Continued SHG

[Briana, Ian]

Mounted 4 mirrors for 780 nm laser (not labelled yet). Need 8-32 x 1/2" screws to finish mounting fifth 780 nm mirror. We don't have the nonlinear crystal/lenses yet so we pretended it existed while we started alignment with the last two 1550 nm mirrors. Unfortunately after exiting the Faraday isolator, beam height was steadily decreasing by the time the light reached the far upper right mirror (see Full Schematic attached yesterday). I realigned up to the waveplate but the Faraday isolator may need to be put in a mount that can translate up and down to maintain constant beam height. Also, based on the board size, there might not be enough room for the previously placed 75 mm lenses in the schematic (focal lengths are too big) so the 50 mm lens distances should work. Also, we may want to collimate the 1550 nm beam before it reaches the lenses before the small nonlinear crystal, but maybe the beam size will be sufficiently small that we don't need to do that. 

The next plan is to put together the 780 nm laser. The L780P010 laser diode has maximum optical power of 12 mW with maximum current 40 mA. We can expect that the current to achieve the maximum optical power is lower than 40 mA. Optimal temperature is probably 20-25 degrees Celsius. Initial laser schematic is uploaded. We will probably need the antistatic wrist bracelet while setting up the laser.  The current limit of the current controller is 500 mA, covering the current supply's max 40 mA. 

If the label maker ever glitches, freezes, flashes when you turn it on, or simply dies, it's probably the battery. We finished work and left around 4:13 pm.

Vapor Cell Layout and Beginning (SHG)

[Briana, Ian, Torrey]

Updated Ian's drawing into a scaled layout with some additional components following Ian and Torrey's comments (attached). The schematic is split into two sections: simple harmonic generation (SHG) for frequency doubling the laser (upper half roughly separated by the first horizontal blue fiber coupling) and the vapor cell locking with counterpropagating beams. The laser, EOM, and Moku elements are not on the breadboard- they will be connected via fiber coupling.

We are currently using the 2 ft by 2 ft black breadboard previously earmarked for LFC. We are keeping it on one board so it can be easily moved to the new lab space. The breadboard is located on the central large table in the corner closest to the tool chest. A vacuum component was moved under the table to make room for the breadboard.

With this set layout, we started building the SHG part using a path off the GQuEST 1550 nm laser to help align the setup. We intend to eventually use the1560 nm laser. The goal of the SHG is to produce a beam with half the wavelength (780 nm). The 4 mirrors used for the 1550 nm SHG setup have blue mounts. We aligned the first two mirrors followed by a half waveplate and the first Faraday isolator. On the other SHG setup in the lab (same table but towards the far wall), a five axis mount was used for the Faraday isolator but we just used a pillar since it was close enough to the beam height above the table (~4 cm). 

There was a problem where the light entering the Faraday isolator did not visibly transmit through despite being in the beam path and also, there were two beams coming out of the first beam dump with different polarizations. The latter occurrence may have been some combination of polarization/alignment error, but it's kind of weird that there were two beams (back reflections maybe?). Anyways, we solved the overall issue by rotating the isolator because the unaligned polarizations caused the entire beam to be dumped (the isolator used, IO-5-1550-HP, is polarization dependent). After rotating it, the isolator now transmits light and there is no significantly visible light coming out of the beam dumps. We should be careful though because one of the isolator's beam exits is now angled upwards with high probability of going into your eye. Beam dumps are not in place yet- there is currently one between the isolator and waveplate just for temporary placement. Also, the arrow on the isolator should point in the direction of propagating light (useful resource for alignment: How to Align an Optical Isolator | Thorlabs Insights - YouTube).

Link to the experiment proposal for context: https://dcc.ligo.org/LIGO-T2400157

More Details regarding Bit Resolution on Biquad Filter

I have attached slides with magnitude bode plots for a 2nd order low-pass filter with critical frequencies of 1 MHz and 10 Hz (and a sampling frequency of 10 MHz). The set of points with lower bits of resolution have been normalized so that they all share the same value (-3 dB) at the respective critical frequency. Slides 2 and 3 show details about the filter with a 1 MHz critical frequency. Behavior of lower resolution deviate as expected with differences becoming more noticeable for values above the critical frequency. The slope for values above 1 MHz is also generally well-behaved, with most being near -40dB/decade as expected for 2nd order filters. 4 bits seems to be a sufficient amount for preserving filter effectiveness. Slides 4 and 5 show details about the filter with a 10 Hz critical frequency. It is worth pointing out that the filter required at least 36 bits of resolution to function; lower values result in a nan error which I am still trying to investigate. It is also worth noting that all the slopes deviate from the expected -40dB/decade, even the float resolution. However, the upside is that the differences plot on Slide 5 shows very little divergence from the float values for lower resolutions, with differences decreasing with more bits.

