In order to setup POP camera and RFPD on the ITMX table, we decided to first work on the IMC and X/Y-arm alignment.
We zeroed IMC WFS outputs and aligned IMC manually to get IMC transmission of 1200 and reflection of 0.35.
We used the new video game tool that moves the pairs of mirrors - PR3 & ETMY, ITMY & ETMY - in common and differential modes. This brought the Y-arm flashing to 0.8. Note that we used the _OFFSET bias values for PR3 & ETMY alignment instead of the _COMM bias values.
We repeated the same procedure of moving the pairs of mirrors - BS & ETMX, ITMX & ETMX - in common and differential modes but manually this time. This brought the X-arm flashing to ~1.0.
Tega has kindly made a summary page for the IMC WFS. Its in a tab on the usual summary pages.
One thing I notice is that the feedback to MC2 YAW seems to have very little noise. What's up with that?
The output matrix (attached) shows that the WFS have very little feedback to MC2 in YAW, but normal feedback in PIT. Has anyone recalculated this output matrix in the past ~1-2 years?
I'm going to read Prof. Izumi's paper (https://arxiv.org/abs/2002.02703) to get some insight.
The output matrix doesn't seem to have any special thing to make this happen. Any ideas on what this could be?
The output matrices have been calculated on Aug 4, 2022 by me. [40m ELOG 17060]
Regarding the noise see [40m ELOG 17061]
With regard to the current IMC WFS design, a SURF student in 2014 (Andres Medina) and Nick Smith made the revision.
The telescope design was described in the elogs [40m ELOG 10410] [40m ELOG 10427] and also T1400670.
One of the biggest challenges in LIGO is reducing the alignment control noise. If you haven't worked on it for at least a few years, it probably seems like a trivial problem. But all versions of LIGO since 2001 have been limited by ASC noise below ~50 Hz.
I think the 40m IMC is a good testbed for us to try a few approaches towards mitigating this noise in LIGO. The following is a list of steps to take to get there:
I think that steps 1-6 are well within our existing experience, but we should do it anyway so as to reduce the IMC beam motion at low frequencies, and also to reduce the 10-100 Hz frequency noise as seen by the rest of the interferometer.
Steps 7-8 are medium hard, but we can get some help from the CSWG in tackling it.
Step is pretty tough, but I would like to try it and also get some help from MLWG and CSWG to address it.
[Paco, Murtaza, JC]
Fixed C1SUS2 Crash from this morning
What we did:
- Attempt to restart only C1SUS2
- Restart All Machines and Burt Restored to Friday @ 7:19 pm
Entering this morning, We were unsure why we were having issues aligning and come to find out that C1SUS2 crashed. Paco attempted to restore by restarted the machine individually and restoring, while this did turn all the machines green on the CDS.MEDM Screen, it did not resolve the issue. So moving forward, please keep in mind that EVEN IF ALL MACHINES SHOW GREEN, OPTICS MAY STILL NOT BE DAMPING.
Next we continued to restart and reboot the old fashioned way.
I attempted to use the ./restartAllModels.sh script in the "/opt/rtcds/caltech/c1/Git/40m/scripts/cds" directory, but there was an error and the message I got said something along the lines of "Restart medm screen and try again". This was weird and all of the machines were already shutdown. So, to bring them back up, I used the ./startAllModels.sh script. When starting up, i was prompted to provide a burtrestore directory, and I inputted /opt/rtcds/caltech/c1/burt/autoburt/snapshots/2023/Sep/22/20:19.
This worked out and IMC came back to nominal alignment. The primary issue we seem to be coming across now is that C1:IOO-WFS2_PIT_OUTPUT is increasing at a steady rate and this is disrupting our alignment.
Rana and I spent some time looking at the IMC demod board earlier today. I will post the details shortly, but there was a label on the front panel which said that the nominal LO level to the input should be -8dBm. The new 29.5MHz routing scheme meant that the LO board was actually being driven at 0dBm (that too when the input to the RF distribution box was attenuated by 5dB).
