Michael and Yuhang

We checked filter cavity alignment yesterday. The pico-motor of PR/BS and END mirrors are moved. The movement of PR is to recover its reference on BS chamber. The movement of BS is to recover its reference in filter cavity transmission camera. The movement of END mirror is to recover the flash.

After these movement, we checked the oplev spectrum as attached in this elog. They look fine. So this time, no touching issue of mirror was found.

Katsuki-san, Marc

This morning we went to PCI to check the ATEC sample.

We found out that the peak visible in the polarization angle measurement is likely due to a dust so we removed it. Except that the sample seems perfectly fine.

We brought a laptop with Kinesis installed and connected it to the motorized HWP.

We got the same issue as with the PC (only jog by 1 deg).

We tried to home it and will check after lunch.

If problems persist we will contact Thorlabs and switch the 2 motorized HWP as the other is working fine.

This entry reports the birefringence measurement of the AZTEC sample with no rotation,

Note that a quite strong earthquake happend right after the measurement with input polarization = 0 deg...

Note that in that case the input polarization angle of 0 deg is made with hwp angle = 346 deg..

While the mean value of the polarization angle measurement seems reasonable, there is now a strange 'peak' on the top right of the sample...

We'll go to PCI to check if the sample is fine.

In order to further investigate the effects of the input beam incident angle we mesured the polarization angle as a function of the input beam polarization angle.

Data are attached to this entry.

Note that :

There are input polarization angle for which the s polarization power transmitted with sample is higher than transmitted s polarization power without sample..

The minimum s polarization seems to be reached for hwp  angle = 71 deg (ie 26 +45 deg) and using this polarization to normalize our signals means that the hwp angle = 341 deg that was assumed to be fully s polarization is in fact 10 deg polarization angle...

We also measured extra polarizations and results are attached to this entry.

Furthermore, I also computed the relationship between the input polarization angle and the polarization angle of the sample (see last figure).

Yuhang and Michael

After discussion in the TAMA filter cavity meeting on the 29th of September, we realigned the OPO, noting the following points:

• The OPO assembly position with respect to the rotation stage has been set up so that rotating one of the upper screws results in OPO yaw rotation, and rotating both upper screws results in X translation (horizontal direction perpendicular to propagation axis). Likewise, rotating one of the lower screws results in OPO pitch rotation, and rotating both lower screws results in Y translation. See figure 1.
• We confirm that the first stage of the OPO alignment starts with the beam entering the curved HR side (figure 2). This was noted with the black dot in figure 4 of 2682, which is closer to the flat faced side of the plastic holder (i.e. the incoupling mirror is mounted on the opposite side of the assembly from the periscope). The rationale is described in Matteo's thesis with respect to the SHG. We do it this way to ensure that the optical axis and OPO central axis are coincident before placement of the incoupling mirror. After placing the incoupling mirror, the cavity can be scanned for removal of higher order modes. Then, when the alignment of OPO/incoupler is confirmed, the assembly can be turned around for the finesse measurement.
• The beam size of the CC/p-pol beam entering from the HR surface of the OPO in the TAMA experiment is nominally 36 um (Marc 936)
• The distance of the beam from the f = 75 mm lens to the beam waist is measured to be 125 mm (2515). It is estimated that the OPO should be about 57 mm from the top of the periscope (figure 3). Note that there is a mistake in the figures of that elog entry - the predicted waist size of 20 um is with the lensing action of the meniscus (incoupler). Without the incoupler, the beam waist should be 25 um, as measured. However, as noted above, this beam size is setup for the OPO cavity finesse measurement, which is performed after the alignment of the incoupler.

The OPO has been aligned by using the camera to make a reference point (figure 4), placing the OPO, then ensuring that the incident and reflected beams overlap. We confirmed reflection/incidence overlap at the AOM exit port and Faraday Isolator. The beam is also centered on the target as indicated by the camera. In the future, we should take note to have the modulators and FI be further away from the OPO cavity for easier alignment of the reflected beam.

One of the wires detached from the DSub unit (figure 5). It appears to be the positive end coming from the OPO thermistor.

Katsuki-san, Marc

After investigating the effect of the AZTEC sample rotation, we decided to try to act on the imaging unit part.

We removed the OD (2 and 3) in front of the PBS in the imaging unit and added similar OD in front of each PD.

We tried to tweak the alignment as best as we could but because we installed the OD with a quite large angle to try to mitigate effects of scattering we don't have a balance powers on the 2 PD.

In order to have a beam small enough on the PDs we had to replace the lens before the PBS by one from the FC Newport lens box (f = 125 mm)

Anyway we started measurement in this new configuration.

In order to investigate the reason of the seemingly uniform theta distribution, we decided to start new measurements with the AZTEC sample rotated.

I marked the position of the holder, rotated the sample and measured an arc length of 3 cm that corresponds to a rotation of the sample by 34.4 deg (radius is 5 cm).

I took several measurements that are reported in the attached figures.

An important thing to notice is that I found I made mistakes in the use of the code.