1550 finesse

Using a PDA05CF2 InGaAs PD in transmission of the filter cavity until it is needed (earmarked for RBQ things I believe).

The 1550 TRANS beam is larger than the small diode on this new PD. Avoiding a lens/ND filter combo for now as there isn't much space. With the beam large it is not saturated.

Took 5 measurements using the same method as described in previous posts except in order to rapidly cut off the incoming light we instead provide a sudden jump on the PDH error signal offset, just above the point where it can't possibly lock anymore. In this case the numbers were:

-Offset required for lock: 2.3 mV

-Error amplitude: 11 mVpp

-Set offset for kick: 8 mV

An example of the data 1550ringdowntransandrefl.png. This looks very similar to the Matt Evans paper measurements Lee linked earlier. Fitting an exponential to this ringdown 1550fitonthedopplerringdown.png yields a finesse of 3300. This is very close to the expected value with the quoted mirror specs. A few nit picks with this so far:

-The decaying exponential should pierce the middle of sine wave if its of the form sin(w*t) * e^-t/tau + C. It clearly doesn't, although is close. The residual (just data - fit) show this not symmetric about 0 residuals.png.

-This is only 1 point, I will process the additional points tomorrow to get an average. 

-The reflection decay is not as clean as the Matt Evans paper and I'm unsure why yet. Would like to get that data working as well. There is additional information to extract.

Testing silicon wafers

[Torrey, Sander, Erin, Briana]

General notes from initial tests of silicon wafer mirrors.

-They came out of the box with visible amounts of dust on them.

-We need a mirror mount that allows for small thickness mirrors. Using the colored mirror mounts in the mean time.

-Turn power down at power distribution to 10 mW coming out of the MISC path. Use this for measuring 1550.

1550 on the wafer mirror

-10.2 mW input / 5.6 mW tranmission - normal incidence

-10.2 mW input / 5.8 mW reflection - 45 AOI

775 on the wafer mirror 

-3.5 mW input / transmission too small to measure - normal incidence

-3.5 mW input / 1.0 mW reflection - 45 AOI

Finesse of 775 cavity

I am going to conclude measurements of the finesse of the 775 filter cavity path. Currently my measurement yields 256.9 +/- 0.21 for the finesse of this path.

Brief recap on how this is done incase you missed the last few posts.

1) Lock the cavity with the 775 light.

2) Have a fast PD (BW > 1/100ns) in reflection and/or transmission of the cavity.

3) (If using REFL ensure channel is DC coupled.) Set the trigger mode on the moku to decaying edge, triggering on either probe point A or probe point B. Zoom in on the time axis to an appropriate window or the resolution won't be high enough to see the decay. The decay time for 775 light is on the order of 100's of nanoseconds.

4) Use the other moku that is power the AOM RF drive, and turn off the drive. 

5) You should see a decaying exponential, save data and repeat if necessary.

I have done this 5 times and saved the data in "\Nextcloud\GQuEST\B102\Output Filter Cavity\multiple_775_ringdowns\*.csv". Attached is a simple script that will parse all .csv files in a directory. Note for glob to work the directory should contain nothing but the data files you want to process. You should also uncomment the plotting lines to ensure your fits are working correctly. 

Now onto the 1550 path!

Hardware latency baseline

Finally was able to run the biquad filter on a KC705 board we had lying around, and I'm measuring a latency of 50ns (5 cycles at 100MHz) rather than the 30ns I was getting on simulations.

Next up I'll try to baseline performance on the Artix 7 FPGA as we're looking to design some boards around it, and of course play around with implementation details to save some cycles or increase clock rate.

New Ringdown Measurement

[Daniel, Sander, Torrey]

A new PD was installed in transmission and reflection for the 775 filter cavity path. These are both high BW 775 optimized PDs. We can then redo the cavity ringdown measurements. 

New ringdown measurement shows a few interesting things. 

1) The initial measurement on the 1811's extra decay term does not show up on these PDs. I simply fit a single exponential to the data and take the decay constant from this.

2) We see this blip - that we now think is associated with the decay of the AOM light, discussed later - only in the REFL PD and not on the trans PD.

3) The REFL and TRANS curves agree to about 1%. Yielding a finesse of 130 and 132 respectively if we believe Matt Evans. Or 260 and 264 if you believe Steck - 7.2.5. I think this is sufficiently accurate measurement of the finesse. Hopefully there is this factor of two and we are much closer to the expected finesse (i.e. not super lossy).

