40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  Cryo Lab eLog, Page 8 of 60  Not logged in ELOG logo
ID Date Author Type Category Subject
  2646   Tue Feb 16 11:30:14 2021 shrutiUpdatePSOMABeat recovered, fiber organization, alignment

I played around with the temperature and current to recover the beat (Attachment 3):

south laser temp: 9.373 kOhm, north laser temp: 9.602 kOhm

The low frequency oscillation still remains.


I decided to re-organize the fiber to make the north and south paths more evident visually. In the process, probably because I could not quite see the fiber coupler and was inserting it incorrectly or I somehow damaged the fiber coupler, the north coupler on the ADAFCB4 did not mode-match adequately so I replaced it with a single coupler borrowed from the west path in Johannes' setup. Now both paths at the output of the fiber show ~2 mW power and the beat is at around -15 dBm. (Attachment 1).


I've replaced the south path fiber to free-space coupler mount and post with an adjustable one to get the beam height to 4 inches. (Attachment 2)



  2645   Mon Feb 15 21:25:00 2021 ranaNoise HuntingLaser"North" current driver having slow current fluctuations

before the current driver went bad, I took the beatnote time series. Its not bad!

I couldnt get the .mat file to load (although Shruti did...), so this is from the CSV file.

It looks better than the usual NPROs, and this is still on the noisy ITC 502 driver. We will need to measure the current noise of that driver and see how it lines up on this plot to figure out how low noise we really need for PSOMA.

And need to redo this measurement with longer duration once I fix one of our custom drivers.


  2644   Mon Feb 15 18:25:19 2021 ChrisComputingDAQcymac to Debian buster

The startup scripts in /opt/rtcds/x1/scripts are obsolete, so I've just now removed them. In the new OS the front end codes and DAQ software are managed by systemd. You can use either the rtcds command, or systemctl to start and stop them.

  2643   Mon Feb 15 14:57:25 2021 aaronNoise HuntingLaserpll

Looks like there are some sites that generate QR codes, which we could print and tape.


  2642   Thu Feb 11 18:51:39 2021 aaronComputingDAQcymac to Debian buster

The startup script was throwing an error that /opt/rtcds/tst/x1/target/x1???/scripts/startupX1rt does not exist (where ??? is the three letter undercase model name). I logged on to nodus to see what that should look like, and made the respective files for x1iop, x1scq, x1cry, and x1siq

sudo /sbin/insmod /opt/rtcds/caltech/x1/target/x1iop/bin/x1iop.ko > /opt/rtcds/caltech/x1/target/c1iop/logs/log.txt 2>& 1 &

After setting the permissions to allow controls and advligorts users to run the script, I'm still having trouble starting all front ends. I've attached the status screens immediately after running /opt/rtcds/scripts/startAll.sh.

  2641   Thu Feb 11 17:24:41 2021 ranaNoise HuntingLaserpll

The "noisy" driver which has 267 n the front panel and S1500270 on the PCB. Is there a way to generate a QR code sticker for the front panel?

  2640   Thu Feb 11 14:44:41 2021 aaronComputingDAQwake up spirou

enter Thu Feb 11 14:44:53 2021

I wanted to access some files on spirou remotely, but it does not wake from sleep on ssh. I found this entry on Debian wiki describing how to enable wake on lan. I followed the steps and confirmed they 'worked' in the manner described on the wiki, but putting spirou to sleep and logging on from gaston returns 'no route to host'. ssh to spirou.local works when spirou is awake, but returns varying errors ('connection timed out' or 'name or service not known', or even elog fragments followed by syntax error) when spirou is asleep.

For now, I've just set spirou to never suspend itself, which will use a bit more power but allow remote work before I figure this out later.

Exit Thu Feb 11 15:42:09 2021

  2639   Thu Feb 11 09:59:31 2021 aaronNoise HuntingLaserpll

No modification to drivers since above elog. We are using S1600246 (2 mA / tick on the coarse current control dial) to drive the N laser, and using the combi controller to drive the S laser. 

  2638   Wed Feb 10 12:39:08 2021 ranaNoise HuntingLaserpll

Has anyone modified any of the drivers since this entry from Aaron in December? Which of these drivers are we know using?

Laser Driver serial number Driver part number known Modifications Voltage limit DC current mon Measurements
East S1600246 D1200719-v4 none no 1V / 10mA (this must be wrong) Zach 2017
West S1500267 D1200719-v3 current doubled no 100 V/A Zach 2017


  2637   Wed Feb 10 09:52:45 2021 shrutiNoise HuntingLaser"North" current driver having slow current fluctuations

The .mat file was the only one I found on the Moku/iPad dated 4 Feb 21. I could open it on Matlab 2019 after re-downloading it from Box: it saved as a 1x1 struct with fields including a table for 'time' and another for 'data' which includes all the frequencies, phases, and amplitudes. The default storage is binary .li and I think Anchal has some python scripts to convert them.

  2636   Tue Feb 9 22:43:29 2021 ranaNoise HuntingLaser"North" current driver having slow current fluctuations

I spent this evening trying to find the source of the slow laser frequency fluctuations I see on the Moku.

The fluctuations are ~150 MHz peak-peak, whereas we had very steady performance before (I can't read the .mat file Shruti sent me; not sure if it got corrupted during transfer or if Moke records in a weird .mat format - I couldn't load it in matlab or python).

Looking at the current outptu of the driver I see a relaxation oscillation with a ~8 s period. This seems to correspond to the observed frequency fluctuations. I bet there's something wrong in the driver.

In the attached scope image, the yellow trace is the DC output of the power meter, the purple trace is the "total current output" of the current driver, and the blue trace is the DC output of the beatnote PD.


Also, the new labeling is a little confusing for me. Its a little hard to follow which path is north/south when the fibers go in and out of the isolators, etc. Probabl have to add more labels.


The new fiber mounts are too high. We should put them on an adjustable post for now, but get some other posts machined so that these give a 4" beam height. Check the Thorlabs CAD drawing.


We really need to scan the temperature of each laser and record the power to see where the mode-hop free zones are, or else we're flying blind.

Also, it would be good to take the TEC outputs on the back of the TEC controllers and route them into the DAQ so that we have these values all the time. Same with the diode currents. Are the DAQ channels being recorded or is the Cymac down?

  2635   Tue Feb 9 16:33:31 2021 aaronComputingDAQcymac to Debian buster

Thanks Chris. The DAC and ADC chassis were both off, so perhaps that was the cause of the red DAC light (indicating bad "FIFO status"). I sent a sine to X1:SIQ-ESD, and got the correct sine from the expected channel -- with an additional -2.2 V offset that I also see on the other DAC channels.

I tried daqd stop --all / start --all, but it just took the ADC and DAC offline. And, the FE are now crashing every minute or so... maybe I started duplicate processes, if they didn't stop correctly. Rebooting cymac didn't help.

