Nishino,

I measured the Open loop trandfer function (OLTF) and other properties of the green locking path. When locking the cavity with mass feedback, there was instability in transmission power at the frequency of 670 Hz. This oscillation cannot be suppressed just by increasing the UGF; the locking was lost. 

To investigate the reason, I measured the OLTF. It is weird that the phase of A*H decreases above ~1 kHz, while its gain is flat (oltfotherplot.png). The dip structure at 670 Hz can also be seen. I also plotted the optical gain (H) in opticalgain.png.

I also measured the power spectrum density of the error and feedback signal. From the error spectrum, one can infer the spectrum of the disturbance out of loop (m/rHz). displacementoutofloop.png shows the ASD and the accumulated RMS noise. At 10 Hz, RMS is ~3 nm. When it is locked, the transmission light oscillates across the full resonance curve. From this phenomena, one can estimate the amplitude of oscillation as:

30 cm (cavity length) * 500 MHz (cavity FSR) / 80 (Finesse) / 600 THz (GR frequency) ~ 3 nm.

which is almost consistent with the plot.

This instability might be caused by a bad bonding of the mirror with piezo.

There was issue with 0deg calibration file (empty..)

Using other files, together with a condition of Jones matrix homogeneity of theta_r > pi - 0.01 and otherwise using semi-homogeneous case for eta < eta_max we can still get the attached plot of the QWP efffective retardation.

With these conditions we use the couples (15/30, 15/45, 30/45).

The result is compatible with LC measurement (88.5 +/-0.6 deg) and with vendor specifications.

I'll repeat the measurements with 0deg input polarization and also compute Jones matrix, full diattenuation/retardation and rotation.

I realigned the beam that should now propagate along (x,y) = with 0.06 deg deviation.

I reinstalled the polarization camera, updated the labview vi and found the qwp sample (previously measured by Hugo) as (x,y) = 322.75,156.1

measurement will be done with 50 averages/medians

calibration map with 1 mm step and 3 mm radius

measurement map with 2 mm step and 9 mm radius

0deg  :

no sample : Thu, Sep 5, 2024 3-07-56 PM

sample : Thu, Sep 5, 2024 3-22-20 PM

15deg

no sample :Thu, Sep 5, 2024 4-19-34 PM

sample : Thu, Sep 5, 2024 4-26-27 PM

30 deg

no sample : Thu, Sep 5, 2024 4-45-38 PM

sample : Thu, Sep 5, 2024 4-52-12 PM

45deg

no sample : Thu, Sep 5, 2024 5-15-55 PM

sample : Thu, Sep 5, 2024 5-21-25 PM

60 deg

no sample : Thu, Sep 5, 2024 5-43-24 PM

sample : Thu, Sep 5, 2024 6-05-04 PM

75 deg

no sample : Fri, Sep 6, 2024 1-47-12 PM

sample : Fri, Sep 6, 2024 1-53-11 PM

The specs of EOM mentioned in the sheet received with the item is different from website.

Input capacitance =14 pF

Amplitude modulated Vpi=508V

Extinction ratio = 18 dB

a possible explanation for the lack of high homogeneity of our measured Jones matrices could be the not collimated enough beam.

I installed a f=-50mm lens as in figure 1.

I installed razor blades for beam characterization

when z = 25mm, blades are at 68 mm (last steering mirror to rotator mount) + 21 mm (rotator thickness) + 58 mm (rotator to blade) = 147mm

horizontal cut of the beam at (x,y,z) = 360,67,25

vertical cut at (x,y,z) = (315,140,25)

After going through the few lenses we have left, I found that placing a f=-50mm between the shutter and already installed lens f=50mm, I could get a somewhat collimated beam across few meters.

The lens is mounted in a removable magnetic mount for quick switch between absorption and birefringence measurement (fig 1).

Using razor blades, I obtained a beam size of about 900 um (in fig 2 which is showing both vertical and horizontal dimensions of the beam).

The EOM is supposed to be driven at some resonant frequency. We have chosen around 7MHz in our case. This makes the 4th harmonic of readout available in the BW of Moku which is 30MHz.

Fig 1 Design of the circuit

Fig 2: Transfer function calculated using Zero

Fig 3: Impedance of the circuit

Fig 4: Smith chart of the circuit, considering 50ohm as characteristic imepdance

This is not the final design. Still requires tuning of some parameters to have good 50 ohm at resonant frequency. 

Nishino,

I measured the response of the new PFD circuit. Input signals are phase-locked sinusoidal waves, generated by Moku:Lab. I scanned the relative phase over 360 degrees. Attached file is the result.

