40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  SUS Lab eLog, Page 34 of 37  Not logged in ELOG logo
IDup Date Author Type Category Subject
  1746   Thu Aug 3 19:03:00 2017 ranaNoise HuntingCracklewhitening

Whitened specgram please to help see small differences. Also your noisy / quiet spectra look like they are contaminated at the low frequencies by the leakage from low frequencies due to the FFT window being too short. Perhaps retry with a 8 second FFT window to see if the effect is different.

  1747   Sat Aug 5 00:31:40 2017 XiaoyueNoise HuntingCracklewhitening

I tried highpass the mich signal with corner frequency at ~ 15 Hz and then take the spectrogram:

 

When whitened, the spectrogram looks pretty stationary. When comparing the noisy and quiet spectra I used 8 second FFT window, we don't see any significant difference:

The 15 ~ 55 Hz non-stationary noise seen in Elog 1745 is indeed a leakage from low frequencies due to FFT window being too short, as Rana suggested. 

Quote:

Whitened specgram please to help see small differences. Also your noisy / quiet spectra look like they are contaminated at the low frequencies by the leakage from low frequencies due to the FFT window being too short. Perhaps retry with a 8 second FFT window to see if the effect is different.

 

  1748   Sun Aug 6 23:46:00 2017 XiaoyueDailyProgressCrackleDiagnose the noise plateau - miscalibration

We have suspected if the noise plateau is caused by miscalibration of the mich signal:

We first improved the lock loop by deleting non-necessary resonant gain filters and sharpening the notch filters that are eating up too much phases. By doing so we were able to improve the phase margin at UGF ~ 60 Hz from 30 to 39 Hz:

Then, I look into the closed-loop transfer function measurement to check on the actuation transfer function calibration. I remeasured and fit for the plant model and implemented the calibration to get a much better match between measured and calibrated CLTF (left is a bad fit for the actuation function, right is an improved fit):

 

  

We see that the noise bump disappears, mainly because of the former miscalibration at lower frequency range 10 ~ 20 Hz, which is causing a fake dip in sensitivity. After recalibration, I compare the current mich sensitivity with the good measurement runs we had back in May. We see the low frequency part is slightly improved, while the high frequency noise indicate a margin to improve the arm symmetry and suppress further the frequency noise.

  1749   Tue Aug 8 23:38:41 2017 XiaoyueDailyProgressCrackleCurrent monitor excessive noise issue

I analyzed the data taken from Aug 6. The demodulation analysis of 22 segments of driving ON and OFF data shows no driving modulated noises. I first noticed that the difference between maximum and minimum of the spectrum in frequency regime 10 ~ 60 Hz is very large, so I looked into the individual spectrums of the 22 segments and found that the noisy ones are all from driving ON (odd number) segments. Notice the spectrum is taken after seismic noise subtraction. I first checked the subtracted mich signal coherence with beam jittering (QPD X and Y) and fround very low correlation. Then I checked the current monitor sum and found high coherence with subtracted mich, as well as an obviously increment of noise during driving ON segments. When discussing with Gabriele, we compared the sample driving ON data with data back in May (2017_05_16 specifically). The current monitor sum with low frequency drive shows significantly noisier behavior than what we had before.

 

We checked that, during the driving ON segments, the coil HF, LF channels are all working as normal. The locking servo is outputing the same level of control signal before. Something must went wrong with the coil Z DAC (cannot be currmon ADC because the noise is going into mich). 

To check, I injected a A=32000, f=0.0317 line excitation to the coil HF Z1 and Z2 channels separately (figure below on the left). Ref0 (green) and Ref1 (brown) curves are the reference spectrum when there's no signal in. Ref2 (magna) and blue curve are the spectra with excitation ON. We see that Z1 currmon is sensing excessive noise in frequency range > 15 Hz. This very likely corresponds to the extra noise seen in both currmon sum and mich signal. To double check, I injected a A = 32000, f = 0.0317 line directly to Z1 currmon channel, with adjusted gain to match the same level of low-frequency ( < 15 Hz) currmon output. The result is shown as the red curve in the figure below on the right -- we don't see any extra noise, which means that coil Z1 HF DAC is causing the problem.

** Note that the spectra below are limited by spectral leakage. However, the difference between currmon Z1 and Z2 outputs already revealved the problem. Just for debugging purpose, I didn't take care of the spectra leakage problem

 

After finding the source of problem, I switched the coil Z1 HF DAC channel (DAC#10 = STP10) with the free DAC port coil A (DAC#0 = STP0). After swapping the channels, I injected the same 0.0317 Hz excitation to coil A channel and observe the same noisy behavior with currmon Z1 output (red curve below, overlapping with the reference noisy one in magna):

I got confused and went to check the downstream. I shuffled a bit the DAC connections to the board in the back of the table, and the problem is solved (red curve below)... The electronics at the back of the table is in open air and is in a not very stable configuration. Maybe a lesson to learn is: be very careful working around the electronics at the back to the table... and always check if everything is working alright after working a lot around that area.....

With everything in a good status, I started a crackling noise measurement with f = 0.0317 Hz,  A = 32000 cts from 12:08 am 0809, logged to crackle_2017_08_08. A brief investigation of the 18 segments of data taken so far shows no excessive noises in mich anymore.