  2634   Tue Feb 9 13:51:29 2021 aaronDailyProgressPSOMAfiber organization

Enter Tue Feb 9 13:51:25 2021 \

Going to Downs to check on our recent orders. We have received the following, which I unpacked from cardboard before bringing into lab. Locations follow part description:

  • From Thorlabs
    • CS1 cable clamps (10 pack, optomechanics drawers)
    • BFCT passive component fiber tray (6, optomechanics drawers)
    • FSR1-P10 1mm jacketed fiber optic storage reel (pack of 10, optomechanics drawers)
    • CMS010 1/4'' cable tie bases (250, tool box)
    • CMS011 releasable cable ties (100, tool box)
    • CMS021 cable tie ID tags (100, tool box)
    • F230APC-1550 fiber collimation package (3, optics cabinet)
    • AD11F fiber collimation adapter for SM1 mounts (3, optomechanics drawers)
    • K6XS 6 axis lockable O1'' kinematic mount (3, optomechanics drawers)
  • from McMaster-Carr
    • 7566K68 Cable tie mounts (3 packs, tool box)
    • 1370N14 neoprene rubber sheet, 3/32'' thick, 50A durometer (3 sheets 6x6'', optomechanics drawers)
    • 9957K12 neoprene rubber balls, 1'' diameter (5, optomechanics drawers)
    • 92196A535 1/4''-20 socket head screws, 3/8'' long (100, tool box)
    • 92196A537 1/4''-20 socket head screws, 1/2'' long (100, tool box)
    • 92141A029 washers for 1/4'' screws, 0.625'' OD (200, for use in narrow fork clamps; these seem to be the wrong size, the inner diameter ID would fit in the clamp but not the outer diameter OD. Still, good to have and we were low. In the tool box.)

I mounted two of the F230APC-1550 collimators in the K6XS kinematic mounts (with AD11F adapter) and replaced the existing fiberports with these new fixed focus fiber-to-free launches. I also mounted some of the fiber components (splitters, faraday isolators) in trays.


There are some sort of optical storage disks in the stereo. The tunes are good, but there's also a bluetooth receiver hooked up to the video port if you like that sort of thing.

  2633   Tue Feb 9 13:29:15 2021 ranaNoise HuntingLaserhow to reduce noise in fibers due to wind and temperature

In principle, the noise in optical fibers should be well predicted by the Wanser formula for thermo-refractive noise. Somewhere out there in the world, people are working on making a fiber that has opposite signs for the dn/dT and the d(dL/L)/dT, but until then TR is the limit.

However, non-fundamental temperature fluctuations are typically bigger than thermodynamic temperature fluctuations in the 0-1 Hz band. Above ~50 Hz, acoustic noise also becomes important.

Here are a collection of links with sort of simple ways to reduce these effects:

  1. IOFFE: putting the fiber in a box; related paper here.
  2. https://www.ppc-online.com/fiber-protection-tubes
  3. asdf


Wondering if we can do something easy for the relative laser noise measurements. Like putting the fiber in a metal tube with foam inside.

  2632   Tue Feb 9 11:14:48 2021 shrutiUpdatePSOMAcavity alignment

I removed some of Zach's optics and a large post from the south west corner of the table and placed the Vitek monitor.

I also placed a HWP and PBS in the transmission path to split the light between the transmission PD and the Vitek camera. Hunting for a suitable DC power supply for the camera.

  2631   Thu Feb 4 22:27:43 2021 ranaUpdatePSOMAcavity alignment

I think the way to go is to go backwards from the cavity:

  1. Adjust the length of the triangle to make the FSR smaller (easier to tune the lasers to a closer resonance)
  2. Keepiing the curved mirror curvature as is, pick the length such that the HOM resonances are far away. In my attached notebook, I do this for the Mariner input mode cleaner. Its based on Shruti's earlier analysis, but with cooler colors. The main to make sure of is that the HOM of the lower sideband don't resonate. Should be easy as long as the transverse mode spacing is > the modulation frequency. If the mode spacing is too much than a HOM of the upper sideband could also resonate.
  3. Based on the waist of that cavity, figure out a simple mode-matching solution with a single lens(?) and the ~1 mm beam that the fiber launcher can produce (check w profiler or razor blade).
  4. We should hook up the little Watec video camera to look at the cavity transmission. Its much more informative than the photodiode. There is a B&W CRT monitor on the optics cabinet that we can use.

I also looked into the monotonic beatnote claim. Turns out that the Moku was set to compare the phase of the beat note relative to a internal 30 MHz signal. So the phase on the display was just integrating away. I set it back to acquire the frequency by itself, and the Moku frequency display now looks like some tiny random noise as it should. Whoever gets to the Moku next, please save the data and send it to me with dropbox.caltech.edu. Whoever knows the Moku's gmail should set up a Dropbox account for it so we can just save there all the time.

The beat note was initially at 174 MHz and I moved the temperature of the lasers around until I got the beat down to 40 MHz. Along the way, I noticed that beat note disappears for some settings of the temperature set point. I think that this is mode-hopping of the lasers, but it seems kind of outrageous if the mode-hop free region is only 100 MHz or so. Is it actually multi-mode behavior (thus making the laser so noisy that its basically broadband) or is there some other reason the beat goes away?

I also got the new laser working (with the beam scan laptop), and it says that its stable. So the next step is to bring over the blue PM fiber that runs to the green monster table and figure out how to get it onto the table in way where it won't immediately be stepped on. Then we're ready for the infinite hat method.

Does anyone know how to calculate what stability we would need for the fiber so as to be able to measure laser noise below the level of the RIO Planex?

  2630   Thu Feb 4 16:19:35 2021 shrutiUpdatePSOMAcavity alignment

I turned on both the lasers and TECs, and cleared out all extra optics/components from around the experiment.

When I re-checked the cavity alignment I saw a large beam off the side of the main beam when the main beam was not blocked, both on the third side (Mc -> Ma). I tweaked around with the alignment quite a bit and even though the beams seemed to overlap along one side (Ma -> Mb). I could not really check the third side properly.

The transmission PD shows that it is receiving power from more than one round trip as checked by blocking the outgoing light from optic Mc, but I'm still not sure if there is interference.

I also measured the beam profiles at different points on the path (will add the data later). but in summary,  the beam seems more or less collimated everywhere with a beam diameter of 2.4 mm.


[Rana, Shruti]

The fiber launch PAF2A-18C seems to have an output waist of 3.49 mm. We are ordering fiber launches with smaller waists, but in the meantime it was previously adjusted to have minimum divergence. Today Rana adjusted the three tip-tilt/ focus knobs to get the smallest waist before the entering the cavity, but it turned out result in the huge divergence within the cavity, then it was re-adjusted to a larger beam with lesser divergence, close to what it was initially.

We also experimented with a lens, but then decided against installing one after seeing that it did not help very much.

The transmission PD reading may be noisy/ somewhat incorrect because of scattered visible light.

QUESTION: Do we have filters for 1550 nm that blocks shorter wavelengths well?

