On wednesday around noon, the North Path got back to stability. I captured this process by going back to the FB data. The process of coming back to stability is not so instantaneous as the other way round. Also, in this process, the path becomes stable, then unsable and stable and so one with the duration od unstability decreasing until it vanishes. Attached are plots of about 14 hours of curucial channels. If anyone has any insights on what might be happening, let me know.

Data

Today at sharp 8:30 am, perfectly fine running experiment went bad again. The North path became buggy again with strong low frequency oscillations in almost all of the loops except the temperature control of vacuum can. The temperature control of beatnote frequency felt a step change and hence went into oscillations of about 65 kHz.

Not sure what went wrong, but 8:30 am might be the clue here. But can't change/test anything until I can go to the lab.

Data

Since this morning atleast, I'm not seeing the North Path unstability (see CTN:2565) and the beatnote is stable and calm at the setpoint. Maybe the experiment just needed some distance from me for few days.

So today, I took a general single shot measurement and even after HEPA filers being on at 'Low', the measurement is the lowest ever, especially in the low-frequency region. This might be due to reduced siesmic activity around the campus. I have now started another super beatnote measurement which would take measurement continuously every 10 min is the transmission power from the cavities look stable enough to the code.

there is a new broad bump though arounf 250-300 Hz which was not present before. But I can't really do noise hunting now, so will just take data until I can go to the experiment.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant post:

CTN:2565: North path's buggy nature NOT solved

Today, magically almost, the North Path was found to be locking nicely without the noise. I was waiting for the beatnote to reach the detector's peak frequency when in about 40 min, it started going haywire again. No controls were changed to trigger any of this and as far as I know, nobody entered the lab. Something is flakey or suddenly some new environmental noise is getting coupled to the experiment. Attached is the striptool screenshot of the incident and the data dumped. In the attached screenshot, channel names are self-explanatory, all units on y axis are mW on left plot (note the shifted region, but same scale of North Cavity Transmission Power) and MHz on the right plot for Beatnote Frequency.

I know for sure that everything until PMC is good since when only PMC is locked, I do not see huge low frequency noise in the laser power transmitted or reflected from the PMC. But whatver is this effect, it makes the FSS loop unstable and eventually it unlocks, then locks again and repeats.

Evidences

It seems like this was never properly solved. On Friday, the same problem was back again. After trying relocking PMC and FSS on the north path without any luck, I switched off the laser to standby mode and after a minute restarted it and the problem went away. I have a strong suspicion that this problem has something to do with the laser temperature controller on the laser head itself. During the unstable state, I see a spike that starts a large surge in error signal of FSS loop which occurs every 1 second! (so something at 1 Hz). The loop actually kills the spike successfully withing 600-700 ms but then it comes back again. I'm not sure what's wrong, but if this goes on and the lockdown is enforced due to Corona virus, I won't even able to observe the experiment from a distance :(. I have no idea what went wrong here.

I found out that since the slow PID of the FSS used Cavity reflection DC level, it is important that this path remains rigid and undisturbed from any testing. In CTN:2559, when Shruti and I were realigning North PMC, we used  BNC-T (which was permanently attached to the rack) to pick-off North cavity reflection DC signal. During this putting on and off the cable, we made the BNC-T itself loose and the connection to acromag card became buggy.

I have fixed this by connecting the acromag cards directly to the cable coming from the table behind the rack. Now connections/disconnections at the front of the rack shouldn't disturb this vital connection. Still, we need to be careful about this from next time. I had no idea what went wrong and was about to start a full-scale investigation into PMC and FSS loops. Thankfully, I figured out this problem before that.

On March 13th around 7:30 pm, I started a super measurement of the beatnote spectrum for over 2 days. The script superBNSpec.py took a beatnote spectrum every 15 minutes for a total of 250 measurements. The experiment was stable throughout the weekend with no lock loss or drift of the beatnote frequency. All data with respective experimental configuration files are present in the Data folder. HEPA filters were on during this measurement.

Analysis and Inference:

• I first plotted time series of how beatnote frequency ASD at 300 Hz varies over the 2 days.
• Remember, this is a median measurement done using function modPSD which uses modWelch function we wrote a while back (See CTN:2399).
• Then I plotted the same for a bunch of frequencies from 200 Hz to 1 kHz.
• It is clear that the time of the day does not have any major or meaningful effect on the beatnote spectrum. It mostly remains the same.
• Finally, I took the median of the 250 ASD measurements.
• I also calculated lower and upper bounds using RMS of difference between 250 upper and lower bounds and the median calculated above.
• All the noise peaks are intact indicating that all noise sources are stationary and REAL (in Will Farr's language).

