Yuheng, Michael

We tried to look at squeezing again to see if our issues went away.

We initially tried replacing the ND filter covering the IRMC reflection photodiode. Currently, we have for screw on ND filters, ND 0.3 (too weak), ND 1.0 (too strong), ND2.0. But no ND 0.5 lying around. I purchased a variable ND filter a while ago so we tried putting it between IRMC reflection and the error signal PD. I could adjust it so that the base level of the reflection spectrum is close to saturation (~ 8V) and the PDH error signal is the usual shape with 1.25 Vpk (3 times what it was before). The attentuation was 1.06/2.81 = 0.377 which is ND 0.42. The threshold was adjusted to the good point and the IRMC was locked. The gain had to be turned down from ~ 10 to about 4 to prevent oscillation. However, the homodyne spectrum did not look so good. At first I guessed this was due to the influence of the variable ND filter, so I removed it and put ND1 back (which is what we were using before). But then it turned out that the homodyne was quite misaligned. I rebalanced the homodyne and the spectrum became ok. Anyway, I should purchase some more screw on ND filters since that variable one is not AR coated for 1064.

I could see the CC error signal that indicates that nonlinear gain is present (roughly, nonlinear gain*sin(green phase mismatch)). It seems stable and not causing problems. So the squeezing output seems fine.

I wanted to demonstrate LO/sqz overlap using AMC, but then the IRMC lock became unstable again. I am not entirely sure what is causing the IRMC lock issue. Staring at it for a while, the PDH error signal sometimes becomes much smaller. It seems to go away after we leave and then turn off the high voltage driver and come back some time later. It could be that the high voltage driver is faulty. Switching for an optimal ND filter might be a temporary fix, but it seems we need a new HVD...

Another issue was that the GRMC kept unlocking. It seems the phase of the PDH error signal is randomly changing in a seemingly not continuous manner. Sometimes it is in perfect I-phase and sometimes it goes to almost Q-phase. I don't know what causes it right now but it could have something to do with the 88.3 MHz lock scheme. 

The TF of ckt with EOM was measured using moku. The resonant freq should be around 394kHz, but I don't see any peak at all (even after zooming in at several frequencies). 

Fig 1 - simulation

Fig 2 - moku measurement

I don't know why the circuit doesnt have a narrow rf peak. The gain magnitude doesn't match simulation and there is no peak. 

The RF part and opamp part are currently disconnected. The response of ckt Fig 1 is shown in Fig 2. 

Although the bandwidth of Op27 is 8MHz, it doesn't behave well, after kHz range. Perhaps I should tweak the feedback resistor value. The cables used to measure are very short (<20cm)

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\Coating measurement\BB1-E03\20241206\Fri, Dec 6, 2024 2-09-14 PM.txt

The 2nd BS was installed to measure coatings

input calibration

infront of 1st BS with only power meter C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Polarization states\Wed, Dec 4, 2024 6-39-05 PM.txt

polarization states after 1st BS C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Polarization states\Thu, Dec 5, 2024 1-10-34 PM.txt

Measurement of the 2nd BS (the BS used to measure coatings)

Front face reflection

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\Beam Splitter Calibration\20241205\frontrefl\Fri, Dec 6, 2024 10-55-55 AM.txt

Back face tranmission

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\Beam Splitter Calibration\20241205\Backfacetra\Thu, Dec 5, 2024 3-15-50 PM.txt

measurement take again after some time

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\Beam Splitter Calibration\20241205\Backfacetra\Thu, Dec 5, 2024 6-35-25 PM.txt

The Back face transmission is the input polarization states

The maximum voltage that can be provided to this circuit should be 0.085V. If the voltage exceeds this, then the OpAmp will be damaged as the current would be more than 20mA. 

So, the safe range for input voltage is from 0-0.085 V. The voltage to crystal will be 0-644V. The half wave voltage is achieved at 0.05V of input.

The PD infront of laser(lightwave 126-1064-100) to just check its stability.

Nishino,

2024.12.04

See

LockAqu/DataLogger.ipynb

Noise2/DataLogger2.ipynb

Fig.1 shows lock acquisition process. IR and GR are phase-locked. First, the main IR is locked to the main cavity. Second, by changing voltage to the piezo on PCM, one finds the optimal position of the PCM which realizes speed meter. Third, by changing the local oscillator frequency for PLL, one makes the fundamental mode of the GR laser resonate in the PCC. Finally, by turning on the loop, the PCC is locked.