  2629   Wed Feb 3 17:56:52 2021 aaronNotesControl SystemPSOMA Temp Controller tuning


Though cymac (and thus automatic data logging) is down, the acromag channels on cominaux are still operational. If the high level library isn't right, an example script for caget- and caput-ing those channels is in gaston:~/CryoQ/scripts/step_response_test.py.

  2628   Wed Feb 3 17:32:29 2021 ranaNotesControl SystemPSOMA Temp Controller tuning

might not be worth the hassle, but it should be possible to get the PID parameters out of the TEC using its serial interface and the python pyserial package. Paco/Radhika are exploring serial interfaces so they may have some code.

  2627   Wed Feb 3 12:45:25 2021 aaronDailyProgressPSOMAstrain relieved connectors

Enter Wed Feb 3 12:34:12 2021

Set up computing environment on spirou. Conda isn't updating

Replacing cable connectors for DC power in PSOMA rack -- existing cables are not strain relieved.

  1. turn of laser current and TEC drives. Power off rack electronics requiring DC power.
  2. remove cables. Locate appropriate connector parts and crimper from EE
  3. Remove housing
  4. Cut the old connector such that there's some insulation remaining to help keep track of pin mapping
  5. Check that the new length has the rubber strain relief fitting into the housing around the fat, outer insulation -- not the interior wire insulation.
  6. Strip the ends of the conductors to the required length for the pins, and crimp on a new pin. Use the right crimper for the wire gauge.
  7. Insert the pins into the connector panel, and close up the housing.
  8. After completion, check that the pin mapping remains the same with a multimeter. Success rate today was an abysmal 50%, cables in need of repair on the workbench.

Before (1) and after (2, opposite cardinality) shots of the open housing attached.

Wed Feb 3 18:30:36 2021

  2626   Wed Feb 3 10:56:54 2021 shrutiUpdatePSOMAcavity locking

I've temporarily removed the phase modulation EOM out of the path to get more power into the cavity.

I have also changed the temperature of the south laser to 7.278 kOhms.


  2625   Wed Feb 3 08:23:12 2021 ChrisComputingDAQcymac to Debian buster

I’ll add a note here for future readers of the log, because previous logs on the subject of cymac1 OS upgrades may have made it appear a little too easy.

If you are reading this because you're having an issue with cymac1 (or another front-end computer), you may be wondering: Should I try upgrading the OS on cymac1 to fix my problem?

The answer to that question is: A resounding no!

An OS upgrade is almost never going to be the solution to any problem you may have with cymac1. In fact, it is almost guaranteed to create many new problems that you will have to solve before cymac1 becomes usable again.

Upgrading the OS on cymac1 should only be undertaken if you know exactly why it is absolutely necessary, AND when you have a lot of free time on your hands to troubleshoot the problems it will inevitably create, AND after you have verified there's a backup of the existing system that you can restore to if needed.

UPDATE: the system is running and acquiring data again (although a DAC status bit remains red). Things that had to be fixed to recover from this upgrade were the following:

  1. update daqdrc config file, environment variables in env/systemd_env, and provide symbolic links to additional config files in /etc/advligorts
  2. get new standalone_edc service running to pull in data from external epics channels
  3. rtcds build and rtcds install all models (some had been built but not installed)
  4. add advligorts user to controls group
  5. add group write permissions to files under /opt/rtcds and /frames
  6. get new local_dc service running, to combine data from all front ends for ingestion by daqd
  7. disable mx_crc_only service that is no longer needed

Also to note, rtcds start was working to get front end codes started, but rtcds stop apparently fails to remove the front end modules sometimes (follow it with rmmod when that happens).

  2624   Tue Feb 2 09:52:19 2021 aaronComputingDAQcymac to Debian buster

enter Tue Feb 2 09:52:17 2021

setting up x1oma model

Last time, I ran into symlinking or kernel issues.

on cymac:

sudo apt-get update

Get an error message: GPG error about out of date signatures in the repo, and/or the necessary public key is not available. The error message points to the apt repo's readme, with instructions to add a line to /etc/apt/sources.list.d/lscsoft.list. I uncomment from that file:

deb http://software.ligo.org/lscsoft/debian stretch contrib

but get the same error messages. some of the errors are for stretch-legacy, others for stretch.

I haven't used the realtime system since last year Chris upgraded cymac1's Debian distro from 8 (jessie) to 9 (stretch), and the slow machines to 10 (buster). So, I figure I'll go through the rtcds wiki for setting up cymac with a fine comb, starting by enabling CDSSoftware repos. Buster is now the fully supported Debian release, so to keep all our machines on the same page I'm upgrading cymac1 to Buster. 

install buster

  • Backed up the recommended repos to spirou
  • remove extra cds/lsc software in /etc/apt/sources.list.d (also backed up to spirou), and replace references to stretch in /etc/apt/sources.list to buster
  • apt update to latest version of stretch buster
  • apt-get upgrade to take care of packages that don't require new packages to be removed or installed, then apt full-upgrade to take care of the rest
  • apt install linux-image-4.19-rt-amd64 (hm... was this necessary or desirable? I ended up installing linux-image-amd64 instead, and only later realized that the cymac setup instructions have a unique kernel--see below. Maybe this is related to the kernel errors I was getting earlier?)
  • reboot

Returning to installing cds software:

make a new /etc/apt/sources.list.d/lscsoft.list and cdssoft.list, same as the old but replacing 'stretch' with 'buster'.

apt-get update
apt-get -y --force-yes install lscsoft-archive-keyring
wget -c http://apt.ligo-wa.caltech.edu/debian/pool/buster-unstable/cdssoft-release-buster/cdssoft-release-buster_1.0.5_all.deb
sudo dpkg -i cdssoft-release-buster_1.0.5_all.deb
sudo apt update
sudo apt upgrade  # ifo set to TST and site set to x1 # this step not necessary
sudo apt install cds-workstation # ERROR: dependencies on cds-crtools, ndscope, and python3-nds2-client can't be resolved. error resolved with correction to buster install steps bolded above
# in /etc/apt/sources.list.d/cdssoft-restricted.list add the line
# deb [signed-by=/usr/share/keyrings/cdssoft-unstable-archive-keyring.gpg] https://$USER:$PASSWORD@apt.ligo-wa.caltech.edu/debian buster-restricted main

setting up cymac

sudo apt install advligorts-cymac # same model directory and user
sudo install linux-image-4.19.0-6-rtcds-amd64-unsigned # ERROR: missing destination file operand, the wiki seems to omit apt
sudo apt install linux-image- unsigned

The menu structure for this new rtcds kernel is "Debian GNU/Linux>Advanced options for Debian GNU/Linux>Debian GNU/Linux, with Linux 4.19.0-6-rtcds-amd64", which I copy into /etc/default/grub, followed by

sudo update-grub

Odd, I'm still seeing the old 4.19.0-14-amd64 when running uname -r... which I uninstalled. Removing it again, repeating update-grub, and reboot did the trick. Installation complete.

