40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  40m Log, Page 76 of 344  Not logged in ELOG logo
ID Date Authorup Type Category Subject
  10782   Thu Dec 11 16:42:12 2014 JenneUpdateElectronicsXend QPD whitening board plan

Here is a little PDF of what I plan to do to both of the transmission QPD whitening boards later today.  The idea is to take away the remote gain slider inputs, and force the gains to always be at +30dB.

The red and blue notes are from Koji's elog 9854, and the green are my plans for today. 

I will cut the traces from the gain slider inputs, and pull the negative input of the AD620 to ground.  The positive input will be connected to the +5 voltage line, with a divider so that the positive input to the AD620 is about 666mV. 

The AD602 will be maxed out at +30dB with anything over 625mV. 

QPDwhiteningModification_11Dec2014.pdf

Unless there are objections, I will start these modifications in an hour or so.  I will also make the Xarm whitening always-on, just like Koji has already done for the Yend.

EDIT, JCD, 12Dec2014:  These are not the modifications that were made.  Please see ____ for actual modifications.

  10788   Fri Dec 12 02:30:25 2014 JenneSummaryGeneralPSL table optical layout

Quote:

 

 I missed to elog this earlier. I have temporarily removed the DC photodiode for GTRY to install the fiber holder on the PSL table. So GTRY will not be seeing anything right now.

 

 After some confusion, I discovered this a few hours ago.

  10789   Fri Dec 12 04:33:49 2014 JenneUpdateASCASS retuned

[Rana, Jenne]

We decided that tonight was the night for ASS tuning. 

We started from choosing new frequencies, by looking at the transmission and the servo control signals spectra to find areas that weren't too full of peaks.  We chose to be above the OpLev UGF by at least a factor of ~2, so our lowest frequency is about 18Hz.  This way, even if the oplevs are retuned, or the gains are increased, the ASS should still function. 

We set the peak heights for the lowest frequency of each arm to have good SNR, and then calculated what the amplitude of the higher frequencies ought to be, such that the mirrors are moving about the same amount in all directions. 

We re-did the low pass filters, and eliminated the band pass filters in the demodulation part of the servo.  The band passes aren't strictly necessary, as long as you have adequate lowpassing, so we have turned them off, which gives us the freedom to change excitation frequencies at will.  We modified the lowpass filter so that we had more attenuation at 2Hz, since we spaced our excitation frequencies at least ~2.5 Hz apart.

The same lowpass filter is in every single demodulator filter bank (I's and Q's, for both length and transmission demodulation).  We are getting the gain hierarchy just by setting the servo gains appropriately. 

We ran ezcaservos to set the demodulation phase of each lockin, to minimize the Q-phase signal. 

We then tuned up the gains of the servos.  Rana did the Y arm, but for the X arm I tried to find the gains where the servos went unstable, and then reduced the gain by a factor of 2.  The Xarm is having trouble getting good alignment if you start with something less than about 0.7, so there is room for improvement.

Rana wrote a little shell script that will save the burt snapshot, if the gains need adjusting and they should be re-saved. 

The scripts have been modified (just with the new oscillator amplitudes - everything else is in the burt snapshots), so you should be able to run the start from nothing and the start from frozen scripts for both arms.  However, please watch them just in case, to make sure they don't run away.

  10794   Fri Dec 12 19:54:21 2014 JenneUpdateElectronicsXend QPD whitening board modified

Okay, I have finished modifying the Xend QPD whitening board, although I will likely need to change the gain on Monday.

Rather than following my plan in elog 10782, I removed the AD602's entirely, and just use the AD620's as the amplifiers.  We don't need remotely adjustable gains, and the AD620s are a less noisy part.

I set the gain to be 30dB using a 1.65k resistor for R_G, which turns out to be too high.  After I installed the board and realized that my counts were much higher than they used to be, I realized that what we had been calling +30dB was in fact +13.2dB. ( I am assuming that the ADC for the gain sliders were putting out a maximum of +10V.  The AD620 used to have a 1/10 voltage divider at the input, and an overall gain of 1, so the output of the AD620 was 100mV.  This goes into pin 16 of the AD602, which has gain of 32*V_set + 10.  Which gives 32*0.1+10=13.2dB.  Ooops.  We've been lying to ourselves. )

Anyhow, before I made the gain realization, I was happily going along, setting the AD620s' gains all to 30dB. I also copied Koji's modification from April of this year, and permanently enabled the whitening filters.

Here is the schematic of what ended up happening.  The red modifications were already in place, and the greens are what I did today.

QPDwhiteningModification_XtransCompleted_12Dec2014.pdf

You can see the "before" picture in my elog Wednesday, elog 10774.  Here is an "after" photo:

IMG_1779.JPG

Here is a spectrum comparing the dark noise of the Xend QPD after modification to the current Yend QPD (which is still using the AD602 as the main instrumentation amplifier).  I have given the Yend data an extra 16.8dB to make things match.

QPD_Xtrans_Ytrans_12Dec2014.pdf

And, here is a set of spectra comparing both ends, dark noise versus single arm lock.  While I'll have to sacrifice a lot of it, there's oodles more SNR in the Xend now.  The Yend data still has the "gain fixing" extra 16.8dB.

QPD_Xtrans_Ytrans_DarkVsLock_12Dec2014.pdf

The Xend quadrant input counts (before the de-whitening filters) now go up to peak values of about 1,000 at single arm lock.  If (optimistically) the we got full power recycling and the arms got to powers of 300, that would mean we would have 300,000 counts, which is obviously way more than we actually have ADC range for.  Currently, the Yend quadrant input counts go as high as 50, which with arm powers of 300 would give 15,000 counts.  I think I need to bring the Xend gain down to about the level of the Yend, so that we don't saturate at full arm powers.  I can't remember right now - are the ends 14-bit or 16-bit ADCs?  If they're 16-bit, then I can set the gain somewhere between the current X and Y values.

 Finally, I added a section of the 40m's DCC document tree for the QPD whitening:  E1400473, with a page for each end.  Xend = D1400414, Yend = D1400415.

  10799   Mon Dec 15 22:30:50 2014 JenneUpdateElectronicsYend QPD modified

Details later - empty entry for a reply.

Short story - Yend is now same as Xend filters-wise and lack of gain sliders -wise.  Both ends have 13.7k resistors around the AD620 to give them gains of ~4.5.

Xend seems fine.

Yend seems not fine.  Even the dark noise spectrum sees giganto peaks.  See Diego's elog 10801 for details on this investigation.

  10801   Mon Dec 15 22:45:59 2014 JenneUpdateElectronicsYend QPD modified

 

 [Jenne, Rana, Diego]

We did some test on the modified QPD board for the Yend; we saw some weird oscillations at high frequencies, so we went and check more closely directly within the rack. The oscillations disappear when the cable from the QPD is disconnected, so it seems something is happening within the board itself; however, looking closely at the board with an oscilloscope in several locations, with the QPD cable connected or disconnected, there is nothing strange and definitely nothing changing if the cable is connected or not. In the plots there are the usual channels we monitor, and the 64kHz original channels before they are downsampled.

Overall it doesn't seem being a huge factor, as the RMS shows at high frequencies, maybe it could be some random noise coming up, but anyway this will be investigated further in the future.

  10804   Tue Dec 16 03:43:09 2014 JenneUpdateLSCPRMI loops need help

[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.

PRMI_55vs165_15Dec2014.pdf

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?

  10810   Wed Dec 17 14:42:13 2014 JenneUpdateLSCPRMI loops need help

EricQ's crazy people filter has been deleted.  I'm trying to lock right now, to see if all is well in the world.

  10818   Fri Dec 19 15:59:49 2014 JenneUpdateLSCLockloss from Wed

I swapped out one of the channels on Q's lockloss plotter - we don't need POP22Q, but I do want the PC drive.  

So, we still need to look into why the PC drive goes crazy, and if it is related to the buildup in the arms or just something intrinsic in the current FSS setup, but it looks like that was the cause of the lockloss that Q and Diego had on Wednesday.

1102929772_PCdriveRailed.png

  10821   Fri Dec 19 18:08:46 2014 JenneUpdateCDSSOS!!! HELP!! EPICS freeze 45min+ so far!

[Jenne, Diego]

The EPICS freeze that we had noticed a few weeks ago (and several times since) has happened again, but this time it has not come back on its own.  It has been down for almost an hour so far. 

 So far, we have reset the Martian network's switch that is in the rack by the printer.  We have also power cycled the NAT router.  We have moved the NAT router from the old GC network switch to the new faster switch, and reset the Martian network's switch again after that.

We have reset the network switch that is in 1X6.

We have reset what we think is the DAQ network switch at the very top of 1X7.

So far, nothing is working.  EPICS is still frozen, we can't ping any computers from the control room, and new terminal windows won't give you the prompt (so perhaps we aren't able to mount the nfs, which is required for the bashrc).

We need help please!

  10824   Fri Dec 19 20:44:23 2014 JenneUpdateComputer Scripts / ProgramsFSS Slow servo moved to megatron

Today Q moved the FSS slow servo over to some init thing on megatron, and some time ago he did the same thing to the MC auto locker script.  It isn't working though.