  1750   Thu Aug 10 22:56:21 2017 XiaoyueDailyProgressCrackleNoise demodulation 'failure'

Today I analyzed the data taken since Aug 9 00:08 am (logged as 2017_08_08). Strangely, there were no modulated noises shown up in either mich or currmon channel at the first place. It's OK we don't see any excited noise in mich but not reasonable if we don't see same level of demodulation noise in current monitor sum -- in the May data we had clear demodulated noise in all frequency components with the same driving frequency (f_drive = 0.0317 Hz), same driving amplitude (A = 32000 cts) current monitor data. I looked into the spectrogram of current monitor data and clearly see the slowly varying upconversion noise (Left figure for the entire 1800 second sample driving ON segment data, Right figure for a zoom in)

 

The quick analysis of the demodulation noise period is ~ 27 s, which corresponds to a frequency of ~ 0.037 Hz. Then I realized that I prescribed a driving frequency of 0.0371 Hz rather than 0.0317 Hz as I should have... This might explain why a higher currmon sum (0808 currd) noise is seen in driving ON segments compared to what we had before (0516 currd) -- because the notch filter at 31.7 mHz in coil HF does not match the actual driving at 37.1 mHz:

Then I applied the 0.0371 Hz demodulation analysis to 2017_08_08 data. The left figure shows mich demodulation results in solid and currmon in dash. We see that the DAC noise nicely explains the 2FQ and 4FQ demodulated results. Right figure shows the subtracted mich - currmon demodulation results. (While writing, I just realized that I didn't use the updated actuation model when projecting the current monitor demodulation results.. will update it tomorrow) At the same time, I started 2017_08_10 run with the correct driving frequency f_drive = 0.0317 Hz.

    

  1751   Sun Aug 13 21:29:06 2017 XiaoyueDailyProgressCrackleResidual current monitor excessive noise

When comparing the 2017_08_10 (Right figure below) data with some old good measurements from May, 2017_05_16 (Left figure below) for example, I noticed that the 2FQ and 4FQ current monitor sum demodulation results have yield negative signs rather than positive ones as before. Also it's clear that in 2017_08_10 measurement run the 2FQ modulated DAC noise dominates the 2FQ mich modulated noise. This indicates that we might have some unresolved current monitor excessive noise that was not there back in May.

 

Recalling the current monitor spectrum analysis for 2017_08_08 measurement run (previous elog), we do have some excessive current monitor sum noise in frequency regime > 10 Hz, which I thought might be caused by the not properly notched driving excitations leaked into the coil Z HF channels. Compared to 0808 data (yellow curve below), 0810 (green curve below) is less noisy with the correct driving frequency matching with the notch filters in coil Z HF channels, but we still have excessive noise at frequency > 20 Hz. I suspect still the coil Z DAC connectors, so I shielded the Z1, Z2 connectors using heat shrink, and reshuffled the cables to make the electronics configuration more stable. Then the noise is largely reduced as shown by 0813 data (red curve below) and we even get a lower DAC noise than before. I checked that when driving is OFF, the current monitor sum has a spectrum as data1 (black curve below) in all cases.

I started another f=0.0317 Hz,  A= 32000 cts test (logged as 2017_08_13 measurement run) today ~ 6 pm to check if we can recover the same current monitor sum demodulation results as before.

  1752   Mon Aug 21 23:54:21 2017 XiaoyueNoise HuntingCrackleNoise budget of Crackle2.1 w/Maraging blades

Using half-hour driving-off segment 2017_08_21 data during gps time [1187378134 1187378134+1800], I generated a noise budget for Crackle2.1 setup with maraging steel blades:

First of all, I did the seismic noise subtraction by estimating the transfer function estimation from the three-axial accelerometer measurements to the michelson signal. The seismic noise contribution is subtracted in fourier space and then converted back to real-time space. The details of seismic noise subtraction is described in Elog 1664.

I generated the current noise budget following the same procedure as described in Elog 1672​ excpet for some minor changes:

  • Laser intensity noise: I think I underestimated the laser intensity noises in former analysis. First of all, I assumes a flat shape of the transfer function and arbitrarily trust only the > 100 Hz measurements. In fact, the coherence between laser power and the mich signal is low everywhere; however for some segment when laser becomes instable, we can see harmonic peaks in intensity measurements that match frequencies well with the peaks shown up in michelson sensitivity. I therefore adjusted the frequency and coherence thresholds (that we use to average for the linear gain factor from intensity noise to mich) based on a match between those peaks -- since we can be pretty confident that those peaks comes from intensity noise. Now the criteria is freq > 1000 Hz and coh > 0.3.
  • Actuation (DAC) noise: I used the measurement directly from the current monitor sum and project the spectrum to mich via the actuation model. 
  • Laser frequency noise: I assumed a free running frequency noise of 100 Hz/rHz * (100 Hz/fr) that is typical for Nd:Yag master laser. We can measure the transfer function from laser frequency noise to mich signal when injecting frequency noise to the laser source. We could safely assume flat shape when trusting measurements with OK coherence (coh>0.6) and project the free running frequency noise to michelson noise.

        

Note in the noise budget, I plot actuation and frequency noise in dash lines because they are rough estimations. The actuation noise is only an upper limit because we are limited by sensing noise at high frequencies, while for the frequency noise we don't have direct measurement and got to make assumptions about the free-running frequency noise of our laser.

Comparing the mich sensitivity (post seismic noise subtraction) and the sum of all investigated noises, we still have some discrepancy that cannot be explained by the current noise studies...

I wrapped the code in <budget_170822.zip>. You will have to download the 0821_2 segment data HERE. Simply run run_noise_budget.m will generate the same noise budget plot shown in this log.

  1753   Tue Aug 22 23:22:46 2017 XiaoyueDailyProgressCrackleCombine multiple demodulation analysis

I combined demodulation results from multiple test runs by taking the demodulation amplitudes measured from each segment, subtracting the projected DAC noise:

ci_min = ci_mich_min - ci_currmon_max

ci_max = ci_mich_max - ci_currmon_mi f

and then do the student-t test on the new aggregated demodulation results over driving ON and OFF segments ci= mean(ci_min, ci_max). Sample combined demodulation results (with DAC noise subtracted) for driving frequency 0.0317 Hz tests in frequency band 30 - 35 Hz is shown below. There are in total ~ 290 1/2-hr segments data combined from three test runs 2017_05_16, 2017_08_10, 2017_08_13.