The only error I'm still seeing is "Skipping acquire of configured file 'main/binary-i386/Packages' as repository 'https://apt.ligo-wa.caltech.edu/debian buster-restricted InRelease' doesn't support architecture 'i386'". This is odd, because I'm running Debian 10 on an amd64 machine, with an amd64 kernel; I last saw this error when configuring cominaux. I think it's because I 'remove'd instead of 'purge'd the old kernels. After purging the non-rtcds kernels, I reboot and... no change in error message. This is not a problem, there is an i386 architecture on cymac1 for running zurich instrument tools, but this shouldn't affect the amd64 architecture; error only occurs for cdssoft-restricted.


model frontends aren't running. Lab snacks are missing. Outlook grim.

  • Run /opt/rtcds/txt/x1/scripts/startAll.sh. Script runs without error, but I still can't connect to cymac1 through nds.
  • rtcsd status tells me all model are not loaded, and the overall status is 'degraded'. Some kernel modules are missing.
  • rtcds start --all... no change
  • Jon set up a cymac in QIL recently
  • After running systemctl reset-failed, systemctl reports overall state 'running', but running s/startAll.sh gives warnings about identical process variable names on multiple servers. I think reset-failed just silences warnings with no change in state. 
  • /etc/advligorts/env has the same environment variables as before, so problem isn't there
  • This entry from Chris suggests to me perhaps rebuilding all models will help, since they were originally compiled with the old rts
    • rtcds make --all
      rtcds install --all
    • no change in systemctl status

  • Investigating the failed systemctl units, I see that rts-daqd is among the units that failed to activate. That's a critical unit, so I inquire with journalctl _PID=1140 (the PID # of the rts-daqd process is listed with systemctl status rts-daqd). The failure is at
    • set master_config="/etc/advligorts/master" 
    • This makes sense, but is surprising--there is no such file, on cymac1 or in my backup! I hadn't heard of this file in any of the rts documentation. The only mention I can find is in a git issue.

It's nearly time for me to hand the lab over to Shruti, so I will stop for now.

exit Tue Feb 2 17:55:24 2021

  2623   Mon Feb 1 18:25:47 2021 shrutiDailyProgressPSOMAcavity locking

Some Observations:

  • Temperature of laser (in kOhms) North: 10.056, South: 10.097
  • Power (in mW) North: 7.02, South: 7.26
  • Beat frequency: ~20 MHz, -50 dBm
  • After changing south laser temperature to 10.985 kOhms, the beat moved to 40 MHz, 4 dBm (!)

Some information:

  • Ring cavity made of :

Ma R~90%, Mb R>96%, Mc R>96% RoC 1m Coastline Optics CO-SAP-1000-250-1.0MP

Ma -> Mb -> Mc -> Ma

  • Photo-diode: PDA20CS


  • Got the beam in the triangular cavity to more or less align, i.e., overlap with itself as checked on one side of the triangle
  • I hooked up the transmission PD in the transmission path of Mc
    • At different temperature settings, blocking the incoming beam (Mb -> Mc) resulted in the power becoming minimum and blocking the beam (Mc -> Ma) resulted in some variable fraction of the the nominal power which I take to be a sign of interference (?)
    • I moved around the temperature using the setpoint knob, but since the movement of the knob was nonlinear which along with a nonlinear change in power and frequency (with temperature) made it hard for me to figure out exactly what was happening in the limited time I spent on this so far

Next steps:

  • Going to explore the laser lock on the Moku:Lab for a while, after which I'll decide whether or not to continue with that or move to mixers + PDH servo box


  2622   Fri Jan 29 09:52:29 2021 karthikNotes PSOMA Temp Controller tuning

[aaron, karthik]

We met around 930, Aaron gave a brief tour of Cryo, QIL, and EE until 10.


Temp Controller starting 10.981 kOhm, Current 5.66 mA ~29.8 MHz-> Drifting high, ~3.6 dBm

  • Temp knob finicky, use Temp Tune input?
  • PID has no recorded values..

10.963 - 88 mhz

10.954 - 114

10.988 - 11.3 -> 36, check VNA - catching sideband -> switch adjust on/off -> 15.61

Run Series 1

Start at 10.981, step down by small random decrements down to 10.95 and then back up again. Record waveforms into .li file

Logging parameters: duration 10s, 15.6kHz, prefix 29_01_KR

PID is default, see pic

Start End File suffix End freq.
10.981 10.978 104334 43
10.978 10.966 104734 80.1
10.966 10.960 104820 98.8
10.960 10.953 104918 123
10.953 10.963 105213 82
10.963 10.976 105314 50.3
10.976 10.995 105353 6.44
10.981 10.984 105539 35.6-26.1
10.984 10.990 105614 7.3

Stopping here, set parameters back to original, before I play with PID etc, will analyze previous data and iterate from there..





  2620   Thu Jan 28 18:03:53 2021 shrutiDailyProgressPSOMAcavity locking

Since the south laser path had the EOM for phase modulation and the north laser path had the intensity modulator, I chose the south for locking to the triangular cavity.

The setup was roughly aligned, but I had to make a few changes:

- The optic labeled 'A' in Attachment 1 had a LH mount which made it nearly impossible to see the cavity reflection. I hunted for a RH mount for the same optic without success, while hunting I decided to switch this R>96% optic to a R~90%.

- The optic labeled 'B' seemed to be a polarizing beam splitter. I switched that to a circular shaped non-polarizing beam splitter.

- I also modified the Y translation knobs of both the fiber launches to make the beam height 4-inches

The image in attachment 1 is modified from elog 2609.

Rana entered the lab and saw that that one of the east table lasers were on and turned it off, after a few minutes I noticed that the south path did not have a beam. Since we did not realize that the south laser was in fact replaced with the beam from the east table, Rana replaced the butterfly mount with one from the cabinet; the cables for the TEC and diode current were also switched out. Once we realized that the south laser was not connected to the current setup, I reconnected it (south laser) replacing the blue fiber from the east table. Both lasers seemed to lase just fine after these changes (after much troubleshooting).

There is no thermal paste on the south laser butterfly mount (may be required later if the PID does not function optimally).

  2619   Thu Jan 28 12:42:14 2021 aaronHowToDAQgrabbing nds data

This is a brief tutorial on grabbing data from the cryo lab using nds2. I'm drawing on Duo's Q measurement script (and and the LIGO DASWG manual and other documentation)

Data in the cryo lab is logged on cymac1, which is only accessible to users within the cryo lab (like the workstations gaston and spirou). I think this means we cannot nds2 without first logging on to a cryo lab computer. Ssh-ing into cymac1 will redirect you to gaston, and the password is the same as for logging into the workstation as controls. 

ssh controls@cymac1.ligo.caltech.edu

A simple python script to grab the current lab temperature with nds2 and print to standard output:

import nds2


conn = nds2.connection('cymac1', 8088)
conn.set_parameter('ALLOW_DATA_ON_TAPE', '1')

for bufs in conn.iterate([TEMP_CHANNEL]):
     temp = bufs[0].data
  2618   Thu Jan 28 11:54:56 2021 PacoNotesEquipment LoanBeam profiler and laptops


Just went into Cryo (~11:54 PT) to drop the profiler and laptops (and chargers). Left them on the left, as you enter the lab, on top of some boxes.

