This morning, I set the alarm levels for DM6 back to their higher threshold values of 7k counts minor and 10k counts major for both 0.3u and 0.5u since doors are back on HAMs 5/6/7.
The levels for the other DMs have already been set back (alog75671), so all LVEA DMs are now back to their Class-10k thresholds.
Matlab2023b is now available from any CDS workstation by running 'matlab2023b' from a shell.
Before the LVEA was transitioned to Operations version of laser safe, I removed the beam dump blocking the CO2X beam from getting to the perescope that was installed while doing the CO2X swap alog 75550.
WP 11696 Betsy and I replaced the two HAM6 +X cameras removed in alog75425. Guillotines also removed will need to be reinserted if camera alignment needs to be adjusted.
Fri Feb 23 08:05:09 2024 INFO: Fill completed in 5min 6secs
Jordan confirmed a good fill curbside.
TITLE: 02/23 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
SEI_ENV state: Maintenance
Wind: 3mph Gusts, 2mph 5min avg
Primary useism: 0.02 μm/s
Secondary useism: 0.26 μm/s
QUICK SUMMARY: 6.3M earthquake from the pacific rolled through overnight. nothing tripped. Slow controls still in red.
Today's activities: - HAM7 pumpdown status: ~8.5E-7 Torr - Corner pumpdown: ~7.4E-7 Torr, ~51 hours after the HV pumping started - EX pumpdown status: 3E-8 Torr. The GV will remain closed during FAC works (1-2 weeks) - At the EX RGA the bakeout temperature is ramping up; in the upcoming days the HAM7 and HAM8 RGAs will also be baked, as per WP #11713 - The inner annulus system of GV5 was valved in to the main annulus system - The hardware was prepared for the newest version Relay Tube's installation, which will be done tomorrow
Nutsinee, Daniel, Naoki
As reported in 75902, the SHG IR injection power was 180mW and the SHG GR power was only ~40mW after PMC install. To increase the SHG IR injection power, we replaced the 50:50 BS to split the SHG path and CLF/seed path with 70:30 BS. After the replacement, the IR power is 400mW before BS, 270mW for SHG path, and 130mW for CLF/seed path. After we optimized SHG alignment with pico and optimized the SHG temperature, the SHG GR power reached 60mW with 230mW SHG IR injection. This is consistent with what we had before in 67172.
The OPO can be locked with LOCKED_CLF_DUAL. The pump ISS control mon is a bit low ~3.8 so we need to align the pump AOM and fiber. We also need to align the seed fiber.
Attached an updated power budget and a SHG scan. After setting the SHG crystal temperature to 34C this morning, the SHG mode mismatch became worse (70% mode matched instead of 80%) on the floor so I reverted to crystal temperature back to ~36C. The normalized current on the SHG BBPD was increase from 0.288mA to ~0.4 mA. SHG common mode board gain is reduced by 3dB to compensate for the factor of 1.4 higher SHG trans. In addition to the BS2 that was replaced with a 30/70 we also replaced BS11 with a 10/90 (R=10%). PMC trans diode sits at the 10% pick off. The diode hasn't been recalibrated.
CLF path has been aligned. AOM1 has about 30% 1st order throughput and AOM2 has about 50% 1st order throughput. At the nominal AOM1 ISS drive point we are capable of sending about 1.5mW to the fiber coupler (we dump most of that power at the moment. Old 1/2 waveplate setting.). We are sending ~10mW into AOM1. Haven't tried to lock the CLF yet.
We also optimized the pump alignment into the fiber. Although that probably wasn't necessary. As Naoki mentioned we were able to lock the OPO with no issues but some optimization required.
Seed has a very small fiber throughput but at least there's something to look at. Optimizing that one tomorrow.
To do list tomorrow:
- Optimize seed and pump fiber throughput. Maybe optimize LO as well if we have the time although we probably won't need this any time soon.
- Lock CLF
- Calibrate PMC trans PD
Ps. I stole the BNC to HV adapter from the OPO PZT driver and forgot to return it...
TITLE: 02/22 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: None
SHIFT SUMMARY: CDS switch upgrade cleanup finished, corner station high voltage supplies were turned back on, SQZ table work and pumpdown continue.