Data

Today, I added a new out-of-loop transmission PD (Thorlabs PDA10CS) for the north path. This will be helpful in future measurements of RIN coupling to beatnote noise. This PD is added at (8, 42) using the dumped light. The optical layout would be updated in a few days. I've also connected Acromag channel for North Transmission DC to this photodiode, so the transmitted power channels and the mode matching percentage channel of North Cavity are meaningful again.

ISS Gain for the Northside has been increased to 2x10000 since half of the light is now being used by the OOL PD.

Today, I added a new out-of-loop transmission PD (Thorlabs PDA10CS) for the south path. This will be helpful in future measurements of RIN coupling to beatnote noise. This PD is added at (1, 40) using the dumped light. The optical layout would be updated in a few days. I've confirmed that this photodiode is reading the same RIN as read earlier in CTN:2555. I've also connected Acromag channel for South Transmission DC to this photodiode, so the transmitted power channels and the mode matching percentage channel of South Cavity are meaningful again.

[Anchal, Shruti]

We realized that the half-wave plates before the EOAMs probably had no real function in the setup and therefore we proceeded to remove the one from the north path at (39,121) aka row 39, column 121 of the Optical layout.

After this was done, we had to re-adjust the quarter wave-plate (39,112) after the EOAM (39,115) to make sure that the EOAM was still functioning about the 50% transmission point. The beam going into the PMC was also re-aligned by adjusting the two mirrors at (32,92) and (37,92). Finally, the mirror at (43,88) was adjusted to align the beam reflecting from the PMC into the photo-diode.

We were able to re-lock the north PMC and north cavity after increasing the power in that path by adjusting some waveplates.

As may be expected, the sign of the ISS feedback had to be inverted. The ISS actuates on the EOAM; removing the half-wave plate would have switched the circularity of the polarization of the beam entering the PBS at (39,110), so the sign of the voltage that would have previously caused the transmission to increase would now cause it to decrease and vice versa.

I think you have to protect this SimPlant board from the HV from the FSS board ?

We added hooked up the plant and measured the in loop transfer function (shown below). The circuit locked in the system... so it works.

Next, we want to measure it with the south FSS to compare the two since anchal tells me he has never seen a direct comparison of the two.

We also need to calculate the gain of the full system. I am trying to do this on my own and get help from Anchal when I run into a problem.

Last night I switched off all the fans in the lab and we have reached the lowest ever recorded beatnote noise.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant posts:

CTN: 2551 : Comparison between Out of loop vs In loop RIN

I took spectrum of Out-of-loop (OOL) photodiode and In-loop (IL) photodiodes with transmitted light from the cavities when ISS is on in both paths at gain value of 2x10000.

Measurement:

• In-loop photodiodes of ISS have been disconnected from Acromag cards as this was injecting noise into the loops.
• So, without witnessing the DC level, we can increase the ISS loops' gain to 2x1000 with HF and LF both set at 300 Hz with 6 dB/octave roll off.
• A major problem in this measurement is the slow drift of DC power level during the time of measurement.
• I was measuring DC value of photodiode averaged over 10 s with an oscilloscope just before the measurement was taken.
• In fact, one can see that between two measurements of the different spans, the DC value has changed enough to show mismatch in South In-Loop RIN.
• The out-of-loop photodiode was measured with sequentially blocking either the north laser or the south laser.
• The same issue with knowing the exact DC level persists here as well.

Inference:

• As expected the out-of-loop noise is higher than the in-loop noise, but I did not expect a factor of ~5 difference.
• However, this method of measuring RIN isn't very faithful.
• Overall for the experiment, it is clear that increasing the gain of ISS further did not lower the beatnote spectrum much.
• This suggests that we have damped down photothermal noise in the experiment below other fundamental noise sources.
• My future efforts would be focused solely to get rid of the peaks in the range of 200 Hz to 1 kHz.
• I see that the sharp peak at 480 Hz has changed into a tri-peak around the same frequency without any changes in the experiment other than ISS.
• This atleast indicates that this peak was not undefeatable, and so must be the remaining ones.

Data

I added a capacitor to c8 in the FSS plant D2000020. This yielded the transfer function shown below. This seems to have pushed the pole lower. 

