Successful PRMIsb locking with REFL165I/Q
My previous entry suggested that somehow the REFL165 signals show reasonable separation between PRCL and MICH, contrary to our previous observation.
I don't know what is the difference now. But anyway I took this advantage and tried to lock sideband resonant PRMI.
REFL165I was adjusted so that the signal is only sensitive to PRCL. Then REFL165I and Q were mixed so that the resulting signal shows.
(Next time, we should try to optimize the Q phase to eliminate PRCL and just use the I phase for PRCL.
At first, I used AS55Q for lock acquisition and then switched the MICH input matrix to REFL165.
Later I found that I can acquire PRMI just turning on AS55Q without turning off REFL165.
The REFL165 MICH signal had an offset of 15cnt. The lock was more robust and the dark port was darker once the MICH input offset was correctly set.
MICH OFS = 0
Turn on AS55Q only / or AS55Q + REFL156I/Q
Once it is locked and all of the FMs are activated, give -15.0OFS to MICH.
Turn off AS55Q.
AS55 WHTN: 21dB demod phase -5.5deg
REFL165 WHTN: 45dB demod phase -156.13deg
AS55Q x1.00 MICH
REFL165I x-0.035 + REFL165Q -0.050 MICH
MICH POP110I 100up/10down / FM Trig FM2/3/6/7/9 35up 2down 5sec delay
PRCL POP110I 100up/10down / FM Trig FM2/3/6/9 35up 2down 0.5sec delay
MICH OFS -15.0 / Gain -10 / Limitter ON
PRCL OFS 0 / Gain -0.02 / Limitter ON
MICH ITMX -1.0 / ITMY +1.0
PRCL PRM 1.0
The attached plot shows the spectra of all the REFL signals with the PRMI SB lock.
We excited the ITMY_LSC with 3000 counts. We used the Masayuki calibration of ITMY (5 nm / count * (1/f^2)) to estimate this peak in the REFL spectra.
To correctly scale the REFL spectra we account for the fact that the DTT BW was "0.187 Hz" and we turn off the "Bin" radio box before measuring the peak height with the cursor.
Since the ITMY motion is 3000 * 5e-9 / (580.1 Hz)^2 = 44.6 pm_peak, we want the DTT spectrum of the REFL spectra to report that too.
i.e. to convert from peak height to meters_peak, we use this formula:
meters_peak = peak_height * sqrt(BW) * sqrt(2)
I *think* that since the line shows up in multiple bins of the PSD, we should probably integrate a ~0.5 Hz band around the peak, but not sure. Need to check calibration by examining the time series, but this is pretty close.
Mystery: why are the REFL_I 3f signals nearly as good in SNR as the 1f signals? The modelling shows that the optical gain should be ~30-100x less. Can it be that our 1f electronics are that bad?
Bonus: notice how we have cleverly used the comb of bounce frequencies around the calibration line to determine that REFL11 is clipping!
Rana and I noticed last week that it looked like the REFL11 beam was clipping. This afternoon, I locked the PRMI with REFL 165 I&Q, and checked the REFL 11 path. The beam looks fine through all of the optics going to the diode, so I just realigned the beam onto the diode using the itty bitty steering mirror. I have not yet checked the change (hopefully improvement) in the REFL11 spectrum.
The ITM oplevs were pretty close to the edge of their ranges, and none of the oplevs have been centered in a while, so I centered ITMX, ITMY, BS, PRM after having done alignment (arms, then PRMI).
I have made a measurement of the PRMI and the DRMI sensing matrices.
Keiko pointed out to me in an email a little while ago that I wasn't zeroing elements in the oscillator drive matrix after using them, so I was effectively driving all the degrees of freedom at once, which is why some of the recent sensing matrices looked a little bullshitty. Anyhow, I put in a few lines to zero that row of the LSC output matrix, so that we don't do that any more.
PRMI sensing matrix:
DRMI sensing matrix, first-ever measurement, after the optic flipping / recent locking success:
Note that we don't have any error bars in the DRMI case, since the IFO fell out of lock during the error bar measurements. So, we got the "real" data from the degrees of freedom, but not extra data for making error bars. Also, the MICH / SRCL coupling hasn't been balanced out in the output matrix yet, but since the notches are engaged in the degrees of freedom during this measurement, that shouldn't be a significant effect.
To get the DRMI sensing matrix measured, I added the SRM's actuator calibration to SensMatDefinitions.py (data from elog 5637). I also created a new file runDRMI_sens, to be the equivalent of runPRMI_sens. In the new runDRMI_sens, I reduced the actuation from the oscillator by a factor of 10. I had several attempts at higher oscillator amplitudes that kept kicking the IFO out of lock.
The DRMI was pretty good, but I wasn't getting ~10s of minutes like Koji was on Friday. I also wasn't able to engage all of the FM triggers that he was. The 10-30 Hz seismic BLRMS is a little higher than a usual night, but other than that, seismic looks pretty quiet.
My settings for the night:
LSC input matrix: +0.1*REFL55Q = MICH, -0.125*REFL11I = PRCL, +1.00*REFL55I = SRCL.
Filter settings: MICH, PRCL, SRCL all had FM4,5 always on. MICH had FM2,3 triggered. PRCL had FM2,3,6 triggered. SRCL had FM2 triggered. In particular, engaging FM 6 for MICH or SRCL made some loud low-ish frequency oscillation. Engaging anything other than FM2 for SRCL kicked the IFO out of lock.
Gains: MICH = -0.800, PRCL = +0.050, SRCL = -0.100
Triggering: All triggered on POP22I, upper = 50, lower = 10 (lower = 25 for SRCL).
FM trigger thresholds: MICH on = 35, off = 2, delay = 2 sec. PRCL on = 35, off = 2, delay = 0.5 seconds. SRCL on = 80, off = 25, delay = 5 sec.
Power normalization: None, for any degree of freedom.
LSC Output matrix: MICH = -0.267 for PRM, +0.50 for BS. PRCL = +1.0 for PRM. SRCL = +1.0 for SRM.
LSC SUS filters: BS, PRM, SRM all had FM1,2,3,6 engaged for the BS, PRM and SRM violin filters, as well as the 3rd order harmonic for one of them.
I tried locking the SRMI, so that I could do the same kind of actuator calibration that Koji did for the PRMI in elog 8816, but was unsuccessful. I checked optickle, and found that for REFL 55 I&Q locking, MICH and SRCL keep the same signs for SRMI as DRMI. Also, for both, the optical response is a factor of ~15 lower for SRMI than DRMI, so the gains should be higher by a factor of 15 for both MICH and SRCL. I think my big problem here is that I don't have anything to trigger on. There isn't any signal to speak of in the POP PDs, with the PRM misaligned. Hopefully we'll have AS110 shortly, and that will help.
I updated the IFO Configure restore scripts to our latest versions of locking. I have also tested them, and restoring the Michelson, PRMI and DRMI all seem to work. (MICH restores to locking with AS55Q. PRMI restores to locking with REFL165 I&Q. DRMI restores to the settings noted above in this entry.) The X and Y arm restores have been working, and I have been using them (semi-)regularly since I announced them in elog 8433 back in April. Still to-do though: Add PRCL ASC to the PRMI up script, and make the dither options work for at least the arms and PRM. (Just need to point the drop down menu options to the new ASS scripts.)
[Koji, Jenne, Manasa, Annalisa, Rana, Nic]
- After we checked the functionarity of the Yarm ALS, both arms were locked with the IR, and aligned by ASS.
