ISC Meeting on 2019/03/11 14:00 - 15:00
Participants: Koji Arai, Masayuki Nakano, Osamu Miyakawa, Yutaro Enomoto, Hiro Yamamoto, Kentaro Somiya, Keiko Kokeyama, Kiwamu Izumi, Yoichi Aso, Naoko Ohishi, Yuta Michimura
Zoom meeting: https://zoom.us/j/6676627462
Next meeting
2019/4/8 (Mon) 14:00-15:00 (JST) (Michimura cannot attend on March 25)
Minutes
Thick OSTM with wrong coating JGW-T1909937
- There were also new ideas that
- use 60mm thick OSTM with big wedge
- use 2inch mirror with doughnut metal
- But we concluded that we should
- make 30mm OSTM with correct coating
- make a suspension for 30mm thick mirror (modify TAMA MC input suspension to put OSEMs?)
- when to replace depends on the sensitivity and timing
- There were also new ideas that
- Where to put OMC shutter?
- no space in between OSTM and OMC
- slow and compact one in between OSTM and OMC?
Agenda
Thick OSTM with wrong coating JGW-T1909937
- Specs
we want: 100mm dia, 30mm thick, p-pol, 45 deg, T=1%, AR<500ppm, 2 deg wedge
we ordered: 100mm dia, 60mm thick, s-pol, 3+/-10 deg, T<100ppm, AR<500ppm, 0.3 deg wedge
measured: p-pol, T=3%, AR<1000ppm (klog #8049)
- verndor calculation: p-pol, 45 deg, T=13%, AR=3%
- Options for what to do
- Order what we want
How much does the mirror cost? [-> Hirose]
- Requires modification of the suspension or borrow one of TAMA/CLIO suspensions
- To modify, it takes 6 weeks after arrival of the parts (~ 20 man-yen). 1/2 to 1/3 of the parts should be replaced [Ohishi]
- Maybe better to modify the 30mm thick mirror to suspend it with the suspension for 60mm thick [Arai]
- Order 60mm thick mirror with
- The mirror could be cheaper
- No modification of the suspension required
- Use TAMA/CLIO MC input mirror
100mm dia, 30mm thick, s-pol, 45 deg, R=99.82%, AR<0.2%, 10 sec wedge (or none)
- measured to be R=97.1% for p-pol, 45 deg [S Nagano]
- Not to replace (not at least before O3)
- Order what we want
- A/I: Calculate ghost beams for 30mm thick case [Somiya]
- Specs
Black tube for AS JGW-T1909863 [Michimura]
- Could be better to use plastic bellows type tube for flexibility?
- In-vac steering mirrors
- For some steering mirrors, p-pol beam
- STM1, STM2, POM1, POM2 around IFI (PYD-20 is used)
- steering mirrors after OFI
- PYD-20 has a transmission
T < 70ppm at 0 deg for s-pol (measured)
T < 25ppm at 30 deg for s-pol (measured)
T < 5ppm at 45 deg for s-pol (measured)
- T ~ 1000pppm at 45 deg for p-pol (specsheet)
- Do we need to order steering mirrors for p-pol?
- Also, how to dump the transmitted beams of steering mirrors?
- For some steering mirrors, p-pol beam
- Polarization of Optical tables
- s-pol BSs have been used for optical tables, but actually REFL and AS/OMC REFL/OMC TRANS beams were p-pol (due to IFI and OFI)
- Beam power reducing scheme for REFL requires p-pol
- not correct! s-pol can also be used [Enomoto]
- Why don't we keep using s-pol BSs for REFL and AS? Any comments?
REFL: JGW-T1809238
- Use reflection of thin film polarizer for beam power reduction
AS: JGW-T1909817
- Put HWP for all the beams
AS optical table layout JGW-T1909817 [Aritomi]
- Procurement done except for 2inch BS and spares.
In-vac layout basically done JGW-T1909618
- A/I: Add OMC DC PD preamp to the layout to check the position. [Somiya]
- Also wind shield for AS table
- TAMA wind shield is done by VIT
Harmonic Generator modification plan JGW-T1909661 [Michimura]
- It outputs x30 instead of x24, and we need a switch and standard Dsub 3pin connector for DC power supply.
- AEL will take care of modification around DC power supply, including rear panel modifications
- Use a chassis without a rear panel and handmade DC power supply cable with a switch for the time being
- Replacement of RF components will be done afterwards, with AEL and MIF.