Fig.2 shows the IR transmission, which is the interference of the first and second circulation light. This reflects the fluctuation of the PCC length for the IR. Red and blue are the ASD when turning the PCC locking on and off. 

The worst mount is attached. Never buy them, as they have the worst designed actuators and the worst possibly designed tap. They don't have anything good about them!!!!

Please don't buy them if you want to save yourself from facing worst hassle!

I think its from newport.

I checked a bit a day later. The IRMC lock works for about 5 minutes and then goes back to being broken. When unlocked, it seems like the error signal peak voltage is fluctating. I'm not sure if it's the amplitude or demodulation phase. When the lock switch is activated, the servo finds the threshold value but doesn't seem to apply sufficient correction to lock to the good point, and then gives up.

I checked once again the green path. SHG now operates normally, outputting 310 mW green (1.88 V on transmission monitor).

GRMC was a bit misaligned seemingly from PZT hysterisis. I adjusted a bit the steering mirror after the MZ and it went back to good alignment - 1460 mV + (60, 34, 34, 22) -> 90.68% mode matching. GRMC demod phase was changed to 230 degrees, error signal 1.74 mVpp.

GRMC/MZ locks fine and outputs 25 mW stable green (312 mV on the transmission monitor) at the usual 4.2 MZ offset.

OPO error signal was 832 mVpp with good phase. ppol to OPO mode matching was 87.43% which is ok. OPO lock is fine.

I checked how much the CC PLL servo can maximally correct using the fast loop. When the setting is far from resonance (~ 32 MHz) and the fast loop is activated, the CC frequency goes to about 14 MHz. i.e. the fast loop maximum correction is about 18 MHz. This is consistent with what I had read previously in Matteo's thesis. It seems the CC PLL behaves well when the laser starts at a detuning of about 15-20 MHz.

I came across the first hitch which is that I cannot see nonlinear gain on the CC1 error signal and very little on CC2 error signal. Green alignment, ppol frequency and OPO temperature are all optimized. I remember that I was using some old short cables but only remembered after I left and was writing this elog. I should check the LEMO cables monitoring CC EPS and make sure they aren't bad. Considering I saw 6 dB squeezing a couple of weeks ago I don't think CC, BAB or ppol lasers suddenly misaligned. For reference, if the green phase shifter is scanning (nonzero scan amplitude on CC1 servo), you should be able to see about 500 mVpp sine wave from the CC1 unlocked error signal, which indicates that the green phase is being cycled between amplification and deamplification of the CC infrared beam.

Then I ran into a worse problem which is that IRMC servo does not work properly. I saw that the reflected power was extremely unstable (about 20-30% fluctuation in output level), and then got worse in that auto lock didn't work at all. I closed CC and ppol, checked main laser current and temperature, turned main laser noise eater off and on, turned NIM rack off and on, PZT off and on, but it didn't fix the issue. The error signal randomly reduced to about half the amplitude of what it was before (200 mVpp). I checked the demodulation LO - it's fine (4.75 dBm). I put the IRMC near resonance and checked the amplified IRMC REFL RF signal - it seems ok and goes up to about -8 to -9 dBm which is what it was before. The error signal size seems fluctating a bit before the error happened, but it's now stuck at half. When the threshold is at the good value, it seems like the servo is trying to force the PZT to the correct position, but it doesn't reach the lock value. A guess might be the interplay between the PZT scan and threshold. Maybe the correction signal is not strong enough to stop the PZT from scanning? I'm not entirely sure but it's just a guess. But I also tried adjusting the PZT scan amplitude and period and it didn't really fix anything. I have also read previously that the IRMC threshold is a bit strange and usually works best at about 90% of the peak value rather than say 50%.

SHG threshold and Auto Lock magically fixed itself and I don't know how or why. It even returned to the same value I had set (9 V). Output green is 310-320 mW, same as before.

I suspect the circuit break is at the marked location since neither the THRESHOLD OUT nor the SHG lock work correctly when the threshold dial is adjusted over the full range. However, THRESHOLD OUT can detect fast changes in the value, so the power supply connection and potentiometer probably still work.