Even though megatron was rebooted, neither script started up automatically.  As Diego mentioned in elog 10823, we ran sudo initctl start MCautolocker and sudo initctl start FSSslow, and the blinky lights for both of the scripts started.  However, that seems to be the only thing that the scripts are doing.  The MC auto locker is not detecting lockloses, and is not resetting things to allow the MC to relock.  The MC is happy to lock if I do it by hand though.  Similarly, the blinky light for the FSS is on, but the PSL temperature is moving a lot faster than normal.  I expect that it will hit one of the rails in under an hour or so. 

The MC autolocker and the FSS loop were both running earlier today, so maybe Q had some magic that he used when he started them up, that he didn't include in the elog instructions?

  10853   Mon Jan 5 18:15:18 2015 JenneUpdateCDSiscex reboot

Rana noted last week that TRX's value was stuck, not getting to the lsc from iscex.  I tried restarting the individual models scx, lsc and even scy (since scy had an extra red rfm light), to no avail.  I then did sudo shutdown -r now on iscex, and when it came back, the problem was gone.  Also, I then did a diag reset which cleared all of the unusual red rfm lights.

Things seem fine now, ready to lock all the things.

  10859   Tue Jan 6 17:41:20 2015 JenneConfigurationCDSDTT doesn't do envelopes??

[Jenne, Diego]

We are working on trying out the UGF servos, and wanted to take loop measurements with and without the servo to prove that it is working as expected.  However, it seems like new DTT is not following the envelopes that we are giving it. 

If we uncheck the "user" box, then it uses the amplitude that is given on the excitation tab.  But, if we check user and select envelope, the amplitude will always be whatever number is the first amplitude requested in the envelope.  If we change the first amplitude in the envelope, DTT will use that number for the new amplitude, so it is reading that file, but not doing the whole envelope thing correctly.

Thoughts?  Is this a bug in new DTT, or a pebkac issue?

  10860   Wed Jan 7 02:54:09 2015 JenneUpdateLSCFiddling with DARM filters

One of the things that we had talked about last night was the totally tiny amount of phase margin that we have in the CARM and DARM loops.  DARM seemed to be the most obnoxious loop last night, so I focused on that today, although the CARM and DARM loops are pretty much identical.

(Q tells me via email that the phase budget has the same ~14 degree discrepancy between what we expect and what we measure as his estimate last night.  However, the Caltech network issues prevented his posting an elog.)

So, we definitely need to figure out where this 14 deg is going, but for now, I wanted to see if I could recover a couple of extra degrees just by modifying the filters.

The original filters do seem to eat a lot of phase:

DARM_design_orig.pdf

The short version of the story is that I didn't leave the filters changed at all.  I reverted back to the last version of the filter file from Monday night, so currently everything is as it was.

I tried increasing the Q of the zeros on the cyan and brown filters, which would sacrifice some gain at ~20 Hz, but hopefully win us 10+ degrees of phase.  This gave me a dip of about a factor of 2 between the new and old filters (all servo filters combined added up to this factor of 2 in magnitude) between ~20Hz - 70Hz. 

When we were locked using DARM for just the Yarm (for the UGF servo commissioning), I took a spectra of the error signal (which was POY) as a reference, then loaded in my new filters.  For the most part, the spectra didn't change (which is good, since the magnitude of the filter didn't change much.).  The spectra was bigger though between 50-70Hz, in kind of a sharp bandpass-looking shape that I wasn't expecting.    I don't know exactly why that's happening.

Anyhow, we tried the new filters once or twice with the full IFO, but kept losing lock.  Since I clearly haven't put in enough thought yet for these (particularly, how much suppression do we really need? what are our requirements???), I reverted back to the filter file from last night.  We continued locking, and checking out the new UGF servo that Diego is elogging about.

  10862   Wed Jan 7 03:04:13 2015 JenneUpdateLSCTRY (thorlabs pd) weird noise

[Jenne, Diego, Rana]

This is a note about work done last night. 

We were starting to lock, and saw glitches in the Thorlabs TRY PD about once every 1/60th of a second.  It is not a sine wave, so it is not 60Hz line noise directly.  It looks like this:

TRY_60Hz_peaks_5Jan2015.pdf

Rana pointed out that this looks like it could be from a power supply that is converting AC to DC. 

We went down to the Yend, and noticed some weird symptoms.  So far, we do not know where the noise is coming from.  Rather, we are just using the QPD for locking.

* The noise comes and goes, particularly if someone is moving around at the end station.

* Moving the Thorlabs power supply farther from the HeNe power supply didn't do much.  Turning off and disconnecting the HeNe supply didn't make the noise go away, so we conclude that it is not the HeNe's fault.

* We suspected the loops of excess cable that were sitting on top of iscey, but moving the coils away from the computer did not make the noise go away.

* We removed a few disconnected BNC cables that were near or touching the end table, but that didn't fix things.

* We disconnected the PD's signal cable and pulled it out of the table enclosure, and then put it back.  Noise was gone when cable was disconnected (good), but it was back after plugging the cable back in.

* The noise still comes and goes, but we don't have to use the Thorlabs PD for locking, so we leave it for another day.

RXA: also moved the Thorlabs power supply to a different power strip and tried putting it closer/farther to the Uniblitz shutter controller. Another suspect is that its some PWM type noise from the doubler crystal temperature driver. Need to try turning off the heater and the Raspberry PI to if it effects the noise.

  10863   Wed Jan 7 03:09:15 2015 JenneUpdateLSCPRFPMI status & IFO status

As a warm-up after the holidays, before the real locking began, I installed 1064nm bandpass filters in front of the transmission QPDs to eliminate the stray green light that is there.

The Yend had threads epoxied to it, so that end should be good.  Steve is going to repeat that for the Xend QPD at some point.  Right now, the filter is just on a lens mount about 2cm away from the PD box aperture, since that's as close as I could get it.

Also, while I was at the Xend, I noticed that the transmission camera is gone.  I assume that it was in the way of Manasa's fiber work, and that it'll get put back somehow, sometime.  She elogged that she had removed it, but I mistakenly thought that it was already replaced.  We don't use that camera much, so I'm not worried.

  10869   Wed Jan 7 14:16:27 2015 JenneUpdateLSCtrans QPDs realigned

Now that both end transmission QPDs have the line filters, I aligned them.

I locked and aligned the IR using the ASS, then went to each end table and put the beam in the center of the QPD.

  10872   Wed Jan 7 15:53:01 2015 JenneUpdateLSCDC PD analog settings exposed

I have added another block to the LSC screen (and made the corresponding sub-screen) to expose the analog settings for the DC photodiodes. 

Note that we have 2 open channels there, which are still called something like "PD2" and "PD3" from olden times.

If we ever chose to use those, we will probably want to change their names, in /cvs/cds/caltech/target/c1iscaux2/LSC_aux2.db and /cvs/cds/caltech/target/c1iscaux/LSC_aux.db

  10876   Thu Jan 8 03:09:07 2015 JenneUpdateLSCToward variable finesse locking

[Jenne, EricQ, Rana]

Tonight we started prepping for an attempt at variable finesse locking. 

The idea is to put in a MICH offset and hold the lock with ASDC/POPDC (so that the offset can be larger than if we were just using RF signals).  This reduces the PRC buildup, which reduces / removes the double cavity resonance problems while reducing the CARM offset. 

  • So.  Today, I pulled out the POP22 razor blade so that we can use the Thorlabs PD as POPDC, without the yaw coupling.  Our other option is to use the POP QPD SUM, but that would require some model changes and more importantly it's not a particularly low noise readout path.
  • We re-set the analog whitening gains for ASDC and POPDC.
    • For ASDC, we want the half-fringe in the PRMI case to be not saturating.  We chose 18dB (it had been the default 0dB).
    • For POPDC, Rana and I saw that it was saturating all the time with the 33dB that it had when the carrier became resonant.  This was never really a problem in the past, but if we use it for normalization, we get glitches that knock us out of lock every time POPDC saturated.  So, now POPDC is at 0dB.  It still occasionally saturates when the PRMI is flashing, but we can't get lower than 0dB without going and putting an ND filter on the PD.
  • We turned off the analog whitening filters and digital unwhitening for both ASDC and POPDC.  We can consider turning them back on later after we have acquired lock if we need them, but we need them off for acquisition.
  • Locked MICH with ASDC/POPDC.  Good.  Stays locked even if PRM is flashing.
  • Locked PRMI with PRCL on REFL33I and MICH on ASDC/POPDC.
  • Locked arms, held off resonance with ALS, lock MICH with ASDC/POPDC. 
  • Failed to lock PRMI with arms held off resonance, using the new scheme (no transition, trying to directly acquire)
  • Locked PRMI on REFL33 I&Q with the arms held off resonance, and tried to transition MICH over to ASDC/POPDC, failed.
  • Confusion about the relative phase between REFL33Q and ASDC.  It looks like it is ~45deg at 100 Hz, or ~90 deg at 375 Hz.  Why isn't it 0 or 180?
  • Went back to PRMI-only, tried to map out fringe by changing MICH offset (tried while MICH was on both REFL33 and ASDC/POPDC).  Not really sure where we are on the fringe.