An elog search revealed this thread, where Koji made some changes to the demod board input attenuators. Rana commented that it isn't a good idea to have the LO input be below 0dBm, so after consulting with Koji, we decided that we will
After implementing these changes, and testing the board with a Marconi on the workbench, I found that the measured power levels (measured with an active FET probe) behave as expected, up till the ERA-5SM immediately prior to the LO (U4 and U6 on the schematic). However, the power after this amplifier (i.e. the input to the on-circuit LO, Minicircuits JMS-1H, which we want to be +17dBm), is only +16dBm. The input to these ERA-5SMs, which are only ~2years old, is -2dBm, so with the typical gain of +20dB, I should have 18dBm at their output. Moreover, increasing the input power to the board from the Marconi doesn't linearly increase the output from the ERA-5SM. Just in case, I replaced one of the ERA-5SMs, but observed the same behaviour, even though the amplifier shouldn't be near saturation (the power upstream of the ERA-5SM does scale linearly).
This needs to be investigated further, so I am leaving the demod board pulled out for now...
The input impedance of the mixer is not constant. As the diode switches, it changes dynamically. Because of this, the waveform of the LO at the mixer input (i.e. the amplifier output) is not sinusoidal. Some of the power goes away to harmonic frequencies. Also, your active probe is calibrated to measure the power across the exact 50Ohm load, which is not in this case. The real confirmation can be done by swapping the mixer with a 50Ohm resistor. But it is too much. Just confirm the power BEFORE the amp is fine. +/-1dB does not change the mixer function much.
Instead, we should measure
- Gain imbalance
of the I/Q output. This can be checked by supplying an RF signal that is 100~1kHz away from the LO frequency and observe I&Q outputs.
This time the test was succesful but I have reverted MC3 f2a filters back to with Q=3, 7, and 10. The inital part of the test is still useful though. I'm attaching amplitude spectral density curves for WFS control points and C1:IOO-MC_F_DQ in the different configurations. The shaded region is the 15.865 percentile to 84.135 percentile bounds of the PSD data. This corresponds to +/- 1 sigma percentiles for a gaussian variable. Also note that in each decade of freqeuncy, the FFt bin width is different such that each decade has 90 points (eg 0.1 Hz bin width for 1Hz to 10 Hz data, 1 Hz binwidth for 10 Hz to 100 Hz and so on.)
The WFS control points do not show any significant difference in most of the frequency band. The differences below 10 mHz are not averaged enough as this was 30min data segments only.
C1:IOO-MC_F_DQ channel also show no significant difference in 0.1 Hz to 20 Hz. Between 20-100 Hz, we see that higher Q filters resulted in slightly less noise but the effect of the filters in this frequency band should be nothing, so this could be just coincidence or maybe the system behaves better with hgiher Q filters. In teh lower frequency band, we would should take more data to average more after shortlisting on some of these f2a filters. It seems like MC1 Q=10 (red curve) filter performs very good. For MC2, there is no clear sign. I'm not sure why MC2 Q=3 curve got a big offset in low frequency region. Such things normally happen due to significant linear trend presence in signal.
I'm not sure what other channels might be interesting to look at. Some input would be helpful.
We ran the f2a filter test for MC1, MC2, and MC3.
The new filters differ from previous versions by a adding non-unity Q factor for the pole pairs as well.
This in terms of zpk is: [ [zr + i zi, zr - i zi], [pr + i pi, pr - i pi], 1] where
We uploaded all these filters using foton, into the three last FM slots on the POS output gain coil.
We ran tests on all suspended optics using the following (nominal) procedure:
C1:IOO-WFS_GAIN to 0.05.
** Excitation = 0.05 - 3.5 Hz uniform noise, 100 amplitude, 100 gain
The triggered code went on at 5:00 am today but a last minute change I made yesterday to increase number of repititions had an error and caused the script to exit putting everything back to normal. So as we came in the morning, we found the mode cleaner locked continuously after one free swing attempt at 5:00 am. I've fixed the script and ran it for 2 hours starting at 8;10 am. Our plan is to get some data atleast to play with when we are here. If the duration is not long enough, we'll try to run this again tomorrow morning. The new script is running on same tmux session 'MCFreeSwingTest' on Rossa
10:13 the script finished and IMC recovered lock.
Thu Mar 11 10:58:27 2021
The test ran succefully with the mode cleaner optics coming back to normal in the end of it. We wrote some scripts to read data and analyze it. More will come in future posts. No other changes were made today to the systems.
I think I have aligned the cavity, including MC1 such that we are seeing flashing of fundamental mode and significant transmission sum value as well.However, I'm unable to catch lock following Koji's method in 40m/16673. Autolocker could not catch lock either. Maybe I am doing something wrong, I'll pickup again tomorrow, hopefully the cavity won't drift too much in this time.