Unfortunately the suspect point is the connection to screw terminal J14 and resistor R104 which is in a very inconvenient location.

Yuheng, Michael

We continued looking at the CC PLL bandwidth using the summing amplifier transfer function measurement.

I saw an old post by Marco Vardaro which says that the nominal bandwidth of the PLL is 40-50 kHz, though maybe the settings were much different back then. Analog devices datasheet says the loop bandwidth is about 20 kHz (fig 1). In this case it was just inspected from the width of the spectral peak, so not particularly accurate.

We took open loop transfer function by checking the coherence in FFT mode (fig 2). In this case, the frequency response is achieved by injecting white noise and taking Ch2/Ch1. We start with a small noise value and increase it until the coherence is approximately 1 over the frequency band of interest. For the CC OLTF the maximum noise excitation is about 30 mV. Also for whatever reason the spectrum analyzer initially gave me frequency response magnitude in dB/rtHz which was a bit odd, I don't know what setting was doing that. It went back to just dB after I switched to swept sine and then back. The unity gain frequency is about 11 kHz and the unity gain phase is +50 degrees (fig 3), which is a bit strange. Judging by this as well as the earlier PLL data it seems there is room to increase the bandwidth. We also checked in swept sine mode for which the optimal excitation amplitude was 15 mV. It gave almost the same result 11.2 kHz UGF.

Afterwards I noticed that this was actually measuring both the fast and slow loop. We repeated the measurement again with only the fast loop. In this case the optimal swept sine amplitude dropped to 10 mV and the UGF very slightly decreased to 10.8 kHz (fig 4, 5).

We used the ppol PLL as a reference check. Looking at the fast loop we see a unity gain frequency of 9.6 kHz with phase 50 degrees. In this case the optimal noise excitation was 150 mV and optimal swept sine was 100 mV (fig 6, 7). Even though the unity gain frequency is the same, it seems the ppol PLL is more robust against unlock. In this case the possible issue might not be control bandwidth but rather the dynamic range of actuation. 

While searching for elogs about PLL bandwidth I came across a previous post by myself and Yuhang. At that time it seems like the correction signal for the CC PLL was too small. This was when we were having small glitch issues and not the major instability. In that case it seems that when the CC PLL was left floating 2 MHz off the setpoint, the fast correction signal was only 200 mV, versus the CC laser PZT tuning coefficient 1 V/MHz. At the time the ppol PLL fast correction signal gave the correct value. When I tried it today, CC detuning of about 7 MHz and ppol detuning of 3 MHz maxed out the fast correction signal.  This is a bit strange so we should check again in detail tomorrow.

Aso, Yuheng, Michael

There is some indication that the CC PLL has a weak lock, so we tried to make an open loop transfer function measurement. Unlike the cavity servos however, the CC PLL servo doesn't have a perturb or monitor ports.

Takahashi-san gave us an old generic summing amplifier (from the original TAMA end mirror) which can be inserted into the CC PLL feedback path to inject swept sine or noise. It has IN, OUT, ADD which function like EPS1, EPS2, PERTURB IN on the cavity servos, and then monitor ports for the input and output signal. There are four sets of amplifiers named after IN/END PITCH YAW.

We checked the frequency response of the summing amplifier. With nothing in ADD, the IN MON, OUT MON and OUT signals should all be the same as IN. We tried first 3 Hz 1Vpp, which was what the function generator was set to. OUT and OUT MON could reproduce the signal but IN MON was attenuated. When the signal was a ramp, IN MON also had some integration, so the summing amplifier is not so useful for low frequency (fig 1, 2). But we don't really care so much about low frequency to check the PLL fast loop. We then tried 10 kHz 1Vpp which was reproduced in all outputs so it seems fine for higher frequency (fig 3). 60 kHz 0.2 Vpp was sent to ADD and it seems to have the proper response for the output and monitor channels (fig 4).

