[Matteo, Eleonora]

In order to lock the PLL, the beating note detected by the fibered PD is sent to the PFD (Phase Frequency Detector) ADF 4002 which provides the error signal for the loop. 

The DC component of the PD signal should be removed and for this purpose we used a transformer. 

The transformer used for the first PLL was soldered by Marco Vardaro and Matteo (pic 1). Yesterday we soldered 3 more trasformer (identical to the first one) to be used for the second PLL and for the two monitor channels. (We want to use the other output of the fiber BS as a monitor and avoid the 90:10 splitter used up to now)

We tried also to use a minicircuit DC block (pic 2) but for some reasons we could not see the beating note on the spectrum analyzer.

We have characterized the TF of the transformer (pic 3), and compared it with that of a minicircuit DC block ( pic 4). 

The behavior is similar but the cut-off frequency seems lower for the DC block. ( Is this the reason why we cannot use it?)

In the two cases the attenuation is negligible but the phase loss is about 40 deg at 20 MHz (Is this relevant for the PPL loop?)

Issues about the characterization

During the characterization we had some problems (probably due to impedance mismatch) which brought to very strange TFs with large peaks and notches at high frequency. At the beginning we were using an oscilloscope and a function generator for the caraterization,  then we swiched to a network analyzer (50 ohm impedence) which gave us more reasonable TFs (shown in pic 3 and 4).

Two days ago, while trying to recover the PLL lock, we found that the spectrum analyzer we used to montor the beating note (HP 8563E - see pic1) was not working properly.

It shows a reduced bandwith: the beating note disappears when its frequency becomes larger than about 130 MHz and the peak amplitude was lower of about 15 dB.

We replaced it with another one (HP E4411B - see pic 2) which seems to work better (correct bandwith and peak height). 

The behaviour seems to suggest a impedance matching problem but since the its settings and the configuration of the setup are exactly the same used before we cannot understand the origin of the problem.

More investiagations need to be done.

Participants: Chien-Ming, Shu-Rong, and Yuhang
Today we succeed in locking the green mode cleaner (GRMC) by using the SR560 with setting the low pass at 3Hz. The S/N of PDH error signal is 300mV/3mV= 100. 

We replaced the end curved mirror of GRMC  to a higher reflectivity one 0.9996 ( the previous one is 0.9993), but the transmission efficiency is still around  52~54% not improved. Fig. 1  and Fig. 2 are the transmission spectrums of the GRMC with previous and current end curved mirror. The reason why we can't have higher transmission efficiency may be because of the impedance mismatching of the GRMC.

We also successfully locked the Mach-Zender, but have not optimized the locking parameters and measured the output power fluctuation of both MC and GRMC. Without locking the Mach-Zender, the ratio of the maximum laser power after Mach-Zender and before green EOM is 15.5mW/23.3mW = 66%.

The last thing is to successfully lock the OPO by using the same SR560 with setting the low-pass at 100Hz and the gain is 100. 

Participaint: Aritomi, Eleonora and Yuhang

Today we finally recovered the PLL. The problem we had during this recovery is listed as following.

1. We didn't record the current value and temperature value of main laser and auxiliary laser. The situation now is listed as following.

2. The spectrum analyzer is broken so that we have only bandwidth of 130MHz. This is too small range and brings diffculty to find the beat note. We found a new one and now it is placed in the place of former spectrum analyzer. 

3. The transformer cannot be replaced by DCclock of minicircuit. Fortunatly, Matteo and Eleonora made quite a lot of new transformers. We guess the DCblock doesn't block enough low-frequency signal.

4. The set up of PLL board still needs to investigated. Up to now, the lock of beat note 20MHz seems fine. However, we need to know how to change set-up if we want to lock it on a different beatnote frequency.

The measurement of birefringence effect is like this. We send both bright alignment beam(s-pol) and p-pol beam inside OPO. We set up temperature of OPO from 7.2kohm to 7.8kohm. At each temperature we change the temperature of p-pol laser so that to make s-pol and p-pol overlap on the oscilloscope. However, this measurement is also not very precise. Because it is very diffcult to know the exact frequency in which two polarization overlap. What we did is just to make the overlap peak as high as possible.