  2617   Wed Jan 27 22:01:16 2021 ranaDailyProgressPSOMAlab improvements

Should we also get some of the 40m kinematic mounts? Maybe get a couple and see if its easier to use.

By 'test equip on the floor' I meant all the stuff under the PSOMA table. That should all go in racks or shelves. Once in awhile the sub-basement labs get a little flooded and we don't want electrical things on the floor for that. Maybe another cabinet on the west wall next to the big grey monster?

  2616   Wed Jan 27 11:58:25 2021 ranaMiscGeneralneeds

You can also use the scripts in https://git.ligo.org/40m/labutils/-/tree/master/moku to directly talk to moku through python or terminal instead of using the Ipad or SD card.


also, I moved Aaron's large phase tracker data file onto the SD card since the RAM is volatile and was nearly full on the iPad. We should really buy a higher capacity SD card too.


  2615   Wed Jan 27 09:11:16 2021 aaronDailyProgressPSOMAlab improvements

enter Wed Jan 27 09:11:13 2021

Facilities mentioned to Paco yesterday that they'd return this morning to complete repairs on the HVAC. I missed him at 9, but around 1pm, Modesto from facilities stopped by to check the air temperature. He measures 63 F at the lab's air intake, and 72 F near the electronics rack. He's checking on our upstream filters to ensure we're getting clean air. Also mentioned that our temperature is centrally controlled, not by the thermostat in the lab -- I was a little confused by this, but sounds like we'll have to contact them to change the temperature. The lab is comfortable for now yes

three corner hat

I'm plotting the three beat note spectra I took Monday them with the help of ctn-scripts. After fixing some improperly set conda permissions on spirou, I didn't untangle its package mess before needing to leave. I've grabbed the files to finish up later.

cable management

I'm putting together a Thorlabs order for cable management and fiber optomechanics -- mounts for our fiber splitters and other pigtailed components.

  • Fiber storage reels (FSR1)
  • fiber component trays, for holding fiber splitters, faraday isolators, etc (BFCT)
  • cable tags (CMS021)
  • Cable tie bases (CMS010) and reusable ties (CMS011)
  • cable straps (for cables that should lay flat) (CS1)

In the meantime, I'm rearranging the electronics rack and cables for increased sanity.

  • Put the L-com BNC panel mount on posts, which may be sturdier than the clamps we had been using
  • Mounting DC power strip to side of electronics rack
    • turn off lasers, then turn off DC power at the Sorensen.
    • Open up power strip to access mounting thru holes on the backside.
    • Mark in pencil the location of 2x holes on the side of the rack, and drill with a #7 gauge bit for #10 rackmount screws (check chart on wall behind EE drill press).
    • Mount the strip's base, then close the power strip as before. Turn on the DC power and lasers.

Fiber launch

The 40m uses these fixed focus collimators (F220APC-1064) in a locking kinematic mount (K6XS) for their fiber launch. Thorlabs also offers similar fixed-focus collimators in 'large beam' (7mm waist for 1550nm) and 'low aberration' varieties. It looks like the triplet lens, low aberration variety uses a slightly different mounting adapter (unthreaded c-clamp, compared with the threaded adapter for F220APC). Thorlabs suggests using epoxy for 'stacked threaded' systems where the internal and external bores both have threads, to avoid inadvertedly loosening one set of threads.

I'm inclined to go with the low aberration version, either T18APC-1550 (3.33 mm waist) or T12APC-1550 (2.27 mm waist). There are also versions of the same with 4.65 and 1.12 mm waist.


  • added a label to the Moku to remind users to plug in ipad after use
  • removing test equipment from floor... we could get something like these electronics cabinets or dolly from McMaster to go under the table. Probably the dolly is more versatile, and we could fit 2 under a table. I've ordered one dolly.


exit Wed Jan 27 15:18:12 2021

  2614   Mon Jan 25 21:59:34 2021 ranaSummaryLasertest laser

I put the new laser we're testing onto the old Q table on the south side. Its not on yet - we need a Windows laptop to use it. I am thinking that we can use the blue fiber that (right now) runs from the cryo cavities setup over to the PSOMA table. If that one is a  FC/APC connector we can just intercompare the lasers via fiber using the 3-way combiner we talked about.

I had some email exchange with facilities. They seem to be trying to debug the situation remotely and the HEPA on the south wall started blowing cold air sometime around 5-6 PM. Let's see how much the temperature changes (I have a StripTool plot running on the Shruthi workstation).

In order to believe our phase noise measurements, it would be good to engineer some wind/acoustic protection for all teh exposed fiber stuff on the north side of the PSOMA table. Anyone have any ideas?


  2613   Mon Jan 25 16:36:24 2021 ranaMiscGeneralneeds

we desperately need:

  1. a sticker on the Moku to make sure the ipad is always left chargin when unused
  2. a rational arrangement and handling of the cables, fibers on the table.
  3. remove all test equipment from the floor under the optical table; maybe get another rack?

also, I moved Aaron's large phase tracker data file onto the SD card since the RAM is volatile and was nearly full on the iPad. We should really buy a higher capacity SD card too.

I am re-running the phase tracker (this time set to frequency instead of amplitude) on the same laser setup. I have tuned the beat frequency to ~30 MHz using the "North" laser temperature.

  2612   Mon Jan 25 13:02:46 2021 aaronUpdatePSOMA 

enter Mon Jan 25 13:02:47 2021

cantilever Q materials

I moved the following materials to measure cantilever Qs from cryo lab to QIL. The are located in bins labelled 'cantilever Q' in a cabinet on the S wall of the QIL lab (see photos).

  • cantilevers
  • HeNe laser and power
  • QPD from the GeNS setup, and power
  • What I think are the clamps Zach was using for cantilever Q measurements.

I also identified in QIL the IR labs cryostat that was used for cantilever Qs, and is now set up for PD measurements. I tried lifting out the cold plate, but felt resistance while opening and stopped. Probably the electrical connections are keeping the cold plate in place, and I would like to check with the user of this cryostat before proceeding.

three corner hat

I am recording spectra of the beat note between three lasers: PSOMA north laser, PSOMA south laser, and cryo cavities East laser. I unplugged the E laser from its fiber-to-free launch (after Faraday isolator), and instead send it to the PSOMA table by fiber. Without changing settings on the cryo cavs TEC or laser driver, the E laser is 8.1 mW when it reaches PSOMA table. When measuring the beat note between the cryo cavs E laser and PSOMA north (south) laser, I send the E laser through the south (north) 90-10 pickoff BS in place of the PSOMA laser. Before recording spectra, I tune the current of the PSOMA N and S lasers until all beat notes are < 20 MHz (there is a few MHz drift during the measurements).