After testing we set up as follows. CC PLL FAST outputs the fast correction signal to the laser PZT. We connected this to IN, then connected OUT to the laser PZT. ADD, IN MON and OUTMON go to Source, Ch2, Ch1 on the SR785 spectrum analyzer (fig 5). The PLL was locked and then we checked the transfer function up to 100 kHz for swept sine. The excitation frequency that could be applied before unlock was quite small, about 10 mV. But we didn't see anything really meaningful. We checked in FFT mode using noise excitation but couldn't get meaningful coherence anywhere in the full span of the spectrum analyzer. I should more carefully check the relevant frequency range and actuation strength of the CC PLL fast signal. The SR785 only goes up to 100 kHz. The other spectrum analyzers we use to monitor the PLL can look at up to 2 GHz but are only single channel.

When we locked the PLL, I compared the time traces with what I saw previously to show the CC glitch problem (fig 6, from before). Only to see that the CC glitches did not appear (fig 7)... I don't have any idea what changed.

The calibration of new BS is shown in plot 1. 

Input pol power has been normalised by the input laser power. 

all other powers have been normalised by both the input laser power and the input power after the LC (which will be referred as input pol power).

Finally, I also normalise the input pol power by itself to have a unit 1 comparison for powers. 

Aso-san showed me how to properly replace the potentiometers in the servo boards. So I hoped it was just the knob that was broken and not the variable resistor itself.

I took a replacement control knob from the elec shop, there's a lot of them.

Before reattaching I rotated the knob to about 5 out of 10, since the threshold was previously stuck at 0 volts, which is the middle of the range setting.

Unfortunately, the threshold out reading is still broken, as is the SHG lock. Threshold out reads zero normally, but when the dial is rotated there is a threshold voltage reading the magnitude of which depends on the rotation speed. So I think the potentiometer and threshold power supply connection work and there is a small circuit break somewhere common to the error signal and Threshold out reading, which is small enough such that the capacitance can detect a fast change in the potentiometer voltage divider.

Previously I had some issue with even basic shot noise spectrum being a bit glitchy, but actually it was just sensitivity to the ceiling lights. The LO shot noise spectrum is fine. Dark noise still has 50 Hz at odd numbered harmonics even with the homodyne plugged into the same power supply as the spectrum analyzer. Maybe I should try spectrum analyzer floating ground

I modified the circuit to match impedance with mokugo. Seems moku go only has impedance of 1M ohm and not 50ohm. So, I added 1Mohm (previously 50ohm) infront of Opamp and 10Mohm (previously 1.1k) on the feedback. I measured the voltage output after the opamp (CH1) and before the EOM (CH2). The EOM was connected to the circuit during this measurement. 

The output doesn't match the expectation. Since I input 0.05V p-p from moku, I should get 0.5V after Opamp circuit (coz its gain is 10). But I get output as 8V both after the opamp circuit and before EOM. The picture is from the oscilloscope. 

Maybe its because I left ground floating when I soldered the SMA cables. 

We sometimes have a power ratio of more than 1 in our measurements. There could be several reasons for this, laser switched off between two measurements, some fluctuations after LC, etc. Hence, another BS (BS014) is added after LC to monitor some fluctuations that were not previously accounted for. I placed the power meter after 7 holes in the reflection of BS, and the camera after 15 holes from the transmission of BS. Sometimes, we require to know the size of the beam as well to take into account the power density. 

The LC voltage change from 0-3.5V, with 0.1V step. Another power meter has been incorporated into the labview, and therefore 2 new columns has been added to the saving file. The last two column have the information from mean and std of power from this power meter. 

1. After the BS with power meter in reflection and camera in transmission

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Polarization states\Fri, Nov 15, 2024 3-15-13 PM.txt

2.  Only the power meter is after LC for this measurement, just to take into account the input power at each voltage. We only care about polarization states after the BS. 

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Polarization states\Sat, Nov 16, 2024 3-40-15 PM.txt

3. Measurement with polarizer at two rotated angles (rotated about its optic axis)

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\GTPC Polarizer\20241105\45 deg\Fri, Nov 15, 2024 4-50-37 PM.txt

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\GTPC Polarizer\20241105\0 deg\Fri, Nov 15, 2024 6-24-29 PM.txt

4. Measurement with HWP at two rotated angles (rotated about its optic axis)

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\Retarder\HWP\20241115\0 deg\Fri, Nov 15, 2024 8-05-03 PM.txt

C:\Users\atama\Dropbox\LC-Experiment\Measurement Data\Birefringence Measurements\Retarder\HWP\20241115\38 deg\Fri, Nov 15, 2024 9-33-43 PM.txt