MICH locked on ASDC normalized by POPDC, PRM and ETMs (and SRM) all misaligned.

MICH offset of -20

MICH input = -0.04*ASDC normalized by 0.1*POPDC.

MICH gain = +5

MICH always triggered on (no triggering for DoF), but FM8 (CLP400) triggered to come on after lock (didn't write down the values).

 

PRMI locked with MICH on ASDC normalized by POPDC, PRCL on REFL33I, ETMs and SRM misaligned.

MICH offset of -10

MICH input = -0.04*ASDC normalized by 0.1*POPDC.

PRCL input = 1*REFL33I

MICH gain = +5

PRCL gain = -0.4 (factor ten times the regular value)

MICH always on, PRCL triggered on POP22.  MICH FM8 and PRCL FM1,2,6,9 triggered on.

Gives POPDC of about 20 counts, POP22 of about 12 counts, ASDC of about 500 counts.

 

Arms held at 3nm, MICH locked on ASDC/POPDC, PRM and SRM misaligned.

MICH offset of -10

MICH input = -0.04*ASDC normalized by 0.1*POPDC.

MICH gain = +5

MICH always on, PRCL triggered on POP22.  MICH FM8 and PRCL FM1,2,6,9 triggered on.

 

Arms held at 3nm, attempt at PRMI lock with MICH on ASDC/POPDC.

Failed.  Tried mostly same MICH gains as arms+mich, and PRCL at 10* normal gain.

 

Arms held at 3nm, PRMI locked with REFL 33 I&Q, attempt at transition to MICH on ASDC/POPDC.

Failed.  At first, I was putting in the TF line at ~375Hz, but we looked at the full transfer function between 100Hz and 1kHz, and there was a weird dip near 300Hz from PRCL-MICH loop coupling.  Here we were seeing that the phase between REFL33Q and ASDC was ~90 degrees.  What?

Tried putting the TF line at ~100 Hz (since MICH UGF is in the few tens of Hz anyway, so 100 is still above that), but still get weird relative phase.  Here it seems to be about 45 degrees when I inject a single line, although it didn't seem like a weird phase when we did the full swept sine earlier.  Maybe I was just not doing something right at that point??

Anyhow, no matter what values I tried to put into the input matrix (starting with REFL33I&Q, trying to get MICH to ASDC/POPDC), I kept losing lock.  This included trying to ramp up the MICH offset simultaneously with the matrix changing, which was meant to help with the PRCL gain change.  Q has since given us MICH and PRCL UGF servos.

 


Tomorrow:

  • Why is there some weirdo phase between REFL33Q and ASDC at 100Hz?  Was I just being a spaz?
  • With PRMI-only, figure out how to transition from REFL33 I&Q over to MICH on ASDC/POPDC.
  • Then hold the arms off resonance, and do the same transition.  (First make sure we're at a good place on the fringe)
  • Lower the CARM and DARM offsets, transition them to RF, engage CARM AO path.
  • Reduce MICH offset, transition to RF.
  • Celebrate (maybe).
  10885   Fri Jan 9 19:18:51 2015 JenneUpdatePSLPMC realigned

A few hours ago I tweaked up the alignment to the PMC.  It was really bad in pitch, and the transmission was down to about 0.711.

  10887   Tue Jan 13 00:42:15 2015 JenneUpdateLSCError signals for MICH with variable finesse technique

In order to know where we should try to make the transition from REFL##Q to ASDC for MICH, I did a quick Optickle simulation to see what the error signals will look like.

The idea is to try to lock the PRMI on a single REFL diode (ex. REFL33 I&Q) with some MICH offset, and then transition over to ASDC.  As soon as we have completed the transition, we can engage the normalization matrix to normalize ASDC by POPDC, and also increase the MICH offset if we want.  Unfortunately, we do not as yet have the ability in our model to independently normalize different error signals, and then blend them, so we have to turn on the normalization after we've transitioned.

Here is the situation for PRMI-only:

You can see that REFL33Q has a slightly wider range than REFL165Q.  It seems like we can perhaps try to make the transition around -15nm or so.  Note that the error signals are not quite symmetric about 0nm, so we can use that to help determine what + and - mean.  We expect that we need to add about 1nm offset to REFL33Q to get a true minimum in ASDC, so the sign of the digital offset that we need will tell us if there is a sign flip or not between the digital offset and this x-axis. 

After we get this to work (hopefully in the next hour or so....), we will want to try the same thing with the arms held off resonance. 

Usually we lock the PRMI at an offset of about 3nm:

However we could do it lower, perhaps around 1nm (which is where we currently are doing our CARM/DARM ALS->IR signals transitions):

At some point, we will arrive at 0nm CARM offset, when we'll want to transition back to RF signals (although probably we could jump straight to a 1f signal, not plotted):

The moral of the story here is that I'm not sure how we were ever successfully locking MICH on REFL165Q, unless my phase-setting in Optickle is way off.  Certainly it looks like we should be sticking with REFL33 for PRFPMI.  Also, since we have an offset in REFL33Q anyway (which we have seen and have commented on before), at 3nm CARM offset it looks like we could try to just do the jump without any extra digital offset.  Here's a zoom of the 3nm situation:

  10891   Tue Jan 13 03:58:27 2015 JenneUpdateLSCTransitioned to ASDC MICH (PRMI and PRFPMI)

[Jenne, Diego, EricQ]

Hopefully there will be more later, but Chiara just went down (network? other?  Q is in there right now looking at it), so this is a so-far-tonight elog.

We have successfully transitioned MICH over from REFL33Q to ASDC in both the PRMI and PRFPMI configurations.  Next up is to start reducing the CARM offset.


Resetting the REFL demod phases

I have been unable to lock the PRMI for more than teeny blips since Thursday.  So, tonight I finally got it to lock with MICH on AS55Q and PRCL on REFL33I, and used that to set the demod phases. 

PRMI, AS55Q and REFL33I MICH PRCL
Input matrix 1*AS55Q 1*R33I
Gain -7 -0.022
DoF trigger POP22I; 80up, 2down POP22I; 80up, 2down
FM trigger FM 2,6,8; 100up, 2down; 0.3sec FM 2,6,9; 100up, 2down; 0.01sec
Normalization n/a n/a
Output matrix 0.5*BS, -0.2625*PRM 1*PRM
UGF servo n/a

n/a

Setting the demod phases, I used an oscillation of 100 cts to PRM, at 400.123 Hz.

REFL 33 demod phase started at 148deg, now 133.2deg.

REFL165 phase started at -105.53, now -172.

No signal in REFL55????  Time series and spectra both look just like noise.  Need to check alignment of beam on PD, or if cables unplugged!!

REFL11 phase started at 16.75, now 18.9deg.

Was then able to lock on REFL 33 I&Q, like normal.

PRMI, REFL33I&Q MICH PRCL
Input matrix 1*R33Q 1*R33I
Gain 2.5 -0.02
DoF trigger POP22I; 80up, 2down POP22I; 80up, 2down
FMs FM 4,5 on FM 4,5 on
FM trigger FM 2,6,8; 100up, 2down; 0.3sec FM 2,6,9; 100up, 2down; 0.01sec
Normalization n/a n/a
Output matrix 0.5*BS, -0.2625*PRM 1*PRM
UGF servo n/a

n/a



Transitioning PRMI from REFL33Q to ASDC

With the PRMI locked on REFL33 I&Q, I found that a MICH offset of -5 counts gives a minimum in ASDC.  From my earlier elog this evening (http://nodus.ligo.caltech.edu:8080/40m/10887), I expect the minimum to be at +1.4nm.  This is only one point though, so I don't know the calibration of the MICH offset yet (we should get this calib during the day by looking at MICH-only).  Anyhow, this informed which side was positive and negative relative to my Optickle plots, so I know that I wanted positive offset in the MICH servo.

I was able to comfortably hold lock at +20 counts.  Looking at a calibration line at 143.125 Hz, I determined that I wanted the matrix element for ASDC to be -0.05.  After I made that transition using ezcastep, I put the POPDC normalization in.  At the time, POPDC was about 151counts, so I put 1/151 in the POPDC->Mich matrix element.

So, here were the final lock parameters.  Note that in PRMI-only, you can acquire lock like this, and with a variety of MICH offsets:

PRMI, ASDC and REFL33I MICH PRCL
Input matrix -0.05 * ASDC 1*R33I
Gain 2.5 -0.02
DoF trigger POP22I; 80up, 2down POP22I; 80up, 2down
FMs FM 4,5 on FM 4,5 on
FM trigger FM 2,6,8; 100up, 2down; 0.3sec FM 2,6,9; 100up, 2down; 0.01sec
Normalization  0.0066 n/a
Output matrix 0.5*BS, -0.2625*PRM 1*PRM
UGF servo n/a

n/a


Locking PRMI part of PRFPMI

Since the PRMI has been fussy, I'm including a brief note on the PRMI settings when the arms are held with ALS off by roughly 3nm.  To get to this point, we just ran the usual carm_cm_up script, and didn't let it run anymore when it asked for confirmation that PRMI was locked.