LOG:
| Start Time | System | Name | Location | Lazer_Haz | Task | Time End |
|---|---|---|---|---|---|---|
| 19:22 | SAF | LVEA is Laser SAFE | LVEA | NO | LVEA is Laser SAFE | 03:59 |
| 16:02 | FAC | Karen, Kim | FCES | n | Tech clean | 17:11 |
| 16:55 | CDS | Jonathan, Tony, Erik | Mech room | n | FMCS, VAC switch work | 18:41 |
| 17:04 | IAS | Jason, Ryan C | LVEA - BG | n | Beam tube surveying | 20:56 |
| 17:35 | - | Betsy, Randy, Travis | LVEA - W | n | Clean room and quad container move | 18:21 |
| 17:35 | VAC | Jordan | EX | n | RGA | 18:41 |
| 17:36 | VAC | Janos | LVEA | n | Checking on turbos | 20:54 |
| 17:51 | OMC | Jeff | CR | n | DCPD meas. | 18:26 |
| 18:22 | SQZ | Julian, Camilla | Opt Lab | n | SHG work | 20:23 |
| 18:24 | VAC | Gerardo, Jordan | FCES | n | Look at purge compressor | 19:37 |
| 18:27 | FAC | Kim | H2 encl | n | Tech clean | 18:28 |
| 19:15 | SQZ | Nutsinee | LVEA - SQZ | LOCAL | SQZ table work | 21:23 |
| 19:17 | FAC | Mitchel | EX | n | Wind fence check | 19:52 |
| 19:42 | VAC | Jordan | EX | n | RGA | 20:00 |
| 20:33 | FIT | Camilla, Jeff | X-arm | n | Jogging | 21:23 |
| 21:16 | PCAL | Tony | PCal Lab | LOCAL | PCal work | 21:31 |
| 21:21 | VAC | Jordan | EX | n | RGA | 21:40 |
| 21:24 | EE | Fil | LVEA | n | Turn on HV at HAM7 | 21:48 |
| 21:25 | IAS | Jason, RyanC | LVEA | n | Beam tube surveying | 23:25 |
| 21:51 | SQZ | Nutsinee, Daniel, Naoki | LVEA - SQZ | LOCAL | SQZ PMC work | Ongoing |
| 22:07 | OPS | Betsy | LVEA | n | Checking on HAM6 cameras and shutters | 22:11 |
| 22:11 | FAC | Randy | LVEA | n | Talk to Jason | 23:09 |
| 22:19 | SQZ | Camilla, Julian, Fil | Optics Lab | LOCAL | SHG work | 23:36 |
| 23:08 | VAC | Gerardo, Jordan | LVEA | n | Checking purge air at GV5 | 23:59 |
| 23:10 | VAC | Travis | MY | n | Looking for pump cart | 23:36 |
| 23:29 | CDS | Fil | FCES, EX, EY | n | Installing SEI chassis, picking up parts | Ongoing |
The following power supplies in the corner station were powered on:
1. PZT High Voltage
2. Fast Shutter High Voltage (field chassis was enabled)
3. SR3 Heater
4. ITMX/Y Ring Heaters
The high voltage bypass on HAM6 and HAM7 was removed, see alog 75805.
With the CDS and Beckhoff work yesterday some large numbers got into the CHILLER_SET_POINT filters which requested a low setpoint for CO2X, the laser faulted and turned of when it got to 15degC at 22:06UTC yesterday. This morning I turned it back on and set H1:TCS-ITMY_CO2_CHILLER_SERVO_GAIN_GAIN from 1 to 0 (as in alog 75715) to stop any feedback form the laser being off gong to the chiller).
TJ and I plan to make sure the H1:TCS-ITMX_CO2_PZT_SERVO_GAIN_SW2R is correctly turned on and off in the Guardian code when the CO2 lasers are taken DOWN to avoid this in future.
This is correctly turned off in DOWN but the Beckhoff work on this day changed the settings (plot) and we did not rerun the DOWN state. TJ has added a detector to TCS_ITM{X,Y}_CO2 guardians to rerun DOWN state if H1:TCS-ITMX_CO2_PZT_SERVO_GAIN_OUTPUT goes >1000. Usually below 100 but has an intergrator that can act strangely on Beckhoff channel changes. Tested and loaded in both X and Y.
Today, we opened the large ISI storage conatiner that housed 4 QUAD Upper Structure assemblies, 2 Baffle Down-tube assemblies, and 2 TransMonSus lower structures. It all looked great! The wires and blades showed no signs of any rusting, the container and parts looked very clean and organized. All clamped down as last left years ago. Nothing found unexpected.
We are working in the West Bay of the LVEA, so the parts are all currenly exposed in the large cleanroom there (which is ON).
I've added a new "TCS TRENDS" button on the sitemap > OPS > WEEKLIES medm. This opens up two monthly ndscope trends (from /opt/rtcds/userapps/release/tcs/h1/templates/ndscope/famis), one for CO2 and the other for HWS. Explanation of the channels plotted below.
Thu Feb 22 08:06:46 2024 INFO: Fill completed in 6min 42secs
TITLE: 02/22 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
SEI_ENV state: Maintenance
Wind: 3mph Gusts, 2mph 5min avg
Primary useism: 0.05 μm/s
Secondary useism: 0.23 μm/s
QUICK SUMMARY: Calm morning, prep for HAM8 work today and some cleanup from CDS switch work yesterday.
Current issues:
Corner station HEPI pump controller EPICS IOC does not connect, we suspect network port issue with h1hpipumpctrll0. This is the reason for the 31 disconnected channels on the EDC.
Camera servers will be rebooted today to get all the digital cameras back online.
FMCS connection to the corner station maintenance building and chiller not operational, mech room switch replacement scheduled this morning.
I just ran a script that sweeps SHG crystal temperature. This test requires TTFSS, PMC, and SHG locked. PLEASE DO NOT TOUCH THE SQUEEZER. PLEASE WALK GENTLY NEAR SQZT0. The duration of the script is ~2hrs. Star time is 16:40.
The script sets the crystal temperature from 29C - 41C at 0.5C increment. We hope to replicate the phase matching curve done by a SURF student Nathan Zhao (P1400193 Fig 23). The script can be found in /opt/rtcds/userapps/trunk/sqz/h1/scripts/SHGphasematch.py
Here's our Sinc function. Not sure what's going on with the dip in the middle but I believe we have been operating at a good place (the right peak, ~36C). Although after the temperature scan the green production didn't come back to where it started. I moved the crystal temperature to the middle of the sinc function to confirm that the conversion efficiency really went down. I then moved the crystal temperature to the left peak of the sinc function (~34 C) and managed to recover 10mW of green.
Something else is drifting as well and it seems to have nothing to do with me. PMC PZT ran out of range this morning so I expanded the scan range down to 0V temporary (the current lock triggered at ~25 V). SHG PZT is still drifitng as I'm writing this alog while the green conversion goes down with it. For now I decidede to park the crystal temperature at 33.78C. Changes accepted in the SDF.
Attached a plot with SHG red transmission.
Kevin, Sheila, Evan, Vicky
Summary: SQZ-OMC mode scans with hot OM2, and PSAMS 120/120 vs. 200/200. From this data, we should get single-bounce SQZ-OMC mode-matching with hot OM2, check SQZ readout losses (AS port throughput), and measure OMC losses via cavity visibility when locked/unlocked to the squeezer beam. With hot OM2, in sqz single bounce, SQZ-OMC mode-matching looks a bit better with PSAMS 120/120 than 200/200.
We'll ask Jennie W. to help us fit these SQZ-OMC mode scans. She can fit the double-peak in the 2xHOM, to give an accurate measure of SQZ-OMC mode-matching with hot OM2 and these two PSAMS settings. Here is just naively calculating mismatch from the relative power in TEM20 (TEM20/(TEM00 + TEM10/01 + TEM20)), and then calculating the total power not in TEM00 (ie 1-TEM00/(TEM00 + TEM10/01 + TEM20)), to get the following estimates on SQZ-OMC mode matching:
PSAMS 120/120, scan: 10/24/23 19:46:53 UTC + 200 seconds.
--> mismatch ~ TEM20/peak_sums ~ 2%. Total incl. mismatch + misalignment: 1-tem00/peak_sums ~ 8%.
PSAMS 200/200, scan: 10/24/23 19:04:57 UTC + 200 seconds.
--> mismatch ~ TEM20/peak_sums ~ 5%. Total incl. mismatch + misalignment: 1-tem00/peak_sums ~ 12%.
We will follow-up with analysis on OMC loss measurements based on cavity visibility, more accurate SQZ-OMC mode mismatches from these scans, and checking single-bounce SQZ powers through the AS port.
---------------------------------------------------------------------------
Notes:
---------------------------------------------------------------------------
Some relevant alogs, as we try to piece together the SQZ-IFO, IFO-OMC, and SQZ-OMC mode matchings:
Thanks to Vicky for helping me update the code to work for SQZ measaurements I had some trouble fitting these in the past as the fitting code was not subtracting off the dark current on the measurements, this doesn't matter so much for mode scans using the PSL as this has a much higher power through the OMC than the SQZ beam (16mA on the DCPDs vs. 0.5 mA on the DCPDs).
For the first measurement taken on 24th October 2023, hot OM2, PSAMS (ZM4 at 120V, ZM5 at 120V).
I used 70s of data taken starting at 1382212031.
See attached plots of the mode scan with identified peaks, and the carrier 02 peaks fitted as a sum of lorentzians.
The blue line shows the data zoomed in to the C02 peak. Th red line shows the sum of lorentzians using the fitted parameters of both centre frequencies, both amplitudes, and the half-width at half-maximum of an individual peak.
The purple line shows the lorentzian sum as a function of the initial fitting parameters.
The fitted mode spacing is 149.665 - 149.153 MHz = 0.512 MHz, which is less than the expected HOM spacing 0.588 MHz from this entry which uses the original measurements by Koiji in Table 25.
The mode-mismatch is 0.0062 + 0.0071 /( 0.0062 + 0.0071 + 0.45) = 2.9 % for the 02 modes with the lower frequency mode (horizontal I think) being higher in magnitude.
Code to do run mode scans is OMCScan_nosidebands6.py and fit the data is in fit_two_peaks_no_sidebands6.py located in labutils/omcscan git reposiotory on /dev branch, ran using labtutils conda enrvironment at labutils gitlab).
Run OMCscan_nosidebands6.py with
python OMCscan_nosidebands6.py 1382212031 70 "PSAMS 120/120, SQZ-OMC 1st scan" "single bounce" --verbose -m -p 0.008 -o 2
And also it is neccessary to hard code in the C02 mode being the 5th largest mode and 01 being the third largest in order to get a good fit as the sidebands are off.
Inside OMCscan_nosidebands6.py
find the module:
def identify_C02(self):
then change the lines shown after:
#set frequency to be that of third largest peak.
to read:
third_larg = np.argsort(self.peak_heights)[-3]#third largest is 01.
fourth_larg = np.argsort(self.peak_heights)[-5]#fifth largest is 02
For the second measurement taken on 24th October 2023, hot OM2, PSAMS (ZM4 at 200V, ZM5 at 200V).
I used 80s of data taken starting at 1382209515.
See attached plots of the mode scan with identified peaks, and the carrier 02 peaks fitted as a sum of lorentzians.
The blue line shows the data zoomed in to the C02 peak. Th red line shows the sum of lorentzians using the fitted parameters of both centre frequencies, both amplitudes, and the half-width at half-maximum of an individual peak.
The purple line shows the lorentzian sum as a function of the initial fitting parameters.
The fitted mode spacing is 149.757 - 149.204 = 0.552 MHz, which is less than the expected HOM spacing 0.588 MHz from this entry which uses the original measurements by Koiji in Table 25.
The mode-mismatch is 0.019 + 0.016 / (0.016 + 0.019 + 0.42) = 0.054 = 5.4 % for the 02 modes with the lower frequency mode (horizontal I think) being higher in magnitude.
Code to do run mode scans is OMCScan_nosidebands7.py and fit the data is in fit_two_peaks_no_sidebands7.py located in labutils/omcscan git reposiotory on /dev branch, ran using labtutils conda environment at labutils gitlab).
Run OMCscan_nosidebands7.py with
python OMCscan_nosidebands7.py 1382209515 80 "PSAMS 200/200, SQZ-OMC 2nd scan" "single bounce" --verbose -m -o 2
And also it is neccessary to hard code in the C02 mode being the 4th largest mode and 01 being the third largest in order to get a good fit as the sidebands are off.
Inside OMCscan_nosidebands7.py
find the module:
def identify_C02(self):
then change the lines shown after:
#set frequency to be that of third largest peak.
to read:
third_larg = np.argsort(self.peak_heights)[-3]#third largest is 01.
fourth_larg = np.argsort(self.peak_heights)[-4]#fourth largest is 02