After the exploration of c1psl, the IMC locking was not functioning well. It looked like the MC autolocker issue.
Now I remember that the autolocker was running on docker. I followed the instruction on the wiki page. https://wiki-40m.ligo.caltech.edu/Computers_andScripts/AlwaysRunningScripts
Remember: Docker is running on optimus / systemctl is running on megatron!
Then: I noticed that MC autolocker heartbeat was blinking even with the docker stopped (!?). I confirmed that autolocker is not running on systemctl
I had no hope who is toggling the heartbeat, but "ps -def" on megatron showed me that "AutoLockMC_LowPower.csh" is running! I was afraid that it is a remnant from vent!? But, is that possible?
The process was killed and the docker locker is running well now.
- IMC was locked.
- Some alignment change in the output optics.
- The WFS servos working fine now.
- You need to follow the proper alignment procedure to recover the good alignment condition.
- Basically followed the previous procedure 40m/16673.
- The autolocker was turned off. Used MC2 and MC3 for the alignment.
- Once I hit the low order modes, increased the IN1 gain to acquire the lock. This helped me to bring the alignment to TEM00
- Found the MC2 spot was way too off in pitch and yaw.
- Moved MC1/2/3 to bring the MC2 spot around the center of the mirror.
- Found a reasonably good visibility (<90%) at a MC2 spot. Decided this to be the reference (at least for now)
SP Table Alignment Work
- Went to the SP table and aligned the WFS1/2 spots.
- I saw no spot on the camera. Found that the beam for the camera was way too weak and a PO mirror was useless to bring the spot on the CCD.
- So, instead, I decided to catch an AR reflection of the 90% mirror. (See Attachment 1)
- This made the CCD vulnerable to the stronger incident beam to the IMC. Work on the CCD path before increasing the incident power.
MC2 end table alignment work
- I knew that the focusing lens there and the end QPD had inconsistent alignment.
- The true MC2 spot needs to be optimized with A2L (and noise analysis / transmitted beam power analysis / etc)
- So, just aligned the QPD spot using today's beam as the temporary target of the MC alignment. (See Attachment 2)
Resulting CCD image on the quad display (Attachment 3)
- To activate the WFS with the low transmitted power, the trigger threshold was reduced from 5000 to 500. (See Attachment 4)
- WFS offset was reset with /opt/rtcds/caltech/c1/scripts/MC/WFS/WFS_RF_offsets
- Resulting working state looks like Attachment 5
PMC seems to have gotten very misaligned over the last 12 hours. I'm going in to align now.
Here is the comparison before and after the fix.
Before the work, the UGF was ~40kHz. The phase margin was ~5deg. This caused huge bump of the frequency noise.
After the LO power increase, I had to reduce the MC loop gain (VCO Gain) from 18dB to 6dB. This resulted 4dB (x2.5) increase of the OLTF. This means that my fix increased the optical gain by 16dB (x6.3). The resulting UGF and phase mergin were measured to be 117kHz and 31deg, respectively.
Now I was curious to see if the PMC err shows reasonable improvement when the IMC is locked. Attachment 2 shows the latest comparison of the PMC err with and without the IMC locked. The PMC error has been taken up to 500kHz. The errors were divided by 17.5kHz LPF and 150kHz LPF to compensate the sensing response. The PMC cavity pole was ignored in this calculation. T990025 saids the PMC finesse is 4400 and the cavity pole is 174kHz. If this is true, this also needs to be applied.
1. Now we can see improvement of the PMC error in the region between 10kHz to 70kHz.
2. The sharp peak at 8kHz is due to the marginally stable PMC servo. We should implement another notch there. T990025 suggests that the body resonance of the PMC spacer is somewhere around there. We might be able to damp it by placing a lossy material on it.
3. Similarly, the features at 12kHz and 28kHz is coming from the PMC. They are seen in the OLTF of the PMC loop.
4. The large peak at 36kHz does not change with the IMC state. This does mean that it is coming from the laser itself, or anything high-Q of the PMC. This signal is seen in the IMC error too.
5. 72kHz, 108kHz, 144kHz: Harmonics of 36kHz?
6. Broad feature from 40kHz to 200kHz. The IMC loop is adding the noise. This is the frequency range of the PC drive. Is something in the PC drive noisy???