- Disengaged the LSC feedback. Approximately aligned the PRM.
- Recorded the current alignment biases. Turned off all of the oplevs.
- Went into the lab, aligned all of the oplevs on the QPDs (except for the SRM).
- Check the locking of the PRMI.
- Once it is locked, go into the lab again and align the POP QPD.
- Check everything of the PRMI LSC/ASC works.
- Misalign PRM by 0.2
- Lock the arm again. Run ASS again.
- Miaslign ETMX.
- Lock the Xarm with green. Adjust the beat freq between 30-50MHz.
- Reset Phase Tracker history.
- Check if there is any offset for the ALS. If there is, adjust it to zero.
- Stabilize the arm with the ALS. We should check the sign of the servo before it is cranked up to the nominal.
- Confirm if the offset FM has LPF (30mHz LPF).
- Run excastep for the ALS offset until we find the TEM00 resonance of the IR
- Record the offset at the resonance.
- Step back by 5 count (=100kHz)
- Started from the offset of -5.
- Aligned the PRM and the PRMI was locked by REFL165I(x0.8)nadQ(x0.2).
- PRM ASC engaged
- Moved the offset to -4 by ezcastep C1:ALS-OFFSETTER2_OFFSET +0.01,100 -s 0.1
ezcastep C1:ALS-OFFSETTER2_OFFSET +0.01,100 -s 0.1
- Moved to -3, -2, -1.5, -1. During the sweep PRCL/MICH gain was tweaked so that the gain is reduced.
Nominal locking gain was PRCL x+2.5/MICH -30 . During the sweep they were +2.2 / -12
PRCL FM2/4/5 ON, Later FM3/6 turned on and no problem.
- Moved to -0.9, .... , and finally to 0.
- Automation of the PRMI+one arm
- PRMI locking with BS/PRM
- Better sensing matrix
- PRMI+two arms
- Use of the DC signals form the transmission monitors. (High power /low power transmon).
AWESOME! You guys rock.
Data retrieved using getdata (30 minutes trend) saved at
Last night, I took sensing matrix data at various different offsets for the Yarm. The sensing matrices I measured were of the PRMI, while the Yarm was (a) Held off resonance, (b) Held at ~50% peak power, and (c) Held on resonance.
The dither lines were clear in the MICH and PRCL spectrum, so I think I'm driving hard enough, but something else seems funny, since clearly the REFL165 I and Q signals were not completely overlapping last night. If they were, we wouldn't have been able to lock the PRMI using REFL 165 I&Q.
Anyhow, here's the data that was taken. Data folder is ...../scipts/LSC/SensingMatrix/SensMatData/
Yarm off resonance, SensMat_PRMI_1000cts_580Hz_2013-07-18_012848.dat
Yarm at ~50% resonance, SensMat_PRMI_1000cts_580Hz_2013-07-18_013937.dat
Yarm on resonance, SensMat_PRMI_1000cts_580Hz_2013-07-18_013619.dat
Hmm. I agree that something was funny.
Let's take the matrix without the arms and confirm the measurement is correct.
So far this afternoon, I have redone the IFO alignment, locked both arms with ALS, moved both arms off resonance, locked PRMI, and started bringing one arm back to resonance.
The alignment was really not good, which I knew yesterday, but the ASS wasn't working yesterday. I hand-did the alignment, and tried locking, which was easier with the slightly better alignment.
I locked both arms with ALS, found the resonances, and then moved them off resonance using Masayuki's scripts.
I then restored the PRM alignment, and locked the PRMI.
I started bringing the Yarm back, but I kept losing lock when I got to about 0.1 transmission.
After losing lock several times, I switched over to looking at the ASS. I have figured out the problem, and fixed it. The ASS for the arms now works again.
Looking at the StripTool plots of the lockin outputs for each arm, it was clear that the "L" traces were their usual size, but the "T" traces, which are demodulated versions of the transmission DC PDs, were tiny. I investigated in the model, and the answer is obvious: both the LSC and the ASS get the transmission information directly from the end sus computers. Since we recently moved the normalization gain for the transmission diodes into the SUS models from the LSC model, this means that the ASS was seeing a differently sized signal than it had in the past.
To fix this, I put a gain into the T_DEMOD_SIG filter banks for all 8 lockins that use info from the transmission DC PDs. I used 1/g , where g is the gain that is in the C1:SUS-ETM#_TR#_GAIN channels. For TRX, that number is -0.003, and for TRY that number is 0.002 . So, in the .snap file that is used when turning on the ASS, I have given the Xarm lockins a gain of -333, and the Yarm lockins a gain of 500. I chose this place, because the only thing that has happened to the signal until this point is a bandpass, so the rest of the servo gains can remain the same.
I tested the ASS, and it works just like it used to. I let it run, and align all of the optics, then I misaligned by a small amount each of the ETMs, saw that the lockin output values changed, and then were servoed back to zero. So, it seems all good.
While Manasa, Jenne, and Masayuki are working on the preparing the interferometer, I write the elog for them.
- 6PM-ish: X and Y arms were was locked. They were aligned with ASS.
- PRMI was locked. The PRM was aligned with ASS.
- Jenne went into the lab and aligned the PRM ASC QPD.
- Jenne also aligned all of the oplev spots except for the SRM.
- 6:40PM Then, Manasa and Masayuki checked the out-of-loop stability of the arms.
The X and Y arms have the rms of 2.2kHz and 600Hz, respectively.
The X arm is significantly worse than the Y arm.
Masayuki saved the plot somewhere in his directory.
- 7:20PM X beat: 41.2MHz, Y beat: 14.8MHz
- 7:22PM PRMI locked POP110 115-120
- 7:30PM Lost lock of everything. Start over. Taking the arm alignment.
- 7:45PM start the 2nd trial. PRMI+one arm ready.
- 8:00PM explosion! Lost lock.
- 8:30PM The Xarm ALS is not stable anymore. It loses the control in ~10sec.
We are investigating the out-of-loop stability of the Yarm ALS.
(i.e. Look at the beat note error signal while locking the Yarm with the IR PDH)
Arms locked in comm and diff with ALS. PRMI locked with REFL33 I&Q while arms off resonance. Having trouble reducing CARM offset, even to get to arm powers of 1.
After Manasa installed the new Xgreen PD, Koji looked at the PSL table alignment with me. I saw only a very weak beatnote with the X BBPD, even though I could see the beatnote on the Y PD from the leakage of the X beam to the Y PD (Yend shutter was closed, so just PSL and X greens were on the table). I had thought that my near-field and far-field alignments were pretty good (actually, I checked them, but didn't feel that I needed to tweak them since Manasa did the alignment this afternoon). Anyhow, it was just a matter of tweaking up the alignment a bit, and then the X beatnote got up to about -25dBm at a few tens of MHz. I am starting to question myself if the other BBPD is broken, or if I just not well enough aligned. Anyhow, the spare is in, we can still have a look at the previous X BBPD, but it may be okay, and it's just me embarrassing myself by not catching an alignment problem.
Anyhow, after the X beat was found, I was able to (on my first try) lock the arms using ALS comm and diff. (I already had a nice strong Y beatnote, so that didn't need finding, other than temp adjustment of the end laser). I ran the carm_cm_up.sh script, and it did everything nicely. I did a quickie check of the phase tracker loop gains, but that should be redone in the morning.
PRMI was a little reluctant to lock, so I played around with the MICH and PRCL gains, but didn't really find any combination that was any better than the usual (+0.8 for MICH, -0.04 for PRCL as I had last night, although I needed to reduce the PRCL gain back to -0.02 to eliminate loop osc).
