I modified the pitch damping filter for the INPUT mirror to reduce the 9 Hz peak that was showing up in the IR lock accuracy plot.

New filter is ACdamp3 and seems to work better.

To reduce the effect of locking accuracy, I reduced the pump green power. First I reduced green power from 26.7mW to 20mW by changing MZ offset, but I couldn't lock GRMC and MZ.

Then I put OD0.2 after GRMC and green power reduced from 26.7mW to 18 mW.

OPO temperature and p pol PLL frequency are 7.173kOhm and 220MHz with 18mW green power.

[Aritomi, Yuhang, Eleonora, Matteo]

By reducing green pump power from 26.7mW to 18mW, we managed to reproduce the last Friday's FDS measurement. The detuning is around 100-130Hz. It seems that squeezing level at low frequency is still limited by locking accuracy.

The parameters are as follows.

sqz_dB = 7;                    % produced SQZ

L_rt = 150e-6;                    % FC losses

L_inj = 0.20;                     % Injection losses

L_ro = 0.11;                      % Readout losses

A0 = 0.05;                         % Squeezed field/filter cavity mode mismatch losses

C0 = 0.05;                         % Squeezed field/local oscillator mode mismatch losses

ERR_L =   13.3e-12;                  % Lock accuracy [m]

ERR_csi = 80e-3;                  % Phase noise [rad]

Strong 50 Hz showed up again in the end oplev signals.  We went to check and tweeked a bit the cables but in all the other configurations that we tried it was worse.

I put a digital notch at 50 Hz in the optical lever signals. 

We found the rapeauto was set to 1/f instead of 1/f^4, we put it back to 1/f^4 and adjust the gain: 

INPUT ATTENUATON : 1.9

PIEZO GAIN: 4 

We measured open loop TF: UFG = 20 kHz, phase margin: 55deg.

It didn't improve the stability problem we have on cavity transmission.

When IR alignment is good, IR trans is around 290 count for 220 uW of IR injection.

When we integrate IR error signal down to 1Hz, locking accuracy is 2.3 Hz.

When we integrate IR error signal down to 0.1Hz, locking accuracy becomes 3.2 Hz which corresponds to 3.4 pm.

[Aritomi, Yuhang, Raffaele, Matteo, Eleonora]

I did a quick fitting of the FDS measuremnt we did last friday (24/01).

We have about 1 dB of FDS sqz at low frequency and 3.5 dB at high frequency.

We see that at some point during the measurement the detuning changed from ~100 Hz to ~70 Hz.

Sqz degradation paramenters used for the fit:

Attached figure is FDS curve with correct detuning (54Hz). With following parameters, frequency at which squeezing crosses shot noise is around 120Hz and frequency at which anti squeezing crosses shot noise is around 45Hz.

sqz_dB = 8;                    % produced SQZ

L_rt = 150e-6;                    % FC losses

L_inj = 0.20;                     % Injection losses

L_ro = 0.10;                      % Readout losses

A0 = 0.05;                         % Squeezed field/filter cavity mode mismatch losses

C0 = 0.05;                         % Squeezed field/local oscillator mode mismatch losses

ERR_L =   8.5e-12;                  % Lock accuracy [m]

ERR_csi = 80e-3;                  % Phase noise[rad]

phi_Hom = [0/180*pi, 10/180*pi, 20/180*pi, 30/180*pi, 40/180*pi, 50/180*pi, 70/180*pi ,90/180*pi];  % Homodyne angle [deg]

det = -54; % detuning frequency [Hz]

Simon, Pengbo

We remove the first PBS of the FI because it will block the reflective beam, so we put another PBS behind the first HWP, then remeasure the beam waist.  

The result shows the waist is about 0.03529 mm at Z_position = 59.96 mm, which is good, our next step will be to recover the birefringence measurement system and test the performance of the new system.

Eleonora, Matteo and Yuhang

When we measure squeezing, we need to lock phase of LO with squeezing. Since the phase of squeezing is influenced a lot by suspended mirrors, the correction signal we send to IR phase shifter is quite large. We want to know how it will influence the measurement of squeezing.

On one hand, the power of IRMC transmission will be modulated. As we reported in entry 1883, we had 10% of IRMC transmission reduction when we drive IR phase shfiter with whole range. We did calculation, if we have 10% power reduction, we will have shot noise reduction of 10log10(0.9), which is 0.46 dB of reduction. This is the maximum reduction while the reduction in the real case should be half of 10log10(0.97),  which corresponds to 0.066dB reduction of shot noise.

On the other hand, there are frequency components of noise sent to IR phase shifter(attached figure 1). We measured the spectrum of homodyne when only lock CC2. (OPO is locked with p-pol, no pump sent to OPO, CC is on resonance inside OPO by tuning PLLp-pol, then let field goes to FC and reflects to homodyne, the 7MHz on homodyne is used to lock CC2) The result is shown in the attached figure 2. We also tried to send sine wave to IR phase shifter, and measured homodyne spectrum. This result is shown in the attached figure 3. These two measurement proves that the low frequency bump has contribution from IR phase shifter noise.

Please check the attached videos.

1st video shows the old broken cable.

2nd video shows the new good cable.

Simon, Pengbo

We move the first lens back to the FI direction with 1 cm, then move the second lens by 0.5 cm to the direction of the input beam.

As can be see from the image, the waist z_position change from 80mm to 70mm, and the radius is smaller than before.

we think by moving the second lens for another 0.5 cm, the waist should exist at the z_position around 60 mm.

Also the waist radius might reduce a little bit.

The measurement is done by measure in DGS with 50 Ohm inserted into the channel.

The unit of measurement is counts, so we need to do calibration. The calbration factor is 0.61V for 1000 counts (see entry #1315). So we have

Noise_ADC = ~ 3e-6  [V/sqrt(Hz)] 

or alternatively

Noise_ADC  = 20*log(3e-6) = -110 dBVrms

This is almost 40dB higher than the noise of Network analizer we always use. So if we use DGS to record data, we will use at least 100 amplification given by pre-amplifier (SR560).

[Aritomi, Eleonora, Matteo]

To reduce LO backscattering, we reduced LO power by a factor of 10. We also increased CC2 gain from 3 to 10.

Attached picture is FDS when CC2 demodulation phase is 20deg and 40deg. There is still large bump below 60Hz.

We thought the bump is changing when CC2 demodulation phase is changed, but it seems that bump itself doesn't change with CC2 demodulation phase.

[Aritomi, Eleonora, Matteo]

To reduce the effect of locking accuracy, we reduced green power from 40mW to 25mW.

OPO temperature: 7.172 kOhm, p pol PLL: 200 MHz

Then we measured FDS. CC2 demodulation phase is 20deg for anti squeezing and 130deg for squeezing.

Simon Pengbo

Today, we continue the measurement of the beam waist. First we changed the position of the first lens, putting it bewteen the FI and the PBS, and make it adjustable.

By changing the position of the first lens, we did two x axis scans. It seems that as the first lens moves away from the FI, the waist become lager and closer to laser bench direction. 

We will do more tests next week.

I re-aligned another double-pass AOM.
The total diffraction efficiency is about 70% which is reasonable value.

Then I modified the position of BS for ISS.
The next step is align the HOM's beams.

In addition, I ordered pillars for composing stage inside the cryostat chamber.
The drawing will be uploaded on wiki.

Given the number Aritomi-san used (which are maybe too optimistic about MM), if we reduce the amount of initial squeezing  from 16 dB to 6 dB. The effect of lock accuracy is much reduced and in priciple we should be able to go below shot noise.