The final student-t test results are shown below:

  1754   Mon Aug 28 20:54:56 2017 XiaoyueSummaryCrackleCrackling noise measurement run #2

Below is a table summarizing the second round of crackling noise measurements

Start time Freq [Hz] A [cts]
2017-07-31 10:30pm 0.0371 32000
2017-08-04 04:10pm 0.0371 32000
2017-08-06 10:20pm  0.0371 32000
2017-08-09 00:08pm  0.0371 32000
2017-08-10 05:38pm 0.0317 32000
2017-08-12 06:12pm 0.0633 32000
2017-08-13 05:45pm 0.0317 32000
2017-08-16 04:50pm 0.0633 32000
2017-08-18 05:58pm 0.0950 32000
2017-08-21 11:43pm 0.1267 32000
2017-08-24 01:04am 0.1584 32000
2017-08-25 04:03pm 0.1900 32000
2017-08-27 02:27pm 0.2217 32000
2017-08-28 04:38pm 0.6330 32000

I combined all the good demodulation results from the tests with driving amplitude A=320000, driving frequency F =

0.0317 Hz (2017_05_16, 2017_08_10, 2017_08_13),

0.0633 Hz (2017_05_18, 2017_08_16),

0.095 Hz (2017_05_30, 2017_08_18).

using the analysis method described in previous elog (Elog 1753), and then do the power-law n = G f^(-k) fitting with exponent k = 3 on the demodulated noises for each driving frequency components as described in Elog 1737, as a way to quantify and compare the noise levels for different driving frequencies. The fitting results are shown in the figure below to the left. The fitted gain G (in logarithm scale) vs. the driving frequency G is plotted in the figure below to the right. We see very consistent demodulated noises at 1FI, 2FI, 2FQ components as before. However, within the error bar, no frequency dependency is discernible.

  1755   Thu Aug 31 22:47:17 2017 XiaoyueDailyProgressCrackleLower driving frequency simulation

The combined demodulation results seem to follow a slightly increasing trend in all demodulation noises over frequencies (see previous elog), though not significant. However, we saw a reverse decreasing trend in the microscopic simulation results (see elog 1729). However, all the simulations are done with driving frequencies faster than the frequencies prescribed in the experiment. In order to compare apple to apple and to investigate more carefully the frequency dependency, we ran a series of low-frequency driving simulation (at the same time we are pushing the driving frequency to its higher end in the experiment, more results come later). An investigation of the demodulated amplitude measured as the power-law fitting gain in the experiment, and the mean demodulated amplitudes in all frequency-bands (since there's no spectrum-frequency dependency) in the simulation, is plotted for each demodulation components versus driving frequencies. Left figure below shows the experimental results, while Right one presents simulation result. 

 .   

The results are different but similar in the way that 1FI, 2FQ have same signs as well as relative magnitude, and all the noise amplitudes (ignoring sign) follow similar increasing trend. I reconstruct the signal using the demodulated amplitudes:

y2= P(1)*sin(2*pi*tt)+P(2)*cos(2*pi*tt)+P(3)*sin(4*pi*tt)+P(4)*cos(4*pi*tt)+P(5)*sin(6*pi*tt)+P(6)*cos(6*pi*tt)+P(7)*sin(8*pi*tt)+P(8)*cos(8*pi*tt);

 

and compare the experiment vs. simulation results on top of drive, using 0.19 Hz ~ 1.2 rad/s demodulation results. Signals are scaled arbitrarily for visualization purpose. 

 

  1756   Sat Sep 2 21:23:55 2017 XiaoyueDailyProgressCrackleCurrent monitor re-calibration

In the high-frequency driving test (2017_08_28) with F_drive = 0.633 Hz, the actuation noise clearly dominates the michelson demodulated signal. The projected demodulated current monitor sum measurements matches with those of mich except for a slight overshoot in 2FI/Q and 4FI/Q components, as shown in the figure below. In the left figure, the solid lines are mich measurements while dashed lines represent projected current monitor noise. The right figure gives the current-monitor-demodulation subtracted mich results. In both figure, red/green indicates a positive/sign sign of demodulation.

  

This sligh mismatch between current monitor noise projection and mich signal motivates us to check back the current monitor calibration. (We have checked the actuation function and optical gain calibration recently so they should be good). I first measured the transfer function SERV_OUT/CURRMON_Z1/2_IN1, which should give directly us the calibration for current monitor input in units of counts as of the locking force in units of uN, and compared them with the current cts2uN calibration filter as shown in the figure below on the left:

  

The magnitude of the measured TF matches well to the digital filter TF; however, the phase was not fitted very well at frequencies higher than 20 Hz. This can be viewed more directly when we measure SERV_OUT/CURRMON_Z1/2_OUT shown in the figure above on the right. There's a phase rotation away from zero starting from frequency ~ 20 Hz, but in the same way in Z1 and Z2 calibration. Since we are looking at spectrum in the end, this delay should not affect our demodulation analysis results, so the mismatch observed between actuation noise and mich noise is not due to the current monitor mis-calibration. 

In any case, I refit the calibration TF from the measurement to improve the fit for the phase rotation at high frequency range:

  

I implemented the new filter from 2017-08-31 9:40 pm.

  1757   Mon Sep 11 01:13:14 2017 XiaoyueDailyProgressCrackleSwitch to highC blades

I want to do another series of measurement with highC steel blades with quasistatic load to ~90% yielding. I started switching the sample in-situ from Friday afternoon. Now I already finished:

  1. Fixing the optical board in place and dismantling the maraging steel setup
  2. Mounting the highC steel blades
  3. Suspending Al blocks
  4. Balancing the board
  5. Centering block OSEMs
  6. Roughly aligned the Mich

Still to do:

  1. Fine tune the Mich
  2. Lock Mich in air
  3. Go into vacuum
  4. re-calibrate compensator, plant, currmon etc...
  5. Start the measurements
  1758   Tue Sep 12 22:33:04 2017 XiaoyueDailyProgressCrackleSwitch to highC blades

I have finished fine-tuning the michelson alignment. The setup is in vacuum now.

The michelson can be easily locked in air or in vacuum, but there is some instabilities at the beginning. With the help from Gabriele today, we diagnosed that a complex structure at ~ 195 Hz is causing the problem. We first lowered the laser beam power to avoid any possible saturation issue, and locked the michelson using low gain. With only lock filter and single-frequency notch filter at ~ 195 Hz engaged, we locked the mich in reasonably stable and low-noise configuration. Then we were able to measure the plant transfer function with extra noise injected at ~195 Hz to achieve a good coherence around the structure. High frequency resolution and long enough averaging also helps. We fitted the structure using a 12th order model, using the TF measurement in frequency range 100 ~ 250 Hz with good coherece coh>0.7:

The loop is nicely compensated using the inverse of the fit. Then I crancked up the laser power back to the same level as we had in the maraing stee configuration (main beam power ~ 17 mW). I changed the servo gain to achieve a UGF ~ 60 Hz, according to the open-loop TF measurements. One strange thing to note is that, the UGF monitor is outputing (constantly) ~ 35 Hz and is not reflecting the in-time change of the UGF. I don't have a clue why this is happening though..

I redesigned the resgain filters according to the low-frequency RMS behavior in error signal. Figure below compares the RMS spectrum before applying the resgains filters in pink and post in green.

Then I recalibrate the actuation function for calibration purpose, and checked the mich signal calibration by measuring the closed-loop transfer functions (CLTF). The measured CLTF (serv_in2/serv_exc) matches very well with the calibrated CLTF (serv_in1/mich_DZ) in magnitude.:

I started a measurements with F_drive = 0.19 Hz, A = 16000 cts. Note the amplitude of 16000 cts gives a similar pk2pk amplitude of ~ 30 um as we had for maraging steel blades.

Quote:

​Still to do:

  1. Fine tune the Mich
  2. Lock Mich in air
  3. Go into vacuum
  4. re-calibrate compensator, plant, currmon etc...
  5. Start the measurements

 

 
  1759   Wed Sep 13 07:46:35 2017 GabrieleDailyProgressCrackleSwitch to highC blades

When the locking loop is changed, you need to recalibrate the UGF monitor. It simply monitors the ratio of the calibration line in the IN1 and IN2 signal, and how this ratio translates to the UGF depends on the loop.

The loop is nicely compensated using the inverse of the fit. Then I crancked up the laser power back to the same level as we had in the maraing stee configuration (main beam power ~ 17 mW). I changed the servo gain to achieve a UGF ~ 60 Hz, according to the open-loop TF measurements. One strange thing to note is that, the UGF monitor is outputing (constantly) ~ 35 Hz and is not reflecting the in-time change of the UGF. I don't have a clue why this is happening though..

 

  1760   Mon Oct 9 22:06:52 2017 XiaoyueDailyProgressCracklePlant TF drifted

The crackle measurements fail autolocking. I remeasured the plant and open-loop transfer function and found the plant structures at high frequency changed.

I re-fit for the compensator and then I can engage the boost again. I also updated the calibration. After recovering the low-noise lock configuration, the low frequency is slightly noisier than before. I am not sure if it's due to scattering or not. I am leaving the setup running a measurement overnight so I can check the data for diagnosis tomorrow.

 

  1761   Tue Oct 17 14:56:53 2017 GabrieleComputingCrackleCoil current monitors

Changed CURRMON_Z1 gain from 1 to 0.854

  1762   Thu Oct 19 23:55:33 2017 XiaoyueDailyProgressCrackleRecover highC measurements

highC measurements became very noisy since ~ Sep 30. I opened the chamber last Friday (Oct 13) and found one Y2 magnet slightly touching. By translating the OSEM I recovered the slow-fringe configuration last Friday. The plant TF changed again. I remeasured and fitted the compensator /calibration TF. The mich is still noisy at the beginning. By recentering the beam referring to both QPD and PD readings helps reduce the noise back to reference level. Right before I called for the day, coil Z2 stopped working... Gabriele helped me resolved the issue on Monday (Oct 16) by magic touching on the coil Z1, Z2 driver connectors. However, the lock became very fragile. The mich fringes became very fast too.

I went back to lab today and found the problem lies probably in Z2 connector again -- firstly I noticed that the Z2 damping filter output is 10x larger than Z1 control output. When reducing the damping gain, the fringes became slower, which means the configuration is still free. I tried touching Z2 connector and occasionally I saw slow and stable fringes. I think it's time to fix permanently the loose connector issue with the coil driver board which is from time to time causing all sorts of problems now -- will do it tomorrow.