After an arm lockloss, I relocked just the PRMI and used awggui to put a line into C1:LSC-PRM_EXC to check the RF PD phasing. I changed REFL33 from 133.5 to 138.5, and REFL 165 from -142.5 to -152.5. I didn't think that REFL11 needed changing, and I didn't check REFL55. I also checked that I could lock PRMI without arms, using both REFL33 and REFL165 - they seemed about the same to me, both stable. REFL33 has 1's in the input matrix, and I was using 0.07's for REFL165. The demod phase adjustment didn't really improve PRMI locking while the arms were held off resonance, even if I moved the arms even farther from resonance (usually we do 3nm, I went out to 5nm to see if that helped - it didn't). I tried REFL165 locking, but that wasn't any good either. I tried using REFL165 with the arms held off resonance, but that didn't seem to catch at all (at least with REFL33 I was getting short lock blips).
Anyhow, of the 3 or 4 times that I caught REFL33 PRMI lock and tried to reduce the CARM offset, only one time did I even get to arm powers of about 1 (CARM digital offset of -0.1, with CARM held on sqrt transmission signals), and then it didn't stay for more than a few tens of seconds. The other few times, it broke lock on the way up to arm powers of 1.
So, carm_cm_up.sh works pretty well, although perhaps the arm powers of 1 offset reduction needs to be a little slower. PRMI doesn't catch and hold lock very easily with REFL33, and even less so with REFL165. It may be useful to try catching lock with REFL11 or 55, and doing a transition over to 3f. No real progress forward, but we're pretty much recovered.
PRMI(sb) lock was recovered
- Stared at the time series data of the REFL demod signals, and decided to use REFL165I&Q for the locking.
- Jiggled the demodulation phase of REFL165 and POP110. Changed the servo gains.
- Finally found a short lock. Further optimized the parameters.
- PRM ASC was turned on by giving the identity matrices for the input and output matrices.
Now just hitting the up button is sufficient to engage the ASC servo.
- Under the presence of the ASC, the PRMI is indefinitely locked as before.
- Reacquisition is also instantaneous. (It acquires even if the ASC is left "on".)
- Actually the lock is somewhat robust even when the PRM ASC is not used.
This is VERY GOOD as we can skip one of the steps necessary for the full lock.
Although, the seismic on Friday night is very quiet.
The spot motion at POP seems to be somewhat pitch/yaw mixed, in stead of previous "totally-dominated-by-yaw" situation.
- We are ready to implement ASS for PRM
Demod phase adjustment
- Shook PRM at 580Hz / 100cnt
- Swept the demod phase of REFL165 such that the PRM peak is minimized in the Q signal
- Open DTT. Measured transfer functions between REFL165I and the Q signals of each PD.
- Minimized the PRCL signal coupling in the signals.
- The resolution of the adjustment was ~1deg.
Locking test with PRM/BS
Tried the lock acquisition only with PRM and BS. (cf. http://nodus.ligo.caltech.edu:8080/40m/8816)
This just worked nicely.
Today's locking parameters:
MC Trans: 17500
POP110I (in lock): 150
PRCL Source: REFL165(I) 106deg / 45dB / Normalization SQRT(10 POP110I) / Input MTRX 1.0
PRCL Trigger: POP110I x 1.0 50up 25down
PRCL Servo: G=+3.5 Acq: FM4/FM5 Opr: FM2/FM3/FM6/FM7
PRCL Actuator: PRM +1.0
MICH Source: REFL165(Q) 106deg / 45dB / Normalization SQRT(0.1 POP110I) / Input MTRX 1.0
MICH Trigger: POP110I x 1.0 50up 25down
MICH Servo: G=-10 Acq: FM4/FM5 Opr: FM2/FM3/FM6
MICH Actuator: (ITMX -1.0 / ITMY +1.0) or (BS 0.5 / PRM -0.267)
In the past, we used to use Stefan's 'ezcademod' or Matt's 'ezlockin' to do auto phase adjustment.
JoeB / Jamie are working on python replacements for these tools, but in the near term possibly I can make a bash script to use ezcaservo and the existing LOCKINs to do this.
My goal for tonight was to lock PRMI,
grasp the current situation by my eye,
and capture some images using Sensoray.
They are done, but what are we going to do to solve the problem? The beam looks terrible than I had expected.
What I did:
1. DC output of POP55 PD was plugged out from 1Y2 rack, so we plugged it in.
2. Aligned POP beam to POP25 PD and moved POP camera position at ITMX table.
3. Mis-aligned PRM and SRM, aligned both arms, aligned FPMI as usual.
4. Mis-aligned PRM and ETMs, aligned MI and locked MI.
5. Aligned PRM, and carrier locked PRMI. PRM alignment was not saved since June 7, so slider values which give good alignment was pretty much drifted (~0.4 in C1:LSC_PRM_(PIT|YAW)_COMM).
6. Took some images of POP, REFL, AS during PRMI lock.
PRMI commissioning plan:
From the beam shape at POP, REFL, and AS, the problem clearly comes from the mode-matching, including clipping, longitudinal mismatch, and alignment mismatch. Koji's idea of flipped-PRM seems reasonable, so I think we should better measure something to prove this.
To prove this,
1. Simulate what the beam look like in POP, REFL, AS if PRM was flipped. Compare them with actual captured images. I need to study on unstable cavities.
2. Calculate power recycling gain and compare.
3. Misalign PRM and capture the image of primary, secondary, ... reflections like Koji did in elog #6421. Measure the beam sizes of these reflections using some image analysis(Python Imaging Library? Is there anyone good at this?) and calculate PRM curvature.
4. Can we do come characterization by making PRM-ITMY cavity? ITMX will be mis-aligned, BS will be the loss port to PRC.
5. Beamspot on POP, REFL, AS looks woblby when PRMI is locked. Why?
6. Open the vaccum chamber and see PRM. Simple.
Any other ideas? I have to lock PRFPMI, at least, by July 13!
To be fair, this is Kiwamu's idea. And nothing is reasonable before it is confirmed quantitatively.
Koji's idea of flipped-PRM seems reasonable, so I think we should better measure something to prove this.
Cycling the vacuum is easy. Why not vent starting Thursday evening and pop the doors on Friday morning? Inspect on Friday and pump on Monday morning.
Lots of work, no solid conclusions yet. In-vac, we aligned MICH and the PRM. Out of vac, we got beam on AS and REFL paths. We can lock MICH, but we're not as happy with PRCL.
To get the beam centered on TT2 in yaw, Koji helped us out and moved TT1 with the sliders a little bit. Then to get the beam centered on PRM and PR2, Koji moved the TT2 sliders a little bit.
Yuta and I then moved PR2 forward a few mm, to keep the optical path length of the PRC approximately (within ~1mm, hopefully) the same as always. After my PR2 optic swapping earlier, the pitch alignment was no longer good. I loosened the screws holding the wire clamp to the optic holder, and tapped it back and forth until the alignment was good. Of course, the screw-tightening / pitch-checking is a stochastic process, but eventually we got it. A small amount of yaw adjustment by twisting the PR2 TT was also done, but not much was needed.
Beam was a little off in pitch at ITMY, so Yuta poked the top of PR3, and that one single poke was perfect, and the beam was very nicely centered on the ITMY target. Beam was getting through BS target just fine. We checked at ITMX, and the beam looked pretty centered, although we didn't put in a target. We didn't do anything to BS while we were in-vac, since it was already good.