11/20~21 Participants: Chien-Ming, Shu-Rong, and Yuhang

Yesterday we measured the SHG output green beam size again by removing the Faraday isolator and placing the beam profiler. (See Fig.1)
The fitting results indicate that the waist is 28 um inside the SHG and 47 mm from the edge of SHG output port. (see Fig.2)
This result is in the case where the SHG cavity concave mirror is not considered.
Yuhang found that he used the incorrect wavelength in his fitting on 11/19.

Today we tried to recover the mode matching of the green mode cleaner (GRMC) by only adjusting the position of the lens L6 and the alignments of mirrors M5, M6, and M7 as showing in Fig.3.
After optimizing, the new position of L6 on the rail moves from 48mm to 41.5mm (Equal to moving 6.5mm towards the EOM)

Fig.4 shows the mode matching transmission spectrum of the MC when the incident beam is p-polarization.
And Fig.5 is s-polarization.

The transmission efficiency of the MC is 77.7% when using p-pol incident beam and 54% when using s-pol beam. This means that if we want to reach 100 mW GRMC output power when using s-pol incident beam, we will need SHG to provide 270 mW green output power.

Participaint: Aritomi, Chienming Eleonora, Shurong and Yuhang

Since I reported the beam parameter has been changed after we did the mode matching of SHG. So we measured the beam coming out again and we decided to recover the beam before BS to recover the situation as before. (The BS we talk here is where the beam is seperated and goes to FC or GRMC seperately)

We put a 150mm lens 9 holes before the first measurement point(3, 17).(now between SHG and measurement point, there is a 200mm 13.8 holes before the first measurement point already there). We dismounted the second lens after SHG to have space measuring beam parameter. The measurement region is between a steering mirror and EOM. The measurement result is shown below.

326, 588, 914, 1165, 1477, 1779, 2090

364, 676, 1018, 1296, 1623, 1921, 2210

The fit result is shown in the first attached figure. Then we use this result to trace back beam to SHG. This trace back process is shown in the attached figure 2. This is an important result, the beam waist size is 36um inside SHG.

Then we used this SHG new waist size and the current set-up to propogate it to the waist around BS. Because we have the measurement we did before to compare. We found the beam waist before was 26um. But now according to the measurement and some necessary simulation, the waist size now is 32um. Since it is close to our good value, we foresee the recover of mode matching by moving only the second lens.

The recover is done by looking at camera image and pd spectrum together. At the beginning, we found the pitch alignment is quite large, we recovered this degree of freedom first. Then we clean and center all the steering mirrors and lenses going to mach-zender. By keeping the TEM00 on camera and maximazing it, the alignment is recovered. Now the green mode cleaner is recovered.

11/19 Participants: Chien-Ming, Shu-Rong, and Yuhang

Today, we finished the telescope of the bright alignment beam, the focal length of L3 is 100 mm and L4 is 150 mm (see Fig.1).
Comparing to the results on 11/14, the beam shape now becomes elliptical after using the Faraday Isolator 2 as shown in Fig. 2 (remove the M4 mirror in this case).
Fig 3. shows the beam shape of the Aux laser 1 coherent control beam (CC beam for short)

Fig 4. shows the transmission spectrum of the OPO cavity by injecting the CC beam. Here we use the Thorlabs InGaAs photodetector to measure the spectrum.

We use the M4 mirror placed on the translation stage to swap the CC beam or the bright alignment beam sending into the OPO.
Fig 5. shows the transmission spectrum of the OPO by injecting the bright alignment beam at the same optical power (60.3 mW) as the CC beam using in Fig. 4
The peak height of TEM00 in Fig.5 bright alignment beam is 8.1 V higher than in Fig.4 CC beam (7.6 V). However, when scaling up the vertical axis, we can see the peaks of other high order modes in the spectrum of bright alignment beam (see Fig. 6) is a little bit worse than in CC beam (see Fig. 7).