PRFPMI, Arms=ALS, PRMI=REFL33I&Q MICH PRCL CARM DARM
Input matrix 1*R33Q 1*R33I -1*alsX+1*alsY 1*alsX+1*alsY
Gain 2.5 -0.012 8 8
DoF trigger POP22I; 80up, 2down POP22I; 80up, 2down n/a n/a
FMs FM 4,5 on FM 4,5 on FM 1,2,3,5,6 FM 1,2,3,5,6
FM trigger FM 2,6,8; 50up, 2down; 0.3sec FM 1,2,6,9; 50up, 2down; 0.01sec n/a n/a
Normalization n/a n/a n/a n/a
Output matrix 0.5*BS, -0.2625*PRM 1*PRM 1*etmx+1*etmy -1*etmx+1*etmy
UGF servo n/a set to 150Hz n/a n/a

With MICH offset of -30 counts, AS port is pretty bright.  ASDC dark offset is set to -475.4 by the LSCoffsets script. with MICH offset = 0, ASDC_OUT is around 300counts.  But, with MICH offset = -30, ASDC_OUT is about 525 counts.  So, I put that 525 counts into the ASDC filterbank offset (so it is now the dark offset + this extra offset), so the ASDC offset is currently around -1,000.  This makes the ASDC signal roughly zero when I am ready to transition MICH over to it.  In principle I should probably set it so the average is the same as the MICH offset, but the noise is so high relative to that offset, that it doesn't matter.

After this, we engaged the CARM and DARM UGF servos.  MICH was gain peaking, so I think we might want to turn that one on too, rather than my by-hand turning down the gain.

The transition has been successful 4 or 5 times with the arms held off resonance at 3nm.  Once, we reduced the CARM offset as low as 1.7 (and had to lower the MICH gain to 1.5), but we were still hearing a woomp-woomp sound.  Not sure what that was from.  At this point, Chiara died, so we lost lock.  After that, we re-acquired lock a few more times, but MC keeps losing it.  We are still able to make the MICH to ASDC transition though, which is good.

The transition won't work if the PRCL UGF servo is not on.  The gain multiplier goes from about 1.1 up to 2.4, so the PRCL gain is certainly changing through the transition.

Diego has written up scripts for the individual UGF servos (look for an elog from him separately), so now the carm_cm_up script goes as far as locking the PRMI on REFL33 I&Q, and then it starts to transition.  PRCL UGF is engaged, MICH offset is set to -30 counts, MICH is transitioned to ASDC, POP normalization engaged, CARM UGF servo turned on, and DARM UGF servo turned on.  There are "read"s in the script before each step, so you can stop whenever you like.

Here's the final configuration for the PRFPMI while the arms are held at 3nm, with MICH on ASDC (so after the transition):

PRFPMI, Arms=ALS, PRCL=REFL33I, MICH=ASDC MICH PRCL CARM DARM
Input matrix 0.27*ASDC 1*R33I -1*alsX+1*alsY 1*alsX+1*alsY
Gain 2.5 -0.015 8 8
DoF trigger POP22I; 80up, 2down POP22I; 80up, 2down n/a n/a
FMs FM 4,5 on FM 4,5 on FM 1,2,3,5,6 FM 1,2,3,5,6
FM trigger FM 2,6,8; 50up, 2down; 0.3sec FM 1,2,6,8,9; 50up, 2down; 0.01sec n/a n/a
Normalization 0.0042*POPDC n/a n/a n/a
Output matrix 0.5*BS, -0.2625*PRM 1*PRM 1*etmx+1*etmy -1*etmx+1*etmy
UGF servo n/a set to 150.001Hz set to 115.1Hz set to 110.1

 

The transition for MICH to ASDC has been successful with the arms held off resonance several times tonight. It's all part of the carm_cm_up script now. I think that if we hadn't lost about an hour of time and our momentum, we would have gotten farther.  I have high hopes for tomorrow night!

  10894   Tue Jan 13 13:22:54 2015 JenneUpdateGeneralAUX Y + PSL beat note at 1064nm: needs work

I'm super excited about this new frequency readbacklaugh, but I'm not sure that it's reliable yetfrown.  Without touching any settings, the readback is currently saying 78.6MHz, and is changing slightly (as is the FSS Slow Temp), so something is working.  However, the beatnote as measured on the spectrum analyzer is 158.2MHz.  So, either the calibration or the tracking or something isn't quite finished being tuned yet.cheeky

It's going to be super awesome when we have this though!!cool

  10900   Wed Jan 14 03:42:31 2015 JenneUpdateLSCThoughts on going forward with variable finesse

[Jenne, Rana]

We tried locking with the variable finesse MICH offset technique again today. 

A daytime task tomorrow will be to figure out where we are in MICH and CARM offset spaces.  This will require some thinking, and perhaps some modelling.

We were using the UGF servos and checking out their step resonses, and had the realization that we don't want the gain multiplication to happen before the offsets are applied, in the case of MICH and CARM.  Otherwise, as the UGF servo adjusts the gain, the offset is changed.  I think this is what ChrisW and I saw earlier on in the evening, when it seemed like the CARM offset spontaneously zoomed toward zero even though I didn't think I was touching any buttons or parameters.  Anyhow, we no longer used the MICH and CARM UGF servos for the rest of the night.  We need to think about where we want the offset to happen, and where we want the UGF servo multiplication to happen (maybe at the control point, with a very low bandwidth?) such that this is not an issue.

Also, I'm no longer sure that the sqrt(I^2 + Q^2) instead of the usual demodulation is going to work for the UGF servos (Q made this change the other day, after we had talked about it).  When the numbers going into the I and Q servo banks are small (around 1e-5), the total UGF servo gets the answer wrong by a factor of 10 or so.  If I made the "sin gain" and "cos gain" 1000 instead of the usual 1, the numbers were of the order 1e-2, and the servo worked like normal.  So, I think we were perhaps running into some kind of numerical error somehow.  We first noticed this when we lowered the DARM excitation by a factor of 10, and the servo no longer functioned.  We should take out this non-linear math and go back to linear math tomorrow.

During the evening tomorrow, we should try locking the PRMI with a large MICH offset, and then leaving CARM and DARM on ALS, and seeing how far we can get.  Is it possible to just jump over to RF signals, since we won't have to worry about the detuned cavity pole?

Tonight, the locking procedure was the same as usual, but stopping the carm_up script before it starts to lower the CARM offset at all.  Only difference was that MICH triggered FMs were 2,3,7 rather than the usual 2,6,8. 

So, assuming you have the IFO with CARM and DARM on ALS held at +3 CARM offset counts (which we think is about 3nm), and the PRMI is locked on REFL33I&Q with no offsets, here's what we did:

  • PRCL UGF servo on
  • MICH offset goes to -20
  • MICH transitions to ASDC (0.27*ASDC, then normalize by POPDC)
  • DARM UGF servo on
  • CARM offset to 1 (arms about 0.25)
  • CARM transition to SqrtInvTrans
  • Lower CARM gain to 4
  • CARM offset to 0.6 (arms about 1)
  • DARM transition to DC transmission
  • Increase MICH offset to about -650 or -670
  • Lower CARM offset, see what happens

Something else to think about:  Should we normalize our DC transmission signals by POPDC, so that the arm powers will change when we change the MICH offset (for a constant CARM offset)?

The best we got was holding for a few minutes at arm powers of 7.5, but since the MICH offset was large and the power recycling was low, this was perhaps pretty far.  This is why we need some calibration action.

Also, earlier today I copied the CARM and DARM "slide" filter module screens so that we have the same thing for MICH.  Now all 3 of these degrees of freedom have slider versions of the filter module screens, which are called from the ctrl_compact screen.

  10903   Thu Jan 15 04:41:01 2015 JenneUpdateLSCThoughts on going forward with variable finesse

[Jenne, Diego]

Life would be easier with the UGF servos working.  As Diego already elogged, we aren't sure why the demod phases are changing, but that is certainly causing the I-signals to dip below zero, which the log function can't handle (there is a limiter before the log, so that the signal can't go below 1e-3).  Anyhow, this is causing the UGF servos to freak out, so we have not been using them for tonight's locking.

Our goal tonight was to see if we could introduce a nice big MICH offset, and then lower the CARM offset while keeping the arms locked on ALS.  We hope to see some kind of sign of a PDH signal in some RF PD. 

In the end, the highest we got to was -460 MICH offset counts, which we think is about 29nm (if our rough calibration is accurate). The MICH half fringe should be 188nm. With this offset, we got down to 0.3 CARM offset counts while locked on ALS.  We think that this is around 300pm, plus or minus a lot.  Note that while yesterday I had a pretty easy time getting to -660 counts of MICH offset, tonight I struggled to get past -200.  The only way we ended up getting farther was by lowering the CARM offset.  Although, as I type this, I realized that last night's work already had a lower CARM offset, so maybe that's key to being able to increase the MICH offset. 

We watched REFL11I and REFL11I/(TRX+TRY) on striptool, but we didn't see any evidence of a PDH signal.  We lost lock when I tried to transition CARM over to REFLDC, but I wasn't careful about my offset-setting, so I am not convinced that REFLDC is hopeless.