- A/I: buy RF components for the replacement [Michimura]
Installed: klog #8095
- Bandpass filter for 135 MHz?
- Rear panel modification on going. Design done. Rear panel arrived at Ando Lab. To be modified at the Machine shop in Hongo.
- High power compatibility of REFL table
- HWP, thin film polarizer, beam dump (water cooled power meter) were used for IMC REFL
- REFL table might not have enough space
- Beam shutter around OMC
Ref: https://dcc.ligo.org/LIGO-T1100208 (PD damaged with ~50 mJ; see Fig 6)
Ref: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=28683 (H1 OMC damage; pulse width ~0.1s)
Ref: https://dcc.ligo.org/LIGO-G1000489 (shutter speed ~1msec necessary)
- Advanced Virgo: three types of shutter
- Slow shutter in front of OMC: open only when locking OMC
- Fast shutter in between the slow shutter and OMC: Uniblitz shutter, 10msec. Digital trigger using OMC TRANS DC PD signals.
- Fast shutter in front of DC PDs: 10msec. Digital trigger using OMC TRANS DC PD signals
- Necessity of fast unlocking of OMC?
- For KAGRA, ~120W at max to OMC on IFO unlocks
current driver for OMC PZT has ~200 usec delay klog #8050
- 1msec shutter prototype already made at Titech. In-vac one to be made by summer 2019 [Somiya]
- Delay of 200usec for the driver sounds too fast. May be too noisy?
- aLIGO uses HV driver and LV driver. One PZT is driven by HV and LV, and the other PZT is driven by HV only. LV (high speed) is used for unlocking OMC.
- We should make OMC drivers ourselves.
- OMC driver delay measurement should be done end-to-end (using beam, OMC locked).
- In-vac RF PD and RF QPD
- collected info of Advanced Virgo design (RF PD from Romain from LAPP, RF QPD from Matteo from Nikhef)
Notes
- Main topics for Phase-2 [Michimura]
- Fix initial alignment scheme for full IFO
- Enomoto?
For XY arm JGW-G1808462
- Finalize ALS scheme and RF AM generation scheme and finalize lock acquisition scheme
- Enomoto?
- ISC modeling, especially ASC modeling, for intermediate IFO configurations
- Tanioka for ASC?
- Cabling diagram
Almost finalized JGW-D1808394
- In-vacuum RF PD and RF QPD
- Izumi will ask LIGO people first.
- Laser welding would be tough (resonant circuit cannot be re-tuned)
- Use similar design to OMC DC PD preamplifier?
- Readout DC PDs and OMC
In-vac amplifier board made JGW-D1809177
Cabling diagram around OMC JGW-G1809012
- Design output optics on optical tables
REFL [Aritomi] (done) Output Tables Optics wiki
POP [done for green; Aritomi for IR]
- Two RF PDs at POP table (POP and SPOP)
- POS [done for green]
Obelisk with a hole and attachement designed JGW-D1909579
- no IR (SPOS moved to AS table)
- TRX/TRY [K. Nagano for both IR and green]
TRX designed and assembled JGW-T1808962
AS [Aritomi]
- Two RF PDs at AS table (AS and SPOS)
20 mm diameter optics post arrived.
- Fix initial alignment scheme for full IFO
- OSTM coating issue
ordered s-pol, AOI 3 deg, T<100ppm instead of p-pol, AOI 45deg, T=1%
measured to be T=3% klog #8049
- under discussion in SEO
- OSTM somehow had to be thick, but Somiya forgot the reason
Sapphire mirror damage with high power klog #7931 [Hirose]
- The average threthold for coating damage is at 92 W/mm^2
- 10 W input at PRM already gives ~13 W/mm^2 at ITM/ETMs.
- Needs to consider this risk when going to higher power.
- Safety guard around output optics
- shield for OMMT1, OMMT2 (~500 mW during observation with 80W input at PRM)
- beam dump for OMC REFL (~500 mW during observation with 80W input at PRM)
- high speed beam shutter, electronics for unlocking OMC (~120W at max on IFO unlocks)
current driver for OMC PZT has ~200 usec delay klog #8050
- 1msec shutter prototype already made at Titech. In-vac one to be made by summer 2019 [Somiya]
Tentative OMC DC QPD Amplifier and procurement status klog #8047 [Somiya]
- Tentative OMC DC QPD amplifier was made with hacked Oplev DC QPD amplifier
- The final one will arrive at around mid-March
DRMI LSC with 3f, 70% SRM JGW-G1909595 [Enomoto]
- DRMI can be done with 3f
- f3 error signals during CARM offset change seem to be more affected by 3f. Strange.