7. The feature at 130kHz. Unknown. Seems not related to IMC. The laser noise or the PMC noise.
Remaining IMC issues:
Done (Nov 23, 2015) - 29.5MHz oscillator output degraded. Possibly unstable and noisy. Do we have any replacement? Can we take a Marconi back from one of the labs?
Done (Nov 23, 2015) - Too high LO?
- Large 36kHz peak in the IMC
- IMC loop shape optimization
- IMC locking issue. The lock streatch is not long.
- IMC PC drive issue. Could be related to the above issue.
Maybe not relevant - PC drive noise?
Well. I thought a bit more and now I think it is likely that this is just the servo bump as you can see in the closed-loop TF.
6. Broad feature from 40kHz to 200kHz. The IMC loop is adding the noise. This is the frequency range of the PC drive. Is something in the PC drive noisy???
I sketched up a quick drawing with estimated length for the IMC reflected beam. This includes the distances and focal length. Distances are from optic to optic.
The ringdown measurements are in progress. But it seems that the MC mirrors are getting kicked everytime the cavity is unlocked by either changing the frequency at the MC servo or by shutting down the input to the MC. This means what we've been observing is not the ringdown of the IMC alone. Attached are MC sus sensor data and the observed ringdown on the oscilloscope. I think we need to find a way to unlock the cavity without the mirrors getting kicked....in which case we should think about including an AOM or using a fast shutter before the IMC.
P.S. The origin of the ripples at the end of the ringdown still are of unknown origin. As of now, I don't think it is because of the mirrors moving but something else that should figured out.
It is HIGHLY unlikely that the IMC mirrors are having any effect on the ringdown. The ringdowns take ~20 usec to happen. The mirrors are 0.25 kg and you can calculate that its very hard to get enough force to move them any appreciable distance in that time.
The huge kick observed in the MC sus sensors seem to last for ~10usec; almost matching the observed ringdown decay time. We should find a way to record the ringdown and the MC sus sensor data simultaneously to know when the mirrors are exactly moving during the measurement process. It could also be that the moving mirrors were responsible for the ripples observed later during the ringdown as well.
* How fast do the WFS respond to the frequency switching (time taken by WFS to turn off)? I think this information will help in narrowing down the many possible explanations to a few.
I extended the ringdown data analysis to the reflected beam following Isogai et al.
The idea is that measuring the cavity's reflected light one can use known relationships to extract the transmission of the cavity mirrors and not only the finesse.
The finesse calculated from the transmission ringdown shown in the previous elog is 1520 according to the Zucker model, 1680 according to the first exponential and 1728 according to the second exponential.
Attachment 1 shows the measured reflected light during an IMC ringdown in and out of resonance and the values that are read off it to compute the transmission.
The equations for m1 and m3 are the same as in Isogai's paper because they describe a steady-state that doesn't care about the extinction ratio of the light.
The equation for m2, however, is modified due to the finite extinction present in our zeroth-order ringdown.
Modelling the IMC as a critically coupled 2 mirror cavity one can verify that:
Where is the coupled light power
is the power rejected from the cavity (higher-order modes, sidebands)
is the cavity gain.
and are the power reflectivity and transmissivity per mirror, respectively.
is the power attenuation factor. For perfect extinction, this is 0.
Solving the equations (m1 and m3 + modified m2), using Zucker model's finesse, gives the following information:
Loss per mirror = 84.99 ppm
Transmission per mirror = 1980.77 ppm
Coupling efficiency (to TEM00) = 97.94%
I translate the results obtained in the previous elog to the IMC 3 mirror cavity. I assume the loss in each mirror in the IMC is equal and that M2 has a negligible transmission.
I find that to a very good approximation the loss per IMC mirror is 2/3 the loss per mirror in the 2 mirror cavity model. That is the loss per mirror in the IMC is 56 ppm. The transmission per mirror in the IMC is the same as in the 2 mirror model, which is 1980 ppm.
The total transmission is the same as in the 2 mirror model and is given by:
where is the coupling efficiency to the TEM00 mode.
I analyze the IMC ringdown data from last night.
Attachment 1 shows the normalized raw data. Oscillations come in much later than in Gautam's measurement. Probably because the IMC stays locked.
Attachment 2 shows fits of the transmitted PD to unconstrained double exponential and the Zucker model.
Zucker model gives time constant of 21.6us
Unconstrained exponentials give time constants of 23.99us and 46.7us which is nice because it converges close to the Zucker model.