We aligned the ITMs so that their beams were retroreflecting back to the BS. After this, we saw nice MICH fringes.
We aligned the PRM so that its beam was retroreflecting.
We checked that we were getting REFL and AS beams out of the vacuum, which we were (a small amount of adjustment was done to AS path steering mirrors).
AS table alignment:
We did a bit of tweaking of the REFL path, and lots of small stuff to the AS path.
The AS beam was coming out of vac at a slightly different place in yaw, so we moved the first out of vac AS steering mirror so the beam hit the center, rather than ~1/3 of the way to the edge. We then aligned the beam through the lens, to the camera, and to AS55. Most significantly, we removed the BS that was just before AS55. This was sending beam to a dump, but it is in place to send beam over to AS110, once we get back to real locking. We measured ~30 microwatts of power going to the AS55 PD, while MICH fringes were fringing.
The REFL path didn't need much, although we had never been going through the center of the HWP and PBS that are used to reduce the power before going to the PDs, so we translated them a millimeter or two.
We see signal on dataviewer for all of the channels that we're interested in....AS55 I&Q, ASDC, REFL11 I&Q, REFLDC (which comes from REFL55).
Locking MICH was very easy, after we rotated the phase of AS55 to get all the good MICH signal in the Q phase. Part of the criteria for this was that the AS55_Q_ERR signal should cross zero when ASDC went to 0. This was done very coarsely, so we need to do it properly, but it was enough to get us locked. We changed the phase from 24.5 to 90 deg.
PRCL has been more of a challenge, although we're still working on it.
On the back face of the Faraday, we see the michelson fringes, but they are not getting through the Faraday's aperture. This implies that we have a poorly aligned michelson, in that the interference between the returning beams from the ITMs is happening at a different place than the original beam splitting. Yuta is working on getting a better MICH right now. EDIT, 10 minutes later.... This seems to be fixed, and the MICH fringes enter the back aperture of the FI, but there is still the PRM refl problem (next paragraph).
Also, when we get the most bright REFL beam, we see that there is some very obvious clipping in the back of the Faraday aperture, and this is matched by a clipped-looking REFL beam on the AS table. We must understand what we have done wrong, such that when the beam is actually going through the Faraday, we see a much dimmer beam. It's possible that there is some clipping happening at that time with the in-vac REFL path....we need to check this. It's not a clipping problem on the AS table - I checked, and the beams are still reasonably well centered on all of the mirrors.
We think that the MICH / REFL beam problems may be that the input pointing is close, but not perfect. We have not confirmed today that the beam is centered on ETMY. We should do this as part of our final alignment procedure before putting on doors.
Plans for tomorrow:
Get POP aligned, especially the camera, so we can see what our intracavity mode really looks like in the PRC. This is probably (in part, at least) due to our having moved PR2 around, so the transmitted beams aren't in exactly the same place.
We think that it's more useful in the short term to check out the PRC, and since the clipping problem with the REFL beam is likely an imperfect input pointing, we want to use the other measured G&H mirror, and do another half-PRC test, with the test mirror in front of the BS. This requires much less perfection in the input pointing, so it should be very quick to set up.
Confirm that PRM oplev is still aligned (turn laser back on first).
Plans for next week:
Perfect the input pointing, by checking the beam position at ETMY. Recheck all corner alignment.
Try again locking PRMI in air. First, confirm ITM and BS oplevs are all aligned.
I feel it's too hasty to use the PRMI.
I support the idea of the half-PRC test, to make an apple-to-apple comparison.
Make haste slowly.
I completely agree with Koji. We definitely should have locked the half PRC first. We were all set up for that. Why go through all this work to align MICH when we haven't confirmed with the half PRC that the flipping is helping us? The first rule of debugging is to only make one change at a time. We have measurements from the half PRC, so we could have made a direct comparison with those to see how things have changed. If we jump the gun we're going to end up wasting more time when we have to back-track.
Also, we never talked about moving PR2 to adjust optical path length, although I can understand why we would think that should be done. My calculations were all done assuming the free-space separation between PRM/PR2 and PR2/PR3 were unchanged. It's possible changing the position is better, but again, it's more work and it changes multiple things at one. I can redo my calculations for this new scenario, but we need to update our drawings with this new configuration. Please note precisely where PR2 has moved to.
We should have just flipped PR2 and that's it. Then we could have run the exact same measurements we had previously. Only then, once we understood this new simple cavity, should we have done further adjustments.
Half-PRC at this time already have two changes from the previous half-PRC; PR2 replaced/flipped and different TM before BS.
PRMI has only one change from the previous PRMI; PR2 replaced/flipped.
This is why I wanted to try PRMI first. But we now recognized that MI alignment (including REFL and AS alignment) is tough without using the arms, I agree that we should try half-PRC first.
I don't exactly know what the situation in the Jamie's calculation, but to make the optical path length the same before and after flipping, PR2 holder have to move about n*t, where n is the substrate refractive index and t is the thickness of the mirror, towards PRM/PR3.
The first rule of debugging is to only make one change at a time.
Also, we never talked about moving PR2 to adjust optical path length,
I completely agree with Koji. We definitely should have locked the half PRC first. We were all set up for that.
I reminded Jamie this morning that we were not, in fact, set up yesterday for a half PRC. I had extracted what was the flat test mirror, to put in as PR2. The test mirror was the better of the 2 G&Hs that we had measurements for, so I had used it as the flat test mirror, but then also wanted it to be the more permanent PR2. After doing the PR2 flip, the IFO was naturally all aligned for PRMI, which is part of why we just did that.
Anyhow, Jamie used his tallness to put the other measured G&H mirror into the mount, and put that in front of the BS. He aligned things such that he saw fringes in the half PRC.
I then aligned POP onto the camera, and onto the PD. Yuta is confirming that we're maximally on the REFL PDs.
We're starting locking in 5 min.
After much tweaking of the alignment using TT1, TT2 and PRM sliders, we were able to get a TEM00 mode locked with the half PRC!
PRCL gain is -0.010
FM4, 5 are always on. FM2,3,6 (boosts and stack res-gains) are triggered to come on after the cavity is locked.
We see a little clipping of POPDC, even though there are 2 BSs in the beam path, to dump 50% and then 67% of the beam. But it's not so much that we can't align.
REFLDC goes from 28.5 to 24.5, so we don't have great visibility.
Please watch our awesome video of the cavity, where we demonstrate that the half cavity is stable:
The cavity is flashing for the 1st 15 sec, then locks. Upper right is REFL, Lower right is POP, Upper left is back of the Faraday, Lower left is MC2F. Note that we definitely see some not so beautiful modes flashing, but most of that is due to the half cavity length and thus greater degeneracy of modes. Jamie is posting a HOM plot presently.
The beam is moving way more than it should be. Right now the PRM oplev is not coming out of the vacuum, since the flat test mirror mount is obstructing it. However, as we saw with other half-cavity tests, turning on the PRM oplev helps, but does not completely eliminate the beam motion. We should consider putting oplevs on one of the passive TTs, at least temporarily, so we know what kind of motion is coming from where.
It seems that the cavity trans looks much better than before. Cool.
At least the optical gain is ~x5 of the previous value. This means what we did was something good.
Looking forward to seeing the further analysis of the caivty...
I fixed up the POP path so that there is no clipping, so that Yuta can take a cavity mode scan.
[Gabriele's work, I'm just spectating]
Annalisa is working on finding the PSL/AUX laser beatnote, so the PSL temp is changing, but Gabriele is still able to lock. Here are some videos:
I got the PRMI transitioned from REFL165 over to REFL55 two times tonight. Also, I had 2 long-ish locks, one 9 minutes, and one 6 minutes. All the other locks were short - less than a minute or two.
I've done some shuffling around of the point in the CARM transition when the anti-boosts (1:20 filters) come on in the CARM A filter bank. I've moved the turn-on of these filters a several gain steps earlier, but I'm not sure that they're in the best place yet. Fiddling with the turn-on of the anti-boosts makes the big CARM oscillations last for longer or shorter - if they last too long, they blow the lock, so we don't want them to get too big.
The PRMI angular feedforward has helped a lot tonight, I think. I've added a line to the up script to enable the output of the OAF after the PRMI is locked, and the down script turns it off again. It's not so great when the PRM isn't aligned, since it's designed to work when the oplev is on, so it should be off unless the PRM is aligned. I tried to get a comparison of off vs. on PRC powers with the arms resonating, but I can't hold the lock for long enough when the OAF is not on to get even one average on my 0.01Hz bandwidth spectrum.
I've turned the arm ASC on a few times, but not every lock. Around 12:34am, I set the offsets when CARM and DARM were on RF signals, and I had hand-aligned the ETMs to minimize the power at the AS port. But, this wasn't a good spot for the next lock - the AS port was much darker with the ASC off for that lock. It would be nice to think about trying some dither alignment, and then maybe resetting the setpoints every lock. I'm using Q's original loop shapes, but as he left them yesterday, only actuating on the ETMs (with Yarm Yaw gain 0.7 rather than 0.9).
The CARM crossover might need more tuning. There's some gain peaking around 400 Hz that goes mostly away if I turn the digital CARM gain down by 2dB. (I'm not using any filters in the CM_SLOW filter bank).
I think that the CARM/DARM transition is more likely to be successful if the FSS slow DC is greater than 0.55ish. So far this is pretty anecdotal, but I think I have more success when it's higher. We should pay attention, and see if our trouble locking later in the nights correlates with smaller FSS slow DC values.
I got the PRMI over to 1f two times, at 1:54am and at 2:25am. I did not re-phase "POP"55 (which is the REFL55 signal), but I did check the values for the input matrix. I needed MICH = 0.01*POP55Q and PRCL = 0.008*POP55I. The first time I lost lock because I turned down the CARM digital gain too much. The second time I forgot to turn down the PRCL gain (I was *actually* using 0.01*POP55I for the PRCL input matrix, but needed to lower the gain from -0.08 to -0.07, which is about the same as just using 0.008 in the input matrix). Anyhow, I think PRCL loop oscillations were the cause of the second lockloss.
Here's a strip chart of my first lock of the night, which was the 9 minute lock. Up until about -6 minutes, I was hand-aligning (including the dip around -7.5 minutes, where I was figuring out which direction to move the ETMs). Around -3.5 minutes there is a significant dip down, that corrected itself. By the time I realized that the power had gone down, and was trying to figure out why, it came back. Maybe the same thing happened at the end of the lock, but it kept getting worse? Self, re-look at this time (around 11:50pm) to find out why the power dips.
My tummy feelings (without any data) make me think that this could be something with ITMX, like Q saw earlier today. Or, maybe ETMX, like we've seen for ages. Anyhow, my tummy feeling says this is an optic pointing problem. I certainly think this might be the same thing we see at the end of many locks, the power going low suddenly. So, it might give a big clue to our locklosses. Maybe.
RXA: I've changed the above text into pink Comic Sans to lend it the appropriate level of gravitas, given its scientific justification.
I have 12 tick marks for times I got all the way to 1f for all 4 degrees of freedom in the PRFPMI. The CARM / DARM transitions now succeed more than they fail, which is nice.
At Q's suggestion, I am turning off all the violin filters in the MC2 path during the CARM transition. This also means that I don't need any of the notches that Den and I put into the CARM_A and CARM_B filter banks last week, which were right at the edges of the violin notches. Anyhow, this seems to make the transition much more likely to succeed. I don't ever use the CM_SLOW FM10 "crossover helper" that Q had to use last night. The violin filters are turned back on after the CARM transition is complete. We don't ever need those other notches.
I checked the REFL165 vs. REFL55 transfer functions for PRCL and MICH, and they are mostly flat. REFL55 seems like it'll give us extra phase for some reason.
I tried setting offsets for PRMI, but they seem to be strongly dependent on arm alignment. I ended up being pretty confused, and since all the REFL signals are pretty close to zero (when CARM/DARM on RF, PRMI on 3f), I have given up on that avenue for tonight.
I think many of my locklosses tonight (lost from the all 1f state) have been fast things, faster than the ADCs can handle. On the lockloss plots that I've looked at, the FSS PC drive is railed at 10V about 200msec before I lose lock. So, something (presumably in the fast CARM path) is making the MC/FSS loop unhappy. I have plugged in the Agilent to the Out2 of the CM board, so that it looks at REFL11. Unfortunately, this is after the input gain slider, so we don't see much until we're locked, but that seems fine. A video camera is pointed at the screen, so that I get real time spectra. It's hard to watch the TV at the same time as everything else, so I haven't witnessed the moment of lockloss in the fast spectrum yet. Be careful when walking down the Yarm. The tripod is partly in the walkway.
Q, I took a few TFs of the total CARM loop, although none of them are particularly good below a few kHz. I can't push hard enough to get coherence, without blowing the lock. TF data is in /users/jenne/PRFPMI/CM_TFs/CM_TFs_2Apr2015/ .
I was worried for a while that, after I transition PRMI to 1f, I hear lots of low frequency rumbling. However, watching the spectra (relative to references taken with CARM and DARM on RF, but PRMI on 3f), the low frequency error and control signals are staying the same for all 4 DoFs, but the high frequency for PRCL and MICH goes down significantly, so it's probably just that the low frequency stuff sounds more obvious, since it's not drowning in high frequency fuzz.
We succeeded in locking PRMI in sideband and carrier.
Measured power recycling gain was ~60 power recycling gain was 4 (edited by YM on Feb 27; see elog #6947), but we have many things we cannot explain.
Here you are, Jamie.
PRMI sideband locked
PRMI carrier locked
I centered POP camera and put attenuator to take these snapshots.
Compare with previous ones.
Aug 17, 2012 elog #7213
Jun 28, 2012 elog #6886
Mar 15, 2012 elog #6421
== MI only ==
MICH: AS55_Q_ERR, AS55_PHASE_R = -12 deg, MICH_GAIN = -7, feedback to BS
== PRMI sideband ==
MICH: AS55_Q_ERR, AS55_PHASE_R = 24.5 deg, MICH_GAIN = -0.05 (acquisition) -> -5 (UGF ~100 Hz), feedback to BS
PRCL: REFL33_I_ERR, REFL33_PHASE_R = -22.65 deg, PRCL_GAIN = 4 (UGF ~120 Hz), feedback to PRM
== PRMI carrier ==
MICH: AS55_Q_ERR, AS55_PHASE_R = 24.5 deg, MICH_GAIN = -0.08 (couldn't measure UGF), feedback to BS
PRCL: REFL33_I_ERR, REFL33_PHASE_R = -22.65 deg, PRCL_GAIN = -0.3 (couldn't measure UGF), feedback to PRM
Power recycling gain:
POPDC was 32 when PRM is misaligned, 25 when PRMI sideband locked, ~2000 when PRMI carrier locked.
This means, power recycling gain is ~60 power recycling gain is ~4 (=POPlocked/POPmis*T_PRM=2000/30*0.06). Expected power recycling gain for PRMI is ~45, when there's no loss (see elog #6947).