  1763   Fri Oct 20 14:01:35 2017 ranaMiscGeneralSR560s removed

There were a couple of SR560 in the lab which I notice have been unplugged for months. Since this slowly degrades the batteries and causes us to spend money/time to replace them, I have moved these to the EE shop and put them on charge.

please plug these in whenever they are not in use

  1764   Mon Oct 23 23:30:37 2017 XiaoyueDailyProgressCrackleRecover highC measurements

I successfully recovered the measurement config on Friday (Oct 20) by carefully the beams according to QPD and PD readings, but clearly the noise degrades over time — I lost the low-noise config very soon after ~ a day. Below red curves show a comparison between quiet (after alignment on Oct 20) and noisy (today) mich spectrum as left vs. right figures, where blue curve is a low-noise reference spectrum:

 
I think there are two most-likely explanataion for the drift:
1. We have some dusty optics in the chamber now. When alignment drifts and the beam hits the scattering objects the measurement becomes noisy.
2. We have slight touching problem again due to highC creep.
For this week I will check first if the low frequency noise is indeed caused by light scattering or not. If so I will open the chamber and clean the optics carefully. If not I am thinking about replacing the 90% highC steel with the 75% ones so we don’t suffer too much from creep, and therefore touching problem.

  1765   Tue Oct 24 23:16:54 2017 XiaoyueDailyProgressCrackleRecover highC measurements

I checked if there's any scattering by investigating the spectrogram (similar anlaysis in Elog 1747):

I don't see very significant nonstationary behavior. The noise does not scale with motion of end mirrors either.

I checked that both QPD and PDs, as well as the plumb line are centered well, which means the alignement didn't change much.. Another possibility is touching: I opened the chamber and found the magnets all free, except X1 magnet was close to the lower end. I recentered it, cleaned all optics --  in fact the two folding mirrors and the first steering mirror have dusts on them. I then closed the chamber. The noise is as low as the reference again (red now, green yesterday). I started running a measurement and will check the robustness of this configuration tomorrow morning.

Quote:

I successfully recovered the measurement config on Friday (Oct 20) by carefully the beams according to QPD and PD readings, but clearly the noise degrades over time — I lost the low-noise config very soon after ~ a day.
I think there are two most-likely explanataion for the drift:
1. We have some dusty optics in the chamber now. When alignment drifts and the beam hits the scattering objects the measurement becomes noisy.
2. We have slight touching problem again due to highC creep.
For this week I will check first if the low frequency noise is indeed caused by light scattering or not. If so I will open the chamber and clean the optics carefully. If not I am thinking about replacing the 90% highC steel with the 75% ones so we don’t suffer too much from creep, and therefore touching problem.

 

  1766   Wed Oct 25 23:30:43 2017 XiaoyueDailyProgressCrackleRebuild connectors

The low-frequeny noise builds up again... I have no clue what could be the problem. The only thing left on my list if to rebuild the connectors for the coil driver /current monitor boards, and I did -- to refresh the memory, the 2-pin connectors we had for the coil driver input and current monitor output does not fit to the 4-pin sockets we have on the PCB board. We had to cut the cage and bond the 2-pin connectors together, which risks short circuiting the four pins actually. In this case, I basically build external 2-pin sockets separately:

<photos here>

I recovered the low-noise config again at the beginning, but as time goes, the noise becomes higher again (checked again the next day Oct26 11am). However one thing seems to be resolved is the current monitor measurements, where the demodulation amplitude in 2FQ component flips sign for the measurements since Sep 30. After rebuilding the connector we again have clean measurements on actuation noises as before -- figure below compares the demodulation analysis no Aug 18 (maraging steel reference), Oct 24 (before fixing connector) and Oct 25 (after fixing connectors). In the figures, please refer to the dashed lines for the current monitor sum demodualtion. Red indicates a positive sign for the demodulation amplitude, while green indicates negative.

 .    

 

  1767   Mon Jun 4 11:15:17 2018 gautamLab InfrastructureEquipmentLoanCrackle lab ---> 40m

Johannes and I have taken the following equipment from the crackle lab to the 40m between Friday, June 1, and Sunday, June 3 2018. 

  • 1 Lightwave M126N-1064-200 NPRO (S/N 259) + controller unit. Measured power output was 320mW.
  • 1 AcoustoOptic Modulator, box was labelled "TNI Isomet AOM 1205C-843"
  • 1 AcoustoOptic Modulator RF driver, box was labelled "AOM Modulator Driver Model 232A-1"
  • 1 broadband EOM (Unlabeled, but looks like the NewFocus)
  • 1 Faraday isolator (Unlabeled)

Apart from the NPRO which was taken from the optical bench, everything else was taken from the storage cabinet by the lab entrance.

  1768   Sat Oct 20 00:38:43 2018 DuoHowTo DAC Noise monitor PCB layers

Background: The design of the DAC noise monitor is in the PCB design stage - I am trying to put the circuit on the PCB board in Altium. We use three power voltages to drive the op amps in the circuit: -15 V. +15 V. We also need power ground and signal ground. This circuit is going to replace a part of a big PCB board with other existing circuits.

Question: What are the layers used by the existing design? The DAC noise monitor needs to fit with other parts, so they have to share the same layers. Is there a PCB layout file for the existing design?

In case of absence, I will start with a signal layer, a -15V power layer, a +15V power layer, a power ground, and a signal ground. I googled a bit and they say the cost will be high and five layers might be more than what we want. Besides, I am not sure about the sequence of the layers either. I will start with this in order to proceed in Altium before we figure out what we need to do:

TOP

Signal layer

+15V power

Power ground

Signal ground

-15V power

BOTTOM

How much sense does it make?

  1769   Tue Oct 23 00:57:13 2018 DuoHowTo DAC Noise monitor PCB layers

I proceeded as described below. The routing is completed. All the signal routing is completed. One thing worries me is that I am afraid the signal ground and power ground is yet separated. I do have two internal planes for signal ground and power ground. Should they be connected to the same power input (so that they are just two planes with the same source)? Altium treats all the ground as one net GND. If the answer to the question is yes, I need to figure out how to get Altium separate them. In Altium, you can specifiy which net you connect to, but I did not figure out how to specify which layer. (Maybe I need to create a separated GND net, like PGND/SGND for that?) 