Fig. 1 The schematic of the telescope for the bright alignment beam.
Fig. 2 Beam measurement of bright alignment beam on 11/15.
Fig. 3 Beam measurement of CC beam on 11/14.
Fig. 4 The transmission spectrum of injecting the CC beam into OPO.
Fig. 5 The transmission spectrum of injecting the bright alignment beam into OPO.
Fig. 6 Scaling up the transmission spectrum of the bright alignment beam to see the high order modes.
Fig. 7 Scaling up the transmission spectrum of the CC beam to see the high order modes.

Participaint: Yuhang and Matteo

Last Friday, we reveived the compnent of alignment infrared mode cleaner. With the help of Yano-san, we cleaned it in ATC by using aceton sonic bath. Then we did the assemble in our clean booth but we found we lost the orings. We will continue this work after we get some orings.

11/15 Participants Chien-Ming, Shu-Rong, and Yuhang

Today we insert the second Faraday Isolator IO-2-YAG (FI 2) into the main laser light path and modify the telescope for SHG.(see Fig. 1)
The transmission of this small aperture (<2mm) isolator is around 85% even if the beam radius inside the isolator is less than 500 um.

We achieve a good mode matching (see Fig. 2) for SHG by using the lens L1 f = 75 mm and lens L2 f= 100 mm.
The output power of 532nm is 148 mW when injecting 426mW 1064 nm light into SHG. This efficiency is consistent with the previous results on 11/9 and 11/12.

However, we did find that the laser power stability of both SHG and Main laser have improved a lot after adding the isolator FI 2.
The SHG 532 nm output power have the ratio of Max/Min is 1.007:1 (see Fig. 3) which is much better than the result 1.088:1 measured yesterday.
The ratios Max/Min of the main laser are both improved as 1.010:1 and 1.012:1 when operating in a higher power and low power. (see Fig. 4 and 5). The power meter is placed behind the BS1.

We believe that such stability should be enough even though it is still somewhat different from the situation of blocking the reflected light from SHG with the ratio of Max/Min is 1.002 :1.(see Fig. 6)

Since the output polarization of the FI 2 is 45 degrees, we use a half waveplate (HWP1) to ensure the s-polarization for BS2 with the split ratio of T/R is 20:80. Otherwise, the BS2 split ratio of T/R is 38:62 when using the p-pol light. The second half waveplate (HWP2) rotates the polarization to p-pol. for serving the SHG.

Fig. 1 A schematic of adding the second isolator FI 2 and modify the telescope for SHG
Fig. 2 The mode matching result of SHG
Fig. 3 SHG output power stability
Fig. 4 Main laser power stability at higher power scale. 
Fig. 5 Main laser power stability at lower power. 
Fig. 6 Main laser power stability when blocking the reflected beam from SHG

Today we tried to test the boards and galvo. 

Since at this moment we don't have a quadrant yet, we tried to send signals through the check point and to see if the galvo give any response. 

We got two galvo and two boards, their situations are below:

Board 1: It is able to give both x and y direction correction signal to the galvo, but the monitor ports for the currents send to galvo both have high frequency oscillation, especially in x direction, the amplitude of this oscilation is very large.

Board 2: Only the y output can drive the coil, for the x direction when we switched on the board, we could see the mirror on the galvo has a sudden move and then goes back slowly to the original position. The monitor signals for both x and y ports are at their maxmum, so they don't change when we send the signals through check point. 

We are going to tune the offset of the board to see if the situation can be changed.

Galvo 1: Working fine. We could hear the vibration sounds from both the motors.

Galvo 2: One of the mirrors is working fine, the other one is a bit loose from the motor, I think after we fasten it, it could be fine.

11/14 Participants: Chien-Ming, Shu-Rong, and Yuhang This morning we set up a telescope for the bright alignment beam. (see Fig. 1)
The focal length of M3 is 150mm. It is located at 65 mm from BS2.
The focal length of M4 is 200mm. It is located at 337.5 mm from M3.

By using the Beam Profiler (Placed on the edge of the optical table), we measure the beam size of the bright alignment beam. (See Fig. 2, a clearer picture will be added soon)
The beam shape is close to circular symmetry and beam size is 2539 um on W-plane and 2342 um on V-plane.
We also check the beam size of Aux laser 1 (CC) by inserting the Mirror M4 placed on the translation stage. (see Fig. 3)
Its shape is ellipse and beam size is 2542 um on W-plane and 2233 um on V-plane.
The size of these two beams is similar.