I reduced POPDC PD gain so that it doesn't saturate.
- We optimized AS55_PHASE_R to -12 deg by looking at MI signal. But somehow, -12 deg didn't work for PRMI.
- Somehow, REFL11_I didn't work to lock PRCL.
- REFL11_Q didn't work to lock MI. SNR not enough?
- We saw POPDC flashing up to ~15000. What is this?
- Carrier lock was not stable, we couldn't hold for more than ~30 sec. It looks like PRM moves too much when PRMI is locked.
- Input pointing is drifting a lot in pitch. I had to re-align TT2/TT1 to the arms every ~1 hour to get good MI alignment. When I tweak TT2/TT1, both TRY and TRX gets better. I think this shows that input pointing is drifting, not the arms.
- redo PRM/BS coil balancing
- optimize REFL33 rotation phase
- stabilize carrier lock somehow
- measure PRC g-factor
It's hard to believe but is AS55Q really almost insensitive to MICH?
Well, anyway, now it is the time to use the automatic demod phase (and input matrix) adjustment.
I am also wondering if I understand / am using the demod phase from the screen correctly. This plot is indicating that MICH is entirely in I, and not at all in Q.
Currently, I take the demod phase, and plot that as the "I" line, then plot the "Q" line 90 degrees away from the I line. Maybe it should be the other way around?
Re: the auto-demod phase, I was starting to wonder about that. For each sensor, can I declare what degree of freedom I want in which quadrature to take priority (ex. MICH goes to REFL55 Q), and set the demod phase to the value that makes that true?
Okay, I think I am finished with the sensing matrix scripts!
I had the syntax for atan2() wrong, so I was calculating the demod phase wrong. Do not trust the phase in any previous elogs!!
Also, the theta=0 axis of the plots are for 0 degree demod phase, but our PDs are not at 0 deg. The measured sensing matrix phase is relative to the current demod phase, not 0 (unless the demod phase for that PD is currently 0degrees). So, now I take that into account. I add the current PD demod phase to the measured sensing matrix phase, so that the plot is actually true.
For interested parties, I have made all of the sensing matrix scripts, and the data folder a subdirectory of the /scripts/LSC folder, since it was starting to get crowded in there. I have moved the 2 data sets that have been collected (21May, 23May) into the new place.
* Save the amplitude and modulation frequency and the current demod phases in the data file. Right now the ampl and mod freqs are included in the title of the data file, but there is no record of what the demod phase was at the time. I need to fix this.
So, really, really, the Sensing Matrix:
Sensing Matrix, units = counts/meter, phase in degrees
PRCL Mag PRCL Phase MICH Mag MICH Phase
AS55 5.485E+08 -43.424 2.679E+09 93.392
REFL11 1.126E+13 -7.168 1.618E+11 135.296
REFL33 2.658E+11 164.973 8.910E+09 -2.226
REFL55 3.012E+11 -75.216 1.210E+10 172.764
Just so we have it, here is the re-analysis with the correct plot and phases for the May 21st data, taken near the violin mode:
Sensing Matrix, units = counts/meter, phase in degrees
PRCL Mag PRCL Phase MICH Mag MICH Phase
AS55 9.048E+11 -22.880 2.927E+12 107.071
REFL11 2.954E+16 -21.628 3.670E+14 123.857
REFL33 2.757E+14 -192.608 9.186E+12 -4.037
REFL55 2.868E+14 -82.690 1.186E+13 170.097
Even though this data was taken near the violin mode (oops!), it is fairly consistent with the stuff taken a few days later at 580Hz (elog 8644).
Neither of these is at all similar to what Kiwamu had measured a year ago (elog 6283), but we have changed many, many things since then. He also includes an Optickle simulation, which is fairly similar to the Koji simulation in the wiki, but neither his measurements nor mine are particularly close to the simulated version. I should think about why this is.
Also, I have fixed up the measurement scripts so that they record all of the relevant current settings / information: Current actuator calibration, current PD demod phases, drive amplitude and drive frequency. The "Analyze Saved Data" script has been updated to read all of this info from the files. If you want to plot / look at any old data, open up SensMatAnalyzeSavedData in /scripts/LSC/SensingMatrix/ and put in the relevant filename that you want (which should be saved in /scripts/LSC/SensingMatrix/SensMatData/)
- We want to add POX11/POY11 in the collection. They may indicates some abnormal asymmetry between two arms (for PRMI).
- We also want to PRCL/MICH after the input matrix. This will be useful when we want to adjust the input matrix to give the optimul
demod phase for the two signals from a single port.
I locked the PRMI and remeasured the sensing matrix, this time at 580Hz. The excellent news here is that the matrix looks quite similar to the one measured the other day, recorded in elog 8632. Yay! I'm not sure why the REFL11 MICH error is so much larger this time around.
Raw data: .../scripts/LSC/SensMatData/sensematPRM_2013-05-23.202312.dat
Sensing Matrix, units = cts/meter, phase in degrees
PRCL Mag PRCL Phase MICH Mag MICH Phase
AS55 5.485E+08 162.424 2.679E+09 25.608
REFL11 1.126E+13 -122.832 1.618E+11 85.296
REFL33 2.658E+11 -87.973 8.910E+09 79.226
REFL55 3.012E+11 -99.534 1.210E+10 12.486
Can you clarify the definition of the phase "0deg" of your plot?
Is "I" the definition of "0deg"? Or does the demod phase of "0deg" define "0deg"?
I want to know if the demod phase of REFL55 is correctly adjusted or not.
With the decent level of the separation, we should be able to keep the decent lock of PRMI with REFL55.
"0 degrees" is 0 degrees of demod phase. I have now added the PD demod phases to the plot:
Just so we have some numbers, I did a by-hand analysis of the PRMI sensing matrix numbers I posted here in the elog the other day. This analysis is ignoring the error bar data.
For each sensor (PD_I or PD_Q), I do loop compensation, since these measurements were taken fairly close to the UGFs of the loops, and notches were not in use at the drive frequency. To do the loop compensation, I multiply the complex value (lockin_I + i*lockin_Q) by (1-G), where G is the (complex) open loop gain of the degree of freedom I'm shaking.
When I'm shaking a single degree of freedom (ex. shaking the PRM to get PRCL information), for each PD_I or PD_Q, we get 2 numbers, the lockin_I and lockin_Q values. I check the phase between the lockin_I and lockin_Q values, since that phase (after loop compensation) should be either 0 or 180, and if it is not, something is wrong.
Of the 16 sensors I measure (where PD_I and PD_Q count as 2 sensors), 11 sensors had phases more than 20 degrees away from either 0 or 180. This is not good, and indicates that something is wrong with my measurement. I suspect that I may not be driving hard enough - I was using an amplitude 4x smaller than the previous value. Next time the PRMI is locked, I will turn on the drive oscillation, and ensure that I can see the line in all of the PD signals.
The results of my quickie analysis script:
Bad REFL11_I_MICH phase! Phase is -82.0185 degrees!
Bad REFL11_Q_MICH phase! Phase is -35.9697 degrees!
Bad REFL33_I_MICH phase! Phase is -134.952 degrees!
Bad REFL55_I_MICH phase! Phase is -79.7997 degrees!
Bad AS55_I_PRCL phase! Phase is -142.6016 degrees!
Bad AS55_Q_PRCL phase! Phase is 90.6194 degrees!
Bad REFL11_I_PRCL phase! Phase is 52.471 degrees!