After testing with Anchal on the FSS box we found that this seems to have done the trick. More to follow.

These measurements take a long time as I take a median over 500 single measurement instances. I was able to increase gain later after having taken in-loop noise measurement, so I didn't repeat it. But what I can do is a take a quick in-loop measurement today with simple averaging and no error bars and post it here for comparison. If we are really interested, I can run overnight measurement for in-loop at this gain as well.

weird - why is the Gain different for in loop and out of loop ?

Attached are the latest transmitted RIN measurements.

Measurement:

• I setup a Thorlabs PDA10CS photodiode at the dumped end of beatnote (15, 13).
• Then, I took out-of-loop RIN measurement of transmitted light by blocking either the North or the South path falling on the beatnote beamsplitter (11, 19).
• I found out that, when the cavity transmission photodiodes are connected to Acromag input, the acromag input injects some noise into the ISS.
• On disconnecting the acromags, I was able to increase gains in ISS loops to 20000.
• A major issue was to measure DC power together with the spectrum to calculate the RIN. I first connected the measured photodetector to Acromag Cavity Transmission Channels (which are empty now) and measured the ratio of DC level with the cavity reflection photodiode.
• Later, using the cavity reflected dc level and this ratio, I was able to estimate the dc level of the spectrum I'm measuring. This way, I ran RIN measurements sequentially for North and South transmitted light.
• The measurement is done by running SR785 repeatedly with no averaging and saving all the curves. 500measurements were attempted, few of them had file transfer errors and were ignored.
• The median, lower bound (15.865% percentile (lower 1-sigma for Normal Distribution)) and upper bound (84.135% percentile (upper 1-sigma for Normal Distribution)) are calculated.

Inference:

• Higher gain improved the noise a little bit more upto 30 kHz.
• Notice that there are 60 Hz harmonics (both odd and even) in the South RIN, while they are not present in the North RIN.

Data

Anchal and I connected the circuit to the EOM and PZT in a closed loop (LIGO-D040105). We looked at the output as seen in the Oscilloscope data and it does not look like the sinusoid that I was expecting. It looks like the sum of a square wave and a misshapen sinusoid. I think that there is some sort of reflection of the signal that is causing this but I don't know where. When we took the spectrum of that data and we saw a large peak at around 2kHz and then harmonics of the signal. (Note there are two spectrums both with the same data just one is on a log-log scale)

We also measured the open-loop transfer function of the circuit and the results are given below. The unity gain frequency is 1240 Hz. This is different from the 2688 Hz that is the location of the first peak in the spectrum. We had thought that maybe these harmonics were a result of the UGF but they don't occur at the same frequency.

We will try to figure out why we have harmonics and what we can do to prevent them. We really have no idea what they are.

Attached are the latest transmitted RIN measurements.

Measurement:

• These are taken with in-loop photodiodes of ISS and hence are mostly Fake News. I'll do an out-of-loop measurement as well.
• The measurement is done by running SR785 repeatedly with no averaging and saving all the curves. 238 measurements were taken this way.
• The median, lower bound (15.865% percentile (lower 1-sigma for Normal Distribution)) and upper bound (84.135% percentile (upper 1-sigma for Normal Distribution)) are calculated.

Data

Measurement at 3 am in the morning today has been the lowest ever recorded beatnote noise. The lasers have been locked for more than a week and the temperature of the cavities is also very. The ISS gains were increased yesterday to 5x1000 on each loop. I've also added RIN measurement and implied photothermal noise.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

The issue isn't visible right now. I'm not sure what changed but with a similar power level, I'm able to increase the gain on North ISS to much higher than before. Currently, I'm taking RIN measurements (still using in-loop detector) to update the noise budget plot which is taking some time as I'm trying to measure the uncertainty in the measurement as well. So, the step response plot would come later.

OLTF

• Open-loop transfer function of the North ISS loop was measured by exciting at port B in SR560 and using mode A-B.
• Measurement was taken as Input A / Output which is equivalent to open loop gain divided by the transfer function of SR560.
• I measured the transfer function of SR560 separately and hence obtained the OLTF.
• The first measurement is actually product of Actuator TF, Plant and Detector and I have plotted it as well.
• This plot shows that a pole of 20 kHz is coming through the plant. It can't be detector or actuator as they have very high bandwidths.
• At the lower end, we couldn't reach to the point where AC coupling sets in. SR560 manual (at page 12) says it is 30 mHz.
• Lower UGF is roughly 5 Hz, which the gain margin from lower ugf must be more than 20 dB.
• Since the power level of laser changes somehow  over day, keeping this OLTF is not really fixed and moves up and down. Hopefully, in the present setting, it doesn't hit oscillations.