In order to check the isolation of the Faraday Isolator (FI: IO-5-1064-HP), We introduce the light from Aux laser 1 to FI. (See Fig. 4)
By using the power meter, we measure and optimize the isolation of FI is 39.4dB which is confirmed to the spec. (38 ~ 44dB) on Thorlabs website.

After optimizing the FI, we block the Aux 1 beam and reinstate the main laser beam to the SHG. (see Fig. 5)
By placing the power meter behind the BS1, we can measure the power fluctuation of the main laser beam.
First, we tune the PZT of the SHG to non-resonance state and obtain the ratio Max/Min is 1.054:1 (see Fig. 6)
which is larger than the ratio of 1.020:1 when SHG is on-resonance (see Fig. 7).
According to this result, we suspect that the power instability is mostly caused by the 1064 nm feedback light from the SHG cavity.

We also measure the ratio of 1.049:1 when increasing the input IR power by 2.8 times ( about 200mW) to SHG. (see Fig. 8).
The generated power of 532 nm is 28.4 mW and the ratio Max/Min is 1.088:1 which is the most serious. (see Fig. 9)

Since we have roughly determined that the interference is coming from the 1064 feedback light, we plan to insert another isolator to the main laser light path.
Fortunately, we found another isolator (IO-2-YAG made by OFR) that is not in use. However, the aperture of this isolator is less than or equal to 2 mm while it can work in high power 750 W/cm².
This means that it can bear the case of the input power is 1 w at 1064 nm with the waist of 250 um inside the isolator where the laser fluence is 509.3 W/cm².

Fig. 1 A schematic of the telescope for the bright alignment beam.
Fig. 2 Beam measurement of bright alignment beam.
Fig. 3 Beam measurement of Aux laser 1 beam.
Fig. 4 A schematic of measuring the isolation of FI
Fig. 5 A schematic of measuring the laser power fluctuation.
Fig. 6 Laser Power fluctuation: after optimizing the FI, the SHG cavity is off-resonance
Fig. 7 Laser Power fluctuation: the SHG cavity is on-resonance
Fig. 8 Laser Power fluctuation: the SHG cavity is on-resonance and increasing 1064 nm laser power
Fig. 9 Laser power fluctuation of the SHG 532 nm output

Participaint: Chienming, Shurong and Yuhang

As we reported, we found the green production has an fluctuation. Following that, we found the similar fluctuation in the power of bright alignment beam and even the beam going to IR mode cleaner. Here we present the result of power monitoring at two different points and in three different situation. The position of monitoring is shown in the attached figure 1. We kept monitoring each case for around 4min.

Montioring point :

Situation 1: Block the beam going to SHG, almost no light going back.

          power fluctuation level is 0.29%

Situation 2: Lock SHG, both infrared and green going back.

          power fluctuation level is 3.9%

Situation 3: Tune SHG off-resonance, only infrared going back.

          power fluctuation level is 4.8%

Monitoring point 2:

situation 1: Lock SHG, both infrared and green going back.

          power fluctuation level is 1.6%

situation 2: Block the beam going to SHG, almost no light going back.

          power fluctuation level is 0.055%

Scanning cavity:

          In the last attached figure, we put the power of point 1 while scanning cavity. As you can see, Cavity resonance causes the drop of main laser power. However, there is one thing strange, when cavity is not resonance, the mainly laser power should be the same since the reflection of SHG is the same. But in the attached figure, we can see the feature caused not only by cavity resonance.  There is also a signal in-phase with the ramp signal. Maybe this is caused by the phase change of light. Need more investigation.

Conclusion: As we can see from the result, the monitoring point is only related to main laser. So we are sure the fluctuation comes from main laser. And we found the fluctuation will disappear only when there is no reflection. We also found the fluctuation will be even larger if there is more infrared reflection. And since even when there is no green relfection, fluctuation exists. We guess green should not be the main reason of this fluctuation.