Summary of the measurements, all taken at the FC 1611 (AC measurements on the Moku):

  • acquisition speed: 125 kHz
  • phaselock bandwidth: 10 kHz
  • coupling: AC, 50 Ohm
  • Moku input range: 1 Vpp
  • Measurement time: 1 m 40 s
  • freewheeling: yes
Pair of lasers DC intensity mon (mW) beat note power (dBm) beat frequency (MHz)
PSOMA north and south 6.0 3 18
PSOMA north, cryo cavs east 4.3 -1 24
PSOMA south, cryo cavs east 3.5 -2 5

I notice that the FC 1611 has a UPC connector on its face... but the rest of our components have APC connectors. I'll watch out for effects from this, and order an adapting fiber. I also see sidebands at the harmonics of the beat note when the note is > 0 dBm (linear regime of 1611 is only 1mW).

The rate-limiting step is transferring data from Moku RAM to the ipad over the Wifi... should set up the Moku on ethernet, and check out ctn-scripts for automating the process.

exit Mon Jan 25 16:10:59 2021

  2611   Thu Jan 21 15:25:13 2021 aaronDailyProgressLab Work 

Enter Thu Jan 21 15:25:08 2021

resealed the HEPA filter

over the lab's air intake with Loctite silicone sealant. The rubber cement I used last time delaminated, and I'm hoping this sealant contours to the rough wall paint and proves more effective.

Historically, sealing the HEPA filter decreased the lab temperature. Attachment 1 shows the lab temperature for the last 17 days -- as you can begin to see, sealing the HEPA actually increased the lab filter. This is unfortunate, but consistent with the HVAC blowing hot air despite being set to cool.


Organizing electronics rack

I relabelled the 'E' and 'W' path to 'North' and 'South'. This should be less confusing, as the lasers' fiber components run parallel from west-to-east (one laser is north of the other). It would be good to continue labeling, including the AC power cables. I started a techmart cart for cable solutions from digikey (+torque wrench, panel mount SMA feedthroughs, to avoid sending RF through BNC, rack mount components).

Bypassed lossy AOM

The South laser is losing >8dB passing through its AOM. I've taken this AOM out of the path, so the South laser now runs from its Faraday isolator directly to its EOM. For symmetry, I've also removed the North laser's EOM, so the north laser passes through an AOM before being sent to a 90-10 beam splitter. With this modification, the power at the laser launch for the north laser is 7.3 mW and the south is 6.2 mW (the AOM is expected to be lossier than the EOM).

For PSOMA we typically think of our signal being an amplitude modulation, while the pump is phase-modulated for locking. With this modification, the north laser is playing the role of 'signal', while the south laser plays the role of 'pump.' Fine for now, but of course eventually we'll want to swap in a higher power laser for the south (pump) path.

I noticed that the latex gloves being used to softly clamp the AOM have begun to stiffen and deteriorate. Touching the glove at all breaks it into small pieces. This is worrying, so I unmounted the AOM and replaced the gloves. I will look into butterfly mounts and soft, non-degradable cable clamps for the future.


  • Pulled aside some bad BNC cables for repair (connector loose, no strain relief).

Exit Thu Jan 21 20:20:27 2021

  2610   Fri Jan 15 09:54:18 2021 aaronHowToPSOMAmode matching

[aaron, rana]

aaron enter about Fri Jan 15 09:54:39 2021. rana about

Aaron wiped down keyboard, doorknob, light switch, other high contact surfaces. Rana pointed out several improvements we should make to the optics and electronics, and we tried to collimate the beam coming out of the fiberport.

We noticed the fiber enters the port at an angle, and tried to correct for this to no avail. Turns out, the reason for the angle is to compensate for the angle on the end of FC/APC connectors. You can see an example in the diagram for Thorlabs' fixed focus FC/APC collimators (as opposed to the FC/PC collimators, where the end is perpendicular to the beam and the fiber enters at normal incidence).

We also were concerned that we were seeing only 1.3 mW out of the fiberport -- but we were measuring the W laser path, so this measurement is consistent with the power output from that laser. The E path has 3.9 mW from the fiberport, which I think is consistent with expected losses from the 10% pickoff and couplers.


accounting for observed power

I'm tracing our optical path and accounting for losses. We should at some point rename the E and W lasers, since they are no longer pumping parallel cavities. Pump and signal would be appropriate; cardinal directions are ambiguous, since the relative orientation of the beams changes before the MZ.

Which laser Power after diode (mW) Before isolator After isolator Before EOAM After EOAM Before EOPM After EOPM Before 10% pickoff After 10% pickoff After fiberport
East 21.5 21.5 18.3 18.2 8.4 8.3 4.6 4.6 3.9 3.9
West 20.4 20.5 18.4 18.4 2.4 2.6 1.5 1.5 1.2 1.2

The measurements showing increasing power across connectors (eg after diode to before isolator on W path) are correct, and I attribute to different beam diameter launching from fiber vs from connector affecting the power meter's reading. Not including connectors, these are the expected losses of the components, with anomalous losses highlighted:

Part Expected Optical loss (without connectors, dB) Observed on E path (dB) Observed W path (dB)
IO-G-1550-APC (isolator) 0.55 0.70 0.45
MXAN-LN-10 (EOAM) 3.5 3.36 8.85
MPX-LN-0.1 (EOPM) 2.7 2.56 2.39
PN1550R2A1 (pickoff) 0.8 0.72 0.97

Obviously we should be getting a lot more power from the W path -- and it's encouraging that the laser may not be at fault! Both lasers are near their nominal output power, and in particular we probably could get no more than another 0.1 mW from the E path (assuming we could recover the extra 0.15 dB at the isolator).

I turn off the W laser to inspect the fiber tips with magnifying glasses (from EE), but nothing visibly dirty. Here is an in-depth document describing inspecting and cleaning fibers. Dry wiping the EOAM's pigtails did not change transmitted power. Perhaps we can try again with a fiber microscope for a better view. The west EOAM itself has one bent pin, and is also suspect.

aligning fiberport

Afterwards, I followed Thorlabs' steps to pre-align the W beam fiberport. The E beam fiberport appears collimated to the extent I can tell by eye, but I'd like to use the beam profiler to make fine adjustments.

quick spectrum

Since the fiber PD is set up, I measured the beat note's frequency noise with Moku's phasemeter. Saved the binary and will plot when I'm with my laptop.

exit Fri Jan 15 18:16:38 2021

  2609   Thu Jan 14 17:16:35 2021 aaronDailyProgressPSOMAfree space setup

enter about Thu Jan 14 17:17:03 2021

Finished laying out the MZ, one cavity, and PDs. Need a couple more beam dumps, and should get 3/4'' posts with height appropriate for DC monitor PDs and PBS mount (as in D0901749).