Bad REFL11_Q_PRCL phase! Phase is 52.2324 degrees!
Bad REFL33_I_PRCL phase! Phase is 52.909 degrees!
Bad REFL33_Q_PRCL phase! Phase is 25.14 degrees!
Bad REFL55_I_PRCL phase! Phase is 52.8113 degrees!
Sensing Matrix, calculated even though most of the measurement data isn't any good:
AS55: MICH = 0.13502, -1.6122deg. PRCL = 0.14993, -2.245deg
REFL11: MICH = 29.6373, -2.6365deg. PRCL = 7376.3206, -2.9098deg
REFL33: MICH = 0.35649, -2.9633deg. PRCL = 69.5133, -3.1302deg
REFL55: MICH = 0.62084, -2.0261deg. PRCL = 75.0214, 3.1176deg
For now forget about the demodulation phase and assume all of the ports are independent.
I want to know the numbers in the following format.
PRCL MICH (unit: cnt/m)
REFL11I: x.xxxEx x.xxxEx
REFL11Q: x.xxxEx x.xxxEx
REFL33I: x.xxxEx x.xxxEx
REFL33Q: x.xxxEx x.xxxEx
REFL55I: x.xxxEx x.xxxEx
REFL55Q: x.xxxEx x.xxxEx
REFL165I: N/A N/A
REFL165Q: N/A N/A
AS55I: x.xxxEx x.xxxEx
AS55Q: x.xxxEx x.xxxEx
PRCL MICH (unit: cnt/m)
REFL11I: x.xxxEx x.xxxEx
REFL11Q: x.xxxEx x.xxxEx
REFL33I: x.xxxEx x.xxxEx
REFL33Q: x.xxxEx x.xxxEx
REFL55I: x.xxxEx x.xxxEx
REFL55Q: x.xxxEx x.xxxEx
REFL165I: N/A N/A
REFL165Q: N/A N/A
AS55I: x.xxxEx x.xxxEx
AS55Q: x.xxxEx x.xxxEx
If you really want to resolve the TF phase difference between the I and Q demod-signals,
you need to look at the transfer functions between the excitation and these ports.
We can't understand what is happening only from the single point measurement.
The PRMI sensing matrix, as measured last Thursday, in a more readable format:
EDIT: DON'T Look at this yet! I forgot to calibrate it! Please hold.....
AS55_I 1.228E-01 5.476E-03
AS55_Q 1.547E-01 1.345E-01
REFL11_I 9.946E+03 8.106E+01
REFL11_Q 2.346E+03 2.707E+01
REFL33_I 9.639E+01 8.717E-01
REFL33_Q 9.707E-01 6.626E-02
REFL55_I 1.040E+02 8.464E-01
REFL55_Q 2.018E+00 5.803E-01
Okay, Calibrated, but forgot to include loop compensation (since notches didn't exist yet):
Sensing Matrix, units = cts/meter
AS55_I 5.024E+08 9.418E+07
AS55_Q 6.328E+08 2.313E+09
REFL11_I 4.068E+13 1.394E+12
REFL11_Q 9.594E+12 4.656E+11
REFL33_I 3.942E+11 1.499E+10
REFL33_Q 3.970E+09 1.140E+09
REFL55_I 4.253E+11 1.456E+10
REFL55_Q 8.252E+09 9.981E+09
After all the work at the LSC rack over the last couple of days, I re-locked the PRMI (ETMs misaligned), and measured the sensing matrix once again. The PRMI was locked using 1f error signals, with AS55_Q as the MICH sensor and REFL11_I as the PRCL sensor. As shown in Attachment #1, the situation has not changed, there is still no separation between the DoFs in the REFL signals. I will measure the MC lock point offset using the error point dither technique today to see if there is something there.
[Gabriele, Rana, Jenne, Jamie, Lisa, Zach]
We tweaked some things after dinner, and our locks got longer (~10sec) and more frequent!
What happened / notes:
* Increased analog gain from 15dB to 27dB for REFL55 I&Q.
* No analog whitening during lock acquisition. (Need trigger + wait so whitening comes on after ~1sec...but this is not our limitation right now).
* Limit MICH and PRCL control to 5000, so that we don't kick optics too much, which makes them take too long to settle.
* ITMX and ITMY Vio2 filters turned off (PRM still has it on) in the SUS-optic_LSC module.
* MICH and PRCL DoFs triggering on POP22I, with levels 200 & 50. FMs 4&5 always on.
* MICH and PRCL FM2 triggering on POP22I with levels 400 & 50.
* MICH gain = -0.200
* PRCL gain = +0.150
* MICH and PRCL normalization using POP22I, with matrix values 0.00160 . This value is ~1/600, where 600 was the peak value of POP22I_ERR.
* REFL 55 phase set back to -15, to minimize PRCL signal in I phase.
* Checked signs for ITMX and ITMY in output matrix for MICH. Lock MICH using only ITMX or ITMY, find sign to hold on the dark fringe for each. +1 for ITMY, -1 for ITMX was correct.
* Tweaked up the oplev servos. See separate elog 8362. May need more tweaking, such as increasing the UGF, engaging 1Hz resonant gain.
* May need better coil actuator balancing on suspensions at 1Hz.
* Found a weird thing in DTT, which went away after closing and reopening, when looking at time series. Sometimes we would see a square wave-like jump in the signals, all signals at the same time, with a frequency of 16.6Hz. This was not present in other data retreival programs, like Jamie's getdata python script.
* We are not sure right now why we are falling out of lock. We need to investigate more signals, to try to figure out what our current problem is.
* Reduced the amount of misalignment with the "misalign" script, to reduce hysteresis.
To Do / ideas:
* Calibrate oplev signals - see if one optic is moving more than others.
* Calibrate ERR and CTRL - look at CTRL in meters, see if cavities are moving around like crazy.
* Calibrate POP22 using something like an AM laser modulation trick into units of PRCL SB gain. Compare with expectation - are we locked optimally, or do we have more power that we can be getting out?
* Try feeding back PRCL CTRL to MC2, to make the laser to follow the power recycling cavity, in hopes of reducing angular motion. Rana tried this quickly with a 1 in the output matrix, but this kicked the MC out of lock - need to try smaller values.
A key step was turning off the whitening filters. With the previous setting (G = 15 dB, white on), the error signals (post anti-whitening) had amplitudes of ~500 counts. This means that they can go as high as (150/15)^2 * 500 = 50000 counts on the ADC.
The purpose of the whitening filter is to match the noise / range of the signal to the ADC. What we would like to do is use the minimum gain so as to make the RFPD electronics noise + shot noise be ~equal to the ADC noise. i.e.
sqrt(V_PD^2 + v_shot^2) * G_white = V_ADC
The RFPD noise is ~3 nV before the internal preamp. The MAX4107 has a gain of 10. There is a factor of 1/2 from the voltage division of the RFPD's 50 Ohm series resistor and the input impedance of the mixer. There is also a power splitter between the PD output and the mixer which gives us a 3 dB loss. The mixer has a conversion loss of ~5-6 dB depending upon the LO level.
V_PD = 3e-9 * (10 * 1/2 * 1/sqrt(2) * 1/2) = 5e-9 V/rHz (this is already bad; the signal coming out of the mixer needs to be amplified by x10 before going out to the whitening board).
In any case, its clear that we need something like 60 dB of gain for the PD noise to match the ADC noise. This is why increasing the whitening gain improves the error signal's SNR, reduces the hash driving the optics, and improves the locking. We should run with 45 dB gain and the switch on whitening after the lock.
