The RF circuit deisgn was truned to obtain impedance of around 50ohm. This will match the impedance of circuit with Moku and not break it, at resonant frequency. Usually, the imepdance mis-matching will cause issue if you use wires longer or same order or magnitude as your resonant frequency, as wave properties start coming into play. 

Our resonant freq is 7MHz. The analysis was done is LTspice.

For our case, we have used an Opamp circuit to match impedance. 

Fig 1 is the circuit diagram

Fig 2 is the transfer function 

Plot a: V(output)/V(input) magnitude and phase

Plot b: V(input1)/I(input1) imepdance at the input (magnitude and phase) respectively

plot c: real(V(input1)/I(input1))

plot d: imaginary(V(input1)/I(input1))

Fig 3: Smith Chart

There seems to be some discrepancy between power observed without sample(HWP) and with Sample(HWP). 

The power (normalised with main laser power) with sample / The power (normalised with main laser power) without HWP is shown in Fig 1 . It basically is the gain of the optical system which is transmitted power divided by the incident power.

Both the measurements were taken in a few days gap(~4-5 days). The normalization should have taken care of the laser power fluctuation. But, for some reason, the gain of the optics is more than 1. It could be arising from some inconsistent laser power increment or some very lossy optics before power meter to observe the main laser power. It could also be that the number lies within the uncertainity of the system.

The delta=det(J^+ J) and max intensity gain plot(of the HWP jones matrix) without considering the power factor in jones vector and without normalization is shown in Fig 2. 

This is a continous work from 3717 (date: 2024.09.06)

I locked the PCC with GR laser by feeding back the signal into frequency actuator of the laser. The locking was stable, no weird behaviour at 667 Hz. The firt png is the open loop gain.

Still 667 Hz peak exists (second png). I guess this peak originates from the piezo actuator, hindering feedback locking at the same time. I will try bonding again.

[Marc, Shalika]

We installed the 30mm diameter sample.

map with sample taken with 2 mm step size and 15mm radius

below files sorted from input linear polarization angles of 0 to 75 with 15 deg increment.

Input_filenames = [ 'Mon, Sep 9, 2024 3-16-23 PM.txt', 'Tue, Sep 10, 2024 10-52-14 AM.txt' , 'Tue, Sep 10, 2024 12-39-57 PM.txt' , 'Tue, Sep 10, 2024 1-20-39 PM.txt' , 'Tue, Sep 10, 2024 2-08-09 PM.txt' ,'Tue, Sep 10, 2024 3-19-48 PM.txt']

Output_filenames = [ 'Mon, Sep 9, 2024 3-40-50 PM.txt', 'Tue, Sep 10, 2024 11-53-20 AM.txt' , 'Tue, Sep 10, 2024 12-46-08 PM.txt' , 'Tue, Sep 10, 2024 1-26-27 PM.txt' , 'Tue, Sep 10, 2024 2-13-50 PM.txt' , 'Tue, Sep 10, 2024 3-19-48 PM.txt']

The spec sheet for the 88 MHz EOM says to apply about 12-15 dBm to get near 0.2 rad modulation depth. I swapped DAC0 (-9 dBm) and DAC1 (8.2 dBm) and plugged DAC1 into +14.1 dB RF amp. I intended to use +14 -12 to achieved 12 dBm RF injection but somehow it only came out of theRF amp as 11 dBm without any attenuator. DAC0 is also amplified ~ 14 dB and sends 4 dBm to the splitter for SHG and IRMC LO.

The SHG has a reasonable error signal and locks, but has oscillation even near zero gain. The IRMC has some visible RF sidebands in the reflection PD (approximately 3% of carrier on each side) but the PDH error signal is quite small. Turning the lock switch on the IRMC servo does nothing, it doesn't even disrupt the spectrum.

I should check internal connections.

The SHG/IRMC EOM modulation is too low. The channel DDS1 DAC0 produces only -9 dBm RF power, which gets amplified to 5 dBm from the RF amplifier rack. The nominal value to be used for the EOM is 25 dBm.

I could see the IRMC PDH error signal with the familar shape from far-detuned RF sidebands, however, the signal strength of the PDH is extremely weak. I suspect this is also the reason behind the SHG/GRMC weak lock, however in the case of the SHG the input IR is 900 mW (vs 3.5 mW for IRMC) so maybe some signal can be seen despite weak modulation.

I think a temporary fix is to use DDS1 DAC1 (SHG/IRMC demod), which outputs 10 dBm, and amplify to send to the EOM. Then amplify DDS1 DAC0 + 14 dB and split into demodulation signal. This will give ~ 2.5 dBm for each demodulation signal.

I had noticed the weird electronics settings problem for the SHG and GRMC servo gain after Mitaka campus was struck by lightning in late July. Perhaps that was also connected to the RF signal generation issues.

Moku was setup to be  interfaced over usb. Connect the green one using localhost:8091

In windows powershell

netsh interface portproxy set v4tov6 listenport=8091 connectaddress=['moku-ip'] connectport=http

PS: The black one has port 8090

Measured noise of the PD and moku. The pd was connected to the CH1 and Moku CH2 was terminated with 50ohm. The Moku:Go spectrum analyzer was used to get the spectrum with 100 avg, and hamming window.

PD: PDA05CF2 : CH1 (red plot)

Moku:Go : CH2 blue in color

Fig 1: With power supply to PD off

Fig 2: With power supply to PD on

after waiting about 1h to let the laser stabilized :

0deg

no sample : Mon, Sep 9, 2024 10-26-56 AM

sample : Mon, Sep 9, 2024 10-32-56 AM

Adding these maps we now have (0/30, 0/45,0/60,15/30,15/45,30/45) couples and effective retardation attached to this entry.

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 =12 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