- Today we spent the morning shift debugging SUS input matrix diagonalization. MC stayed locked for most of the 4 hours we were here, and we didn't really touch any controls.
It seems like the AO path gain stages on the IMC Servo board work just fine. The weird results I reported earlier were likely a measurement error arising from the fact that I did not disconnect the LEMO IN2 cable while measuring using the BNC IN2 connector, which probably made some parasitic path to ground that was screwing the measurement up. Today, I re-did the measurement with the signal injected at the IN2 BNC, and the TF measured being the ratio of TP3 on the board to a split-off of the SR785 source (T-eed off). Attachments #1, #2 shows the result - the gain deficit from the "expected" value is now consistent with that seen on other sliders.
Note that the signal from the CM board in the LSC rack is sent single-ended over a 2-pin LEMO cable (whose return pin is shorted to ground). But it is received differentially on the IMC Servo board. I took this chance to look for evidence of extra power line noise due to potential ground loops by looking at the IMC error point with various auxiliary cables connected to the board - but got distracted by some excess noise (next elog).
I am running some tests on the IMC servo board with an extender card so the IMC will not be locking for a couple of hours.
With the input matrix, coil ouput gains and F2A filters loaded as in 16091, I tested the suspension loops' step response to offsets in LSC, ASCPIT and ASCYAW channels, before and after applying the "new damping gains" mentioned in 16066 and 16072. If these look better, we should upload the new (higher) damping gains as well. This was not done in 16091.
Note that in the plots, I have added offsets in the different channels to plot them together, hence the units are "au".
We repeated the same test with IMC unlocked. We had found these gains when IMC was unlocked and their characterization needs to be done with no light in the cavity. attached are the results. Everything else is same as before.
Edit Tue May 4 14:43:48 2021 :
Overall, I would recommend setting the new gains in the suspension loops as well to observe long term effects too.
Following the conclusion, we are reverting the suspension gains to old values, i.e.
While the F2A filters, AC coil gains and input matrices are changed to as mentioned in 16066 and 16072.
The changes can be reverted all the way back to old settings (before Paco and I changed anything in the IMC suspensions) by running python scripts/SUS/general/20210602_NewIMCOldGains/restoreOldConfigIMC.py on allegra. The new settings can be uploaded back by running python scripts/SUS/general/20210602_NewIMCOldGains/uploadNewConfigIMC.py on allegra.
Unix Time = 1622676038
GPS Time = 1306711256
In 16087 we mentioned that we were unable to do a step response test for WFS loop to get an estimate of their UGF. The primary issue there was that we were not putting the step at the right place. It should go into the actuator directly, in this case, on C1:SUS-MC2_PIT_COMM and C1:SUS-MC2_YAW_COMM. These channels directly set an offset in the control loop and we can see how the error signals first jump up and then decay back to zero. The 'half-time' of this decay would be the inverse of the estimated UGF of the loop. For this test, the overall WFS loops gain, C1:IOO-WFS_GAIN was set to full value 1. This test is performed in the changed settings uploaded in 16091.
I did this test twice, once giving a step in PIT and once in YAW.
Attachment 1 is the striptool screenshot for when PIT was given a step up and then step down by 0.01.
Attachment 2 is the striptool screenshot when YAW was given a step up and down by 0.01. Note the difference in x-scale in this plot.
We measure the IMC transfer function using SR785.
We hook up the AOM driver to the SOURCE OUT, Input PD to CHANNEL ONE and the IMC transmission PD to CHANNEL TWO.
We use the frequency response measurement feature in the SR785. A swept sine from 100KHz to 100Hz is excited with an amplitude of 10mV.
Attachment 1 shows the data with a fit to a low pass filter frequency response.
IMC pole frequency is measured to be 3.795KHz, while the ringdowns predict a pole frequency 3.638KHz, a 4% difference.
The closeness of the results discourages me from calibrating the PDs' transfer functions.
I tend to believe the pole frequency measurement a bit more since it coincides with a linewidth measurement done awhile ago Gautam was telling me about.
I think of trying to try another zero-order ringdown but with much smaller excitation than what used before (500mV) and than move on to the first-order beam.
Also, it seems like the reflection signal in zero-order ringdown (Attachment 2, green trace) has only one time constant similar to the full extinction ringdown. The reason is that due to the fact the IMC is critically coupled there is no DC term in the electric field even when the extinction of light is partial. The intensity of light, therefore, has only one time constant.