So.  Tonight, we didn't make any major locking progress (the MC started being fussy for about an hour, right after I ran the LSC offsets script, just before we started locking in earnest).  However, we have some ideas from talking with Rana about directions to go:

* Can we transition CARM over to REFL11I, and then engage the AO path?

* Then, while the MICH offset is still large, can we transition DARM over to POX or POY, actuating on a single arm?  If CARM is totally suppressed, this is DARM-y.  If CARM doesn't have the AO path yet, this is halfsy-halfsy, but maybe we don't care.

* Then, can we lower the MICH offset and transition back to a REFLQ signal?

* Separately, it seemed like we kept losing PRC lock due to PRC motion.  If the MICH offset is very large, are we sideband-limited at the POP port, such that we can use the POP DC QPD?  Is it even worth it?

 

MICH calibration:

A single mirror (ITM) moving by lambda/2, in the MICH-only situation is the full range, from bright to dark fringe.  So, half fringe should be lambda/4, or about 133nm.  If we are thinking about pushing on the BS, there's an extra factor of sqrt(2), so I think the half fringe should be at 188nm of BS motion.

When we had MICH locked on ASDC/POPDC, we put in a line at 143.125Hz, at 20 counts to (0.5*BS-0.2625*PRM), so a total of 10 counts to the BS at 143Hz.  Given the BS calibration in http://nodus.ligo.caltech.edu:8080/40m/8242, this is 10.1pm of actuation.  We saw a line in the error signal of 0.1 counts, so we infer that the MICH error signal of ASDC/POPDC has a calibration of 94pm/count. This number was invariant over a few different MICH offsets, although the ones I measured at were all below about 100 counts of MICH offset, so maybe this number changes as we start to get farther from the MICH dark fringe.

 

IFO left flashing (all mirrors aligned except SRM) in case anyone wants fresh data for that.

  10904   Thu Jan 15 14:28:14 2015 JenneUpdateLSCLSC model change idea

Something that kind of drives me crazy with our current LSC model setup is that I can't make "finished" error signals before blending them.  The blending happens before the normalization matrix, and there is no place to put an offset to help match a new error signal to the current offset.  So.  While I'm sure this is not going to be immediately popular, here's a cartoon of a proposed model change to the LSC. 

The most important difference between this and the ramping matrix that is used at the sites is that you can put in offsets before the blend.  Also useful is the fact that the normalization can happen before the blend.  This proposal would make the LSC input matrix  and the normalization matrix have twice as many rows, and add an extra "selector matrix" just before the triggering at the error point of the loops. 

I've only drawn one degree of freedom in my cartoon, but assume that they all have the same capability (maybe we don't have to do XARM, YARM and MC this way).  One row is currently being used for the error signal, while the other row is just used to prep a new singal.  For a first transition (say, ALS to DC transmission), maybe the ALS signals are on row 1, and the DC trans is on row 2.  Once the transition is complete, row 1 is available to prep for the next transition (such as AS55Q).

 

Thoughts?  Is there a better way to achieve what I'm going for here?

  10905   Thu Jan 15 18:06:34 2015 JenneUpdateComputer Scripts / ProgramsInstalled kerberos on Rossa

I have installed kerberos on Rossa, so that I don't have to type my name and password every time I do an svn checkin, since I'm making some modifications and want to be sure that everything is checked in before and afterwards. 

I ran sudo apt-get install krb5-user.  I didn't put in a default_realm when it prompted me to during install, so I went into the /etc/krb5.conf file and changed the default_realm line to read default_realm = LIGO.ORG

Now we can use kinit, but we must (as usual) remember to kdestroy our credentials when we're done.

As a reminder, to use:

> kinit albert.einstein

Password for albert.einstein@LIGO.ORG: (type your pw here)

When you're finished, run

> kdestroy

The end.

  10907   Thu Jan 15 18:30:18 2015 JenneUpdateComputer Scripts / ProgramsInstalled kerberos on Rossa

 

Quote:

WARNING: since the workstations are all shared user, if you forget to kdestroy the next user can commit under your user ID.  It might be good to set the timeout to be something much shorter than 24 hours, like maybe 1, or 2.

Good call.  I added a line ticket_lifetime = 3600, which should make it destroy the credentials after an hour.

  10910   Fri Jan 16 03:31:35 2015 JenneUpdateLSCLSC model change implemented

Okay, it has taken me almost exactly 12 hours (with a dinner break), but I have implemented this change.

Everything was svn-ed before I did things, and then again afterward.

Here is the "before" screenshot of the LSC model:

And here is afterward:

If you look extra carefully, you will see that it matches my proposal from http://nodus.ligo.caltech.edu:8080/40m/10904 .  I have made the change for DARM, MICH, PRCL, SRCL and CARM.  I did not alter XARM, YARM or MC.  Also, the CESAR stuff was taken out of the CARM area, since this is now a more generalized version of the same thing.


I have also checked and modified all of the scripts that I could think of, as well as all of the ifoconfig burt .req and .snap files that I could think of.  I also ran the carm_cm_up.sh script once, and it seems to work fine.  All of the transition scripts that are listed below (which are the only ones used currently in the sequence) now use the new error signal blending scheme.

Scripts:

  • Lock_ALS_CARMandDARM.py
  • ALSfindIRresonance.py
  • ALSwatch.py
  • carm_cm_down.sh
  • carm_cm_up.sh
  • CheckPRMIlock.py
  • Transition_MICH_REFL33Q_to_ASDC.py
  • Transition_CARM_ALS_to_TransInvSqrt.py
  • Transition_DARM_ALS_to_DCtrans.py
  • UGFup.py
  • UGFdown.py

Burts (listed are the .req files, but I also checked the .snap files, hand-editing the matrix element numbers where needed if I wasn't in the right config to do a save):

  • C1configure_Yarm.req
  • C1configure_YarmALS.req
  • C1configure_Xarm.req
  • C1configure_XarmALS.req
  • C1configure_CARM.req
  • C1configure_DARM.req
  • C1configure_PRM_forCARMdarm.req
  • C1configure_PRM_SBres.req
  • C1configure_PRM_Carr.req
  • C1configure_PRY.req
  • C1configure_DRM.req
  • C1configure_SRM.req

I also modified the screens for the input matrix and for the normalization matrix.  I created a new screen for the final blending matrices (which are all 2x1's), and I also modified the LSC_OVERVIEW screen. 

The input matrix and normalization matrix screens have colored bars that tell you whether a row is in use or not.  If the background to the row is the blue of the whole screen, that row is not being used.

The LSC screen has new hand-modified Kissel Buttons.  I wanted to show the total PD error signal that is being used, regardless of what row (A or B) it is on at that time.  So, I have collapsed the rows so that DARM_A and DARM_B are overlapped, even though they are actually rows 1 and 2 of the matrix.  The PD should only show up green on the LSC screen if that row is in use (so, if you are prepping a row, but aren't using it yet, you won't see those elements in the matrix).  Anyhow, the point is that the LSC overview part of things shouldn't look any different than before.


Brain not working anymore now that it's ~4am, but I need to rethink and recheck to make sure that the PD whitening triggering is still okay and working.  Or maybe we can remove it, and just include that in the scripts, as Rana has been suggesting for ages.  Thoughts for tomorrow.

  10913   Fri Jan 16 18:09:09 2015 JenneUpdatePSLPMC autolocker not running?

Have we been running the PMC autolocker lately?  I can't remember, and I also can't find where it might be running.  It's not on megatron, either in the crontab or Q's new /etc/init place.  It's also not on op340m. 

Anyhow, what prompted this was that the PMC transmission has been incredibly fuzzy today.  On the StripTool it looks like it was fine until about -7 hours ago, when it lost lock.  Then Diego relocked it around -3 hours ago, and it's been fuzzy ever since. It was unlocked again for about 15 minutes about 45 minutes ago, and when I relocked it, it was even more fuzzy.

The FSS slow is almost exactly zero, the PMC's PZT is not at the edge of the range, the FSS PC drive RMS has been both high and low, and the PMC fuzz level has just been consistently high.  I was checking the parameters in the PMC phase shifter screen, and looked up the autolocker to see what the nominial values are supposed to be. 

For the RFADJ value, the autolocker sets it to 2.0, and after it locks increases it up to 4.5.  I found the value at 2.0, and the autolocker isn't running, which made me wonder if the autolocker died sometime after it set the value low, but before it could detect lock and reset the value to high.  (Actually, after lock it sets the value to whatever is in the channel C1:PSL-STAT_PMC_NOM_RF_ADJ, which is 4.5).

Anyhow, I manually set the RFADJ value to 4.5, and the PMC transmission immediately improved. 

 

EDIT, 8pm, JCD:  Rana reminded me that he attached a screenshot back on 30Dec2014 (http://nodus.ligo.caltech.edu:8080/40m/10849), which I had ignored earlier because the parameters weren't written in text.  My bad.  Anyhow, after the New Year's tune-up, the RFADJ should be 6.0.  I have set it so, and also changed the C1:PSL-STAT_PMC_NOM_RF_ADJ chan to be 6.0.