Cleaned up to more easily work with two people tomorrow.

exit Thu Jan 14 19:25:15 2021

  2608   Wed Jan 13 11:07:24 2021 aaronDailyProgressPSOMAfiber + free 3 corner hat

Picked up Thorlabs and Newport packages from Downs, enter lab Wed Jan 13 11:07:36 2021

The items that arrived today are:

  • FC 1611 photoreceiver, plus 2 power cables
  • cell phone lens kit
  • FC mating sleeve brackets (double only)
  • FiberPort free space couplers, plus mount
  • LB1 beam blocks
  • 90-10 (x2), 50-50 (x1), and 99-1 (x1) fiber beamsplitters
  • rectangular optics mounting bases
  • Remaining ultima mirror mounts (left handed)

Attachment 1 is a week long hour-trend of the temperature at the cymac electronics rack.

I disassembled most of the free space PLL setup, and replaced with fiber components (2x 90-10 BS, 1x 50-50 BS, leading to 1611 and one open ended fiber). On the 90% paths of the first BS, I'm coupling the fibers to free space with the fiber ports.

I also moved the mach zehnder and cavity over, and swapped most of the optic posts-and-forks to post-and-base. Still fiddling with positions / spacings that allow adjustment.

Exit around Wed Jan 13 19:16:39 2021

  2607   Wed Jan 6 10:47:18 2021 aaronDailyProgressPSOMADC to PSOMA rack, realtime model

enter Wed Jan 6 10:47:10 2021 

Making a new cable to run from the Sorensens' rack over to PSOMA rack

  • Grabbed from EE: an appropriately chonky 3-wire cable (AWG 12, same as the old cable), the 18V 'ligo style' connectors (not needed), and some lugs. Confirm the cable is long enough
  • open up the power strip, take a photo to make sure I get the wire mapping right.
  • Remove the old (short) cable, crimp the lugs (and insulator jacket) onto the new cable and install it on the power strip.
  • Run the cable along the cable tray to the Sorensen. Turn off the power supply, then attach the lugs on that end with a 10-32 screw/washer/nut
  • Turn on the supplies, confirm +- 18 V at the power strip. All good, returned parts to EE.

PSOMA realitime model

I figured I'd initialize our realtime model for PSOMA. First, copied over x1cry.mdl (cryo cavities model) into x1oma.mdl. When I build this model, I get the following error

Linux source does not match currently running kernel.

The RCG expects the linux source to be at /usr/src/linux.  For modern
distros this should be a symlink to the source installed as part of
the kernel header package for the running kernel.  For this system:

/usr/src/linux -> /usr/src/linux-headers-3.2.88-rtcds

Please create/update this link, or set the RCG_IGNORE_KERNEL_RELEASE
environment variable to bypass this check

 I haven't seen this before, do I need to update something? I haven't made recent changes to cymac, so I would be surprised if it's running a newer version of linux / rtcds than the existing source (and only one version of rtcds exists in /usr/src anyway). Indeed, when I ask for the currently running kernel (uname -r), it is '3.2.88-rtcds'; and the link at /usr/src/linux points to /usr/src/linux-headers-3.2.88-rtcds. I would expect that the source "linux-headers-3.2.88-rtcds" matches the kernel '3.2.88-rtcds'. I'm setting the ignore flag for now, but would like to know what this means.

Setting ignore to '1' or 'True' does not avoid the error message. I don't see mention of this message on the 40m or other elogs, so I'm a bit stuck. Since the realtime model isn't critical right now anyway, that's all for now.

exit Wed Jan 6 16:27:10 2021

  2606   Tue Jan 5 16:43:44 2021 aaronLab InfrastructureTemperature Sensingfresh temperature sensor

Here is the 10 min trend for the past day. The temperature fluctuates by up to ~0.3 F over short times, and drifted by only ~0.5 F on average throughout the day. The lab is running consistently hot.

  2605   Mon Jan 4 10:46:01 2021 aaronLab InfrastructureTemperature Sensingfresh temperature sensor

Entered Mon Jan 4 10:45:56 2021 .

Temperature in the lab is 89 F at the sensor by the electronics rack. I had been planning to put our AD590s on a board today, unnecessary equipment is already off but I also turned off cominaux and the Sorensen DC supplies. There is hot air coming from HVAC.

I spent some time gathering components and checking out AD's notes on the sensor. I'm using the same components as Andrew/Johannes' old circuit, and putting it on one of Anchal's "generic op amp / diff amp" boards (D1900129-v2). After checking out the old circuit, I realized we still have it in the cryo lab -- it's just missing the AD590. I put the AD590 on a breadboard, made a few connectors, and confirmed that I get a sensible reading (83 K near the center of the room, a couple degrees hotter in/by the electronics rack where the sensor is located).

I noticed that the AC connection on one of our Tenma power supplies is loose (wobbling the cable switches the supply off/on). It's a dual supply, but since I wouldn't want to use it elsewhere I'm powering the AD590 with it (doesn't need to move, out of the way above the rack).

Here's the temperature (second trend) over 20 min, will post the overnight minute trends tomorrow.

Picked up the mess, changed the dust mats, exit

  2604   Wed Dec 23 11:13:57 2020 aaronThings to BuyPurchasesoptomechanics arrived

Entered lab around Wed Dec 23 11:14:29 2020 to bring in optomechanics from Newport, step stool from McMaster, and a few other items for around the lab.

  2603   Thu Dec 17 10:23:08 2020 aaronNotesPSOMAlab entry

enter Thu Dec 17 10:24:44 2020

Just deskwork today.

exit Thu Dec 17 16:54:59 2020

  2602   Mon Dec 14 10:32:51 2020 aaronUpdate  

Enter Mon Dec 14 10:32:49 2020

Met with Shruti and finalized the fiber components we want to order.

Alignment procedure has apertures on the E beam path, one before the mixing BS and one after.

  1. Use apertures to determine a path for the E beam
  2. Place the PD such that signal from the E beam is maximized
  3. Align W path to the PD using the penultimate aperture. Because the penultimate aperture and PD position are shared with the E beam, the beams should be copropagating.
  4. Add short focal length lens before the PD.
  5. adjust alignment to maximize beat note.
    1. maximize DC level for E and W with the other path blocked.
    2. maximize beat
    3. repeat

W path laser is down to 25 uW at the PD. Also noticed that the y-alignment depends sensitively on the position of the final lens. The PD height is not the same as the height of the apertures, or the lens is off-center.

Exit Mon Dec 14 15:36:59 2020

  2601   Thu Dec 10 10:05:03 2020 aaronNoise HuntingLaserpll

enter Thu Dec 10 10:04:54 2020


Doesn't the phase meter just read out the noise even with no locking? I thought that was going to be the magic.

For locking, the mixer readout is in units of phase and the laser current modulation gives a proportional frequency modulation with no frequency wiggles until > 1 MHz. So it should phase lock with no integrator, but I'm not sure if the free running noise will drive it out of the phase lock or not. I wonder if its possible to use the phase meter as an error signal. It would be much easier to lock frequency instead of phase via a mixer.

Hm, hadn't tried the phasemeter application. I'll check it out now... if I understand your second comment, you're saying because

f = \frac{d\phi}{dt} \\ \implies \int f dt = \Delta \phi

an error signal proportional to phase is already integrating the frequency error. Makes sense, but does 'easier to lock frequency instead of phase via a mixer' follow, or is that unrelated? 