Fitting this curve (Attachment 3) gives a time constant of 18us, a bit too small (gives a pole of 4.3KHz). I think a smaller extinction ringdown will give a cleaner result.
A little more information about the IMC loop tweak...
I increased the overall IMC loop gain by 4dB, and decreased the FAST gain (which determines the PZT/EOM crossover) by 3dB. This changed the AO transfer function from the blue trace to the green trace in the first plot. This changed the CARM loop open loop TF shape from the unfortunate blue shape to the more pleasing green shape in the second plot. The red trace is the addition of one super boost.
Oddly, these transfer functions look a bit different than what I measured in March (ELOG 11167), which itself differed from the shaping done December of 2014 (ELOG 10841).
I haven't yet attempted any 1F handoff of the PRMI since relocking, but back when Jenne and I did so in April, the lock was definitely less stable. My suspicion is that we may need more CARM supression; we never computed the loop gain requirement that ensures that the residual CARM fluctuations witnessed by, say, REFL55 are small enough to use as a reliable PRMI sensor.
I should be able to come up with this with data from last night.
Well, green looks better than blue, but it makes the PCDRIVE go high, which means its starting to saturate the EOM drive. So we can't just maximize the phase margin in the PZT/EOM crossover. We have to take into account the EOM drive spectrum and its RMS.
Also, your gain bump seems suspicious. See my TF measurements of the crossover in December. Maybe you were saturating the EOM in your TF ?
Lets find out what's happening with FSS servos over in Bridge and then modify ours to be less unstable.
I've been thinking about the IMC WFS. I want to repeat the sort of analysis done at LLO where a Finesse model was built and some inferences could be made about, for example, the Gouy phase separation b/w the sensors by comparing the Finesse sensing matrix to a measured sensing matrix. Taking the currently implemented output matrix as a "measurement" (since the IMC WFS stabilize the IMC transmission), I don't get any agreement between it and my Finesse model. Could be that the model needs tweaking, but there are several known issues with the WFS themselves (e.g. imbalanced segment gains).
Building the finesse model:
Some notes about the WFS heads:
Update 215 pm 5/6: adding in some comments from Rana raised during the meeting:
This is the doc from Keita Kawabe on why the WFS heads should be rotated.
OK so the QPD segments are in the "+" orientation when the 40m IMC WFS heads are mounted at 45 deg. I thought "+" was the natural PIT/YAW basis but I guess in the the LIGO parlance, the "X" orientation was considered more natural.
MC WFS Demod board needs some attention.
Tomislav has been measuring a very high noise level in the MC WFS demod output (which he promised to elog today!). I thought this was a bogus measurement, but when he, and Paco and I tried to measure the MC WFS sensing matrix, we noticed that there is no response in any WFS, although there are beams on the WFS heads. There is a large response in MC2 TRANS QPD, so we know that there is real motion.
I suspect that the demod board needs to be reset somehow. Maybe the PLL is unlocked or some cable is wonky. Hopefully not both demod boards are fried.
Please leave the WFS loops off until demod board has been assessed.
After the CDS upgrade team called for a day (their work TBD), I took over the locked IMC to check how it looked like.
The lock was robust but the IMC REFL spot and the WFS DC/MC2 QPD were moving too much.
I wondered if there were something wrong with the damping. I thought MC3 damping seemed weak, but this was torelable level.LR
During the ring down check of MC2, I saw that the OSEM signals were glitchy. In the end I found it was LR sensor which was glitchy and fluctuating.
I went into the lab and pushed the connectors on the euro card modules and the side connectors as well as the cables on the MC2 sat amp.
I came back to the control room and found the MC2 LR OSEMs had the jump and it seems more stable now.
I leave the IMC locked & WFS running. This sus situation is not great at all and before we go too far, we'll work on the transition to the new electronics (but not today or next week).
By the way the unit of the signals on the dataviewer didn't make sense. Something seemed wrong with them.
The IMC and the IMC WFS kept running for ~2days. 👍
I wanted to look at the trand of the IMC REFL DC, but the dataviewer showed that the recorded values are zero. And this slow channel is missing in the channel list.
I checked the PSL PMC signals (slow) as an example, and many channels are missing in the channel list.
So something is not right with some part of the CDS.