  10915   Fri Jan 16 20:01:32 2015 JenneUpdateLSCLSC model change implemented

Nope, I used the script.

Yesterday's changes were mostly to the generateLSCscreen/C1LSC_OVERVIEW_INPUT_MATRIX.adl sub-screen.  The UGF servos were added earlier in the week to the LSC screen in the generateLSCscreen/C1LSC_OVERVIEW_SERVOS.adl sub-screen.

  10917   Sat Jan 17 01:10:36 2015 JenneUpdateLSCSome locking, may need to modify UGF part again

I have been playing with the IFO tonight.  Mostly, I wanted to make sure that all of the scripts for the carm_cm_up sequence were working, and they seem to all be fine. 

I also turned on all 4 UGF servos.  My big ah-ha moment for the night is that the excitation is multiplied by the gain multiplier.  This means that if the UGF servo is multiplying by a small number (less than 1), the excitation will get smaller, and could get small enough that it is lost in the noise.  Now the error signal for the UGF servo is very noisy, and can dip to zero.  Since we can't take the log of zero, there are limiters in the model, but they end up giving -80dB to the error point of the UGF servo.  This makes it all freak out, and often lose lock, although sometimes you just get a weird step in the UGF servo output. 

Anyhow, we need to be mindful of this, and offload the UGF servos regularly.  I think the better thing to do though will be to divide the excitation amplitude by the gain multiplier.  This will undo the fact that it is multiplied by that number, so that the number of counts that we put into the excitation amplitude box is what we expect. 

  10923   Tue Jan 20 15:09:01 2015 JenneUpdateLSCLSC model change implemented

 

Quote:

Brain not working anymore now that it's ~4am, but I need to rethink and recheck to make sure that the PD whitening triggering is still okay and working.  Or maybe we can remove it, and just include that in the scripts, as Rana has been suggesting for ages.  Thoughts for tomorrow.

LSC whitening triggering was not working, because of the implementation of the double-rows for the input matrix.  I have modified the c-code that looks at the input matrix and triggers, and decides when to turn on the PD whitening, so that it now works.

  10924   Tue Jan 20 19:32:06 2015 JenneUpdateSUSSUS Drift restore scripts

I made little scripts to go with the sus driftmon buttons, that will servo the alignment sliders until the susyaw and suspit values match the references on the driftmon screen. 

  10925   Tue Jan 20 20:03:17 2015 JenneUpdateLSCALS lock not staying?

The Xarm ALS has been a little funky today. 

First, the green and the arm-axis would not stay co-aligned.  I'm not sure which was moving (although neither ITMX nor ETMX seemed to be moving very much according to their oplevs and OSEMs).  I went to the Xend table and jiggled the mounts for the steering optics, in case one was loose or something.  None were.  However, the transmission quit jumping around by a factor of 2 after that.   The beatnote alignment on the PSL table was also bad, so I tweaked that alignment up for the Xarm.  There were some not connected cables and fibers blocking the access to the X beatnote area, so those are up on top of the PSL table.

Anyhow, I haven't locked the individual arms, but I cannot hold the lock with CARM/DARM.  The CARM output keeps hitting the threshold for the locking watchdog, which turns off the lock.  Obviously I could just increase this threshold, but the right thing to do is figure out why the Xarm ALS is so noisy today, and why it wants to output such a large control signal to maintain the lock.

  10927   Wed Jan 21 15:27:47 2015 JenneUpdateLSCFixed LSC model bug

This problem with the CARM loop last night was the fault of a bug that I had put into the LSC model last week.  When I gave the input matrix and normalization matrix double rows, I had put the goto tags for the CARM normalization matrix rows backwards.  So, even though I thought I was not normalizing CARM, in fact I was normalizing by POPDC, which was near zero since the PRM was misaligned.

Anyhow, found, fixed, currently locking, and all seems well.

  10929   Thu Jan 22 03:21:24 2015 JenneUpdateLSCLocks with large MICH offsets

[Jenne, Diego, EricQ]

Tonight we worked on the acquisition sequence (including re-re-re-commissioning the UGF servos, hopefully for the last time...) for the PRFPMI with large MICH offsets. 

The procedure is all in the carm_up script, as far as things work.

We had some locklosses, but they were mostly due to non-carefulness on my part during the transitions between error signals, or the UGF servos getting upset because the oscillator peaks had gotten lost in the noise.  The one that I show here is our very last one of the night, where we are hitting the rails for the MICH signal, which is then causing the other loops to have to do weird things to try to compensate, and they lose lock.

Here also is a StripTool shot during that lock stretch.  I was in the middle of increasing the MICH offset to 75% of the fringe.  The yellow trace (called MICH_B_MON) is ASDC/POPDC normalized so that it always goes 0-1.  I was pleased to see that perhaps REFL11I and AS55Q are turning over, although as Q will tell us in a more detailed elog tomorrow, having a large MICH offset does weird things and moves the DARM zero-point.  So, maybe we aren't actually anywhere awesome yet. 

After some MICH offset, the maximum arm power is always going to be about 50, so arm powers of 8 or 10 equates to 100 pm.  We didn't get there tonight while on IR signals.


The locking sequence is now something like this:

  • Lock carm and darm on ALS, find resonances, move to 3 counts (roughly 3nm) offset.
  • Set PRMI up to acquire on REFL33I and ASDC/POPDC at 25% MICH fringe.  (After a while, I assume perhaps because the alignment is no longer tip-top, I have been by-hand reducing the MICH offset from -700counts which is 25% to -200counts, and then immediately putting it back to -700 after the PRMI acquires.)
  • Engage all 4 UGF servos
  • Reduce the CARM offset a bit, to 1.0 count, which gives arm powers of about 0.4 (with 50 being the max possible)
  • Transition CARM from ALS to sqrtInvTrans
  • Transition DARM from ALS to DC trans:  (TRY-TRX)/(TRX+TRY)
  • Reduce the oscillator amplitudes of the UGF servos
  • Reduce the CARM offset to powers of about 1
  • Ramp to 50% MICH fringe

After this, we tried a few times to lower the CARM offset, but kept losing lock, I think because the UGF servos went crazy.  The final lock, shown above, we lost because the MICH output was hitting the rails.

The problem with the MICH servo right now is the low SNR of the POPDC being used to normalize ASDC.  The control output is enormous, even if we have the 400Hz lowpass on.  We need to rethink our MICH servo, starting with a lower UGF, so that we're not injecting all this sensing noise all over the place. 


For tomorrow:

  • Re-look at MICH loop, to prevent sensing noise injection.
  • How does the large MICH offset affect our zero points for CARM and DARM?  Can we stay on DC transmission signals through 30 or 100 pm?
  • What to do next?  One or two of the locklosses were because the CARM detuned double cavity pole wasn't de-Q-ed enough, so still hit 0dB and created an unstable unity gain point.  Can we go to higher MICH offset, maybe 75%? 
  • Still need to figure out where our missing phase is for our LSC loops.  CARM and DARM are short on phase, and we could definitely use some more.  So, I will work on trying to give us filters that don't eat too much phase, but we still need to find that missing ~14 deg. 
  10933   Fri Jan 23 02:11:40 2015 JenneUpdateLSCCARM filters modified slightly

[Jenne, Diego]

One of tonight's goals was to tweak the CARM filters, so that we could engage the lowpass filter, to avoid the detuned double cavity pole resonance disturbing the CARM loop.

I increased the Q of the zeros in the FM3 boost so that it eats fewer than the original 18 degrees of phase at 100 Hz.  We kept losing lock though, so for each lock I backed off on the Q a little bit.  In the end, the filter eats 9 degrees of phase at 100 Hz.  I also moved the lowpass from 700 Hz to 1kHz, although that doesn't change the phase at 100 Hz very much.

We modified the carm_up script re: PRMI locking a little bit.  The PRMI is not so enthusiastic about locking immediately at 25% MICH fringe, so I backed that off.  It now acquires lock at a few percent, and then ramps up the offset.  Also, the MICH FM6 bounce roll filter is now turned on after lock is acquired, effectively giving it an extra second or two of delay beyond the rest of the filters.

We were able several times to get to some MICH offset and turn on the lowpass filter, but starting to reduce the CARM offset makes us lose lock.  I think the problem is that the UGF servo demod phase is changing as we are changing offsets, filters and error signals.  We see that the I-phase is servoed successfully to 0dB, but that the Q-phase is starting to move around by 30 degrees or more.  We either need to monitor this much more closely, and add the changing demod phases to the carm_up script, or we need to go back to the sum-of-squares situation that we had last week.  Note that we saw failures with that method for a completely separate reason:  we were getting too close to the limiters, which cause the UGF servos to glitch and the outputs jump by a significant amount.  So, the issues that we were seeing last week when we had the sum-of-squares were a different thing, that we noticed and understood later.

Anyhow, nothing too exciting and glorious tonight, but progress has been made.

Also, from some Mist simulations that Q did, Diego made a sweet plot that is now posted in the control room, so we can translate arm power to CARM offset, at various MICH offsets. 