---------------------------------------------------------

An interesting aside that came along with this measurement. We believe we know what this blip is on the REFL side of this measurement. Note the approximate time it takes the light coming from the AOM to turn off - AOMturnofftime. This measurement is made in the same way with no cavity locking, just using the first cavity mirror as an HR. This time frame roughly corresponds to the blip being seen on the cavity ringdown measurements - new_ringdown_measurement_zoomed.png. The explanation is as follows: (Image to refer to - rn_image_picker_lib_temp_be29ac5f-b54e-4fc9-849e-b251590347bc(1).jpg). The beginning of the blip is the incoming E field (E_in) beginning to rapidly decay. The REFL pd is reading out the quantity |E_in + Ecav|^2, where E_cav is a negative value because the two signals are 180 degrees out of phase as defined by our PDH lock. Thus, as E_in decays this total quantity decreases. The minimum point of the blip is the point in the E_in decay where E_in = E_cav, maximally destructive (I suspect if this was done at a higher resolution the blip minimum would be 0). After the minimum, the blip rises back up because the light built up in the cavity starts leaking out, increasing E_cav. At the top of the blip E_in = 0 and |E_in + Ecav| is purely from the cavity leaking.

3D Printed New OFC Alignment Tools for use with the Thorlabs VRC4D05

I 3D Printed two new OFC Alignment Tools for use with the Thorlab VRC4D05. Because the Bambu Labs printer makes horizontal holes somewhat deformed, I printed a new pair with more circular holes. See the attached SolidWorks and STL file for one chirality of the pairs. I printed the file and its mirror image. See attached photo.

I still need to purchase and attach two of the VRC4D05 onto the front face of each part.

SM1 Tapping Piezo Bases for In Vacuum Piezoss

From 3D Hubs, I have some parts that will hold a mirror and viton o-ring using an SM1 mount in an assembly that will hold a piezo to actuate the length of the cavity. 3D Hubs did not tap the hole, they just made the 1.011" diameter minor diameter pilot hole. For a square-shaped part for the GQuEST Output Filter Cavity, I previously used a 2D Mill to hold and tap the part. Since this part has cylindrical symmetry, I used a lathe.

I place the part all the way in the back of the jaws so that the tap would be perpendicular to the hole. I used the Thorlabs SM1 tap and Anchorlube as my lubricant. Anchorlube is approved by LIGO to machine because it is easy to clean. The tap is mounted on a spring guide. To move the tap clockwise, I used a 20 mm ratchet wrench. See attached photo. To move the tap counterclockwise every so often to break up the chips, I used a 11/16 in open ended wrench. I made sure to tap all the way through such that I can easily spin the tap clockwise with my hand before retracting. I then used an SM1RR to ensure the threads went all the way deep. I then washed the lubricant off with water and dried the parts.

The next step is to clean and bake the parts before installing the cleaned and baked 1/4-20 helicoils. 

775 Light leakage through 1550 output coupler

Reported 775 light leaking through what should be 775 HR mirrors (<1 ppm @ 775, i.e. the 1550 input/output couplers) in this post. Confirmed it is not scatter coming from anywhere else but through the 1550 output coupler. This does not explain why it is not gaussian. Most likely some kind of scatter from inside the cavity. Could potentially be from the ghost being observed on the 775 camera. Could also just be the intercavity alignment is not optimal and the beam is clipping on exit.

New 775 FC REFL PD

I have taken a thorlabs PDA10A2 from the recent large thorlabs purchase to use as the REFL PD on the 775 FC path. Previously the PD in use was an 1811 which has very low responsivity at this wavelength. Set it up, realigned, and we have a much better PDH signal (more prominent sidebands, larger PDH error amp/locked RMS ratio, etc).

Moving the Plant to Environmental Shaping Filter for Riccati Solver

In the Buzz code we move the plant to shape the seismic noise without shifting the controls noise is:

\[ S_c = \left |\frac{EP}{1-KP}\right |^2  S_\mathrm{env} + \left |\frac{MKP}{1-KP}\right |^2  S_\mathrm{meas}  \]

where \(K\) is the controller, \(P\) is the plant, \(E\) is the environmental noise shaping filter, \(M\) is the measurement noise shaping filter, \(S_c\) is the controls noise PSD, and  \(S_\mathrm{meas} \) and \(S_\mathrm{env} \) are the unshaped PSDs of the respective noise. If we move the plant to the env noise shaping filter such that \(E'=PE\),  \(P=1\), and  \(K'\) is a new controller calculated for this system then the controlls noise becomes

\[ S_c' = \left |\frac{EP}{1-K'}\right |^2  S_\mathrm{env} + \left |\frac{MK'}{1-K'}\right |^2  S_\mathrm{meas}  \]

In order for these two ststments to be equlivilant

\[ K'=KP \]

I think this means there needs to be some sort of elimination of the \(P\) from the \(K'\) since what we are looking for is \(K\) and not \(K'\).