Witness detector?

The beatnote detector SN101's DC output has electronically railed. This might be due to a disconnection inside which may or may not be intentional. I don't think I should meddle around with that detector so close to the result. I will instead replace the other 1611 detector in the blocked port of beam splitter (11,19) with a thorlabs PDA10CS and use that for RIN measurement. But since, the loops were in modd of being stable today, I decided to go ahead and take the measurement with current settings. This will also work as in-loop measurement for comparison later.

Data

do some step response and swept sine and post plots

I'm currently struggling with a problem in North ISS which I think needs to be documented here. Here's the synopsis:

The problem:

• After running for the weekend with high gains and gain peaking at ~45 kHz, I found this week that the North ISS was overloading.
• The power level reaching the cavities change over the course of the day by +- 20-30% usually (still a mystery to me), so I assumed this could just be because of increased laser power saturating the SR560.
• But what I have found is that except for some random instances, I can't keep the gain high enough in this ISS path.
• At the start, I thought this problem is at least reset-able by unlocking the North FSS and locking the cavity back. But later, I've found that it is not working either now.
• The defining feature is, as I slowly increase the gain (giving ample time for the loop to settle), at the first point where I see any sign of overload, it happens at a rate of ~300 mHz, that's once every 3  seconds!
• I checked at the photodiode output at this gain setting and I see the loop gets unstable and for about half a second every 3 seconds or so.
• On increasing the gain further, this overload just becomes a constant condition.
• My first guess is that somehow the lower unity gain frequency is becoming unstable. But the mystery is:
  • Why did this not happen before?
  • Why is this not seen in the South Path at all? It is healthy as ever.
  • What really changed during the weekend?
  • And why once in a while, it was actually working?
• The only difference in the SR560 settings for the two loops is that the South path has 'invert' ON while North path does not. At first, I was perplexed by this too. South needed 'invert' to be stable and the North did not. Maybe this is the reason?
• I have ruled out any failure in SR560 by switching it with another one and seeing the exact same phenomenon.
• Remaining suspects are a damaged BNC cable/connector, damaged photodiode or EOAM (I wish not). However, some bad control loop configuration is my primary suspect.

I would like it if anyone has any comments on any of the points above or suggestions to tackle this problem. I can make some measurements and post them on requests.

1. too much gain peakin!
   1. 0 dB - thats no fun
   2. 20 dB - too much cowbell !!
   3. 10 dB - ahhhh, that's nice....
2. in-loop performance is "fake news": u need to have an unbiased reporter

After installing the new ISS, the noise is even lower in lower frequencies. Still no change in the noisy peaks at 400 Hz and around 800 Hz. There is also no difference in noise floor above 200 Hz.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant elog post:

CTN: 2542: Installed New ISS on both paths using SR560s

I've installed new ISS consisting of one SR560 per path.

ISS SR560 Details:

Measurements:

• I took transfer function by exciting at port B and using mode A-B on the SR560.
• The transfer function was measured as Actuation Signal (output of SR560) over Error Signal (input to SR560).
• This transfer function should be equal to open-loop gain over the SR560 transfer function.
• So, I took another measurement of the transfer function of SR560 itself not connected to the loop.
• This gave me total open-loop transfer function.
• I used as the transfer function of EOAM in units of relative intensity change per V (1/V).
• I used the coupling factor (See CTN:2528) and DC power level of the laser after transmission to get the detector transfer function in units of V.
• Using the above, I calculated the plant transfer function for each path and then used that to calculate the closed-loop transfer function.
• To compare, I also calculated noise suppression before and after switching on ISS.
• These two separate methods match very well. At 300 Hz, intensity noise is suppressed by 40 dB.
• There is no other witness detector after transmission to use as an out-of-loop photodiode, neither is there any space to put a new one.
• There is oscillation happening at UGF  fo around 40 kHz in both paths. I can work on it if this looks like a problem in the final beatnote or if it seems to saturate FSS loops. Currently, I saw nothing alarming.

Data

After switching out a bad opamp we have clearer results.

Next steps: moving the real value closer to the predicted value and determine the voltage the circuit can handle.