Linear range is wider than the spec sheet of AD9901.

As the C directory of PCI PC is full I moved the Dropbox folder to the D directory.

I modified the robocopy script accordingly.

Don't forget to update the saving folder for all PCI labview scripts to avoid crash.

This night it took roughly 2h to perform this automatic back-up.

There is also now /purge option that will remove deleted files from the source folders.

I removed the /v and /tee options.

The initial version did not work so I had to modify the file (the 'robocopy_test.bat' on PCI pc desktop).

It is working properly so I set up the daily back-up of the dropbox to the NAS BIGFOOT folder.

I used the following parameters :

/e : copy folders and sub-folders

/copy:DATSO : copy data, attributes, timestamps, security and ower infos of all files

/r:5 : retry 5 times to run

/w:5 : wait 5s before new trial

/MT:64 : multi-thread copy

/tee : display output in console (maybe can be erased for the automatic back-up)

/log+:Z:\group\tama300\MIR\Shareslog_ShareName_%date:~-10,2%-%date:~7,2%-%date:~-4,4%.txt : create daily log file

/v : verbose

After going through few curved mirrors in TAMA it seems a convenient combination could be a pair of RoC = 0.3m awith R = 99% and cavity length of 0.55m.

All curved mirrors I could find are stored in the large dessicator.

The attached plot shows that we should get about 2mm radius beam 75cm before the cavity where we could place our birefringent sample.

I set up on PCI a test robocopy script that should automatically copy the folder 'C:\Users\tama3\Dropbox\PCI\robocopy_test' to the NAS folder '\group\tama300\MIR' everyday at 2 in the morning.

I wrote a small script 'robocopy_test' saved on pc desktop and set up a recurring task from windows task scheduler 'Robocopy script'

I'll play a bit with it during this week and if everything goes smoothly I'll set up the back-up of PCI and BIGFOOT pc.

I wanted to investigate the weak error signal of the GRMC and its somewhat sensitive response to acoustic noise. Actually I found the SHG has a bit of the same behaviour. I'm not sure what happened, but recently I noted I had to turn the control loop gain way down to avoid oscillation. I probably should have checked the open loop transfer function of both but I got sidetracked by some other thoughts. Anyway it's a bit mysterious why the ideal gain went down for the SHG and GRMC.

Since the SHG temperature was not optimized in a long time, I changed it from 3.163 (174 mW green) to 3.192 (318 mW green). Actually this seems like a bit more, it was supposed to be 280 mW max. Input IR is about 777 mW directly in front of the SHG, compared to 680 mW 3151. Overall SHG efficiency is still 41% though. However, I later noticed that the IR Sensor type power meter (Thorlabs S-121C) measures 655 mW IR at the SHG, while the stick type sensor (S-130C) measures 777 mW. I did a small profile of the temperature vs green.

The green error signal amplitude is a bit weak. However, the green lock PD is very sensitive. Since it locks and outputs stabilized green power to the OPO, maybe I should just keep going and see if the OPO works properly. It seems quite surprising that we can lock despite using 88 MHz IR sidebands that are going through three dichroic mirrors. The 88 MHz signal at GRMC_REFL_RF is about -14.7 dBm. The green lock PD has bandwidth 100 MHz so the frequency doubled green sidebands don't show up in the PD spectrum.

Nishino,

I measured the beam profile of the green laser before the PCM. The beam widths are {0.091, 0.111} mm for {x, y}, the positions are at z={320, 257} mm from the starting point (z=0).

z=0 is 900 mm away from the PCM. Using these values, I derived the position of two lenses, f150 and f400, at z = 134 mm and 566 mm, achieving the best mode-matching of 93%.

We can save data from Moku using labview "Complete Integration for MokuGo". We can currently save the voltage data. Averaging exists for this as well.

This MokuGo has 2 channels and we can save both channels simultaneously if required.

First I tried Connecting NI-SCX1-1000 to labview. Seems this component has been discontinued and there are hence several issues connecting with labview. Bascially, I could't figure it out and I gave up very quickly.

Next, I connected the PD to Moku. 

PD output range is 0-10V. Moku input range is ±25V. It seems for most of PD interfacing, its better and simpler to use Moku. 

I installed Moku drivers to be used in Labview. Moku is now available in "Functions/Liquid Instruments Moku" during VI development.

Moku is connected to Labview using usb connections as mentioned in the liquid instrument instrucions page

On a side not, Moku has two applications, "Waveform Generator" and "Arbitrary Waveform Generator". The former one can generate signal upto ±5V and 20MHz(moku specs), whereras the latter one only can generate upto ±2.5V and 10MHz(much lower than specs). Its better to use "Waveform generator" to be able to easily meet specs. 
 