Starting from the code available in the PC the modifications for our current setup are :

DC gain = 10

AC gain = 1000

DC is s - pol

AC is p pol

While the gain were correct, the previous figures were made with inverted AC and DC !

Also, the incident polarization angle was wrongly estimated (ie no normalization).

Note that to estimate the delta_n distribution it has to be multiplied by 1e6 to properly compute variance and mean.

Now all these modifications have been implemented for these figures.

Furthermore, the HWP working condition could be recovered a bit and all the reported measurements are made with the HWP between 341 deg ( s-pol) and 26 deg (p pol).

As a brief summary of the results :

- s0 and s1 absolute values are now similar

- delta n and theta increase with the incident polarization angle

Raffaele suggested to put time series together with spectrum of BAB PDH signal. The spectrum was measured in elog2573.

The last time of FC aligned was about one month ago. However, the alignment work took less than 20min today. In addition, the BAB alignment to FC didn't drift away too much. Although I didn't check higher order modes, the BAB transmission was about 430 counts.

After optimizing BAB transmission to about 460 counts, I took time series and put together with spectrum as attached figure.

There are now 4 SHINKOSHA evaluation plates inside the PCI clean room.

I marked the 7,11 and 14 on the side close to the ingot position marking.

new measurement with 30 deg polarization finished.

The results (especially of the theta distribution) seems to agree with our hypothesis that the previous measurement with input polarization 30 deg was something like 30+90 deg.

I attached to this entry several results extracted from measurements with incident polarizations 90 deg (ie s-pol), 60 deg, 45 deg and 30 deg. They are made based on a python code kindly shared by Simon.

It seems this sample is more uniform than the SHINKOSHA #7.

The s0 component is far smaller while the s1 component is larger. A possible explanation could be (from MIR meeting discussion) that the birefringence appearing here is not due to stress but to the crystal axis orientation with respect to the pump beam optical axis.

The delta n distribution seems to be quite identical for all incident polarizations and the offset might be due to the cutting angle of this sample.

A possible explanation for the offset on the theta distribution for incident polarization of 30 deg could be explain by having rotating the incident polarization by more than 90 deg.

Indeed, the HWP software now is stuck at 15 deg whatever the polarization...

I started again this 30 deg incident polarization while only decreasing the value of the HWP angle.

Yuhang and Michael

We had some trouble aligning the OPO inside the holder (initially without the input coupler). A photo is shown in figure 1. The OPO assembly is contained on the white plastic mount and has a small ~ 5mm hole at both the entrance and exit. The mount past the OPO assembly is a beam splitter that goes to a camera and photodetector. The two mirrors on the bottoms left and right are the steering mirrors. However, these are used to control the horizontal and vertical alignment of the beam before placing the OPO. Our approach was to constrain the cavity axis prior to placing the OPO - we make sure the beam stays at constant height, send it to a camera and then mark the position of the beam on the TV (fig 2). The OPO cavity is placed so that the entry hole has the input beam centered and the exit hole aligns with the previous mark on the TV screen. Then, in theory, we only need one degree each of pitch and yaw (i.e. a rotation stage) to align the OPO to the cavity axis and match the incident and reflected beams. An illustration is shown in figure 3. As seen with the arrangement of the periscope, a yaw misalignment of the reflected beam when inspected at the prompt reflection translated to a pitch misalignment of the reflected beam when inspected past the periscope (fig 4).

Using the alignment setup described, we couldn't meet the alignment conditions. We tried again by moving the OPO assembly (and the OPO  curved HR surface) closer to the centre of rotation of the rotation stage. Actually, after all of this we found that the camera wasn't very stable on the mount as well, so we fixed it properly.

Initially we put the modulators on the edge of the table so that their connecting cables wouldn't get in the way during the experiment. However, in retrospect, it would have been better to put the OPO assembly on the edge of the table since the alignment of the OPO is the most delicate task.

This morning I put back the correct lockin parameters.

I started a measurement so that the polarization angle is close to 45 degrees.

This time, a new trouble with the HWP appeared : I can change the polarization angle (quite visible from the relative change of the p pol and s pol photodiodes) but the value on the angle value on the Kinesis software does not change.

I'll investigate this issue after all required measurements for this sample are finished.

I did a series of measurements while changing the polarization angle by 90 degrees.

I did not recorded the required max/min of s and p polarizations to have accurate calibrations of the measurements.

So today I restarted measurements with incident p pol, s pol and started a 30 degrees polarization angle measurement.

However, there was another WIndows update... As it is not possible (yet?) to fully control the 2 lockin amplifiers from remote, I can not restart the measurement for today.

Furthermore, the HWP2 controller is not recognized by the software after acting strangely in the past days (only jog by 1 degree otherwise it resets the current angle to 0 and still move by 1 degree...)

For the new measurement a possibility could be to start by an absorption measurement before switching to birefringences ones.

This would assure 100 % that the orientation of the sample during all measurements is the same.

Yuhang and Michael

We assembled the input mirror assembly for the OPO cavity as per 812. The only thing to mention here is the orientation of the input coupler convex surface (fig 4 and 5).