Even better would be to modify the LT1128 input stage of the card to have the single stage of fixed whitening as we did for iLIGO. Then we can have triple whitening in lock.
We have achieved PRMI locks of the order ~5 seconds! Here is an example lock:
This was actuating MICH on the ITMs (+1 for ITMY, -1 for ITMX in the output matrix), and PRCL on PRM (+1).
PRCL gain was +1, MICH gain was -10.
PRCL signal was normalized with POP22I with a matrix value of 0.003 . (No normalization of MICH).
Both PRCL and MICH were triggered on POP22I with high thresh of 200, low of 50. MICH and PRCL FM2 (integrators) were triggered on POP22I with thresh of 400 and low thresh of 50. FMs 4 and 5 were on for both MICH and PRCL always.
We zoomed in on the MICH_OUT signal, and the instability looks like it is around 300 Hz. We aren't sure what this is. I think this is a similar frequency to an oscillation that Yuta saw, but I'll have to check the old elogs.
PRM and BS SUS_LSC_POS filter banks both have notches between 1280-1290Hz. The ITMs do not have this "Vio2" filter.
This is one of the first locks with the new triggering of both the INPUT and OUTPUT of the control filter banks. I modified the lsc model before lunch.
To do: Where is this 300Hz coming from, and what can we do about it? Why are we losing lock? It's not due to the oscillation - maybe too much afternoon seismic? Steve says he went next door and the rock monster / river is on medium/high.
[Rana, Gabriele, Jenne, Jamie, Lisa, Rana]
We have tuned the oplev servos for PRM, BS, ITMX, ITMY. For each, we measured the servo transfer function. Most had a UGF ~ 3Hz. For those, we increased the gain by a factor of 2, and engaged the 3.3Hz resonant gains. The other case, such as PRM yaw, the gain was already okay, we just needed to engage the resonant gain. We also checked the new phase margin, and for some of them switched the elliptic lowpass to 50Hz rather than 30 or 35.
Before and afters:
We need to, as a last check, look at the spectra before and after to ensure that no modes (like bounce or roll) are newly excited.
Tonight PRMI was locked on REFL55 I&Q for PRCL and MICH with POP110I as a trigger and power normalizer.
I could see power fluctuations and beam motion on the POP camera very much the same as for carrier. The difference is that carrier stays for hours while sidebands for a few minutes.
I&Q analog gains were set to 15 dB. Relative phase was set to 25 degrees by looking at I and Q components when the cavity goes through the resonance. Q should be 0.
Phase rotation was measured by exciting PRM at 20 Hz and minimizing this line at REFL55_Q. I stopped at 33 degrees.
I compared power fluctuations of PRCL when it was locked on carrier (POP_DC) and on sidebands (POP110_I).
Time series of POP110_I during one of the locks
The first thing I looked at tonight was locking the PRMI on REFL 165.
I locked the PRMI (no arms), and checked the REFL 165 demod phase. I also found the input matrix configuration that allowed me to acquire PRMI lock directly on REFL165. After locking the arms on ALS, I tried to lock the PRMI with REFL 165 and failed. So, I rechecked the demod phase and the relative transfer functions between REFL 165 and REFL 33. The end of the story is that, even with the re-tuned demod phase for CARM offset of a few nanometers, I cannot acquire PRMI lock on REFL 165, nor can I transition from REFL 33 to REFL 165. We need to revisit this tomorrow.
For the PRMI-only case, I ended up using 0.1's in the input matrix, and I added an FM 1 to the MICH filter bank that is a flat gain of 2.2, and then I had it trigger along with FM2.
I turned this FM1 off (and no triggering) while trying to transition from REFL33 to REFL165 in the PRFPMI case, but that didn't help. I think that maybe I need to remeasure my transfer functions or something, because I could put values into the REFL165 columns of the input matrix while REFL33 was still 1's, but I couldn't remove (even if done slowly) the REFL33 matrix elements without losing lock of the PRMI. So, we need to get the input matrix elements correct.
I also recorded some time series for a quick RAM investigation that I will work on tomorrow.
I left the PRM aligned, but significantly misaligned both ITMs to get data at the REFL port of the RAM that we see. I also aligned the PRMI (no arms) and let it flash so that I can see the pk-pk size of our PDH signals. I need to remember to calibrate these from counts to meters.
Raw data is in /users/jenne/RAM/ .
I have not tried any new DARM signals, since PRMI wasn't working with 3f2.
We should get to that as soon as we fix the PRMI-3f2 situation.
[Jenne, Rana, Diego]
After deciding that the Yend QPD situation was not significant enough to prevent us from locking tonight, we got started. However, the PRMI would not acquire lock with the arms held off resonance.
This started some PRMI investigations.
With no arms, we can lock the PRMI with both REFL55 I&Q or REFL165 I&Q. We checked the demod phase for both Refl 55 and 165. REFL55 did not need changing, but REFL165 was off significantly (which probably contributed to the difficulty in using it to acquire lock). I didn't write down what REFL165 was, but it is now -3 degrees. To set the phase (this is also how Rana checked the 55 phase), I put in an oscillation using the sensing matrix oscillators. For both REFL165I and 165Q, I set the sensing matrix demod phases such that all of the signal was in the I phase (so I_I and Q_I, and basically zero in I_Q and Q_Q). Then, I set the main PD demod phase so that the REFL165Q phase (the Q_I phase) was about zero.
Here are the recipes for PRMI-only, REFL55 and REFL165:
Both cases, actuation was PRCL = 1*PRM and MICH = (0.5*BS - 0.2625*PRM). Trigger thresholds for DoFs and FMs were always POP22I, 10 up and 0.5 down.
REFL55, demod phase = 31deg.
MICH = 2*R55Q, gain = 2.4, trig FMs 2, 6, 8.
PRCL = 12*R55I, gain = -0.022, trig FMs 2,6,9.
REFL165, demod phase = -3deg.
MICH = -1*R165Q, gain = 2.4, trig FMs 2,6,8.
PRCL = 2.2*R165I, gain = -0.022, trig FMs 2,6,9.
These recipes assume Rana's new resonant gain filter for MICH's FM6, with only 2 resonant gains at 16 and 24 Hz instead of a whole mess of them: elog 10803. Also, we have turned down the waiting time between the MICH loop locking, and the filters coming on. It used to be a 5 second delay, but now is 2 sec. We have been using various delays for the PRCL filters, between 0.2s and 0.7s, with no particular preference in the end.
We compared the PRCL loop with both PDs, and note that the REFL 165 error signal has slightly more phase lag, although we do not yet know why. This means that if we only have a marginally stable PRCL loop for REFL55, we will not be stable with REFL165. Also, both loops have a non-1/f shape at a few hundred Hz. This bump is still there even if all filters except the acquisition ones (FM4,5 for both MICH and PRCL) are turned off, and all of the violin filters are turned off. I will try to model this to see where it comes from.
To Do list:
Go back to the QPDY situation during the daytime, to see if tapping various parts of the board makes the noise worse. Since it goes up to such high frequencies, it might not be just acoustic. Also, it's got to be in something common like the power or something, since we see the same spectra in all 4 quadrants.
The ASS needs to be re-tuned.
Rana was talking about perhaps opening up the ETMX chamber and wiggling the optic around in the wire. Apparently it's not too unusual for the wire to get a bit twisted underneath, which creates a set of places that the optic likes to go to.
Diego is going to give us some spectra of the MC error point at various levels of pockel's cell drive. Is it always the same frequencies that are popping up, or is it random?