Note that I also reported that the WFS plot in the above refered previous elog has the value like 1e39
Thanks for pointing out that EPICS data collection (slow channels) was not working. I started the service that collects these channels (standalone_edc, running on c1sus), and pointed it to the channel list in /opt/rtcds/caltech/c1/chans/daq/C0EDCU.ini, so this should be working now.
controls@c1sus:~$ systemctl status rts-edc_c1sus
● rts-edc_c1sus.service - Advanced LIGO RTS stand-alone EPICS data concentrator
Loaded: loaded (/etc/systemd/system/rts-edc_c1sus.service; enabled; vendor preset: enabled)
Active: active (running) since Sun 2022-10-09 23:30:15 PDT; 10h ago
Main PID: 32154 (standalone_edc)
├─32154 /usr/bin/standalone_edc -i /etc/advligorts/edc.ini -l 0.0.0.0:9900
1. The response of the IMC WFS board was measured. The LO signal with 0.3Vpp@29.5MHz on 50Ohm was supplied from DS345. I've confirmed that this signal is enough to trigger the comparator chip right next to the LO input. The RF signal with 0.1Vpp on the 50Ohm input impedance was provided from another DS345 to CH1 with a frequency offset of 20Hz~10kHz. Two DS345s were synced by the 10MHz RFreference at the rear of the units. The resulting low frequency signal from the 1st AF stage (AD797) and the 2nd AF stage (OP284) were checked.
Attachment 1 shows the measured and modelled response of the demodulator with various frequency offsets. The value shows the signal transfer (i.e. the output amplitude normalized by the input amplitude) from the input to the outputs of the 1st and 2nd stages. According to the datasheet, the demodulator chip provides a single pole cutoff of 340kHz with the 33nF caps between AP/AN and VP. The first stage is a broadband amplifier, but there is a passive LPF (fc=~1kHz). The second stage also provides the 2nd order LPF at fc~1kHz too. The measurement and the model show good agreement.
2. The output noise levels of the 1st and 2nd stages were meausred and compared with the noise model by LISO.
Attachment 2 shows the input referred noise of the demodulator circuit. The output noise is basically limited by the noise of the first stage. The noise of the 2nd stage make the significant contribution only above the cut off freq of the circuit (~1kHz). And the model supports this fact. The 6.65kOhm of the passive filter and the input current noise of AD797 cause the large (>30nV/rtHz) noise contribution below 100Hz. This completely spoils the low noiseness (~1nV/rtHz) of AD797. At lower frequency like 0.1Hz other component comes up above the modelled noise level.
3. Rana and I had a discussion about the modification of the circuit. Attachment 4 shows the possible improvement of the demod circuit and the 1st stage preamplifier. The demodulator chip can have a cut off by the attached capacitor. We will replace the 33nF caps with 1uF and the cut off will be pushed down to ~10kHz. Then the passive LPF will be removed. We don't need "rodeo horse" AD797 for this circuit, but op27 is just fine instead. The gain of the 1st stage can be increased from 9 to 21. This should give us >x10 improvement of the noise contribution from the demodualtor (Attachment 3). We also can replace some of the important resistors with the thin film low noise resistors.
Summary: The demodulator input noise level was improved by a factor of more than 2. This was not as much as we expected from the preamp noise improvement, but is something. If this looks OK, I will implement this modification to all the 16 channels.
The modification shown in Attachment 1 has actually been applied to a channel.
Attachment 2 shows the measured and expected output signal transfer of the demodulator. The actual behavior of the demodulator is as expected, and we still keep the over all LPF feature of 3rd order with fc=~1kHz.
Attachment 3 shows the improvement of the noise level with the signal reffered to the demodulator input. The improvement by a factor >2 was observed all over the frequency range. However, this noise level could not be explained by the preamp noise level. Actually this noise below 1kHz is present at the output of the demodulator. (Surprisingly, or as usual, the noise level of the previous preamp configuration was just right at the noise level of the demodulator below 100Hz.) The removal of the offset trimmer circuit contributed to the noise improvement below 0.3Hz.
more U4 gain, lesssss U5 gain
ELOG of the Wednesday work.
It turned out that the IMC WFS demod boards have the PCB board that has a different pattern for each of 8ch.
In addition, AD831 has a quite narrow leg pitch with legs that are not easily accessible.
Because of these, we (Koji and Rana) decided to leave the demodulator chip untouched.
I have plugged in the board with the WFS2-Q1 channel modified in order to check the significance of the modification.