The Moku phasemeter does produce a nice power spectrum. Here it is up to 200 Hz, I'm working with Anchal's ctn-scripts and pymoku to get the higher frequencies. 

Still odd that the beat amplitude is so small. Let's check:

quantity Power of E beam @ 1611 (power meter) Power of W laser @ 1611 (power meter) 1611 DC voltage from E beam 1611 DC voltage from W beam DC voltage gain responsivity @ 1550 (approx) Expected DC voltage due to E beam Expected DC voltage due to W beam
value 307.8 uW 65.7 uW  -1.55 V  -400 mV 10 V/mA 1 A/W -3.08 V -660 mV

Looks like neither beam is producing the expected photocurrent, but because the error is not the same factor for both beams I suspect alignment / beam size. I'm aligning with some apertures to avoid smearing the beam on lenses. Aligning each beam led to more power, but my procedure doesn't simultaneously align both beams. 

exit Thu Dec 10 15:11:30 2020

  2600   Sun Dec 6 19:22:10 2020 ranaNoise HuntingLaserpll

Doesn't the phase meter just read out the noise even with no locking? I thought that was going to be the magic.

For locking, the mixer readout is in units of phase and the laser current modulation gives a proportional frequency modulation with no frequency wiggles until > 1 MHz. So it should phase lock with no integrator, but I'm not sure if the free running noise will drive it out of the phase lock or not. I wonder if its possible to use the phase meter as an error signal. It would be much easier to lock frequency instead of phase via a mixer.


  2599   Fri Dec 4 10:36:18 2020 aaronNoise HuntingLaserpll

Entered lab, then grabbed a spool of cable from EE, started elog Fri Dec 4 10:37:52 2020

thought about filters. The narrowest line I managed (yes really) is in the attached screenshot. I amplified +40 dB with Agilent 8447A before the splitter.

exit Fri Dec 4 16:14:19 2020

  2598   Thu Dec 3 11:57:52 2020 ranaNoise HuntingLaserW laser oscillating

Definitely always use 50 Ohm for all inputs and output with RF frequencies. The Moku should be able to drive the current modulation input of the diode driver to directly phase lock based on the 1611 signal with no amplifiers. A 1 mW beat should give you ~1 Vpp, which is a few dBm.

  2597   Thu Dec 3 10:18:07 2020 aaronNoise HuntingLaserpll

Entered somewhat before Thu Dec 3 10:18:07 2020

finishing up the PLL. I still need to set an appropriate gain for the LO, but in the meantime I'll try to use the Moku's laser lock feature


This is pretty straightforward. Moku has an internal oscillator and lets you control the LP (corner frequency) and controller filter (proportional gain, integrator frequency, differentiator frequency, integrator saturation level, differentiator saturation level). I'm driving the E laser HF and LF inputs from the Moku outputs. Quickly acquire a lock and play around with filter settings for a while.

exit Thu Dec 3 12:30:47 2020

  2596   Wed Dec 2 13:37:55 2020 aaronNoise HuntingLaserW laser oscillating

Enter Wed Dec 2 13:37:57 2020

I'd like to know whether the problem is in the W laser driver (S1500207) or the diode (104987). I swap the E laser drive over to the W laser diode, so am driving diode 104987 with driver S1600246. I still don't see the E laser driver oscillating, but the W diode power still reaches only 1.38 mW before falling off (ramping up current at T_set). Just in case, I also drive the W diode using the E driver but the cable from the original W driver (checking that the cable isn't the cause); no oscillation.

Entered EE to obtain grounding strap (Wed Dec 2 14:10:12 2020). On second thought, putting off diagnosing the W laser drive and just continuing with ITC 502. I did get a look inside, and don't actually see evidence that R33 was changed, just a added resistor from R30 to V_ref mentioned on the schematic.

Found this elog helpful for making sure the pinout for ITC 502 is compatible. Oddly, I can't set the laser diode current limit higher than 10mA, though the range of the device is +- 200 mA. I suspect the left 2 digits of the current display are just inoperable, because I can turn the adjustment pot in either direction until the digits stop moving. I set the current limit at 149 mA by counting by 10, and will double check the eventual current from the monitor. the diode definitely is nearing end of life, I'm only getting 1mW at nominal setpoint. I was squeezing a little extra out of the custom laser driver, but probably at the expense of the diode.

With the thorlabs driver, we're down to 1 peak at last!

Looking at the spectrum today, maybe the low frequency cutoff I noticed yesterday is just a frequency offset issue?

switching over to Moku... wow this is a user friendly device.

  • beat note drifts by MHz over minute timescales
  • FWHM ~ 140 kHz
  • 35 dB above noise floor

Phaselock setup

  • Photodiode output to splitter (ZFSC-2-4+)
    • Long cable to 83 dB (!!?) gain amplifier to mixer (ZAY-1B) LO input. Mixer requires +23 dBm LO.
    • short cable to [gain] to mixer (ZAY-1B) RF input

While choosing the gain for producing an LO from the 1611 output, I tried to measure the absolute power with the Moku. When I change the range from 1 Vpp to 10 Vpp, the overall peak height changes by almost 10 dB. I wondered if this were just a clipping issue, so I hooked up the Moku function generator output to the channel 2 input to see if the behavior is consistent. The beat note is 300 mVpp at the laser settings I'm currently using, which should be about 0 dBm dissipated into 1 MOhm. Sourcing 300 mVpp from the Moku into its channel 2 input gives the correct power reading, but the 300 mVpp from the 1611 reads -25 dBm with 10 Vpp range and -34 dBm with 1 Vpp range. What's going on?

  • Probably just outside the linear range of the 1611, which saturates at 1mW input power. I'm seeing some unexpected sidebands (>> 9MHz), so nonlinearity makes sense. We're still well below the 10 mW damage threshold, but should add back the 90-10 pickoff mirrors now that alignment is done.
  • After adding the 90-10 (power) BS, the AC output of the 1611 has ~3 mVpp (amplitude). nice
  • The beat is now -53 dBm. That's still below what I'd expect for a 3 mVpp signal; Moku thinks it's seeing a 1.45 mVpp
    • I want to attribute this to some setting of the oscilloscope or Moku, but it's also not consistent with my expectation. moku shows 6 mVpp when measuring its 3 mVpp output with 1 MOhm input impedance, suggesting there's a factor of 2 I'm missing somewhere; on the other hand, the signal from the 1611 needs a factor of 1/2 to be consistent with the oscilloscope measurement. Awkward.
  • started these measurements at 50 Ohm input impedance for the SA, but when results didn't make sense I repeated at 1 MOhm. Should have stuck with 50.

I would have to string together eg 2x SR560 to get enough gain. I guess that's fine, because the noise floor of the SR560 is O(nV/rtHz)... but it gives me the feeling of doing something wrong, and I'm hungry so will pick it up tomorrow.

exit Wed Dec 2 18:42:15 2020

ELOG V3.1.3-