We also took some CARM loop measurements with the new filters.  We have a little more phase than we used to, which is nice.  These traces don't have the lowpass engaged, since I was trying to see how far we could get without it.  We lost lock right after the second measurement, but I think that was to do with the UGF servos.

  10937   Fri Jan 23 18:08:10 2015 JenneUpdateCDSEPICS freezes

So, I neglected to elog this yesterday, but yesterday we had one of those EPICS freezes that only affects slow channels that come from the fast computers.  It lasted for about 5 minutes.

Right now, we're in the middle of another - it's been about 7 minutes so far. 

Why are these happening again?  I thought we had fixed whatever was the issue.

EDIT:  And, it's over after a total of about 8 minutes.

  10938   Fri Jan 23 19:38:02 2015 JenneUpdateLSCcarm_cm_up supports both signs of CARM offset

A small change, but now the carm_up script supports both sides of the CARM offset.  After the arms are locked with ALS it asks for a "+" or a "-", which indicates which sign of digital CARM offset will be added.  In the past, we have been primarily using the "+" sign.
 

  10939   Mon Jan 26 13:07:28 2015 JenneUpdatePEMSeismometers back in nominal places

I have just put the seismometers back into their nominal positions, on the concreted slabs.  The T-240 is in the vertex, and the 2 Guralps are at the end stations.

The vertex location doesn't have a spaghetti pot right now.  There is an aluminum support for cable trays that is welded to the supports under the beam tube that is in the way.  The pot looks like it will fit barely, if it were slid totally horizontally into place.  However we can't do that with the seismometer in place.  I'll chat with Steve this afternoon about our options. 

Since I don't know that we are planning on ever putting a cable tray on the inside of the beamtube, perhaps we can cut ~6 inches of this piece away. 

  10941   Mon Jan 26 21:10:04 2015 JenneUpdateModern ControlWaking up the OAF

I had a look at the OAF model today. 

Somehow, the screens that we had weren't matching up with the model.  It was as if the screens were a few versions old.  Anyhow, I found the correct screens in /userapps/oaf/common/medm, and copied them into the proper place for us, /userapps/isc/c1/medm/c1oaf.  Now the screens seem all good.

I also added 2 PCIE links between the OAF and the SUS models.  I want to be able to send signals to the PRM's pitch and yaw.  I compiled and restarted both the oaf model and the sus model.

The OAF model isn't running right now (it's got the NO SYNC error), but since it's not something that we need for tonight, I'll fix it in the morning.


My thought for trying out the OAF is to look at the coherence between seismic motion and the POP DC QPD when the PRMI is locked (no arms).  I assume that the PRM is already handled in terms of angular damping (local and oplev), so the motion will be primarily from the folding mirrors.  Then, if I can feedforward the seismometer signal to the PRM to compensate for the folding mirrors' motion, I can use the DC QPD as a monitor to make sure it's working when we're PRMI-only locked, or at low recycling gain with the arms.  But, since I'm not actually using the QPD signal, this will be independent of the arm power increase, so should just keep working.

Anyhow, that's what my game plan is tomorrow for FF.  Right now the T-240 is settling out from its move today, and the auto-zero after the move.

  10942   Tue Jan 27 04:11:21 2015 JenneUpdateLSCSmall tweaks to the locking

[Jenne, Diego, EricQ]

We did a series of small things that may have helped with the locking, although we didn't actually get anywhere closer in CARM offset.

  • Removed the demodulation phase from the UGF servos.
    • We don't care about the phase value, just the magnitude.
    • Since we are using these through transitions between error signals, as well as with different filters coming on and off, the demodulation phase isn't constant, so the UGF servo is getting the wrong answer, and was throwing us out of lock.
    • The problems that Rana and I saw last time we had the sum of the sqrt were later discovered to be attributable to losing the peak in the noise, and hitting saturation limiters in the model, so not the fault of the sum of the squares.
    • I don't actually take the square root.  As Koji pointed out, the very next thing that happens to the signal is mag2dB, which is usually 20*log10(mag).  To compensate for the fact that I'm not taking the square root, I just use 10*log10(mag).  This removes one element of non-linearity, and leaves it at about the same number of square-ings as the complex division.
    • The excitation and measurement still happen after the multiplication.
    • The UGF screens will be updated tomorrow to reflect the change.
  • Added the new error signal rows to the LSC model's DAQ list.  So, now DARM_A_ERR and DARM_B_ERR are both acquired (and the same for CARM, MICH and PRCL).  This is to allow us to look at _DQ channels with dataviewer and DTT without having to clear the testpoints all the time.
    • We were running into too many channels being recorded, so we are not keeping the SRCL A & B signals right now.  Also, the CESAR signals that are no longer in use have been removed from the DAQ list.
  • Added a comb to the ASDC and POPDC signals, to remove the 60Hz harmonics from the MICH error signal.
    • The harmonics of the 60Hz line were dominating by more than an order of magnitude the RMS for the MICH control signal.  We couldn't afford too much phase at a few tens of Hz, so we do not notch out the original 60Hz line, although it isn't as big as, say, the 180Hz line.  So, I think that the notches are for the 2nd, 3rd, 4th and 5th order harmonics of 60Hz.  This significantly improved the RMS of the MICH output signal.
  • Lowered the MICH UGF slightly, from 48Hz to 41Hz. 
    • We wanted to go lower, to maybe 30Hz-ish, but we don't have the phase margin.  The roll mode notch is in the way, so we compromised at 41Hz.  With the comb mentioned above, the MICH control signal looks much more reasonable, and we're not injecting oodles of sensing noise into the BS.
    • Together with the comb, this ensures that we are not constantly railing the MICH output limiter, which lost us lock several times.
  • Increased the POPDC analog whitening gain from 0dB to 18dB.  We will certainly saturate POPDC when the carrier is resonant, but we were hoping to improve our SNR in our MICH error signal.  It helped noticeably, although we'll have to think about the fact that if it saturates while we're trying to acquire, the AS/POP composite signal won't be any good, and we'll kick the optics.  Also, we may have to lower the gain again as the carrier comes in to resonance.  Anyhow, something to think about.  Right now the gain is left at 18dB, and the dark offsets are set to match.
  • We may have been using the wrong side of the MICH offset, according to Q's plots.  We determined tonight that we want a (-) in the MICH gain, although the offset values can stay the same.  Even though we tried this, we didn't really get any farther in CARM offset reduction.
  • We took 2 sets of CARM and DARM loops.  They are both at 50% MICH offset, although I don't remember which sign the first one had.  The second one, at arm powers of 1.4, definitely had the new negative MICH gain. 
    • The first loop was taken at 50% MICH offset (don't remember which sign), and arm powers of about 1.15.
    • The second loop was taken at 50% MICH offset with negative gain, and arm powers of about 1.4.
    • While these numbers are not so different, maybe the first one was at roughly 300pm, and the second was at roughly 200pm, the loop shapes change dramatically.
    • The phase goes flat and we lose some of the phase margin.  Also, the magnitude is starting to get wiggly at lower frequencies.
    • See the transfer functions attached below.
  • While we were sitting at about arm powers of 1.4, with the 50% MICH offset with the negative sign in the gain, we set the REFL 11 demod phase.  It was 18.9deg (I don't remember from when), but we set it to 2.9deg to minimize the PRCL actuation in the Q-phase.  Oscillator was at 443Hz, about 10 counts. 
    • The coherence between the sqrtInvTrans signal and REFL11I looked pretty good from a few Hz to a few tens of Hz. 
    • It was late, so we decided to try transitioning over to REFL11I, and failed at the first very baby step.  So.  Ooops.
    • We should look more carefully at the TF between our current CARM signal and REFL11I.
    • We should also seriously consider using a normalized RF signal.  The SNR in the transmission PDs is just fine, although POPDC isn't a perfect choice at such high MICH offsets.
  10943   Tue Jan 27 14:00:05 2015 JenneUpdateModern ControlPRMI locked

PRMI is locked on sideband, starting ~4 minutes ago, to collect ASC/seismic data for feedforward.

 

  10945   Tue Jan 27 17:58:21 2015 JenneConfigurationTreasureWelcome, Donatella!

Welcome to your new abode, Donatella!

  10947   Wed Jan 28 03:01:24 2015 JenneUpdateLSCTransitioned DARM to AS55Q

[Jenne, Diego]

Tonight we were able to transition DARM from DC transmission signals to AS55Q/(TRX+TRY).  That's about as far as we've gotten though.