As a confirmation for what I reported in entry 1090, I measured the spectrum of the in-loop PD at different powers of the pump. In the plot, we can see the peak at the chopper frequency.

This confirms that there is stray light going to the PD. We are processing the purchase of new filters: additional long-pass at 1250nm

11/12 Participation: Chien-Ming, Shu-Rong, and Yuhang

Following the SHG telescope set up last Friday, we try to improve its mode matching by adjusting the position of Lens 1 (that is the distance to the input waist which is located at the output end of the EOM). The focal length of Lens1 (L1) =200 mm and L2 = 125 mm (see Fig. 1)

First, we change the position of M1's clamping fork to increase the moving space of L1. However, this cause the alignment of M1 slightly deviated. We spend a lot of time recovering the alignment, but it is worse until we find a 3-Adjuster mirror mount to replace the original M1 mount (2-Adjuster).

Then we install the L1 on the optical rail to optimize the mode matching. Last Friday's L1 position was about 138 mm from the EOM output port, and L2 was about 675 mm from EOM output port. After optimization today, the new position of L1 from EOM is 115 mm and L2 is 674 mm. The SHG scanning spectrum is attached in Figure 3. You can see the mode matching is improved to 95.8%.Although the TEM03 mode shown on the scope is almost disappeared after optimization, and the peak of TEM02 mode also drops compared to last Friday. However, the conversion efficiency of SHG output power is still the same as the result of last Friday.

We also slightly change the temperature of the SHG crystal, but can't find a better result. So we decide to stop here and start to set up another telescope of the light leading to the OPO(bright alignment beam).

We also tested with a LASER current of 1.34A. We can have maximum green power of 237+/- 6 mW.

Problem: Now the green production has a power fluctuation from 137~147mW when SHG's incident IR power is 420mW.

Fig. 1: The modified position of the SHG telescope.

Fig. 2: The pattern of the remained TEM02 mode obtained by the CCD camera, even its peak of transmitted signal showing on the Scope is lower than that of last Friday.

Fig. 3: SHG scanning spectrum

Fig. 4: The simulation of SHG telescope in the update position(now the waist zise of this simulation agrees with Chienming's calculation result)

After checking the alignment of both the probes, I tried to measure a LMA coating (that absorbs a few ppm).
I increased the pump power up to 1W (980mW) rotating the HWP in the IPC (so without changing the laser current the power is immediately stable).

The looking at the scan on the screenshot attached we can see that there is a large constant-phase signal.

After removing the sample it was clear that it is stray light from the pump because the phase is -22deg.

In front of the PD there is a long-pass filter 1250nm that has OD 5.5 @1064, but it is not enough.  The transmission at 1310nm is 85%.
Probably the fastest solution could be to put 2 filters together in the same SM1 attached at the PD, but I'm afraid of internal reflections effects.

I made the profile of the pump, the HeNe and the 1310nm laser.

I used the blade on the translation stage to cut the beam and measured the transmitted power with the power meter connected to the labview software that makes scans. Then I fitted each scan with the erf function and then I fitted the profile for each laser beam.

The blade 0mm position is at 75mm from the optical board, and at 172mm from the mount of the last focusing lens of the pump.

I upload the previous measurements of the profiles (first 2 plots) and the new one (third plot), after replacing the 1310nm laser. After changing the fiber, the waist size didn't change but the waist position moved a few cm.

Today I tried to swap the input of fiber for coherent control PLL, and did the alignment. I achieved coupling ratio of more than 50%. So this means they are not broken.

Participaint: Chienming, Shurong and Yuhang

Situation before changing telescope: there is a higher order Laguerre-Gaussian mode appeared in the spectrum while we were scanning SHG. The conversion efficiency of SHG is 13%.

Motivation: Have more green production for the measurement of non-linear gain measurement of OPO. Decide to improve mode-matching situation. Certainly, more coupling of infrared into SHG will produce more green.