WFS performance before the modification
Attachment 1 shows the PSD of WFS2-I1_OUT calibrated to be referred to the demodulator output. (i.e. Measured PSDs (cnt/rtHz) were divided by 8.9*2^16/20)
There are three curves: One is the output with the MC locked (WFS servos not engaged). The second is the PSD with the PSL beam blocked (i.e. dark noise). The third is the electronics noise with the RF input terminated and the nominal LO supplied.
This tells us that the measured PSD was dominated by the demodulator noise in the dark condition. And the WFS signal was also dominated by the demod noise below 0.1Hz and above 20Hz. There are annoying features at 0.7, 1.4, 2.1, ... Hz. They basically impose these noise peaks on the stabilized mirror motion.
WFS performance after the modification
Attachment 2 shows the PSD of WFS2-Q1_OUT calibrated to be referred to the demodulator output. (i.e. Measured PSDs (cnt/rtHz) were divided by 21.4*2^16/20)
There are three same curves as the other plot. In addition to these, the PSD of WFS2-I1_OUT with the MC locked is also shown as a red curve for comparison.
This figure tells us that the measured PSD below 20Hz was dominated by the demodulator noise in the dark condition. And the WFS signal is no longer dominated by the electronics noise. However, there still are the peaks at the harmonics of 0.7, 1.4, 2.1, ... Hz. I need further inspection of the FWS demod and whtening boards to track down the cause of these peaks.
ELOG of the work on Thursday
Gautam suggested looking at the preamplifier noise by shorting the input to the first stage. I thought it was a great idea.
To my surprise, the noise of the 2nd stage was really high compared to the model. I proceeded to investigate what was wrong.
It turned out that the resistors used in this sallen-key LPF were thick film resistors. I swapped them with thin film resistors and this gave the huge improvement of the preamplifier noise in the low frequency band.
Attachment 1 shows the summary of the results. Previously the input referred noise of the preamp was the curve in red. We the resistors replaced, it became the curve in magenta, which is pretty close to the expected noise level by LISO model above 3Hz (dashed curves). Unfortunately, the output of the unit with the demodulator connected showed no improvement (blue vs green), because the output is still limited by the demodulator noise. There were harmonic noise peaks at n x 10Hz before the resistor replacement. I wonder if this modification also removed the harmonic noise seen in the CDS signals. I will check this next week.
Attachment 2 shows the current schematic diagram of the demodulator board. The Q of the sallen key filter was adjusted by the gain to have 0.7 (butter worth). We can adjust the Q by the ratio of the capacitance. We can short 3.83K and remove 6.65K next to it. And use 22nF and 47nF for the capacitors at the positive input and the feedback, respectively. This reduces the number of the resistors.
I have implemented the modification to the demod boards (Attachment 1).
Now, I am looking at the noise in the whitening board. Attachment 2 shows the comparison of the error signal with the input of the whitening filter shorted and with the 50ohm terminator on the demodulator board. The message is that the whitening filter dominates the noise below 3Hz.
I am looking at the schematic of the whitening board D990196-B. It has an VGA AD602 at the input. I could not find the gain setting for this chip.
If the gain input is fixed at 0V, AD602 has the gain of 10dB. The later stages are the filters. I presume they have the thick film resistors.
Then they may also cause the noise. Not sure which is the case yet.
Also it seems that 0.7Hz noise is still present. We can say that this is coming from the demod board but not on the work bench but in the eurocard crate.
Today I did a lot of steps to eventually reach to WFS locking stably for long times and improving and keeping the IMC transmission counts to 14400. I think the main culprit in thw WFS loop going unstable was the offset value set on MC_TRANS_PIT filter module (C1:IOO-MC_TRANS_PIT_OFFSET). This value was roughly correct in magnitude but opposite in sign, which created a big offset in MC_TRANS PIT error signal which would integrate by the loops and misalign the mode cleaner.
WFS offsets tuning
WFS loops UGF tuning
OLTF measurements were done using this diaggui file. The measurement file got deleted by me by mistake, so I recreated the template. Thankfully, I had saved the pdf of the measurements, but I do not have same measurement results in the git repo.
Today, I worked on WFS loop output matrix for PIT DOFs.
Sun Dec 4 17:36:32 2022 AG: IMC lock is holding as strong as before. None of the control signals or error signals seem to be increasing monotonously over the last one hour. I'll continue monitoring the lock.
Mon Dec 5 11:11:08 2022 AG: IMC was locked all night for past 18 hours. See attachment 4 for the minute trend.