RF PD and RF QPD tuning [Michimura]
- The number of RF PDs and RF QPDs necessary are as follows:
- RF PD
- f1, f2: in-air 6pcs, in-vac 5pcs
- in-air: REFL, IMMT2T, POP, POX, POY, AS
- in-vac: REFLVAC x2, POPVAC, POXVAC, POYVAC
- f3-f1, f3-f2: in-air 1pc (needs design)
- in-air: REFLNRSB
- 2*f1, 2*f2: in-air 2pcs (needs design)
- in-air: SPOP, SPOS
- 3*f1, 3*f2: in-air 1pc (needs design)
- in-air: REFLTHREEF
- f1, f2: in-air 6pcs, in-vac 5pcs
- RF QPD
- f1,f2: in-air 6pcs, in-vac 6pcs
- in-air: REFL A/B, POP A/B, AS A/B
- in-vac: REFLVAC A/B, POPVAC A/B, AS A/B
- f1,f2: in-air 6pcs, in-vac 6pcs
AEL (Shimode-san) will take care of f1 and f2 PD/QPD tuning (9pcs of RF PD and 15pcs of RF QPD) JGW-L1909582
- A/I: Design in-vac housing for RF PD and RF QPD
- A/I: Design 3f, 2f (and f3) RF PD resonant/notch circuit
- 3*f2 (135 MHz) would be tough due to the capasitance of 2mm PD. Replace PD to silicon one?
Double peak design used in aLIGO RF PDs should be implemented in future KAGRA RF PD (see, also, LHO alog #13881)
LIGO-T1100402: ISC RF Photodetector Design: LSC & WFS
- LIGO-T1400196: ISC RFPD Resonant Circuit Optimizer Code
- LIGO-T1200506: RFPD Build Status and Locator
- LIGO-T1300488: Guide to Troubleshooting aLIGO RFPDs
- LIGO-T1300315: Notes on RFPD Signal Chain Measurements
- LIGO-T1300506: aLIGO RFPD Spot Check Procedure
AS port layout JGW-T1909618 [Somiya,Michimura]
- The power of OMC REFL will be quite high (~500 mW with 80 W input, carrier TEM00 is ~20 mW).
- Put a pick-off mirror inside the vaccum chamber to reflect 99 % of the beam to dump, and extract 1% transmission in-air.
- The power will be even higher in IFO unlocks.
- A/I: High speed beam shutter, Analog circuit for unlocking OMC, Protectors for OMMTs and OSTM.
Ref: https://dcc.ligo.org/LIGO-T1100208 (PD damaged with ~50 mJ)
Ref: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=28683 (H1 OMC damage; pulse width ~0.1s)
- A/I: Check AR reflectivity of viewports [Hirose]
-> ~1% at 45 deg
- A/I: Check if OMC TRANS is extractable from OMC mirrors [Somiya]
- A/I: Buy a pick-off mirror [Somiya]
- The power of OMC REFL will be quite high (~500 mW with 80 W input, carrier TEM00 is ~20 mW).
- Mach-Zehnder modulator [Yamakoh]
- Not sure if requirement calculation using Optickle is reasonable or not
- Requirement calculated directly from the TF from MZI displacement noise to DARM and that calculated using TFs from MZI displacement to multiple f1/f2 PM/RAM amplitude/phase noise, multiplied by TFs from multiple f1/f2 PM/RAM amplitude/phase noise to DARM do not match.
- A/I:
Analytically calculate TFs from multiple f1/f2 PM/RAM amplitude/phase noise (LIGO-T1500559) to DARM and compare them with Optickle.
- Analytically calculate TFs from MZI displacement to multiple f1/f2 PM/RAM amplitude/phase noise
- A/I: Look into requirement calculations again
- Phase noise for f1 RAM, which comes from asymmetry of 1st MZI, gives the most stringent requirement
- Displacement noises of 1st and 2nd MZI do not meet the Phase 2 requirement.
Plan for O3
Agree to do 3f in O3 (Minutes of the Commissioning Meeting 20190108)
- Technical review on Jan 22
- Displacement noise, modulation depth, residual amplitude modulation, phase noise, long term stability
- Disassemble 1st MZI and 2nd MZI? Just block one of the arms (loose 7/8 of the power)?