Here is a summary file with the schematics and PCB design: NoiseMonitor.pdf

Also, this is the link to the Wiki page, with more details about this work: https://wiki-40m.ligo.caltech.edu/Electronics/NoiseMonitor

Quote:

Background: The design of the DAC noise monitor is in the PCB design stage - I am trying to put the circuit on the PCB board in Altium. We use three power voltages to drive the op amps in the circuit: -15 V. +15 V. We also need power ground and signal ground. This circuit is going to replace a part of a big PCB board with other existing circuits.

Question: What are the layers used by the existing design? The DAC noise monitor needs to fit with other parts, so they have to share the same layers. Is there a PCB layout file for the existing design?

In case of absence, I will start with a signal layer, a -15V power layer, a +15V power layer, a power ground, and a signal ground. I googled a bit and they say the cost will be high and five layers might be more than what we want. Besides, I am not sure about the sequence of the layers either. I will start with this in order to proceed in Altium before we figure out what we need to do:

TOP

Signal layer

+15V power

Power ground

Signal ground

-15V power

BOTTOM

How much sense does it make?

 

  1770   Wed Oct 24 15:01:50 2018 DuoHowTo DAC Noise monitor PCB layers

As per Chris's suggestions, I replaced the capacitors with surface mount ceramic capacitors, doubled the trace width to 0.5mm and adjusted the routings accordingly. New PCB layout is attached.

  1771   Wed Oct 24 21:55:45 2018 DuoHowTo DAC Noise monitor PCB layers

I forgot to connect the outputs of U1 and U2. It is fixed. I also run the design rule check and verified that all the connections are made. I separated the power ground and signal ground as well. The summary PDF is updated below.

  1772   Sat Nov 24 20:09:14 2018 DuoMisc Noise monitor PCB assembly completed

Progress: the board and components arrived and assembled. Some obvious mistakes are fixed on the next version in Altium.

Next: how to test the board? i.e. How to connect the test instruments (such as spectrum analyzer, DC power supply) to the board? We need connector converters (from BNC to headers female & from BNC to 9 pin D shape male). Or do we have better ways to test it?

Note: Altium footprints for WIMA capacitors are created. Altium test point component is created. These might be useful in the future.

  1773   Sun Nov 25 19:25:37 2018 ranaHowToElectronicsNoise monitor PCB assembly completed

you can just use some BNC clip doodles (mini grabbers, etc). Go directly from the test equip (scopes, analyzers) to the pins on the board. Or if you are able to mount the D-sub connectors, you can use a breakout board. Can borrow from the 40m if you don't have them in WB.

  1774   Fri Dec 7 18:27:27 2018 DuoMisc Noisemon problem identified and another order placed

Progress: The reason why the board from oshpark did not work is found. The board has 6 layers, but Oshpark only make 2 or 4 layer boards. They just ignored two layers (the two ground layers) so there is no ground at all on the board.

Some known issues is fixed in the new board (capacitor footprint, connector in the wrong direction). The new board will arrive next Wednesday.

Some good quality connectors are made - next board will be ready to test once arrived.

Next: I plan to put other components into Altium by Wednesday. 

  1775   Wed Dec 12 21:50:44 2018 DuoDailyProgress New board arrived and other components on the board placed in Altium

The new board arrived this afternoon. I tried the connections - it has enough layers and is grounded. I will assemble the board tomorrow. 

In the meanwhile, I have put other unchanged components on the board into Altium, not quite finished (put them in schematics but not PCB, gives me error when importing changes). I will prioritize assembling the new board.

A picture of the board is attached.

  1776   Sat Dec 15 15:30:44 2018 DuoDailyProgress Noise in the last Low pass sallen key filter

Progress:

1. Board assembled

2. One design error found and fixed in the instrumental amplifier. Now the instrumental amplifier is working

Issue:

Noise above 100Hz (pass band 20-100Hz), as shown in the transfer function in the picture. 

The noise comes from the last stage of the circuit: the low pass sallen key filter. The first two high pass stages works well.

(structure of the circuit: differential input - passive filter - instrumental amplifier - high pass - high pass - low pass - output)

I have tried

1. Checking the connections - the connections are good

2. Replacing the opamp - did not work

  1777   Mon Dec 17 11:20:32 2018 ranaDailyProgressElectronicsSR785 netgpibdata
  1. add photo of stuffed board
  2. add time series of output with input terminated
  3. check for internal saturations
  4. use the software from Craig to download and plot the SR785 data
  1778   Thu Jan 3 15:33:40 2019 DuoDailyProgress New, Full noisemon completed

Here is a full version of the noisemon, with four channels and the power regulator. I did the routing again since the previous routing 1) did not leave enough space for connectors/other components; 2) Altium does not transfer properly from the schematics to the PCB layout when expanding to 4 channels.

  1779   Thu Jan 3 16:24:22 2019 DuoDailyProgress Noise in the last Low pass sallen key filter

The reason for this problem was found. The gain of the sallen key filters was too high. There is an intrinsic limit of the sallen key filters - they cannot have a gain more than a certain value. Otherwise, they will be unstable. See this TI document  for details.

Quote:

Progress:

1. Board assembled

2. One design error found and fixed in the instrumental amplifier. Now the instrumental amplifier is working

Issue:

Noise above 100Hz (pass band 20-100Hz), as shown in the transfer function in the picture. 

The noise comes from the last stage of the circuit: the low pass sallen key filter. The first two high pass stages works well.