Here are the details:

  • Set the ASDC->MICH matrix element such that the MICH fringes were 0-1.  For some reason this number seems to change by ~10% or so each night.
  • Followed main carm_cm_up script, although stopped at MICH offset of 25% (mostly because I forgot to let it go to 50% - no fundamental reason)
  • So, MICH was at 25% (with a + for the gain accidentally, even though we decided yesterday that - was better), arm powers were about 1.1 or so.
  • Took transfer functions driving DARM and looking at normalized AS55Q, and driving CARM looking at normalized REFL11I.
    • There is still not a lot of coherence in the CARM->REFL11I case, so we decided to stick with DARM for starters.
    • The TF between DARM and AS55Q looked really nice!
  • Prepared the unused DARM error signal row, including an offset before the blend matrix.
  • Transitioned over to normalized AS55Q.
    • This left the DARM servo filterbank with a zero digital offset.
    • But, the error signal had an offset before it got to the DARM filter bank.
      • This offset does not have any ramping (I don't know how to do that when building a model), so it's not as nice for reducing an offset.
      • Maybe we can, after transitioning to the new signal, move the offset to the DARM servo filterbank?
  • Was reducing the DARM offset so that we were at the true AS55Q zero crossing.
  • Saw that the UGF servo lines were starting to get a bit lost in the noise, so I increased the DARM's amplitude.
    • I don't know if the UGF servo was already too far gone and increasing the SNR couldn't recover it, or if I was driving too hard and directly kicked myself out of lock.  Either way, we lost lock.

The carm_cm_up script now should get all the way to this point by itself, although occasionally the PRMI part will lose lock (but not the arms), so you have to go back to the 3nm CARM offset and relock the central part.  I have written a little "relockPRMI.sh" script that lives in ..../scripts/PRFPMI/ that will take care of this for you.  

We were able to transition DARM to AS55Q a total of 3 or so times tonight.  The first time was with the + MICH gain, and the rest of the times were with - MICH gain.  The carm_up script now asks for a sign for the MICH gain just after asking for a CARM offset sign. 

I think that we understand all of our locklosses from these states.  Twice (including the time described above) the UGF lines got lost in the noise, and the UGF servos went crazy.  Once the PRCL loop rang up, because a filter that wasn't supposed to be on was on.  This was a terrible filter that I had made a long time ago, and was never supposed to be part of the locking sequence, but it was getting turned on by a restore script, and kept eating our phase.  Anyhow, I have deleted this terrible boost filter so it won't happen again (it was called "boost test", which may give you an idea of how non-confident I was in its readiness for prime-time).  The last time of the night I must not have been quite close enough in CARM offset, so we should probably check with a TF before making this last jump.

Diego wrote a nifty burt restoring script that will clear out all the elements of the input matrix and the normalization matrix for a row that you tell it (i.e. DARM_A will clear out all the elements in the first row of those 2 matrices).  This is useful for the switches back and forth between the _A and _B signals, to make sure that everything is in order.  So, I now run those clear scripts, then put in the elements that I want for the next step.  For example, DARM initially locks with ALS using the A row.  Then, DARM transitions to the B row for DC transmission.  Then, I prepare the A row for AS55Q locking, and I don't want any elements accidentally left from the ALS signal.  It lives in ..../scripts/LSC/InputMatrix/

 

Thoughts for tomorrow:

Daytime re-commission the Xarm ASS.

Nighttime try to get back to DARM on AS55Q and push farther forward. 

 

  10959   Fri Jan 30 02:57:03 2015 JenneUpdateModern ControlFirst try with PRCL ASC Wiener filtering

[Aside - How do you rotate plots in the new elog?  It's showing them correctly in the attachments list below the entry, but not in the body of the log :( ]

I tried a round of PRCL ASC Wiener filtering today, but something wasn't right.  I was able to either make the cavity motion worse, or completely throw the cavity out of lock.  Making it less noisy didn't happen.

I took only 9 minutes of data the other day, since the PRMI didn't want to stay locked while it was daytime.  So, this wasn't a whole lot to train on.  But, even so, I designed some Wiener filters.  The plots with the designs show the calculated Wiener filter ("Wiener") and the result from vectfit ("Fit"). Below the bode plot is the coherence between that witness (seismometer direction) and the degree of freedom (QPD channel).  The fits were weighted by the choherence, so you can see that in the areas where the coherence was good, the fit was good.  Elsewhere, it's not so great.

 

Using these filters, and assuming a Cheby1 2nd order lowpass filter at 30Hz, I predicted the following residuals:

After discovering that these filters didn't work, I went rogue and also put in a high pass filter at 0.1 Hz, but that didn't really change anything.

Here is a plot of what happened in Yaw.  The Wiener filters' gains were all 0.3 here, which made the cavity motion larger, but not so large that it lost lock.  The filters ought to have gains of 1 - the Wiener calculation should figure out the gains appropriately, if I've given it enough information.  Here, as in the prediction plots above, red is the reference with the Wiener off, and black is with the Wiener filters on.  Black is supposed to be below red, if the filters are working.  Blue is the estimate of the angular motion that is being fed forward to the PRM, and you can see that at least the general shape is correct.  I do need to figure out what the resonance in the blue trace is from - it's at the same frequency as a peak in the T-240's spectrum (that I didn't save).  I suspect the cable might be touching the spaghetti pot on the inside, and making a mechanical short to pot vibrations.

Anyhow, more work to be done.  I left the PRMI locked for a while this afternoon, starting at 5:15ish, so I'll see tomorrow how long of a lock stretch I was able to capture for training. 

 

  10962   Sat Jan 31 01:34:12 2015 JenneUpdateLSCNot able to engage AO path

Nothing earth-shattering today.

A few things of note:

  • I checked the coherence (no lock present, just noise) between REFL11_I_IN1 and CM_SLOW_OUT, which are meant to be the same thing when only the REFL1 path is allowed through the CM board.
    • However, at first, there was very little coherence - small band, and only about 0.7 or so.
    • I went inside and jiggled the cables, and also toggled the whitening switches for REFL11 and the CM_SLOW, and after that I had excellent coherence. 
      • I didn't take a coherence spectrum in between those activities, but since the cable connections were all solid, I believe that it may have been a sticky slider -esque problem, and the CM whitening wasn't matched between the analog and digital.
    • I tried two or three times to engage the AO path, but I always lost lock before I was getting any meaningful gain through.
      • I took some TFs remotely with the SR785, but they're totally noise.  I don't even know which sign of the CM board is correct, so no real knkowledge added there, from today.
  • The ~400Hz ringing that we have been seeing, we have been blaming on the PRCL loop.  However, just before my last lockloss I saw gain peaking around 400Hz in the CARM input spectrum. I didn't see if it was also there in the PRCL spectrum.  So, either it is coupling from PRCL to CARM, or CARM itself.  But I think we need to see if we can eek out a little more phase at higher frequencies for both of those loops.  I  just looked, and about 16 seconds before the last lockloss, I see the PRCL loop coupling into the CARM loop.
    • Since yesterday, we have been lowering the PRCL UGF using the servos to be about 70Hz, to give us more gain margin at the high end of the phase bubble, but we still see this ringing. 
  • Two or three times my arm power buildup at 0 CARM offset, 25% MICH offset was at 20 or 21.  Then, a few other times I was only getting about 10 (which is what we have been seeing the last few days.)
    • I'm running the ASS between each lock, although I'm not running it for a full ~2 minutes or so usually. 
    • I think that the reason I was able to get to such high arm powers was excellent alignment, so maybe it's worth sitting and waiting for the ASS to have a full 2 or 3 minutes between locks, even if the ASS error signals look zero-ed out.
    • This is still a factor of 2 lower than we expect for 25% MICH offset, but the whole factor of 5 isn't explained by some mysterious loss.  At least half of it is alignment.
  • The PRCL ASC feedforward still isn't working, but I'd like to try Q's trans qpd ASC soon, with the full lock.  I think the system is ready, but scripts are not, so Q has to be here to run it.

See first plot below for the PRCL->CARM coupling just before lockloss.  The second plot is the corresponding lockloss, where the PRCL loop is starting to see that oscillation again, and it's just barely starting to get into CARM. 

  10963   Mon Feb 2 12:24:27 2015 JenneUpdateASCPOP QPD centered

After aligning the PRC, I centered the POP QPD.

  10967   Tue Feb 3 04:07:49 2015 JenneUpdateModern ControlMeasured PRM -> POP QPD TFs

Today (after centering the POP QPD), I measured the PRM to POP QPD transfer functions.  I am suspicious that this was part of my problem last week.  Since most of the angular noise is coming from the folding mirrors, but I can't actuate on them, I need to know (rather, the Wiener calculator needs to know) how actuating on the PRM will affect the spot at the POP QPD.

For the plots below, I have cut out any data points that have coherence less than 0.95. I may want to go back and fill in a little bit some of the areas (particularly around 3Hz) that I had trouble getting coherence in.

Using these to prefilter my witness data, I am failing to calculate a Wiener filter.  I have tried the Levinson algorithm, as well as brute-forcing it, but I'm too close to singular for either to work.  I am able to calculate a set of Wiener filters without the prefiltering, or with a dummy very simple prefilter, so it's not inherently in the calculators.  Separately, I can plot my vectfit-ed actuator TFs, and I can convert them to a discrete fiilter with the bilinear transform, and then use sosfilt to filter some white noise data, which comes out with the shape I expect.  So.  It's not inherently the filters either.  More work to do, when it's not 4am.

Here are the measured actuator transfer functions.  They were measured as usual with DTT, but since the measurement kept getting interrupted (MC unlock, or I wanted to add more integrations or more cycles), these are several different DTT files stitched together.  In both cases I am acuating PRM's ASC[pit, yaw] EXC point, and looking at the POP QPD.

 

ELOG V3.1.3-