ILM Sample

average = 50
air r = 3 ; sample r = 14
center x= 323.59; y= 156.7210 ; z=60

azi=0

air : Tue, Aug 6, 2024 9-04-34 AM.txt

sample :Tue, Aug 6, 2024 9-13-19 AM.txt

azi=45

air : Tue, Aug 6, 2024 11-14-14 AM.txt

sample :Tue, Aug 6, 2024 11-21-40 AM.txt

azi=15

air : Tue, Aug 6, 2024 1-32-48 PM.txt

sample : Tue, Aug 6, 2024 1-38-53 PM.txt

azi=30

air : Tue, Aug 6, 2024 3-46-54 PM.txt

sample : Tue, Aug 6, 2024 3-58-21 PM.txt

azi=60

air : Tue, Aug 6, 2024 5-53-21 PM.txt

sample :Tue, Aug 6, 2024 5-59-05 PM.txt

azi=75

air : Wed, Aug 7, 2024 7-19-44 AM.txt

sample : Wed, Aug 7, 2024 7-28-12 AM.txt

I borrowed Ratchet and couple of hex bits for today's work. I will return in by end of the day.

we still have som issue ith 50 average so we try with 100

QWP
average = 100
air r = 3 ; sample r = 5
center x= 323.59; y= 154.375 ; z=60

azi = 0

air ; Mon, Aug 5, 2024 10-03-06 AM

sample : Mon, Aug 5, 2024 10-14-33 AM

azi = 15

air ;Mon, Aug 5, 2024 10-46-54 AM

sample : Mon, Aug 5, 2024 10-57-28 AM

azi =30

air ;Mon, Aug 5, 2024 11-28-22 AM

sample : Mon, Aug 5, 2024 11-39-02 AM

azi = 45

air ; Mon, Aug 5, 2024 12-14-31 PM.txt

sample: Mon, Aug 5, 2024 12-26-25 PM.txt

azi=60

air : Mon, Aug 5, 2024 12-56-13 PM.txt

sample :Mon, Aug 5, 2024 1-07-41 PM

azi = 75

air : Mon, Aug 5, 2024 4-05-47 PM.txt

sample ; Mon, Aug 5, 2024 4-16-33 PM.txt

I managed to recover the "good" error signal of GRMC. It turns out that it actually did have something to do with the optical part, or rather the SHG output oscillation. I didn't turn down the gain on the SHG quite enough, so residual power oscillation made its way to the green error signal port (GRMC REFL RF), visible on the GRMC EPS2 OUT when the MZ is not locked to stabilize power. After adjusting appropriately, the GRMC error signal became the usual PDH with about 1 Vpk. I locked GRMC/MZ then checked the output power with the power meter - it sits fairly stably at 26.3 mW when the MZ servo is set to 4.2 offset, which is consistent with our usual work. The GRMC output power also had some oscillation at the nominal good gain - I had to reduce it from 3 to 1.4. While I am not certain yet if this is a byproduct, the GRMC output power is extremely sensitive to acoustic noise. Tapping foot makes a wave in the output. I don't entirely know if it is SHG, MZ, phase shifter or GRMC. Maybe can try also tweaking MZ gain and see if that fixes it or makes other things worse.

Initially I couldn't see error signal for IRMC, but the cable was disconnected. However, the error signal is not normal. It looks like a low detuning/transmission signal for PDH rather than the more familiar high detuning reflective form. I checked that it goes to zero when blocking IRMC REFL so it wasn't a misconnection to a transmission PD. I tried connecting some longer cables to change the demodulation phase but it doesn't change much. Then I found that the IRMC DEMOD LO signal was changed from DDS1 DAC1 (has a splitter which also goes to SHG DEMOD) to DDS2 DAC0, which right now is redundant (previously used for GRMC EOM which was removed from the locking scheme). This has a good intention since we don't have to bother with fudging the cable length or using an IQ demodulator. I checked the LO strength and DDS2 DAC0 is quite weak at -9.7 dBm. I checked the splitter channel DDS1 DAC1 (4.5 dBm) and tried using it as IRMC DEMOD LO but the error signal was still quite bad. IRMC will also not lock. 

I don't know whether or not to be concerned about DDS2 signal strength. GRMC DEMOD and the main 88 MHz EOM all come from weak signal ports but are RF amplified to about 0-5 dBm level. There are some RF amplifier ports +14 dBm that are unused, labelled "NO USE". I didn't check if that means they were broken or just unused. Next time.