- Future plans?
- Future plans
- Make monolithic MZM?
- Make a wind shield?
- Not sure if requirement calculation using Optickle is reasonable or not
- MZM plans, list of measurements to be done [Control room meeting with Yamakoh, Kokeyama, Guiguo, Izumi, Enomoto, Miyakawa, Michimura]
- Jan 29: MZM alignment with new posts
- Jan 29-Feb 1:
- Displacement noise measurements, coherence measurements for each optic, calibration of accelerometer signal to MZM displacement noise
- KOACH on/off, precision air-con on/off tests
- Preparations for long term stability measurement (80MHz mixer for modulation index measurement, temperature, polarization monitors)
- Feb 4-Feb 8 ~ Mar:
- Preparations for long term stability measurement continued (PSL to IOO rack cabling, IQ demodulator cabling etc.)
- IMC frequency noise measurement with MZM
- Mar 14-16: Test run for long term stability measurement
- Mar 23-24: Long term stability measurement
- Mar 30-31: Final chance for long term stability measurement
- Mar 30: Remove EOM and measure displacement noise
- April: Disassemble MZM
- IMC frequency noise measurement without MZM
- Any time:
- Measure actual phase difference between EOMs.
- Offline subtraction of noises using accerelometer signals etc.
- Commissioning can be done with MZM blocked until we need high power (July), but we will disassemble MZM once we have all the measurements done.
- OMC status
OMC was locked from the digital system klog #6846
- DC PD Preamp SN1 was broken, but replaced
- transfer function of DC PD should be measured within a few % for calibration
- DC QPD amplifier not arrived yet
Calculations on DC power and circuit noise -> Somiya & Arai
Koji's calculation: DCPDCalculation
- Electronics status
In-vac DC PD amplifier board finalized JGW-D1809177
- In-air DC QPD amplifier board not arrived yet
ITMY inhomogeneity JGW-T1808715
- We should check impact on sensitivity and locking as soon as possible
- For both O3 and full bKAGRA
- There was a discussion whether we should re-optimize OMC for BRSE or DRSE without homodyne angle optimization.
- ASC simulation by FINESSE show worse signal when ITMY inhomogeniety is added (a few % degradation in shot noise?)
- higher order contents, especially odd-order modes are worse
- But still need to check the Finesse code (alignment loop not closed yet)
- FINESSE calculation on ASC was not reliable since Gouy phase tuning was not done
- Working on Gouy phase tuning
- We should check impact on sensitivity and locking as soon as possible
RF generation using Mach-Zehnder interferometer [Kohei Yamamoto] JGW-G1707309 JGW-G1807655 JGW-G1808029
- simulation on imperfections (MZ arm asymmetry, MZ lock, EOM phase difference control etc.)
- calculate requirements on MZI displacement, EOM modulation phase difference, EOM imbalance
- Requirements derived from full Optickle simulation and those derived from transfer function between displacement of MZI and sideband power kind of matched after implementing RAM into calculation for the latter. However, we need more investigation on the following issues:
- Modulation depth for RAM generating EOM changes the requirement.
- Spectrum shapes of the requirements for mid-fringe do not match. Maybe coming from accuracy of the mid-fringe?
- Unknown frequency response of the transfer function from the displacement of MZI and sideband power
- Check if caluculated RF oscillator amplitude noise is relative one or not with simple Optickle model
- calculations for requirements on delay line lock displacement noise, parallel EOM MZ lock displacement noise, EOM modulation phase difference, EOM imbalance on-going
Relative f1 PM and f1 AM also needs to be adjusted (CQG 31, 095003 (2014)). Another EOM for f1 or relative amplitude adjuster between two EOMs?
- Requirement on carrier power loss?
- Comment from Stefan Ballmer
- Similar calculation was done in LIGO in the past. Displacement noise requirement for MZ was tough (~1e-14 m/rtHz).
- Any in-air interferometer suffer from jitter and frequency noise which cannot be suppressed.
Ref: LIGO-T040119
- RF oscillator noise requirements for MZI should also be calculated
- Other long standing tasks
- Real Low Noise VCO, also for FSS
- Fiber noise cancellation for ALS
- DC QPD holder
- IMMT1 and IMMT2 transmission
- Slow channels for monitoring things (PLC based? Beckhoff? Acromag?)
See Minutes20170410 and Minutes20190225 for other agendas.