After installing the preliminary ISS, which I'll change tomorrow as per Rana's suggestions, we see some reduction in the beatnote noise in the lower frequency region. I think I should also have an estimate curve for the coupling of laser intensity noise into the final result. I can maybe make some sort of transfer function measurement from actuation on intensity to the beatnote frequency itself using moku.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant elog post:

CTN: 2538 : Installed ISS on both paths using SR560s

CTN:2539 : Rana's suggestion on ISS.

you don't need 6(!!) SR560s - just 1 for each loop:

• AC coupled,
• 1st order high pass at 300 Hz,
• 1st order low pass at 300 Hz
• Aim for a UGF of ~10 kHz

Then use multi-res spectra and check out the out-of-loop noise (with loop on/off). The in-the-loop noise is always an underestimate.

I've installed ISS on both paths using 3 SR560s each. Preliminary feedback is setup to get a stable loop. More optimization with TF analysis can be done further.

Path changes:

• I replaced the unmarked waveplate in the north path right after the EOAM at (112, 39) with a zeroth-order quarter waveplate.
• I adjusted these quarter waveplates at (112, 39) and (90, 24) to the half transmission through the PBS when EOAM inputs are shorted.
• I followed the notes in Antonio's post (See CTN:1593 and CTN:1604).

ISS:

• ISS is formed with three SR560s in series.
• First SR560 is AC coupled, has a 2nd order zero at 1 Hz and gain of 100 at Low noise Mode.
• The second SR560 is DC coupled and has a pole at 1 kHz (North) and 3 kHz (South) with gain 1 and High Dynamic Reserve setting.
• The third SR560 is also DC coupled and has a pole at 300 kHz with gain 1 and High Dynamic Reserve setting.

Measurement:

• I took the measurement of the spectrum from the output of transmission photodiodes.
• I used the calibrated coupling factor I measured earlier to convert the measured voltage to watts of power right after the cavities. (See CTN:2528).
• Measurement configuration files is attached.
• We see about a factor of 10 improvement in the intensity noise between 100Hz to 1 kHz.
• Note that the bad spikes notes in the beatnote spectrum at 480 Hz and around 900 Hz are present in the north transmission spectrum as well.
• Effect on the beatnote noise would be evident after tonight's measurement at 3 am. In my preliminary measurements, I see some reduction in noise below 200 Hz but nothing much.

Data

The PCB came in and I have assembled it. The preliminary look at the transfer function on the AG4395A shows that there is a problem. The function is very noisy and the signal is very low. I will go through and verify all the solder points are fully connected and generally debug the circuit. A quick measurement of the transfer function on the SR785 of the low frequency (10Hz-100kHz) showed good results earlier so I fear something has come loose. Note the low-fq measurement was not recorded properly. so I have no graph for it. 

I think this is a small fix that needs to be made. 

The graphs below show the EOM and PZT path TF measurements on the AG4395A. They make it clear that something is not connected. This is just a progress report.

With some of the changes done recently, the beatnote noise has lowered in the low-frequency region indicating reduction in scatter.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant elog post:

CTN:2530 : Increased sampling rate to 125kSa/s; lowest noise in higher frequencies

CTN:2533 : blocked NF1811 with hex beam dump

CTN:2535 : BN Detector was saturated. Reduced laser powers.

CTN:2531 : Further iterated back and forth to optimized FSS Gains.

I measured dark noise of the beatnote detector reaching moku and its effect on measured beatnote frequency noise.

Measurement

• I used Andrew's low noise preamplifier to measure this noise.
  
• I first took the transfer function through the preamplifier so that everything later can be referenced back to its input.
• I measured measurement setup noise by shorting the input to the pre-amplifier.
• Then I measured dark noise after the 20 dB coupler (the cable that I connect to moku) when the detector was completely blocked with a beam dump.
• The noise was converted to precieved beatntoe frequency noise by using peak to peak voltage of beatnote signal. (Noise*freq*2*pi/Vsigpkpk)
• The noise came a little less than Tara's measurement. I'm updating my noise budget model with this now.

Data

I have blocked the unused output port of the beam splitter before our beatnote detector with a hex beam dump at (13, 19). This was being used for broadband detector NF1811. We don't need it now.

I found that the beatnote detector was actually saturating and the output was not a good pure sinewave. I've reduced the laser powers reaching the intensity to avoid that so that 20 dB coupled output of beatnote remains around 200 mVpkpk. Following is the summary of changed settings:

However, the beatnote did not change because of these changes showing that moku is strictly sensitive to zero crossings of the acquired signal rather than its shape near the edges.

I took beatnote measurements and spanned gain values in the PMC to see if stability issues in PMC loops can be affecting the FSS downstream.

Method

• I used BNspec.py with default settings (10 kHz bandwidth at 15.625 kSa/s for 60s) while spanning the parameter space of PMC gains.
• The beatnote frequency was robustly within 2 kHz of 27.34 MHz during this whole measurement.
• The overall experiment was run by spanParamSpace.py script.
• First I spanned North gains using the last optimum values found for South Side and tried to span the area towards which we found minima earlier.
• Next, I used the new minima found for North side and spanned region in South side.
• This time, I used sum of the ASD of beatnote frequency between 300 to 600 Hz to determine improvement in noise overall.
• Still, I would say, the improvement is hardly more than 5% over large areas of parameter space, meaning mostly BN noise is independent of these gains.

Final result

Data

I have increased the sampling rate of moku to 125 kSa/s (fastest allowed) for the frequency-time series acquisition. After reading matermost chat of Sean Leavey etc, I felt I might have downconverted aliased noise as earlier sampling rate was just 15.625 kSa/s. This didn't change the noise below 1 kHz but we see some improvement above 1 kHz so I'll keep this from now on. The measured time series files are not around 900 Mb, so I'm not saving them and only keeping the psd calculated using modifiedPSD.py script.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant elog post:

CTN:2521 and replies: Beanote Spectrum vs FSS Gain Values 

CTN:2528: Cleaned up table; Installed hex beam dumps

The PMC error signals have some weird broadened oscillations in them:

Above, pink is North and green is South PMC Error Signal taken from Mixer Out port.

Spectrum measurement:

• I took the spectrum using SR785 of the Mixer Out port of the two PMC loops.
• Next, I measured the cross-spectrum density of the two error signals.
• I wasn't sure though if the peak in the cross-spectrum is just because higher magnitude signal on both error signals, so I divided it by the magnitude of measured FFT of the two signals to get a unitless cross spectrum in the third plot.
• I'm not sure though how to interpret the cross-spectrum in either case.
• The oscillations could be a manifestation of some common-mode noise on the table (maybe some mechanical oscillation).
• Or there could be a systematic bad UGF around 480 Hz.
• In any case, can we actually relate the bump seen in beatnote noise near 400 Hz to these oscillations?

Data

Today I cleaned up the table, removed Scott's RFAM measurement setup and installed hex beam dumps on the input rejection of faraday isolators.

Recalibration

• I measured the laser power at (50, 26) and (50, 29) when the lasers were detuned from the cavities using a power meter. Then I updated the ratio of DC readout of SN010 and SN009 with these power levels to infer reflected power in mW correctly.
• Similarly, I measured the laser power at (9, 26) and (9, 29) right after the cavities and updated ratio with dc readout of cavity transmission photodiodes at (4, 37) and (9, 34) to infer transmitted power correctly.
• I also adjusted horizontal tilt of periscope top mirrors at (46, 26) and (46, 29) to improve modematching to about 68% on North side and 75% on South side (numbers calculated by ratio of transmission to reflected power).

Table cleanup and new beam dumps

• Not complete, but I removed a lot of unused optics, beam dumps, clamps etc from the table.
• I removed the photodiodes I put in before and after PMC to measure the intensity noise in the north path and its origin (CTN:2502).
• I dismantled Scott's RFAM measurement set up (CTN:2377) on the South path which used pick off from the input rejection of the faraday isolator at (58, 19).
• I have now installed new hex beam dumps at the input rejection of the faraday isolators on path path at (62, 16) and (65, 27).

Board Plant Circuit should come in on Feb. 18th. It is really taking its time. 

I took beatnote measurements and spanned gain values in the PMC to see if stability issues in PMC loops can be affecting the FSS downstream.

Method

• I used BNspec.py with default settings (10 kHz bandwidth at 15.625 kSa/s for 60s) while spanning the parameter space of PMC gains.
• The beatnote frequency was robustly within 2 kHz of 27.34 MHz during this whole measurement.
• Overall experiment was run by spanParamSpace.py script.

Inference

• I calculated the integrated total beatnote frequency noise in the frequency range 200 Hz to 1 kHz where the total noise is expected to be dominated by coating Brownian noise.
• Unstable regions in the parameter space are clearly visible.
• Ideally, the noise in beatnote frequency should have very minimal coupling with the gain values in PMC loops.
• This has been found true indeed for the south side but not so much on the North side.
• Very clearly there is a corridor from -3 dB to 8 dB in NPMC gain where noise remains low. We generally use 7 dB value.
• But this might just be too bad PMC loop causing light amplitude noise in the output.
• There is a bulge at 400 Hz in the beatnote freqeuncy noise. I can also see a broadened oscillation in PMC error signals around that frequency on both sides.
• I'm afraid that maybe the North PMC is causing big mode mismatch at 400 Hz causing the noise in the beatnote frequency. Or is that not possible?

Data

I took beatnote measurements and spanned gain values (COM and FAST) on South side to see the variation in beatnote with them.

Method

• I used BNspec.py with default settings (10 kHz bandwidth at 15.625 kSa/s for 60s) while spanning the parameter space of NFSS gains.
• Every time the gain values are changed, a gainCycle.py is done and script waited for 10s before attempting beatnote measurement to let loops settle down.
• The beatnote frequency was robustly within 2 kHz of 27.34 MHz during this whole measurement.
• Overall experiment was run by spanParamSpace.py script.

Inference

• I calculated the integrated total beatnote frequency noise in the frequency range 200 Hz to 1 kHz where the total noise is expected to be dominated by coating Brownian noise.
• Unstable regions in the parameter space are clearly visible.
• But as to the stable region, once the loop gets onto that, the noise does not change much, similar to North side.
• So here as well, it seems like the beatnote noise is largely independent of the gain settings.
• I'm unsure if I should look at this and the last measured data sets differently. Maybe I'm missing something.

Data

Yesterday I took beatnote measurements and spanned gain values (COM and FAST) to see the variation in beatnote with them.

Method

• I used BNspec.py with default settings (10 kHz bandwidth at 15.625 kSa/s for 60s) while spanning the parameter space of NFSS gains.
• Every time the gain values are changed, a gainCycle.py is done and script waited for 10s before attempting beatnote measurement to let loops settle down.
• The beatnote frequency was robustly within 2 kHz of 27.34 MHz during this whole measurement.
• Overall experiment was run by spanParamSpace.py script.

Inference

• I calculated the integrated total beatnote frequency noise in the frequency range 200 Hz to 1 kHz where the total noise is expected to be dominated by coating Brownian noise.
• Interestingly, the beatnote noise in this range is not so much dependent on the NFSS loop gain values.
• I notice a largely flat heatmap with no significant local minima.
• This indicates that beatnote is indeed not limited by North NPRO residual frequency noise in this region at least.
• I'm repeating this measurement today while scanning SFSS loop gains.

Data

Today we have measured the lowest beatnote spectrum till now. This happened because I set the FSS gain values to lower than the maximum I could reach.

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Relevant Elog Post:

CTN:2522: Beanote Spectrum vs FSS Gain Values

CTN:2518: NFSS Boost Stage

CTN:2514: SN010 (South RFPD) Notch Improved

CTN:2512: SN009 (North RFPD) Notch Improved!

Measurement

• Everything is almost the same as the last measurement.
• This time, if you notice from orange to yellow curve, I had to manually lower down the gain on the south common path by 12 dB instead of 3 dB to ensure the south path doesn't go into oscillations.
• From there onwards, the gains were reduced normally by 3 dB in each step/

Inference

• The beatnote spectrum indeed drops down to lowest when the gains are 3 dB below the maximum gains where I keep the loops.
• Maybe keeping a large gain margin is important. But I need help to understand this result.
• Also, it is kind of clear that the laser noise starts showing up from the yellow curve onwards only. So in current gain settings, the noise floor is probably due to something else.
• Another weird fact I saw was that the RMS of error signal taken at "Mixer" port which is the TP5 point in LIGO-D040105-C_CTN_North increases significantly only after the gain values of the loop is decreased by more than 15 dB (to -1 dB each on FAST and COM).
• The same thing can't be said for the south path where I see a gradual increase in this error signal RMS value as the gains are reduced.
• If someone has a better idea of understanding these results or has some suggestions on further tests or a combination of parameter changes I should do, please let me know.

Data

In a discussion with Craig sometime back, it was brought up what happens when I lower the gains of the FSS loops. So today I did a test which lowers the Common and Fast Gain values on the FSS boxes by 3 dB in each step and sees what happens to the beatnote.

Measurement

• The measurement was done solely by bnvsFSSgains.py script present in the Data folder.
• Before each measurement, the gain values are stepped and the script waits for 5 s before measurement is attempted.
• Measurement is done through the usual mokuPhaseMeterTimeSeries.py script which is used for all beatnote measurements of the lab. Recently I was able to make it more robust with some discussion with liquid instruments application engineer.
• All default settings were used except for duration which was kept to 10s to keep reasonable time series data file size.
• SN101 detector's RF out coupled through a 10 dB directional coupler was used to make the measurement (as usual).
• I ensured by watching the error signal on an oscilloscope that none of the FSS loops went into oscillation during the measurement. Its is hard to say for lower gain settings though if that happened or not.

Inference

• The beatnote spectrum noise does not start going up until after the gains have been reduced significantly by around 12 dB at each stage.
• Also, weirdly the beatnote spectrum was actually lower for 2 particular settings of gain which were midway. There the beatnote spectrum is only 2 times as much as estimated!
• I'll repeat this measurement again to make sure this optimal gain setting region was not a coincidence due to some other parameter out of my control.
• But since we do not see a clear scaling of beatnote noise with lowering of FSS gains, I think it is safe to infer that we are actually not limited by residual NPRO noise as has been the thesis for quite some time.

Data

Added all documents to the LIGO document for the FSS Plant Circuit LIGO-D2000020-v1. The updated board and schematic are also below with the zip file of all the Eagle files.

The board was ordered from JLCPCB. The current status is "In Production." Its production will take longer because of the Chinese new year but its shipping should not because it is being shipped with DHL. The low-noise thin-film resistors were ordered from Digikey and should arrive on Friday 01/31/2020.

The next step is assembling the circuit and putting it in a box. I can't assemble the circuit until the parts arrive but I can get a box and add the power supply. This will make the production faster once the board arrives.

Once the board is assembled its transfer functions will be measured and I will work with Anchal to implement it. We are almost done.

After the Notch improvement, I took 20 dB coupled outputs of the RF out ports of the FSS RFPDs, SN009 (North) and SN010 (South), when cavities were locked or unlocked.

Measurement:

• Configuration files for all measurements are present in the Data directory.
•  ZFDC-20-5-S+ 19.5 dB directional coupler was used.
• 19.5 dB was added to all measurements to correct back from coupled value.
• Note that in case of measurement with unlocked cavities, attenuation of 10 dB was used for North side as it was overloading Ag4395A. hence different noise floors.

Inference:

• No significant oscillations above 100 MHz, in particular around 300 MHz or 350 MHz.
• When unlocked, the second harmonic is 50.7 dB lower than the first harmonic for North side and 53.7 dB lower for South side.
• However, when locked, the second harmonic is only 18.1 dB lower than the first harmonic for North side and 6.9 dB lower for South side.
• Last measurement was taken in CTN:2470 where the difference between first and second harmonics during lock was ~12 dB for North and ~17 dB for South.
• But I don't have very good confidence with the last measurement as it was foolishly taken in Noise mode and had different IF Bandwidth than the present measurement.
• With -50 dBc SFDR (Spurious Free Dynamic Range) of MAX4107, I expect that the distortion signal generated at the 1-Omega frequency would be -90.95 dBm for North and -96.5 dBm for South.
• This suggests that the non-linear distortion component in the error signal would be about 59 dB lower than the actual signal for North side and 53 dB lower than the actual signal for South side.

Data

Relevant elog post:

CTN:2470 FSS Diagnostics - RFPD RF Ouput under inspection

Today Koji came to the lab to help me out with the FSS and give me few tips.

We made some changes to the boost stage at U7 in North FSS Box board D040105-C. All the changes were made by Koji personally. He replaced Koji showed me how to solder wires and SMD components so that I can do it better next time. We soldered the wires for the boost switch, replaced the resistor R29 with a thin film 5.6 kOhm resistor and replaced the capacitor (however the older capacitor was fine too). Overall, we made the arrangement of the components neater in the stage.

We figured that the only way to measure the transfer function of this stage when boost is on is to provide some reasonable offset as well so that the opamp doesn't saturate at DC. Hence, I do not have a good measurement of the stage before the changes were made. But after the touch-up, the stage is very close to the expectation as shown in the plot.

Data

Relevant elog posts:

CTN:2384 : Mentioned that something is wrong with the boost stage. But It was just the issue of measuring it wrong.