(structure of the circuit: differential input - passive filter - instrumental amplifier - high pass - high pass - low pass - output)

I have tried

1. Checking the connections - the connections are good

2. Replacing the opamp - did not work

 

  1780   Fri Jan 4 22:18:53 2019 DuoDailyProgressElectronicsUpdates

Photo attached in attachment 1.

The times series output is shown in attachment 2 (Attached picture since I cannot get data from the oscilliscope, which requires floppy disk data transfer). There is an 87mV/rtHz oscillation at about 1.4MHz (op amp oscillation?).

I tested the noise with SR785, both time and frequency domains, in attachment 4 and 5. In time domain, I only see the 60Hz noise, not the 1.4MHz one (maybe because SR785 does not reach that high frequency). In frequency domain, noise in the passband is generally less than 10uVrms/rtHz. With a gain of 125, 10uV/rtHz corresponds to roughly 100nV/rtHz.

Attachment 3 is the transfer function, which is as we wanted, with a gain 2.5 less since this version does not have the last stage.

Internal saturation: what input do we use to test it?

Note: the noise FFT measurement has a lot of time dependence. It fluctuates a lot. Also sometimes (just a few hours before this measurement), I cannot reproduce the noise measurement mysteriously - it gives me much higher noise.

Quote:
  1. add photo of stuffed board
  2. add time series of output with input terminated
  3. check for internal saturations
  4. use the software from Craig to download and plot the SR785 data
  1781   Sun Jan 6 15:37:39 2019 DuoDailyProgress Isolated test of the oscillating op amp

It seems there can be multiple reasons for an op amp to oscillate. I wanted to identify the nature of the oscillation.

I want to isolate one stage and see what is going on. I used the extra empty board and assembled the last stage on it. Putting in nothing at all (the input is GND), I get a signal of 5.792MHz, 321mV at the output.

Now that the problem is even more clear, I will keep looking into this.

  1782   Tue Jan 8 22:47:07 2019 DuoDailyProgress Isolated test of the oscillating op amp

After a few days of struggling (and essential help from Chris), mystery is resolved. Fortunately, the oscillation does not have much to do with my circuit design. It is caused by the RLC resonance formed by 1) the inductance of the parallel wires + 2) capacitance of the signal ground plane and the power ground plane.

As is seen in the picture, I twisted the two grounding wires together (reduce the inductance) and the oscillation is gone.

You can also connect the planes on the board (removing the capacitance) and the oscillation will disappear as well.

Quote:

It seems there can be multiple reasons for an op amp to oscillate. I wanted to identify the nature of the oscillation.

I want to isolate one stage and see what is going on. I used the extra empty board and assembled the last stage on it. Putting in nothing at all (the input is GND), I get a signal of 5.792MHz, 321mV at the output.

Now that the problem is even more clear, I will keep looking into this.

 

  1783   Sun Jan 20 16:00:37 2019 DuoDailyProgress Update on the noisemon board

Two more oscillations problems are resolved, and there is no more oscillations. In the time series (the inputs are terminated), we see only the 60Hz noise.

- Some big bypass capacitors are used to regulate the power.

- A small capacitor is attached to the negative feedback loop in the second HP filter.

New board/components arrived. I will assemble and test them immediately.

  1784   Fri Feb 1 12:35:13 2019 not DuoDailyProgress Update on the noisemon board

Duo's noisemon has been in the EE shop/cryo lab for testing.  It is a drop-in replacement for the existing monitor board, including both noisemon and Vmon/Imon/RMSmon circuits for all four channels.

Duo is still working on a log entry summarizing the performance of the new board vs simulation.  This entry shows some measurements of the performance of the new board vs the old board.

Attachments:

  1. Transfer functions of the new noisemon board, in coil driver state 1.
  2. Noise spectra of the new noisemon board, in coil driver state 1, with and without DAC noise input.
  3. Transfer function of an old noisemon board, in coil driver state 1.
  4. Noise spectra of an old noisemon board, in coil driver state 1, with (REF traces) and without (live traces) DAC noise input.
  1785   Tue Feb 5 20:55:34 2019 DuoSummary Noisemon test results

Based on the test results posted, I did the following analysis:

1. Compared measured transfer function to the LISO calculations. Attachment 4 and 6. The transfer functions match well with LISO.

2. Compared measured noise at the output to the LISO calculations. Attachment 1 and 3. The noise is more than LISO calculations by roughly a factor of 2, but I think it is expected - there is coil driver noises (amplified more than 300 times). Also, LISO uses ideal resistors, considering that the noise here is dominated by resistor noise. We also have plots of the noise spectrum with DAC noises injected. In this case, the noise in the passband (20 - 100Hz) is much more, suggesting that the board noise is dominated by the DAC noise.

3. Compared input-referenced measured noise to DAC noise. Attachment 2 and 5. We divided the noise by the transfer function and compare it directly with the DAC noise model. We can see that, in the passband, the board noise is about a factor of 10 less than the DAC noise (channel 2 and 4 has more noise; the signal is polluted by the ADC). 

4. A simple calculation based on the transfer function comparing the ADC noise and the amplified DAC noise. 

DAC noise > 300nV/rtHz. Passband amplification > 50dB > 300. Amplified DAC noise in the passband > 90uV/rtHz, compared to ADC noise 4uV/rtHz.

FYI: 1. I cannot attachment PDF plots directly since it will stuck the elog server. I put some PNG plots, but PDF plots can still be found in the compressed files.

2. Also, channel 2 is more noisey. It comes from ADC not the noisemon.

  1786   Fri Feb 15 18:38:00 2019 ranaNoise HuntingElectronicsNoiseMon nonlinearity?

Before making a wide deployment, we should also test the latest noisemon circuit for downconversion.