We then connected the interface electronics for controlling the Peltier, thermistor and piezo. The Peltier, thermistor and piezo each have two connecting wires that must be soldered onto whatever means we are using to connect them to their controllers. We soldered these onto a PCB connector. The Peltier and thermistor are connected to a DSub-9 cable going to the Thorlabs temperature controller (Peltier +ve pin4, -ve pin 5, thermistor +ve pin 2, -ve pin 3). The piezo wires were soldered to a LEMO F port.The connecting wires had to be soldered outside the cleanroom - the OPO cavity and input coupler assembly were placed in plastic zip lock bags and sealed so that only the wires were coming out (fig 6). This way they could be protected from fumes and dust.

The Peltier and thermistor were tested using the temperature controller. We saw that we could stabilise the resistance to ~10 kOhm. The piezo also made the expected high-pitched noise when it was actuated (~ 2 kHz test signal).

There might have been some issue with this measurement as after 35h it did not finished..

Today Yuhang kindly went to PCI to reset the parameters (sensitivity = 1 mV for both lockin amplifiers and input A for the one connected to the PC).

We restarted the measurement with HWP = 46 deg (ie incident polarization is p).

The measurement finished and I started with HWP = 91 deg (ie incident polarization s)

Analysis to follow.

Today I tried to check if I could improve the shielding of the PBS transmission photodiode but it seemed to be already fine like this.

Then, following Simon advices, I tweaked both QWP and HWP at the DC photodiode (s pol) minimum.

In the end, the s pol  is minized for HWP = 46 degrees.

I took some values of AC and DC in different configurations :

without laser

AC= 0.6 V
DC = 0.065 V with lockin = 0.2 uV (sensitivity 30uV)
DC = 0.076 V with lockin = 0.007 mV (sensitivity 1mV)

with laser (~30 mW ie HWP1 = 20 deg) & HWP = 46 deg

AC= 607 V
DC = 0.099 V with lockin = 0.01 uV (sensitivity 1mV)

with AZTEC sample :

AC= 378 V
DC = 0.326 V with lockin = 0.032 uV (sensitivity 30uV)

The AZTEC sample orientation is the same as for the absorption measurement (black dot at the top).

I used the AC (p pol) to find the AZTEC sample center and got Y_center = 327.72 mm and Y_center = 122.825 mm. I didn't tune the Z position which is now 53mm.

The measurement started with Median filter  = 1, average filter = 0, wait time = 500 ms, map radius = 20 mm and step size = 0.1 mm and HWP = 46 degrees (ie p pol maximum).

The measurement should last 4h30. But before leaving PCI room I touched by mistake the lockin sensitivity... so one or two points might be wrong...

I will start other polarizations over the holidays.

Marc, Matteo

Last Friday, we decided to slightly move the optical board hosting the birefringence readout part in order to be able to close the enclosure to reduce possible effects of scattered light.

The flipping mirror to switch to absorption measurement is not reinstalled but it should be straightforward. We checked that it is still possible to move the absorption imaging unit translation stage on all its range with this configuration. nm path but the sensor was too small.

We tried to use the photodiode for the 1310 nm path but the sensor seemed too small (1mm diameter vs 3.6 for the ones we installed). So we used a thorlabs power supply labelled 'broken' for our Thorlabs photodiode and it seemed to work fine (maybe just too noisy?)

In any case we plan to buy 2 new photodiode and 3 new powersupply.

We connected these 2 photodiodes to the lockin amplifier and could get signals.

With 2 optical density (2 and 3) there was no saturation.

We moved the HWP angle between the 2 lenses and confirmed a maximum and minimum for the photodiode in reflection of the PBS separated by 45 degrees.

_________

Today we started by changing the imaging unit PBS post by a more stable one and realigned the 2 photodiodes. This time, the photodiodes have a small angle to limit back-scattering.

Then, we installed walls on the birefringence board before putting the large enclosure and closing its 4 walls.

Then, we tuned the QWP angle so to be at a maximum on one of the 2 photodiode.

Then, we did a scan of the motorized HWP  with incident pump power around 30 mW) and results are reported in the attached figure.

min p pol = -0.16 mV

max p pol = 0.606 V

min s pol = 0.12 mV

max s pol = 0.588 V

where the offset without pump beam is already taken into account (p-pol = 0.6 mV, s-pol = 12 uV)

I guess we could try to add more shielding for the p pol photodiode (ie in transmission of the PBS).

Notes :

Be careful when acting on the QWP as its holder is not so stable !

Only use jog function for motorized HWP otherwise it will reset the current angle value to 0

You're welcome.

The input power is 30 mW and can indeed be tweaked with the HWP.

Also, the lockin amplifier settings have not been changed yet.

For the photodiode powersupply we will take the one used for the 1310 nm beam I think.

For the PSDs, one is now used in FC experiment so only one is available.

It seems that with the 100 mm focal length lens, the beam is small enough for the photodiode. And we can also tune their gain.