So we removed all the lens set by yuefan and implement the telescope we designed before. I am so sorry that I didn't ask Matteo to buy some new rails for telescope, so we put only one lens on the rail. Now the situation is we put 200mm at hole between (26, 7) and (26,8)(closer to 7) while put 125mm at hole (18,17). The lens on the rail is 125mm one. See attached figure 1 and 2.

Situation after changing telescope: The mode matching now is shown in the attached figure 3. We can see from the plot that mode-matching now is 94%. So there is still possibility to increase mode matching and further increase conversion efficiency. We need to note here that the movement of the second lens mainly change the position of waist while the movement of the first lens change mainly the waist size. And since we know the measurement result of yuefan, the waist size should be 54um. We can improve the mode matching further easily if we have another rail. Then we can easily move two lenses together and achieve a good beam waist size and position together. After did the mode matching we measure the power of infrared going inside SHG, which is 419.5mW(shown in attached figure 4). Then we locked SHG, while the invert is set as 'on'(shown in attached figure 5) and the gain is set as full gain(shown in attached figure 6). Then we measured the green production, now the green power is 101mW(shown in attached figure 7).

Situation after changing temerature: As suggested by Chienming, the mode-matching change will change phase matching situation. We increased temperature and measured the green power generation. The result is shown in attached figure 8. We found the best phase matching temperature is between 3.151 and 3.147kOm, which is smaller than before. This mode-matching difference of 20% bring optimal temperature difference of 0.2K. Now the best temperature is  around 331.4K. Now the green power can reach 147mW(as shown in attached figure 9) Then we lock SHG again, we found now the alignment is quite sensitive. Even we touch a little bit mirror mount, we will degrade the green power by several mW. This means we may need a better mirror mount(so this can be somthing be improved in the future). And now the tranmission voltage is 1.46V(as shown in the attached figure 10). This is a little bit higher than the peak value of scanning as we expected. And also now we change temperature to 3.151kOm(as shown in attached figure 11). This means conversion effciency of 35% now. As pointed out by manufacture, it can reach 45%. So we still can improve it anyway. The good thing is we have enough green as we want.

Problem we found and solved:

1. One of the lenses is with a wrong coating. This can expalin the strange ratio of power we found before. But anyway we removed it.

2. The telescope for matching SHG's transmission into PD. The beam size was very large and collimated with using a lens of 100mm. We replaced it with a 75mm lens as shown in attached figure 12 . But now the pd saturates, we put a ND filter (ND = 1).

Problem found but not solved:

1. stray light: the stray light hit on the mount of mirrors or lenses. Maybe this is something we should consider in the future.

2. alignment after SHG is changed as shown in attached figure 13.

Additional check and work needs to be done:

1. check the beam shape before EOM(for filter cavity)

2. Buy a new rail and improve alignment further more.

3. Replace mirror mount for the two steering mirrors in front of SHG by two very stable mirrors.

4. Align the path after SHG.

I engaged a control loop on the S1FC1310PM laser.

I used the modulation input of the laser controller. I used the SR560 as a servo. I set the low pass filter at 30Hz and gain 2000.

First I measured the noise of a 50 Ohm terminator to check the spectrum analyzer noise floor.
Then I measured the spectra of the 2 PDs (in-loop and out-of-loop), without laser (dark noise) and with the laser on.
The other day there were some structures on the out-of-loop PD that we didn't understand at the time, then I found that the beam was not well centered on the PD, so after I centered (maximizing the DC) the structures disappeared.
Then I closed the loop and measured the spectra again. The signal at 380Hz in the out of loop PD reduces by 10dB (about a factor of 3).

I confirmed the noise reduction by checking the lockin output with and without control loop. The 2 plots have the same axis scale, so the reduction is more clear.
The noise now is 1.3 ppm*W

I measured the actuator TF (plant) and fitted it with a zpk model: 2 single poles at 7kHz and 30kHz.
Then I modeled a servo TF and plotted the measured open loop TF. There is a factor of 2 of discrepancy with the model because the oscilloscope was connected to the modulation input, since they have the same input inpedance, the measured TF dropped by a factor of 2. But when I closed the loop the oscilloscope was not connected, so the actual OLTF is the dashed blue line on the plot.

A reference procedure for the filter cavity alignment: