lockloss, and save it somewhere.
* Fix up Ygreen modematching on the PSL table. The X green spot is quite similar on the camera to the corresponding PSL green spot. However the Y green spot is not at all the same as its PSL green spot.
Slightly updated Game Plan. Mostly, Q is continuing to check out the Xend PDH box saturation, and I am thinking on what our requirements are for ALS, and thus for the green PDH boxes.
Q put the X PDH box back, so that I could try locking, and remember which end is up after a week away.
I am unable to hold ALS comm/diff for any length of time. Only once today did I hold it through the FM3 boost turn-on. So, I looked at the individual arms.
Xarm, even though it's the one that Q is seeing this saturation problem with, seems fine.
Yarm however is having trouble holding lock for more than a few minutes at a time. The green beam stays locked to the arm for ~infinity, so I'm not so worried about the PDH box right now. If I look at the error and control points of the ALS digital servo, the Yarm is much more noisy above about 20 Hz. Something that I might think of for this kind of mismatch at higher frequencies is poorly matched whitening / dewhitening, or none at all for the Yarm, however this doesn't look like that to me. Based on the shape of the spectra, I don't think that we're running into ADC noise. For this plot, both arms are individually locked with ALS feeding back to the ETM, gain magnitude of 15 (Xarm gets a minus sign because of our temperature / beatnote moving direction convention), FMs 1,2,3,5,6 on. Something that seems critical for getting the Yarm to have the FM3 boost without losing lock is having the SLOW temperature servos on for a little while so that the PZT output (as monitored on the temp servo screen) for the end lasers fluctuate around zero. Right now, both beatnotes are at about 62MHz, with an amplitude of about -31dBm.
I still need to do a somewhat more thorough investigation of what might be causing the Yarm locklosses. Is the length-to-angle decoupling worse for ETMY than for ETMX? Am I moving the arm length so far that the PZT can't follow within its actuation limits? Does the Yend PDH box have a similar saturation to the Xend box, but somehow (a) worse, and (b) not as obvious so we didn't suspect it before?
I need to put this plot into calibrated units, and also include the low frequency monitor that we have of the PDH error point (all of which are _DQ channels).
Things to do:
* Figure out Xend PDH box saturation issue. Is Yend seeing same saturation in the variable gain amplifier? We have 3 spares of these chips in the Plateau Tournant Bleu, if we need them.
* Check Yarm ALS stability. (NB: The arms have been individually locked for the last 15 min or so while I've been writing, so maybe letting the slow servo settle is the key, and this is not something that needs work).
* Get CARM on DC Trans, DARM on AS55Q (after arm powers of about 1). Can we see good REFL DC dip? Should we try using just the transmission PD signal as the error signal for the CM board, if we aren't close enough to resonance to use REFL DC?
From EricQ's simulations reported in elog 10390, we want to transition from ALS comm to DC transmission signals around 500 pm. However, around 100 pm, the DC transmission signals have a sign flip, so we don't want to have the ALS swing that close to the CARM resonance. So. We want to be at about 500 pm, and not touch 100 pm. So, we don't want our peak ALS motion to go beyond ~400 pm. Which means that we need to have less than about 40 pm in-loop RMS, to avoid hitting 400 pm. This is an ALS requirement, but since the analog PDH box is what forces the end laser to follow the arm cavity, and thus give us information about the arm length fluctuations, the PDH residual noise is part of our sensor noise for the full ALS. So, we need to have the PDH in-loop RMS be less than 40 pm, integrated from a few kHz down to at least 30 mHz. Recall that above the ALS UGF (of about 200 Hz), the sensor noise will be suppressed by 1/f, so we should take that into account when we are looking at the PDH error signal, before we calculate the RMS motion.
Q also measured the in-loop error signal with the current Yend PDH box in elog 10430, and it looks like most of the RMS is coming from a few hundred Hz. I designed a hack to the PDH board boost that has a zero at about 2kHz, and a gain of 30 at DC, so that we will win by squishing all that RMS. Also, it shouldn't be too aggressive, so we should be able to leave it on all the time, and still acquire lock of the green laser to the arm, without having to do triggering.
The board schematic is at DCC D1400294. The boost is also called the "integrator stage", although it will no longer be a simple integrator.
EDIT, JCD: This cartoon is not correct for the non-boosted state, doesn't include effect of R16.
Okay, went back to the drawing board with Rana and Koji on PDH box stuff.
Currently (at least for the Yend), in the boost OFF state, we have an overall gain of about 50. This is crazy big. Also, the zero in the "transfer function stage" is around 1kHz, however our green cavity pole is (calculated) to be around 20 kHz. Since these are supposed to cancel but they're not, we have a wide weird flat region in our loop TF.
So. I calculated the changes to the TF stage that I'll need so that I have an increase of about 20 in DC gain, kept the pole at the same ~20Hz, but moved the zero way out to 18kHz. I also calculated the changes needed for the integrator stage to make it effective at much higher frequency than it was designed for. Now the pole is at 75 Hz, and the zero will be at 1.6kHz, and the high frequency gain will stay pretty close to the same with and without the boost.
Planned new TF stage:
Planned boost stage (with and without boost activated):
New boost stage only, so you can see the phase:
The schematic, modified to show my planned changes (which I will put in the DCC after I make the changes):