  1. Measure the noise output with no DAC signal
  2. Measure the noise output with 0.1x the reference DAC signal (low noise ETM drive)
  3. Drive a line at high frequency in addition to the reference signal
  4. adjust the ampltiudes of the high frequency drive and the reference signal independently and look at how the 20-100 Hz noise changes.
  1787   Tue Feb 19 18:11:43 2019 DuoDailyProgress Tuesday report

Some clean up work on the noisemon is done.

1. Added compensate capacitor.

2. Added mounting holes.

3. Added DCC number. https://dcc.ligo.org/LIGO-D1900052

4. Renumbered the components.

5. Added 0 ohm resistor between power ground and signal ground.

6. Added more test points for the voltage monitor and current monitor.

7. Increased schematics font size.

Next I will create the Bill Of Materials. I need to assemble the manufacturer information and put meaningful and consistent descriptions for the components.

  1788   Mon Apr 1 20:25:20 2019 DuoDailyProgress Changes on noisemon being made

Attachment: new PCB schematics with all the changes made.

We (Chris and I) had a conversation with Rich last week and the following work on the noisemon board has been suggested:

1. Name of power nets: +VCC to +15, -VCC to -15. +V to +18; -V to -18, making it clearer what the power is.

2. Fix the off-grid problems of the schematics.

3. Draw the circuits on the schematics in the standard way. (Rich gave me a bunch of snippets that shows the standard way to draw the circuits, like how to draw a sallen key filter)

4. Ground the shells of the D connectors.

5. Add 1 Ohm resistors at the inputs of the power regulators

6. Use polymer tantalum with at least 35V rating. Previously we are not using polymer ones. Rich said the ones (non-polymer ones) we were using burn and explode sometimes.

7. Add "No error checking" for those pins not being used (e.g. unused op amp pins)

8. Disassemble the "repeat()" in the sheet symbol. Making four sheet symbols and connect them directly to the connectors.

9. Change the outputs of the current monitor, noise monitor and the voltage monitor to differential. Previously we had one of the pins of the D connectors pairs grounded. Now we add a differential driver at the end. It doubles the gain and the range.

Rich said my PCB routing was OK, so all the changes can be reflected on the schematics. I have made all the changes on the schematics (I do have the previous version). The current schematics is attached.

However, "#9 change the outputs to differential" requires a lot more space and the current PCB routing does not have enough free space between the components. Thus, this requires routing the whole PCB again, which is what I am working on now.

  1789   Wed Apr 17 16:34:37 2019 DuoDailyProgress Noisemon update

We added differential drivers at the outputs of all the monitors. After that the routing becomes impossible at the output connectors.

I replaced the signal ground with an additional signal layer, and reversed the order of the channels on the layout.

Having exhausted all the possible routing tricks, I finally managed to connect the whole board.

  1790   Sat Jun 8 15:52:48 2019 DuoMisc Arrival date of noisemon

We ordered the board from Screaming Circuits and chose to provide the component ourselves. However, the parcel we ordered from Verical was lost by Fedex on its way to Screaming Circuits. The original delivery date was delay from late May to June 13.

Once the board arrives, we will test the board - TF, noise etc. Any others?

  1791   Fri Jul 12 13:45:39 2019 DuoNoise HuntingElectronicsNoisemon board test plan

1. Transfer functions of 1-4 channels, compare with simulations.

2. Noise of 1-4 channels, compare with simulations.

3. If feasible, nonlinearity.

4. Functionality of fast current, slow current and  voltage monitor channels.

All test results will reply to this post.

  1795   Wed Jul 17 13:35:02 2019 DuoDailyProgress Unknown issue

I connected the DAC to ADC direclty (picture 1) and send a sine signal into the DAC. However, I did not get the sine signal back from the ADC. I sent the signal in X1:CRY-DITHER_W_MOD_EXC, channel 9 of DAC and expect the signal from X1:CRY-E_REFLDC_IN1, channel 16 of ADC. However, picture 2 shows what I get: a constant signal around 4400 counts.

  1796   Thu Jul 18 11:50:43 2019 DuoNoise HuntingElectronicsNoisemon board test plan

Noisemon + Coil driver TF and noise data. This is raw data measured from the lab. TF data is in counts dB. It needs to be converted to volts dB (+6.02dB). The noise is in ADC counts: 2^16 counts is 40V.

Quote:

1. Transfer functions of 1-4 channels, compare with simulations.

2. Noise of 1-4 channels, compare with simulations.

3. If feasible, nonlinearity.

4. Functionality of fast current, slow current and  voltage monitor channels.

All test results will reply to this post.

 

  1797   Thu Jul 18 16:44:30 2019 DuoNoise HuntingElectronicsNoisemon board test plan

Analyzed Noisemon + Coil driver TF. The coil driver is in LP_OFF and ACQ_OFF status. The TFMeasured.txt file in the zip is in counts (ADC over DAC) so we need to add 20log10(2) to convert it to volts.

Quote:

Noisemon + Coil driver TF and noise data. This is raw data measured from the lab. TF data is in counts dB. It needs to be converted to volts dB (+6.02dB). The noise is in ADC counts: 2^16 counts is 40V.

Quote:

1. Transfer functions of 1-4 channels, compare with simulations.

2. Noise of 1-4 channels, compare with simulations.

3. If feasible, nonlinearity.

4. Functionality of fast current, slow current and  voltage monitor channels.

All test results will reply to this post.

 

 

  1798   Fri Jul 19 23:55:19 2019 DuoNoise HuntingElectronicsNoisemon board test plan

Noise compared with LISO.

In the region we care about noise (20 - 100 Hz), we can see it matches well with LISO calculations.

But why not at all frequencies ??

ELOG V3.1.3-